leandrof/linux
1
0
Fork 0
forked from Minki/linux
Code Issues Pull Requests Packages Projects Releases Wiki Activity
6a7d64e3e0
linux/drivers/net/ethernet/intel/e1000/e1000_main.c

5222 lines
143 KiB
C
Raw Normal View History

Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/*******************************************************************************
e100, e1000, ixgb: update copyright header and remove LICENSE This update to the copyright header adds the mailinglist, and aligns it with the kernel licensing as well as remove the offending 'all rights reserved'. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:14 +00:00
Intel PRO/1000 Linux driver
Copyright(c) 1999 - 2006 Intel Corporation.
This program is free software; you can redistribute it and/or modify it
under the terms and conditions of the GNU General Public License,
version 2, as published by the Free Software Foundation.
This program is distributed in the hope it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
more details.
e100, e1000, ixgb: update copyright header and remove LICENSE This update to the copyright header adds the mailinglist, and aligns it with the kernel licensing as well as remove the offending 'all rights reserved'. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:14 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
You should have received a copy of the GNU General Public License along with
e100, e1000, ixgb: update copyright header and remove LICENSE This update to the copyright header adds the mailinglist, and aligns it with the kernel licensing as well as remove the offending 'all rights reserved'. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:14 +00:00
this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
The full GNU General Public License is included in this distribution in
the file called "COPYING".
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
Contact Information:
Linux NICS <linux.nics@intel.com>
e1000: Version bump, contact fix, year string change Add the sourceforge project mailinglist to the contact information. Bump version to 7.0.38-k2 Update copyright string with the new year. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-04-15 02:05:31 +00:00
e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
*******************************************************************************/
#include "e1000.h"
e1000 linkage fix ia64: drivers/built-in.o(.text+0xd9a72): In function `e1000_xmit_frame': : undefined reference to `csum_ipv6_magic' Cc: Auke Kok <auke-jan.h.kok@intel.com> Cc: "David S. Miller" <davem@davemloft.net> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-11-14 15:35:03 +00:00
#include <net/ip6_checksum.h>
e1000: Add support for the CE4100 reference platform This patch adds support for the gigabit phys present on the CE4100 reference platforms. Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-06 14:29:54 +00:00
#include <linux/io.h>
Add appropriate <linux/prefetch.h> include for prefetch users After discovering that wide use of prefetch on modern CPUs could be a net loss instead of a win, net drivers which were relying on the implicit inclusion of prefetch.h via the list headers showed up in the resulting cleanup fallout. Give them an explicit include via the following $0.02 script. ========================================= #!/bin/bash MANUAL="" for i in `git grep -l 'prefetch(.*)' .` ; do grep -q '<linux/prefetch.h>' $i if [ $? = 0 ] ; then continue fi ( echo '?^#include <linux/?a' echo '#include <linux/prefetch.h>' echo . echo w echo q ) | ed -s $i > /dev/null 2>&1 if [ $? != 0 ]; then echo $i needs manual fixup MANUAL="$i $MANUAL" fi done echo ------------------- 8\<---------------------- echo vi $MANUAL ========================================= Signed-off-by: Paul <paul.gortmaker@windriver.com> [ Fixed up some incorrect #include placements, and added some non-network drivers and the fib_trie.c case - Linus ] Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-22 20:47:17 +00:00
#include <linux/prefetch.h>
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
#include <linux/bitops.h>
#include <linux/if_vlan.h>
e1000: Add support for the CE4100 reference platform This patch adds support for the gigabit phys present on the CE4100 reference platforms. Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-06 14:29:54 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
char e1000_driver_name[] = "e1000";
[PATCH] drivers/net/e1000/: possible cleanups This patch contains the following possible cleanups: - make needlessly global code static - #if 0 the following unused global functions: - e1000_hw.c: e1000_mc_addr_list_update - e1000_hw.c: e1000_read_reg_io - e1000_hw.c: e1000_enable_pciex_master Signed-off-by: Adrian Bunk <bunk@stusta.de> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2005-10-30 15:53:34 +00:00
static char e1000_driver_string[] = "Intel(R) PRO/1000 Network Driver";
e1000: fix screaming IRQ VMWare reports that the e1000 driver has a bug when bringing down the interface, such that interrupts are not disabled in the hardware but the driver stops reporting that it consumed the interrupt. The fix is to set the driver's "down" flag later in the routine, after all the timers and such have exited, preventing the interrupt handler from being called and exiting early without handling the interrupt. CC: Anupam Chanda <anupamc@vmware.com> CC: stable kernel <stable@kernel.org> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-11-21 17:54:21 +00:00
#define DRV_VERSION "7.3.21-k8-NAPI"
e1000: sparse warnings fixes Fix sparse warnings and problems from e1000 driver. Added a sparse fix for the module param array index -- Auke Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-10-29 17:46:19 +00:00
const char e1000_driver_version[] = DRV_VERSION;
static const char e1000_copyright[] = "Copyright (c) 1999-2006 Intel Corporation.";
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* e1000_pci_tbl - PCI Device ID Table
*
* Last entry must be all 0s
*
* Macro expands to...
* {PCI_DEVICE(PCI_VENDOR_ID_INTEL, device_id)}
*/
drivers/net/: use DEFINE_PCI_DEVICE_TABLE() Use DEFINE_PCI_DEVICE_TABLE() so we get place PCI ids table into correct section in every case. Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-01-07 11:58:11 +00:00
static DEFINE_PCI_DEVICE_TABLE(e1000_pci_tbl) = {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
INTEL_E1000_ETHERNET_DEVICE(0x1000),
INTEL_E1000_ETHERNET_DEVICE(0x1001),
INTEL_E1000_ETHERNET_DEVICE(0x1004),
INTEL_E1000_ETHERNET_DEVICE(0x1008),
INTEL_E1000_ETHERNET_DEVICE(0x1009),
INTEL_E1000_ETHERNET_DEVICE(0x100C),
INTEL_E1000_ETHERNET_DEVICE(0x100D),
INTEL_E1000_ETHERNET_DEVICE(0x100E),
INTEL_E1000_ETHERNET_DEVICE(0x100F),
INTEL_E1000_ETHERNET_DEVICE(0x1010),
INTEL_E1000_ETHERNET_DEVICE(0x1011),
INTEL_E1000_ETHERNET_DEVICE(0x1012),
INTEL_E1000_ETHERNET_DEVICE(0x1013),
INTEL_E1000_ETHERNET_DEVICE(0x1014),
INTEL_E1000_ETHERNET_DEVICE(0x1015),
INTEL_E1000_ETHERNET_DEVICE(0x1016),
INTEL_E1000_ETHERNET_DEVICE(0x1017),
INTEL_E1000_ETHERNET_DEVICE(0x1018),
INTEL_E1000_ETHERNET_DEVICE(0x1019),
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
INTEL_E1000_ETHERNET_DEVICE(0x101A),
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
INTEL_E1000_ETHERNET_DEVICE(0x101D),
INTEL_E1000_ETHERNET_DEVICE(0x101E),
INTEL_E1000_ETHERNET_DEVICE(0x1026),
INTEL_E1000_ETHERNET_DEVICE(0x1027),
INTEL_E1000_ETHERNET_DEVICE(0x1028),
INTEL_E1000_ETHERNET_DEVICE(0x1075),
INTEL_E1000_ETHERNET_DEVICE(0x1076),
INTEL_E1000_ETHERNET_DEVICE(0x1077),
INTEL_E1000_ETHERNET_DEVICE(0x1078),
INTEL_E1000_ETHERNET_DEVICE(0x1079),
INTEL_E1000_ETHERNET_DEVICE(0x107A),
INTEL_E1000_ETHERNET_DEVICE(0x107B),
INTEL_E1000_ETHERNET_DEVICE(0x107C),
INTEL_E1000_ETHERNET_DEVICE(0x108A),
[PATCH] e1000: Added hardware support for PCI express, 82546GB, and 82571 Fiber Added 82571 fiber to WOL fix for dual port adapters. Added support for 82546GB (Quad Copper). Added PCIe typedef for x2, igp cable length 115, and extended TX CTRL registers. Added parity error detection and PCIe CTRL registers. Added EEPROM config registers. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:21 +00:00
INTEL_E1000_ETHERNET_DEVICE(0x1099),
INTEL_E1000_ETHERNET_DEVICE(0x10B5),
e1000: Add support for the CE4100 reference platform This patch adds support for the gigabit phys present on the CE4100 reference platforms. Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-06 14:29:54 +00:00
INTEL_E1000_ETHERNET_DEVICE(0x2E6E),
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* required last entry */
{0,}
};
MODULE_DEVICE_TABLE(pci, e1000_pci_tbl);
e1000: remove unused code and make symbols static Signed-off-by: Nicholas Nunley <nicholas.d.nunley@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:34 +00:00
int e1000_up(struct e1000_adapter *adapter);
void e1000_down(struct e1000_adapter *adapter);
void e1000_reinit_locked(struct e1000_adapter *adapter);
void e1000_reset(struct e1000_adapter *adapter);
int e1000_setup_all_tx_resources(struct e1000_adapter *adapter);
int e1000_setup_all_rx_resources(struct e1000_adapter *adapter);
void e1000_free_all_tx_resources(struct e1000_adapter *adapter);
void e1000_free_all_rx_resources(struct e1000_adapter *adapter);
[PATCH] drivers/net/e1000/: possible cleanups This patch contains the following possible cleanups: - make needlessly global code static - #if 0 the following unused global functions: - e1000_hw.c: e1000_mc_addr_list_update - e1000_hw.c: e1000_read_reg_io - e1000_hw.c: e1000_enable_pciex_master Signed-off-by: Adrian Bunk <bunk@stusta.de> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2005-10-30 15:53:34 +00:00
static int e1000_setup_tx_resources(struct e1000_adapter *adapter,
e1000: remove unused code and make symbols static Signed-off-by: Nicholas Nunley <nicholas.d.nunley@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:34 +00:00
struct e1000_tx_ring *txdr);
[PATCH] drivers/net/e1000/: possible cleanups This patch contains the following possible cleanups: - make needlessly global code static - #if 0 the following unused global functions: - e1000_hw.c: e1000_mc_addr_list_update - e1000_hw.c: e1000_read_reg_io - e1000_hw.c: e1000_enable_pciex_master Signed-off-by: Adrian Bunk <bunk@stusta.de> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2005-10-30 15:53:34 +00:00
static int e1000_setup_rx_resources(struct e1000_adapter *adapter,
e1000: remove unused code and make symbols static Signed-off-by: Nicholas Nunley <nicholas.d.nunley@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:34 +00:00
struct e1000_rx_ring *rxdr);
[PATCH] drivers/net/e1000/: possible cleanups This patch contains the following possible cleanups: - make needlessly global code static - #if 0 the following unused global functions: - e1000_hw.c: e1000_mc_addr_list_update - e1000_hw.c: e1000_read_reg_io - e1000_hw.c: e1000_enable_pciex_master Signed-off-by: Adrian Bunk <bunk@stusta.de> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2005-10-30 15:53:34 +00:00
static void e1000_free_tx_resources(struct e1000_adapter *adapter,
e1000: remove unused code and make symbols static Signed-off-by: Nicholas Nunley <nicholas.d.nunley@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:34 +00:00
struct e1000_tx_ring *tx_ring);
[PATCH] drivers/net/e1000/: possible cleanups This patch contains the following possible cleanups: - make needlessly global code static - #if 0 the following unused global functions: - e1000_hw.c: e1000_mc_addr_list_update - e1000_hw.c: e1000_read_reg_io - e1000_hw.c: e1000_enable_pciex_master Signed-off-by: Adrian Bunk <bunk@stusta.de> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2005-10-30 15:53:34 +00:00
static void e1000_free_rx_resources(struct e1000_adapter *adapter,
e1000: remove unused code and make symbols static Signed-off-by: Nicholas Nunley <nicholas.d.nunley@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:34 +00:00
struct e1000_rx_ring *rx_ring);
void e1000_update_stats(struct e1000_adapter *adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
static int e1000_init_module(void);
static void e1000_exit_module(void);
static int e1000_probe(struct pci_dev *pdev, const struct pci_device_id *ent);
net/intel: remove __dev* attributes CONFIG_HOTPLUG is going away as an option. As result the __dev* markings will be going away. Remove use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit. Signed-off-by: Bill Pemberton <wfp5p@virginia.edu> Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Bruce Allan <bruce.w.allan@intel.com> Cc: Carolyn Wyborny <carolyn.wyborny@intel.com> Cc: Don Skidmore <donald.c.skidmore@intel.com> Cc: Greg Rose <gregory.v.rose@intel.com> Cc: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com> Cc: Alex Duyck <alexander.h.duyck@intel.com> Cc: John Ronciak <john.ronciak@intel.com> Cc: Tushar Dave <tushar.n.dave@intel.com> Cc: e1000-devel@lists.sourceforge.net Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-03 14:24:25 +00:00
static void e1000_remove(struct pci_dev *pdev);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
static int e1000_alloc_queues(struct e1000_adapter *adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
static int e1000_sw_init(struct e1000_adapter *adapter);
static int e1000_open(struct net_device *netdev);
static int e1000_close(struct net_device *netdev);
static void e1000_configure_tx(struct e1000_adapter *adapter);
static void e1000_configure_rx(struct e1000_adapter *adapter);
static void e1000_setup_rctl(struct e1000_adapter *adapter);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
static void e1000_clean_all_tx_rings(struct e1000_adapter *adapter);
static void e1000_clean_all_rx_rings(struct e1000_adapter *adapter);
static void e1000_clean_tx_ring(struct e1000_adapter *adapter,
struct e1000_tx_ring *tx_ring);
static void e1000_clean_rx_ring(struct e1000_adapter *adapter,
struct e1000_rx_ring *rx_ring);
[E1000]: Secondary unicast address support Add support for configuring secondary unicast addresses. Unicast addresses take precendece over multicast addresses when filling the exact address filters to avoid going to promiscous mode. When more unicast addresses are present than filter slots, unicast filtering is disabled and all slots can be used for multicast addresses. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-11-14 04:54:59 +00:00
static void e1000_set_rx_mode(struct net_device *netdev);
e1000: use work queues E1000 is using several timers that in a follow on patch will need to acquire the rtnl_lock in order to be safe. This patch moves the timer bodies into work queues which will allow the next patch to add rtnl_lock. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-22 18:22:17 +00:00
static void e1000_update_phy_info_task(struct work_struct *work);
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
static void e1000_watchdog(struct work_struct *work);
e1000: use work queues E1000 is using several timers that in a follow on patch will need to acquire the rtnl_lock in order to be safe. This patch moves the timer bodies into work queues which will allow the next patch to add rtnl_lock. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-22 18:22:17 +00:00
static void e1000_82547_tx_fifo_stall_task(struct work_struct *work);
intel: convert drivers to netdev_tx_t Get rid of some bogus return wrapping as well. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-08-31 19:50:55 +00:00
static netdev_tx_t e1000_xmit_frame(struct sk_buff *skb,
struct net_device *netdev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
static struct net_device_stats * e1000_get_stats(struct net_device *netdev);
static int e1000_change_mtu(struct net_device *netdev, int new_mtu);
static int e1000_set_mac(struct net_device *netdev, void *p);
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 13:55:46 +00:00
static irqreturn_t e1000_intr(int irq, void *data);
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
static bool e1000_clean_tx_irq(struct e1000_adapter *adapter,
struct e1000_tx_ring *tx_ring);
[NET]: Make NAPI polling independent of struct net_device objects. Several devices have multiple independant RX queues per net device, and some have a single interrupt doorbell for several queues. In either case, it's easier to support layouts like that if the structure representing the poll is independant from the net device itself. The signature of the ->poll() call back goes from: int foo_poll(struct net_device *dev, int *budget) to int foo_poll(struct napi_struct *napi, int budget) The caller is returned the number of RX packets processed (or the number of "NAPI credits" consumed if you want to get abstract). The callee no longer messes around bumping dev->quota, *budget, etc. because that is all handled in the caller upon return. The napi_struct is to be embedded in the device driver private data structures. Furthermore, it is the driver's responsibility to disable all NAPI instances in it's ->stop() device close handler. Since the napi_struct is privatized into the driver's private data structures, only the driver knows how to get at all of the napi_struct instances it may have per-device. With lots of help and suggestions from Rusty Russell, Roland Dreier, Michael Chan, Jeff Garzik, and Jamal Hadi Salim. Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra, Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan. [ Ported to current tree and all drivers converted. Integrated Stephen's follow-on kerneldoc additions, and restored poll_list handling to the old style to fix mutual exclusion issues. -DaveM ] Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-03 23:41:36 +00:00
static int e1000_clean(struct napi_struct *napi, int budget);
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
static bool e1000_clean_rx_irq(struct e1000_adapter *adapter,
struct e1000_rx_ring *rx_ring,
int *work_done, int work_to_do);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
static bool e1000_clean_jumbo_rx_irq(struct e1000_adapter *adapter,
struct e1000_rx_ring *rx_ring,
int *work_done, int work_to_do);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
static void e1000_alloc_rx_buffers(struct e1000_adapter *adapter,
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
struct e1000_rx_ring *rx_ring,
[PATCH] e1000: Added cleaned_count to RX buffer allocation Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:19 +00:00
int cleaned_count);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
static void e1000_alloc_jumbo_rx_buffers(struct e1000_adapter *adapter,
struct e1000_rx_ring *rx_ring,
int cleaned_count);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
static int e1000_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd);
static int e1000_mii_ioctl(struct net_device *netdev, struct ifreq *ifr,
int cmd);
static void e1000_enter_82542_rst(struct e1000_adapter *adapter);
static void e1000_leave_82542_rst(struct e1000_adapter *adapter);
static void e1000_tx_timeout(struct net_device *dev);
WorkStruct: Pass the work_struct pointer instead of context data Pass the work_struct pointer to the work function rather than context data. The work function can use container_of() to work out the data. For the cases where the container of the work_struct may go away the moment the pending bit is cleared, it is made possible to defer the release of the structure by deferring the clearing of the pending bit. To make this work, an extra flag is introduced into the management side of the work_struct. This governs auto-release of the structure upon execution. Ordinarily, the work queue executor would release the work_struct for further scheduling or deallocation by clearing the pending bit prior to jumping to the work function. This means that, unless the driver makes some guarantee itself that the work_struct won't go away, the work function may not access anything else in the work_struct or its container lest they be deallocated.. This is a problem if the auxiliary data is taken away (as done by the last patch). However, if the pending bit is *not* cleared before jumping to the work function, then the work function *may* access the work_struct and its container with no problems. But then the work function must itself release the work_struct by calling work_release(). In most cases, automatic release is fine, so this is the default. Special initiators exist for the non-auto-release case (ending in _NAR). Signed-Off-By: David Howells <dhowells@redhat.com>
2006-11-22 14:55:48 +00:00
static void e1000_reset_task(struct work_struct *work);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
static void e1000_smartspeed(struct e1000_adapter *adapter);
e1000: De-inline functions to benefit from compiler smartness De-inline functions to benefit from compiler smartness Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-04-15 02:04:52 +00:00
static int e1000_82547_fifo_workaround(struct e1000_adapter *adapter,
struct sk_buff *skb);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
static bool e1000_vlan_used(struct e1000_adapter *adapter);
net: introduce and use netdev_features_t for device features sets v2: add couple missing conversions in drivers split unexporting netdev_fix_features() implemented %pNF convert sock::sk_route_(no?)caps Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-11-15 15:29:55 +00:00
static void e1000_vlan_mode(struct net_device *netdev,
netdev_features_t features);
e1000: fix vlan processing regression This patch fixes a regression introduced by commit "e1000: do vlan cleanup (799d531)". Apparently some e1000 chips (not mine) are sensitive about the order of setting vlan filter and vlan stripping/inserting functionality. So this patch changes the order so it's the same as before vlan cleanup. Reported-by: Ben Greear <greearb@candelatech.com> Signed-off-by: Jiri Pirko <jpirko@redhat.com> Tested-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-03-20 18:10:01 +00:00
static void e1000_vlan_filter_on_off(struct e1000_adapter *adapter,
bool filter_on);
net: vlan: prepare for 802.1ad VLAN filtering offload Change the rx_{add,kill}_vid callbacks to take a protocol argument in preparation of 802.1ad support. The protocol argument used so far is always htons(ETH_P_8021Q). Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-04-19 02:04:28 +00:00
static int e1000_vlan_rx_add_vid(struct net_device *netdev,
__be16 proto, u16 vid);
static int e1000_vlan_rx_kill_vid(struct net_device *netdev,
__be16 proto, u16 vid);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
static void e1000_restore_vlan(struct e1000_adapter *adapter);
e1000: fix CONFIG_PM blocks e1000_suspend is called even when !CONFIG_PM. The non-PM code inside of it is properly #ifdef'd. This fixes the compiler warnings when !CONFIG_PM. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:36 +00:00
#ifdef CONFIG_PM
NET/e1000: Fix powering off during shutdown Prevent e1000 from putting the adapter into D3 during shutdown except when we're going to power off the system, since doing that may generally cause problems with kexec to happen (such problems were observed for igb and forcedeth). For this purpose seperate e1000_shutdown() from e1000_suspend() and use the appropriate PCI PM callbacks in both of them. Signed-off-by: "Rafael J. Wysocki" <rjw@sisk.pl> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-15 17:43:24 +00:00
static int e1000_suspend(struct pci_dev *pdev, pm_message_t state);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
static int e1000_resume(struct pci_dev *pdev);
#endif
e1000: add shutdown handler back to fix WOL Someone was waaay too aggressive and removed e1000's reboot notifier instead of porting it to the new way of the shutdown handler. This change broke wake on lan. Add the shutdown handler back in using the same method as e100 uses. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-05-23 20:35:57 +00:00
static void e1000_shutdown(struct pci_dev *pdev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#ifdef CONFIG_NET_POLL_CONTROLLER
/* for netdump / net console */
static void e1000_netpoll (struct net_device *netdev);
#endif
[PATCH] e1000: Make the copybreak value a module parameter Allow the user to vary the size that copybreak works. Currently cb is enabled for packets < 256 bytes, but various tests indicate that this should be configurable for specific use cases. In addition, this parameter allows us to force never/always during testing to get full and predictable coverage of both code paths. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:40:39 +00:00
#define COPYBREAK_DEFAULT 256
static unsigned int copybreak __read_mostly = COPYBREAK_DEFAULT;
module_param(copybreak, uint, 0644);
MODULE_PARM_DESC(copybreak,
"Maximum size of packet that is copied to a new buffer on receive");
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
static pci_ers_result_t e1000_io_error_detected(struct pci_dev *pdev,
pci_channel_state_t state);
static pci_ers_result_t e1000_io_slot_reset(struct pci_dev *pdev);
static void e1000_io_resume(struct pci_dev *pdev);
netdev: make pci_error_handlers const Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2012-09-07 16:33:15 +00:00
static const struct pci_error_handlers e1000_err_handler = {
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
.error_detected = e1000_io_error_detected,
.slot_reset = e1000_io_slot_reset,
.resume = e1000_io_resume,
};
e1000: implementation of the multi-queue feature Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:03:23 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
static struct pci_driver e1000_driver = {
.name = e1000_driver_name,
.id_table = e1000_pci_tbl,
.probe = e1000_probe,
net/intel: remove __dev* attributes CONFIG_HOTPLUG is going away as an option. As result the __dev* markings will be going away. Remove use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit. Signed-off-by: Bill Pemberton <wfp5p@virginia.edu> Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Bruce Allan <bruce.w.allan@intel.com> Cc: Carolyn Wyborny <carolyn.wyborny@intel.com> Cc: Don Skidmore <donald.c.skidmore@intel.com> Cc: Greg Rose <gregory.v.rose@intel.com> Cc: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com> Cc: Alex Duyck <alexander.h.duyck@intel.com> Cc: John Ronciak <john.ronciak@intel.com> Cc: Tushar Dave <tushar.n.dave@intel.com> Cc: e1000-devel@lists.sourceforge.net Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-03 14:24:25 +00:00
.remove = e1000_remove,
e1000: keep .suspend and .resume driver methods in CONFIG_PM Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:19 +00:00
#ifdef CONFIG_PM
Fix common misspellings Fixes generated by 'codespell' and manually reviewed. Signed-off-by: Lucas De Marchi <lucas.demarchi@profusion.mobi>
2011-03-31 01:57:33 +00:00
/* Power Management Hooks */
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
.suspend = e1000_suspend,
e1000: add shutdown handler back to fix WOL Someone was waaay too aggressive and removed e1000's reboot notifier instead of porting it to the new way of the shutdown handler. This change broke wake on lan. Add the shutdown handler back in using the same method as e100 uses. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-05-23 20:35:57 +00:00
.resume = e1000_resume,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#endif
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
.shutdown = e1000_shutdown,
.err_handler = &e1000_err_handler
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
};
MODULE_AUTHOR("Intel Corporation, <linux.nics@intel.com>");
MODULE_DESCRIPTION("Intel(R) PRO/1000 Network Driver");
MODULE_LICENSE("GPL");
MODULE_VERSION(DRV_VERSION);
intel: make wired ethernet driver message level consistent (rev2) Dan Carpenter noticed that ixgbevf initial default was different than the rest. But the problem is broader than that, only one Intel driver (ixgb) was doing it almost right. The convention for default debug level should be consistent among Intel drivers and follow established convention. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-03-13 06:04:20 +00:00
#define DEFAULT_MSG_ENABLE (NETIF_MSG_DRV|NETIF_MSG_PROBE|NETIF_MSG_LINK)
static int debug = -1;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
module_param(debug, int, 0);
MODULE_PARM_DESC(debug, "Debug level (0=none,...,16=all)");
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
/**
* e1000_get_hw_dev - return device
* used by hardware layer to print debugging information
*
**/
struct net_device *e1000_get_hw_dev(struct e1000_hw *hw)
{
struct e1000_adapter *adapter = hw->back;
return adapter->netdev;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/**
* e1000_init_module - Driver Registration Routine
*
* e1000_init_module is the first routine called when the driver is
* loaded. All it does is register with the PCI subsystem.
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static int __init e1000_init_module(void)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
int ret;
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
pr_info("%s - version %s\n", e1000_driver_string, e1000_driver_version);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
pr_info("%s\n", e1000_copyright);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
drivers/net: Remove deprecated use of pci_module_init() From: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Michal Piotrowski <michal.k.k.piotrowski@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-08-19 21:48:59 +00:00
ret = pci_register_driver(&e1000_driver);
[PATCH] e1000: Make the copybreak value a module parameter Allow the user to vary the size that copybreak works. Currently cb is enabled for packets < 256 bytes, but various tests indicate that this should be configurable for specific use cases. In addition, this parameter allows us to force never/always during testing to get full and predictable coverage of both code paths. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:40:39 +00:00
if (copybreak != COPYBREAK_DEFAULT) {
if (copybreak == 0)
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
pr_info("copybreak disabled\n");
[PATCH] e1000: Make the copybreak value a module parameter Allow the user to vary the size that copybreak works. Currently cb is enabled for packets < 256 bytes, but various tests indicate that this should be configurable for specific use cases. In addition, this parameter allows us to force never/always during testing to get full and predictable coverage of both code paths. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:40:39 +00:00
else
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
pr_info("copybreak enabled for "
"packets <= %u bytes\n", copybreak);
[PATCH] e1000: Make the copybreak value a module parameter Allow the user to vary the size that copybreak works. Currently cb is enabled for packets < 256 bytes, but various tests indicate that this should be configurable for specific use cases. In addition, this parameter allows us to force never/always during testing to get full and predictable coverage of both code paths. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:40:39 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return ret;
}
module_init(e1000_init_module);
/**
* e1000_exit_module - Driver Exit Cleanup Routine
*
* e1000_exit_module is called just before the driver is removed
* from memory.
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void __exit e1000_exit_module(void)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
pci_unregister_driver(&e1000_driver);
}
module_exit(e1000_exit_module);
e1000: rework driver hardware reset locking After studying the driver mac reset code it was found that there were multiple race conditions possible to reset the unit twice or bring it e1000_up() double. This fixes all occurences where the driver needs to reset the mac. We also remove irq requesting/releasing into _open and _close so that while the device is _up we will never touch the irq's. This fixes the double free irq bug that people saw. To make sure that the watchdog task doesn't cause another race we let it run as a non-scheduled task. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:28 +00:00
static int e1000_request_irq(struct e1000_adapter *adapter)
{
struct net_device *netdev = adapter->netdev;
e1000 endianness annotations Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-11 19:49:39 +00:00
irq_handler_t handler = e1000_intr;
e1000: Fix msi enable leak on error, don't print error message, cleanup pci_enable_msi failure is a normal event so we should not print any error. Going over the code I spotted a missing pci_disable_msi() leak when irq allocation fails. The whole code also needed a cleanup, so I combined the two different calls to pci_request_irq into a single call making this look a lot better. All #ifdef CONFIG_PCI_MSI's have been removed. Compile tested with both CONFIG_PCI_MSI enabled and disabled. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Cc: H. Peter Anvin <hpa@zytor.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-05-16 08:49:46 +00:00
int irq_flags = IRQF_SHARED;
int err;
e1000: rework driver hardware reset locking After studying the driver mac reset code it was found that there were multiple race conditions possible to reset the unit twice or bring it e1000_up() double. This fixes all occurences where the driver needs to reset the mac. We also remove irq requesting/releasing into _open and _close so that while the device is _up we will never touch the irq's. This fixes the double free irq bug that people saw. To make sure that the watchdog task doesn't cause another race we let it run as a non-scheduled task. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:28 +00:00
e1000: Fix msi enable leak on error, don't print error message, cleanup pci_enable_msi failure is a normal event so we should not print any error. Going over the code I spotted a missing pci_disable_msi() leak when irq allocation fails. The whole code also needed a cleanup, so I combined the two different calls to pci_request_irq into a single call making this look a lot better. All #ifdef CONFIG_PCI_MSI's have been removed. Compile tested with both CONFIG_PCI_MSI enabled and disabled. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Cc: H. Peter Anvin <hpa@zytor.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-05-16 08:49:46 +00:00
err = request_irq(adapter->pdev->irq, handler, irq_flags, netdev->name,
netdev);
if (err) {
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(probe, "Unable to allocate interrupt Error: %d\n", err);
e1000: Fix msi enable leak on error, don't print error message, cleanup pci_enable_msi failure is a normal event so we should not print any error. Going over the code I spotted a missing pci_disable_msi() leak when irq allocation fails. The whole code also needed a cleanup, so I combined the two different calls to pci_request_irq into a single call making this look a lot better. All #ifdef CONFIG_PCI_MSI's have been removed. Compile tested with both CONFIG_PCI_MSI enabled and disabled. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Cc: H. Peter Anvin <hpa@zytor.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-05-16 08:49:46 +00:00
}
e1000: rework driver hardware reset locking After studying the driver mac reset code it was found that there were multiple race conditions possible to reset the unit twice or bring it e1000_up() double. This fixes all occurences where the driver needs to reset the mac. We also remove irq requesting/releasing into _open and _close so that while the device is _up we will never touch the irq's. This fixes the double free irq bug that people saw. To make sure that the watchdog task doesn't cause another race we let it run as a non-scheduled task. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:28 +00:00
return err;
}
static void e1000_free_irq(struct e1000_adapter *adapter)
{
struct net_device *netdev = adapter->netdev;
free_irq(adapter->pdev->irq, netdev);
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/**
* e1000_irq_disable - Mask off interrupt generation on the NIC
* @adapter: board private structure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_irq_disable(struct e1000_adapter *adapter)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
ew32(IMC, ~0);
E1000_WRITE_FLUSH();
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
synchronize_irq(adapter->pdev->irq);
}
/**
* e1000_irq_enable - Enable default interrupt generation settings
* @adapter: board private structure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_irq_enable(struct e1000_adapter *adapter)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
ew32(IMS, IMS_ENABLE_MASK);
E1000_WRITE_FLUSH();
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
[PATCH] drivers/net/e1000/: possible cleanups This patch contains the following possible cleanups: - make needlessly global code static - #if 0 the following unused global functions: - e1000_hw.c: e1000_mc_addr_list_update - e1000_hw.c: e1000_read_reg_io - e1000_hw.c: e1000_enable_pciex_master Signed-off-by: Adrian Bunk <bunk@stusta.de> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2005-10-30 15:53:34 +00:00
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_update_mng_vlan(struct e1000_adapter *adapter)
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
struct net_device *netdev = adapter->netdev;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
u16 vid = hw->mng_cookie.vlan_id;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u16 old_vid = adapter->mng_vlan_id;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
if (!e1000_vlan_used(adapter))
return;
if (!test_bit(vid, adapter->active_vlans)) {
if (hw->mng_cookie.status &
E1000_MNG_DHCP_COOKIE_STATUS_VLAN_SUPPORT) {
net: vlan: prepare for 802.1ad VLAN filtering offload Change the rx_{add,kill}_vid callbacks to take a protocol argument in preparation of 802.1ad support. The protocol argument used so far is always htons(ETH_P_8021Q). Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-04-19 02:04:28 +00:00
e1000_vlan_rx_add_vid(netdev, htons(ETH_P_8021Q), vid);
e1000: Fix AMT losing connectivity when switching VLAN in passive mode Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:17:55 +00:00
adapter->mng_vlan_id = vid;
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
} else {
adapter->mng_vlan_id = E1000_MNG_VLAN_NONE;
}
if ((old_vid != (u16)E1000_MNG_VLAN_NONE) &&
(vid != old_vid) &&
!test_bit(old_vid, adapter->active_vlans))
net: vlan: prepare for 802.1ad VLAN filtering offload Change the rx_{add,kill}_vid callbacks to take a protocol argument in preparation of 802.1ad support. The protocol argument used so far is always htons(ETH_P_8021Q). Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-04-19 02:04:28 +00:00
e1000_vlan_rx_kill_vid(netdev, htons(ETH_P_8021Q),
old_vid);
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
} else {
adapter->mng_vlan_id = vid;
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
}
}
[PATCH] e1000: Fix ASF/AMT for 8257{1|2|3} controllers The 82573 controller required different logic than 82571|2 controllers. Corrected the reset logic for 8257{1|2|3} controllers. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:30 +00:00
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_init_manageability(struct e1000_adapter *adapter)
e1000: consolidate managability enabling/disabling Several bugs existed in how we handle manageability issues all over the driver. This patch consolidates all the managability release and init code in two single functions and call them from appropriate locations. This fixes several BMC packet redirect issues and powerup/down hiccups. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:56:10 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: consolidate managability enabling/disabling Several bugs existed in how we handle manageability issues all over the driver. This patch consolidates all the managability release and init code in two single functions and call them from appropriate locations. This fixes several BMC packet redirect issues and powerup/down hiccups. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:56:10 +00:00
if (adapter->en_mng_pt) {
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
u32 manc = er32(MANC);
e1000: consolidate managability enabling/disabling Several bugs existed in how we handle manageability issues all over the driver. This patch consolidates all the managability release and init code in two single functions and call them from appropriate locations. This fixes several BMC packet redirect issues and powerup/down hiccups. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:56:10 +00:00
/* disable hardware interception of ARP */
manc &= ~(E1000_MANC_ARP_EN);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(MANC, manc);
e1000: consolidate managability enabling/disabling Several bugs existed in how we handle manageability issues all over the driver. This patch consolidates all the managability release and init code in two single functions and call them from appropriate locations. This fixes several BMC packet redirect issues and powerup/down hiccups. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:56:10 +00:00
}
}
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_release_manageability(struct e1000_adapter *adapter)
e1000: consolidate managability enabling/disabling Several bugs existed in how we handle manageability issues all over the driver. This patch consolidates all the managability release and init code in two single functions and call them from appropriate locations. This fixes several BMC packet redirect issues and powerup/down hiccups. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:56:10 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: consolidate managability enabling/disabling Several bugs existed in how we handle manageability issues all over the driver. This patch consolidates all the managability release and init code in two single functions and call them from appropriate locations. This fixes several BMC packet redirect issues and powerup/down hiccups. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:56:10 +00:00
if (adapter->en_mng_pt) {
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
u32 manc = er32(MANC);
e1000: consolidate managability enabling/disabling Several bugs existed in how we handle manageability issues all over the driver. This patch consolidates all the managability release and init code in two single functions and call them from appropriate locations. This fixes several BMC packet redirect issues and powerup/down hiccups. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:56:10 +00:00
/* re-enable hardware interception of ARP */
manc |= E1000_MANC_ARP_EN;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(MANC, manc);
e1000: consolidate managability enabling/disabling Several bugs existed in how we handle manageability issues all over the driver. This patch consolidates all the managability release and init code in two single functions and call them from appropriate locations. This fixes several BMC packet redirect issues and powerup/down hiccups. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:56:10 +00:00
}
}
e1000: FIX: be ready for incoming irq at pci_request_irq DEBUG_SHIRQ code exposed that e1000 was not ready for incoming interrupts after having called pci_request_irq. This obviously requires us to finish our software setup which assigns the irq handler before we request the irq. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-03-06 16:57:21 +00:00
/**
* e1000_configure - configure the hardware for RX and TX
* @adapter = private board structure
**/
static void e1000_configure(struct e1000_adapter *adapter)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct net_device *netdev = adapter->netdev;
e1000: rework driver hardware reset locking After studying the driver mac reset code it was found that there were multiple race conditions possible to reset the unit twice or bring it e1000_up() double. This fixes all occurences where the driver needs to reset the mac. We also remove irq requesting/releasing into _open and _close so that while the device is _up we will never touch the irq's. This fixes the double free irq bug that people saw. To make sure that the watchdog task doesn't cause another race we let it run as a non-scheduled task. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:28 +00:00
int i;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[E1000]: Secondary unicast address support Add support for configuring secondary unicast addresses. Unicast addresses take precendece over multicast addresses when filling the exact address filters to avoid going to promiscous mode. When more unicast addresses are present than filter slots, unicast filtering is disabled and all slots can be used for multicast addresses. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-11-14 04:54:59 +00:00
e1000_set_rx_mode(netdev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_restore_vlan(adapter);
e1000: consolidate managability enabling/disabling Several bugs existed in how we handle manageability issues all over the driver. This patch consolidates all the managability release and init code in two single functions and call them from appropriate locations. This fixes several BMC packet redirect issues and powerup/down hiccups. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:56:10 +00:00
e1000_init_manageability(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_configure_tx(adapter);
e1000_setup_rctl(adapter);
e1000_configure_rx(adapter);
[PATCH] e1000: Added cleaned_count to RX buffer allocation Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:19 +00:00
/* call E1000_DESC_UNUSED which always leaves
* at least 1 descriptor unused to make sure
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* next_to_use != next_to_clean
*/
[PATCH] e1000: Fix adapter structure and prepare for multique fix Fix adapter structure to handle multiple queues and prepping the driver for full multiple queue support, some changes are ifdef'd our unless you define CONFIG_E1000_MQ. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:39 +00:00
for (i = 0; i < adapter->num_rx_queues; i++) {
[PATCH] e1000: Added cleaned_count to RX buffer allocation Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:19 +00:00
struct e1000_rx_ring *ring = &adapter->rx_ring[i];
[PATCH] e1000: Added copy break code Improves small packet performance with large amounts of reassembly being done in the stack. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:30 +00:00
adapter->alloc_rx_buf(adapter, ring,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
E1000_DESC_UNUSED(ring));
[PATCH] e1000: Fix adapter structure and prepare for multique fix Fix adapter structure to handle multiple queues and prepping the driver for full multiple queue support, some changes are ifdef'd our unless you define CONFIG_E1000_MQ. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:39 +00:00
}
e1000: FIX: be ready for incoming irq at pci_request_irq DEBUG_SHIRQ code exposed that e1000 was not ready for incoming interrupts after having called pci_request_irq. This obviously requires us to finish our software setup which assigns the irq handler before we request the irq. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-03-06 16:57:21 +00:00
}
int e1000_up(struct e1000_adapter *adapter)
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: FIX: be ready for incoming irq at pci_request_irq DEBUG_SHIRQ code exposed that e1000 was not ready for incoming interrupts after having called pci_request_irq. This obviously requires us to finish our software setup which assigns the irq handler before we request the irq. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-03-06 16:57:21 +00:00
/* hardware has been reset, we need to reload some things */
e1000_configure(adapter);
clear_bit(__E1000_DOWN, &adapter->flags);
[PATCH] e1000: Fix mulitple queues Fixed stats when using multiple queues. When multiple queues are enabled, log a message in syslog. Fixed memory allocation for multiple queues. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:41 +00:00
[NET]: Make NAPI polling independent of struct net_device objects. Several devices have multiple independant RX queues per net device, and some have a single interrupt doorbell for several queues. In either case, it's easier to support layouts like that if the structure representing the poll is independant from the net device itself. The signature of the ->poll() call back goes from: int foo_poll(struct net_device *dev, int *budget) to int foo_poll(struct napi_struct *napi, int budget) The caller is returned the number of RX packets processed (or the number of "NAPI credits" consumed if you want to get abstract). The callee no longer messes around bumping dev->quota, *budget, etc. because that is all handled in the caller upon return. The napi_struct is to be embedded in the device driver private data structures. Furthermore, it is the driver's responsibility to disable all NAPI instances in it's ->stop() device close handler. Since the napi_struct is privatized into the driver's private data structures, only the driver knows how to get at all of the napi_struct instances it may have per-device. With lots of help and suggestions from Rusty Russell, Roland Dreier, Michael Chan, Jeff Garzik, and Jamal Hadi Salim. Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra, Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan. [ Ported to current tree and all drivers converted. Integrated Stephen's follow-on kerneldoc additions, and restored poll_list handling to the old style to fix mutual exclusion issues. -DaveM ] Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-03 23:41:36 +00:00
napi_enable(&adapter->napi);
e1000: delete non NAPI code from the driver Compile-tested only. Signed-off-by: Francois Romieu <romieu@fr.zoreil.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:38 +00:00
[PATCH] e1000: Enable polling before enabling interrupts Enable polling before enabling interrupts -- avoids (in NAPI mode) entering the ISR and returning without doing any work because polling is not enabled. [romieu@fr.zoriel.com] Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_main.c net-drivers-2.6/drivers/net/e1000.new/e1000_main.c
2005-04-29 02:39:30 +00:00
e1000_irq_enable(adapter);
e1000/e1000e/igb/ixgb: do not use netif_wake_queue un-necessarily It was pointed out that the Intel wired ethernet drivers do not need to wake the tx queue since netif_carrier_on/off will take care of the qdisc management in order to guarantee the correct handling of the transmit routine enable state. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-21 18:42:05 +00:00
netif_wake_queue(adapter->netdev);
[PATCH] e1000: No-delay link detection at interface up Currently after an interface up, the link state is detected 2 seconds later when the first watchdog timer runs. This patch changes that by triggering the hardware to generate a link-change interrupt from the up() function instead. This has the result that the link state gets detected immediately and without races. This has the potential to speed up booting since a normal distribution boot process waits for a link before DHCP is attempted. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:42:34 +00:00
/* fire a link change interrupt to start the watchdog */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(ICS, E1000_ICS_LSC);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return 0;
}
e1000: Make PHY powerup/down a function In relation to the irq work done earlier we also move the PHY powerup and powerdown functions into separate functions and move the calls to _close and _open, making the PHY stay in it's power state as long as the device is _up. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:32 +00:00
/**
* e1000_power_up_phy - restore link in case the phy was powered down
* @adapter: address of board private structure
*
* The phy may be powered down to save power and turn off link when the
* driver is unloaded and wake on lan is not enabled (among others)
* *** this routine MUST be followed by a call to e1000_reset ***
**/
e1000: Explicitly power up the PHY during loopback testing. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-16 20:31:33 +00:00
void e1000_power_up_phy(struct e1000_adapter *adapter)
e1000: Make PHY powerup/down a function In relation to the irq work done earlier we also move the PHY powerup and powerdown functions into separate functions and move the calls to _close and _open, making the PHY stay in it's power state as long as the device is _up. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:32 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u16 mii_reg = 0;
e1000: Make PHY powerup/down a function In relation to the irq work done earlier we also move the PHY powerup and powerdown functions into separate functions and move the calls to _close and _open, making the PHY stay in it's power state as long as the device is _up. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:32 +00:00
/* Just clear the power down bit to wake the phy back up */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (hw->media_type == e1000_media_type_copper) {
e1000: Make PHY powerup/down a function In relation to the irq work done earlier we also move the PHY powerup and powerdown functions into separate functions and move the calls to _close and _open, making the PHY stay in it's power state as long as the device is _up. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:32 +00:00
/* according to the manual, the phy will retain its
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* settings across a power-down/up cycle
*/
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_read_phy_reg(hw, PHY_CTRL, &mii_reg);
e1000: Make PHY powerup/down a function In relation to the irq work done earlier we also move the PHY powerup and powerdown functions into separate functions and move the calls to _close and _open, making the PHY stay in it's power state as long as the device is _up. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:32 +00:00
mii_reg &= ~MII_CR_POWER_DOWN;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_write_phy_reg(hw, PHY_CTRL, mii_reg);
e1000: Make PHY powerup/down a function In relation to the irq work done earlier we also move the PHY powerup and powerdown functions into separate functions and move the calls to _close and _open, making the PHY stay in it's power state as long as the device is _up. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:32 +00:00
}
}
static void e1000_power_down_phy(struct e1000_adapter *adapter)
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: handle manageability for pci-e adapters at PHY powerdown When powering down the PHY (if WoL is disabled) we should only check copper PHY's and handle PCI-E adapters differently. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:54 +00:00
/* Power down the PHY so no link is implied when interface is down *
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
* The PHY cannot be powered down if any of the following is true *
e1000: Make PHY powerup/down a function In relation to the irq work done earlier we also move the PHY powerup and powerdown functions into separate functions and move the calls to _close and _open, making the PHY stay in it's power state as long as the device is _up. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:32 +00:00
* (a) WoL is enabled
* (b) AMT is active
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* (c) SoL/IDER session is active
*/
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (!adapter->wol && hw->mac_type >= e1000_82540 &&
hw->media_type == e1000_media_type_copper) {
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u16 mii_reg = 0;
e1000: handle manageability for pci-e adapters at PHY powerdown When powering down the PHY (if WoL is disabled) we should only check copper PHY's and handle PCI-E adapters differently. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:54 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
switch (hw->mac_type) {
e1000: handle manageability for pci-e adapters at PHY powerdown When powering down the PHY (if WoL is disabled) we should only check copper PHY's and handle PCI-E adapters differently. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:54 +00:00
case e1000_82540:
case e1000_82545:
case e1000_82545_rev_3:
case e1000_82546:
e1000: Add support for the CE4100 reference platform This patch adds support for the gigabit phys present on the CE4100 reference platforms. Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-06 14:29:54 +00:00
case e1000_ce4100:
e1000: handle manageability for pci-e adapters at PHY powerdown When powering down the PHY (if WoL is disabled) we should only check copper PHY's and handle PCI-E adapters differently. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:54 +00:00
case e1000_82546_rev_3:
case e1000_82541:
case e1000_82541_rev_2:
case e1000_82547:
case e1000_82547_rev_2:
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (er32(MANC) & E1000_MANC_SMBUS_EN)
e1000: handle manageability for pci-e adapters at PHY powerdown When powering down the PHY (if WoL is disabled) we should only check copper PHY's and handle PCI-E adapters differently. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:54 +00:00
goto out;
break;
default:
goto out;
}
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_read_phy_reg(hw, PHY_CTRL, &mii_reg);
e1000: Make PHY powerup/down a function In relation to the irq work done earlier we also move the PHY powerup and powerdown functions into separate functions and move the calls to _close and _open, making the PHY stay in it's power state as long as the device is _up. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:32 +00:00
mii_reg |= MII_CR_POWER_DOWN;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_write_phy_reg(hw, PHY_CTRL, mii_reg);
e1000: convert mdelay to msleep With the previous commit, there are several functions that are only ever called from thread context, and are able to sleep with msleep instead of mdelay. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:46 +00:00
msleep(1);
e1000: Make PHY powerup/down a function In relation to the irq work done earlier we also move the PHY powerup and powerdown functions into separate functions and move the calls to _close and _open, making the PHY stay in it's power state as long as the device is _up. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:32 +00:00
}
e1000: handle manageability for pci-e adapters at PHY powerdown When powering down the PHY (if WoL is disabled) we should only check copper PHY's and handle PCI-E adapters differently. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:54 +00:00
out:
return;
e1000: Make PHY powerup/down a function In relation to the irq work done earlier we also move the PHY powerup and powerdown functions into separate functions and move the calls to _close and _open, making the PHY stay in it's power state as long as the device is _up. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:32 +00:00
}
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
static void e1000_down_and_stop(struct e1000_adapter *adapter)
{
set_bit(__E1000_DOWN, &adapter->flags);
e1000: Prevent reset task killing itself. Killing reset task while adapter is resetting causes deadlock. Only kill reset task if adapter is not resetting. Ref bug #43132 on bugzilla.kernel.org CC: stable@vger.kernel.org Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-05-17 01:04:50 +00:00
/* Only kill reset task if adapter is not resetting */
if (!test_bit(__E1000_RESETTING, &adapter->flags))
cancel_work_sync(&adapter->reset_task);
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
cancel_delayed_work_sync(&adapter->watchdog_task);
cancel_delayed_work_sync(&adapter->phy_info_task);
cancel_delayed_work_sync(&adapter->fifo_stall_task);
}
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
void e1000_down(struct e1000_adapter *adapter)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: fix close race with interrupt this is in regards to http://bugzilla.kernel.org/show_bug.cgi?id=12876 where it appears that e1000 can leave its interrupt enabled after exiting the driver. Fix the bug by making the interrupt enable paths more aware of the driver exiting. Thanks to Alan Cox for the poke and initial investigation. CC: Alan Cox <alan@lxorguk.ukuu.org.uk> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:59:22 +00:00
struct e1000_hw *hw = &adapter->hw;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct net_device *netdev = adapter->netdev;
e1000: fix close race with interrupt this is in regards to http://bugzilla.kernel.org/show_bug.cgi?id=12876 where it appears that e1000 can leave its interrupt enabled after exiting the driver. Fix the bug by making the interrupt enable paths more aware of the driver exiting. Thanks to Alan Cox for the poke and initial investigation. CC: Alan Cox <alan@lxorguk.ukuu.org.uk> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:59:22 +00:00
u32 rctl, tctl;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: driver state fixes (race fix) We were plagued by our interrupt handler posting a watchdog event which could occur when our adapter was going down in case a late packet arrived just before e1000_down() finished. This caused the watchdog timer to start after the NIC was down and keep rescheduling it every N seconds. Once the driver unloaded it would panic. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:54:02 +00:00
e1000: fix close race with interrupt this is in regards to http://bugzilla.kernel.org/show_bug.cgi?id=12876 where it appears that e1000 can leave its interrupt enabled after exiting the driver. Fix the bug by making the interrupt enable paths more aware of the driver exiting. Thanks to Alan Cox for the poke and initial investigation. CC: Alan Cox <alan@lxorguk.ukuu.org.uk> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:59:22 +00:00
/* disable receives in the hardware */
rctl = er32(RCTL);
ew32(RCTL, rctl & ~E1000_RCTL_EN);
/* flush and sleep below */
e1000: use netif_tx_disable we can use netif_tx_disable now because LLTX has been removed. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:17:01 +00:00
netif_tx_disable(netdev);
e1000: fix close race with interrupt this is in regards to http://bugzilla.kernel.org/show_bug.cgi?id=12876 where it appears that e1000 can leave its interrupt enabled after exiting the driver. Fix the bug by making the interrupt enable paths more aware of the driver exiting. Thanks to Alan Cox for the poke and initial investigation. CC: Alan Cox <alan@lxorguk.ukuu.org.uk> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:59:22 +00:00
/* disable transmits in the hardware */
tctl = er32(TCTL);
tctl &= ~E1000_TCTL_EN;
ew32(TCTL, tctl);
/* flush both disables and wait for them to finish */
E1000_WRITE_FLUSH();
msleep(10);
[NET]: Make NAPI polling independent of struct net_device objects. Several devices have multiple independant RX queues per net device, and some have a single interrupt doorbell for several queues. In either case, it's easier to support layouts like that if the structure representing the poll is independant from the net device itself. The signature of the ->poll() call back goes from: int foo_poll(struct net_device *dev, int *budget) to int foo_poll(struct napi_struct *napi, int budget) The caller is returned the number of RX packets processed (or the number of "NAPI credits" consumed if you want to get abstract). The callee no longer messes around bumping dev->quota, *budget, etc. because that is all handled in the caller upon return. The napi_struct is to be embedded in the device driver private data structures. Furthermore, it is the driver's responsibility to disable all NAPI instances in it's ->stop() device close handler. Since the napi_struct is privatized into the driver's private data structures, only the driver knows how to get at all of the napi_struct instances it may have per-device. With lots of help and suggestions from Rusty Russell, Roland Dreier, Michael Chan, Jeff Garzik, and Jamal Hadi Salim. Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra, Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan. [ Ported to current tree and all drivers converted. Integrated Stephen's follow-on kerneldoc additions, and restored poll_list handling to the old style to fix mutual exclusion issues. -DaveM ] Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-03 23:41:36 +00:00
napi_disable(&adapter->napi);
e1000: delete non NAPI code from the driver Compile-tested only. Signed-off-by: Francois Romieu <romieu@fr.zoreil.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:38 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_irq_disable(adapter);
e1000: Remove Multiqueue code until we have support for MSI-X in our hardware Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:16:38 +00:00
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* Setting DOWN must be after irq_disable to prevent
e1000: fix screaming IRQ VMWare reports that the e1000 driver has a bug when bringing down the interface, such that interrupts are not disabled in the hardware but the driver stops reporting that it consumed the interrupt. The fix is to set the driver's "down" flag later in the routine, after all the timers and such have exited, preventing the interrupt handler from being called and exiting early without handling the interrupt. CC: Anupam Chanda <anupamc@vmware.com> CC: stable kernel <stable@kernel.org> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-11-21 17:54:21 +00:00
* a screaming interrupt. Setting DOWN also prevents
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
* tasks from rescheduling.
e1000: fix screaming IRQ VMWare reports that the e1000 driver has a bug when bringing down the interface, such that interrupts are not disabled in the hardware but the driver stops reporting that it consumed the interrupt. The fix is to set the driver's "down" flag later in the routine, after all the timers and such have exited, preventing the interrupt handler from being called and exiting early without handling the interrupt. CC: Anupam Chanda <anupamc@vmware.com> CC: stable kernel <stable@kernel.org> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-11-21 17:54:21 +00:00
*/
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
e1000_down_and_stop(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
adapter->link_speed = 0;
adapter->link_duplex = 0;
netif_carrier_off(netdev);
e1000_reset(adapter);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
e1000_clean_all_tx_rings(adapter);
e1000_clean_all_rx_rings(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: make e1000_reinit_safe local Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-10-21 07:50:57 +00:00
static void e1000_reinit_safe(struct e1000_adapter *adapter)
e1000: fix occasional panic on unload Net drivers in general have an issue where timers fired by mod_timer or work threads with schedule_work are running outside of the rtnl_lock. With no other lock protection these routines are vulnerable to races with driver unload or reset paths. The longer term solution to this might be a redesign with safer locks being taken in the driver to guarantee no reentrance, but for now a safe and effective fix is to take the rtnl_lock in these routines. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-22 18:22:42 +00:00
{
while (test_and_set_bit(__E1000_RESETTING, &adapter->flags))
msleep(1);
e1000: convert to private mutex from rtnl The e1000 driver when running with lockdep could run into some possible deadlocks between the work items acquiring rtnl and the rtnl lock being acquired before work items were cancelled. Use a private mutex to make sure lock ordering isn't violated. The private mutex is only used to protect areas not generally covered by the rtnl lock already. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:51 +00:00
mutex_lock(&adapter->mutex);
e1000: fix occasional panic on unload Net drivers in general have an issue where timers fired by mod_timer or work threads with schedule_work are running outside of the rtnl_lock. With no other lock protection these routines are vulnerable to races with driver unload or reset paths. The longer term solution to this might be a redesign with safer locks being taken in the driver to guarantee no reentrance, but for now a safe and effective fix is to take the rtnl_lock in these routines. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-22 18:22:42 +00:00
e1000_down(adapter);
e1000_up(adapter);
e1000: convert to private mutex from rtnl The e1000 driver when running with lockdep could run into some possible deadlocks between the work items acquiring rtnl and the rtnl lock being acquired before work items were cancelled. Use a private mutex to make sure lock ordering isn't violated. The private mutex is only used to protect areas not generally covered by the rtnl lock already. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:51 +00:00
mutex_unlock(&adapter->mutex);
e1000: fix occasional panic on unload Net drivers in general have an issue where timers fired by mod_timer or work threads with schedule_work are running outside of the rtnl_lock. With no other lock protection these routines are vulnerable to races with driver unload or reset paths. The longer term solution to this might be a redesign with safer locks being taken in the driver to guarantee no reentrance, but for now a safe and effective fix is to take the rtnl_lock in these routines. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-22 18:22:42 +00:00
clear_bit(__E1000_RESETTING, &adapter->flags);
}
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
void e1000_reinit_locked(struct e1000_adapter *adapter)
e1000: rework driver hardware reset locking After studying the driver mac reset code it was found that there were multiple race conditions possible to reset the unit twice or bring it e1000_up() double. This fixes all occurences where the driver needs to reset the mac. We also remove irq requesting/releasing into _open and _close so that while the device is _up we will never touch the irq's. This fixes the double free irq bug that people saw. To make sure that the watchdog task doesn't cause another race we let it run as a non-scheduled task. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:28 +00:00
{
e1000: fix occasional panic on unload Net drivers in general have an issue where timers fired by mod_timer or work threads with schedule_work are running outside of the rtnl_lock. With no other lock protection these routines are vulnerable to races with driver unload or reset paths. The longer term solution to this might be a redesign with safer locks being taken in the driver to guarantee no reentrance, but for now a safe and effective fix is to take the rtnl_lock in these routines. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-22 18:22:42 +00:00
/* if rtnl_lock is not held the call path is bogus */
ASSERT_RTNL();
e1000: rework driver hardware reset locking After studying the driver mac reset code it was found that there were multiple race conditions possible to reset the unit twice or bring it e1000_up() double. This fixes all occurences where the driver needs to reset the mac. We also remove irq requesting/releasing into _open and _close so that while the device is _up we will never touch the irq's. This fixes the double free irq bug that people saw. To make sure that the watchdog task doesn't cause another race we let it run as a non-scheduled task. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:28 +00:00
WARN_ON(in_interrupt());
while (test_and_set_bit(__E1000_RESETTING, &adapter->flags))
msleep(1);
e1000_down(adapter);
e1000_up(adapter);
clear_bit(__E1000_RESETTING, &adapter->flags);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
void e1000_reset(struct e1000_adapter *adapter)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u32 pba = 0, tx_space, min_tx_space, min_rx_space;
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
bool legacy_pba_adjust = false;
e1000: fix flow control thresholds when testing the jumbo frames with pages patch, the stats would show rx_missed errors (dropped packets) even when connected to a link partner with flow control enabled. this indicates that for this MTU (9000) the flow control thresholds are not adjusting correctly. In fact, before this change, the FCRTH (xoff threshold) is 36864 when the fifo size is only 40000, with 9000 byte MTU. fix it so that we at least have room for one frame after we send the xoff. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:45:01 +00:00
u16 hwm;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Repartition Pba for greater than 9k mtu
* To take effect CTRL.RST is required.
*/
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
switch (hw->mac_type) {
[PATCH] e1000: Fix PBA allocation calculations Assign the PBA to be large enough to contain at least 2 jumbo frames on all adapters. This dramatically increases performance on several adapters and fixes TX performance degradation issues where the PBA was misallocated in the old algorithm. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:39:45 +00:00
case e1000_82542_rev2_0:
case e1000_82542_rev2_1:
case e1000_82543:
case e1000_82544:
case e1000_82540:
case e1000_82541:
case e1000_82541_rev_2:
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
legacy_pba_adjust = true;
[PATCH] e1000: Fix PBA allocation calculations Assign the PBA to be large enough to contain at least 2 jumbo frames on all adapters. This dramatically increases performance on several adapters and fixes TX performance degradation issues where the PBA was misallocated in the old algorithm. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:39:45 +00:00
pba = E1000_PBA_48K;
break;
case e1000_82545:
case e1000_82545_rev_3:
case e1000_82546:
e1000: Add support for the CE4100 reference platform This patch adds support for the gigabit phys present on the CE4100 reference platforms. Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-06 14:29:54 +00:00
case e1000_ce4100:
[PATCH] e1000: Fix PBA allocation calculations Assign the PBA to be large enough to contain at least 2 jumbo frames on all adapters. This dramatically increases performance on several adapters and fixes TX performance degradation issues where the PBA was misallocated in the old algorithm. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:39:45 +00:00
case e1000_82546_rev_3:
pba = E1000_PBA_48K;
break;
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
case e1000_82547:
[PATCH] e1000:Fix Packet Buffer Allocation logic for 82547_rev_2 Fix Packet Buffer Allocation logic for 82547_rev_2 controller Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_main.c net-drivers-2.6/drivers/net/e1000.new/e1000_main.c
2005-04-29 02:44:14 +00:00
case e1000_82547_rev_2:
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
legacy_pba_adjust = true;
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
pba = E1000_PBA_30K;
break;
[PATCH] e1000: Fix PBA allocation calculations Assign the PBA to be large enough to contain at least 2 jumbo frames on all adapters. This dramatically increases performance on several adapters and fixes TX performance degradation issues where the PBA was misallocated in the old algorithm. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:39:45 +00:00
case e1000_undefined:
case e1000_num_macs:
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
break;
}
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
if (legacy_pba_adjust) {
e1000: fix flow control thresholds when testing the jumbo frames with pages patch, the stats would show rx_missed errors (dropped packets) even when connected to a link partner with flow control enabled. this indicates that for this MTU (9000) the flow control thresholds are not adjusting correctly. In fact, before this change, the FCRTH (xoff threshold) is 36864 when the fifo size is only 40000, with 9000 byte MTU. fix it so that we at least have room for one frame after we send the xoff. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:45:01 +00:00
if (hw->max_frame_size > E1000_RXBUFFER_8192)
[PATCH] e1000: Fix PBA allocation calculations Assign the PBA to be large enough to contain at least 2 jumbo frames on all adapters. This dramatically increases performance on several adapters and fixes TX performance degradation issues where the PBA was misallocated in the old algorithm. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:39:45 +00:00
pba -= 8; /* allocate more FIFO for Tx */
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (hw->mac_type == e1000_82547) {
[PATCH] e1000: Fix PBA allocation calculations Assign the PBA to be large enough to contain at least 2 jumbo frames on all adapters. This dramatically increases performance on several adapters and fixes TX performance degradation issues where the PBA was misallocated in the old algorithm. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:39:45 +00:00
adapter->tx_fifo_head = 0;
adapter->tx_head_addr = pba << E1000_TX_HEAD_ADDR_SHIFT;
adapter->tx_fifo_size =
(E1000_PBA_40K - pba) << E1000_PBA_BYTES_SHIFT;
atomic_set(&adapter->tx_fifo_stall, 0);
}
e1000: fix flow control thresholds when testing the jumbo frames with pages patch, the stats would show rx_missed errors (dropped packets) even when connected to a link partner with flow control enabled. this indicates that for this MTU (9000) the flow control thresholds are not adjusting correctly. In fact, before this change, the FCRTH (xoff threshold) is 36864 when the fifo size is only 40000, with 9000 byte MTU. fix it so that we at least have room for one frame after we send the xoff. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:45:01 +00:00
} else if (hw->max_frame_size > ETH_FRAME_LEN + ETH_FCS_LEN) {
[PATCH] e1000: Fix PBA allocation calculations Assign the PBA to be large enough to contain at least 2 jumbo frames on all adapters. This dramatically increases performance on several adapters and fixes TX performance degradation issues where the PBA was misallocated in the old algorithm. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:39:45 +00:00
/* adjust PBA for jumbo frames */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(PBA, pba);
[PATCH] e1000: Fix PBA allocation calculations Assign the PBA to be large enough to contain at least 2 jumbo frames on all adapters. This dramatically increases performance on several adapters and fixes TX performance degradation issues where the PBA was misallocated in the old algorithm. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:39:45 +00:00
/* To maintain wire speed transmits, the Tx FIFO should be
e1000: fix flow control thresholds when testing the jumbo frames with pages patch, the stats would show rx_missed errors (dropped packets) even when connected to a link partner with flow control enabled. this indicates that for this MTU (9000) the flow control thresholds are not adjusting correctly. In fact, before this change, the FCRTH (xoff threshold) is 36864 when the fifo size is only 40000, with 9000 byte MTU. fix it so that we at least have room for one frame after we send the xoff. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:45:01 +00:00
* large enough to accommodate two full transmit packets,
[PATCH] e1000: Fix PBA allocation calculations Assign the PBA to be large enough to contain at least 2 jumbo frames on all adapters. This dramatically increases performance on several adapters and fixes TX performance degradation issues where the PBA was misallocated in the old algorithm. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:39:45 +00:00
* rounded up to the next 1KB and expressed in KB. Likewise,
e1000: fix flow control thresholds when testing the jumbo frames with pages patch, the stats would show rx_missed errors (dropped packets) even when connected to a link partner with flow control enabled. this indicates that for this MTU (9000) the flow control thresholds are not adjusting correctly. In fact, before this change, the FCRTH (xoff threshold) is 36864 when the fifo size is only 40000, with 9000 byte MTU. fix it so that we at least have room for one frame after we send the xoff. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:45:01 +00:00
* the Rx FIFO should be large enough to accommodate at least
[PATCH] e1000: Fix PBA allocation calculations Assign the PBA to be large enough to contain at least 2 jumbo frames on all adapters. This dramatically increases performance on several adapters and fixes TX performance degradation issues where the PBA was misallocated in the old algorithm. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:39:45 +00:00
* one full receive packet and is similarly rounded up and
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* expressed in KB.
*/
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
pba = er32(PBA);
[PATCH] e1000: Fix PBA allocation calculations Assign the PBA to be large enough to contain at least 2 jumbo frames on all adapters. This dramatically increases performance on several adapters and fixes TX performance degradation issues where the PBA was misallocated in the old algorithm. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:39:45 +00:00
/* upper 16 bits has Tx packet buffer allocation size in KB */
tx_space = pba >> 16;
/* lower 16 bits has Rx packet buffer allocation size in KB */
pba &= 0xffff;
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* the Tx fifo also stores 16 bytes of information about the Tx
e1000: fix flow control thresholds when testing the jumbo frames with pages patch, the stats would show rx_missed errors (dropped packets) even when connected to a link partner with flow control enabled. this indicates that for this MTU (9000) the flow control thresholds are not adjusting correctly. In fact, before this change, the FCRTH (xoff threshold) is 36864 when the fifo size is only 40000, with 9000 byte MTU. fix it so that we at least have room for one frame after we send the xoff. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:45:01 +00:00
* but don't include ethernet FCS because hardware appends it
*/
min_tx_space = (hw->max_frame_size +
sizeof(struct e1000_tx_desc) -
ETH_FCS_LEN) * 2;
e1000: ROUND_UP macro cleanup in drivers/net/e1000 E1000_ROUNDUP macro cleanup, use ALIGN Signed-off-by: Milind Arun Choudhary <milindchoudhary@gmail.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-04-27 20:55:29 +00:00
min_tx_space = ALIGN(min_tx_space, 1024);
[PATCH] e1000: Fix PBA allocation calculations Assign the PBA to be large enough to contain at least 2 jumbo frames on all adapters. This dramatically increases performance on several adapters and fixes TX performance degradation issues where the PBA was misallocated in the old algorithm. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:39:45 +00:00
min_tx_space >>= 10;
e1000: fix flow control thresholds when testing the jumbo frames with pages patch, the stats would show rx_missed errors (dropped packets) even when connected to a link partner with flow control enabled. this indicates that for this MTU (9000) the flow control thresholds are not adjusting correctly. In fact, before this change, the FCRTH (xoff threshold) is 36864 when the fifo size is only 40000, with 9000 byte MTU. fix it so that we at least have room for one frame after we send the xoff. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:45:01 +00:00
/* software strips receive CRC, so leave room for it */
min_rx_space = hw->max_frame_size;
e1000: ROUND_UP macro cleanup in drivers/net/e1000 E1000_ROUNDUP macro cleanup, use ALIGN Signed-off-by: Milind Arun Choudhary <milindchoudhary@gmail.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-04-27 20:55:29 +00:00
min_rx_space = ALIGN(min_rx_space, 1024);
[PATCH] e1000: Fix PBA allocation calculations Assign the PBA to be large enough to contain at least 2 jumbo frames on all adapters. This dramatically increases performance on several adapters and fixes TX performance degradation issues where the PBA was misallocated in the old algorithm. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:39:45 +00:00
min_rx_space >>= 10;
/* If current Tx allocation is less than the min Tx FIFO size,
* and the min Tx FIFO size is less than the current Rx FIFO
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* allocation, take space away from current Rx allocation
*/
[PATCH] e1000: Fix PBA allocation calculations Assign the PBA to be large enough to contain at least 2 jumbo frames on all adapters. This dramatically increases performance on several adapters and fixes TX performance degradation issues where the PBA was misallocated in the old algorithm. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:39:45 +00:00
if (tx_space < min_tx_space &&
((min_tx_space - tx_space) < pba)) {
pba = pba - (min_tx_space - tx_space);
/* PCI/PCIx hardware has PBA alignment constraints */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
switch (hw->mac_type) {
[PATCH] e1000: Fix PBA allocation calculations Assign the PBA to be large enough to contain at least 2 jumbo frames on all adapters. This dramatically increases performance on several adapters and fixes TX performance degradation issues where the PBA was misallocated in the old algorithm. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:39:45 +00:00
case e1000_82545 ... e1000_82546_rev_3:
pba &= ~(E1000_PBA_8K - 1);
break;
default:
break;
}
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* if short on Rx space, Rx wins and must trump Tx
* adjustment or use Early Receive if available
*/
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
if (pba < min_rx_space)
pba = min_rx_space;
[PATCH] e1000: Fix PBA allocation calculations Assign the PBA to be large enough to contain at least 2 jumbo frames on all adapters. This dramatically increases performance on several adapters and fixes TX performance degradation issues where the PBA was misallocated in the old algorithm. Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:39:45 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(PBA, pba);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* flow control settings:
e1000: fix flow control thresholds when testing the jumbo frames with pages patch, the stats would show rx_missed errors (dropped packets) even when connected to a link partner with flow control enabled. this indicates that for this MTU (9000) the flow control thresholds are not adjusting correctly. In fact, before this change, the FCRTH (xoff threshold) is 36864 when the fifo size is only 40000, with 9000 byte MTU. fix it so that we at least have room for one frame after we send the xoff. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:45:01 +00:00
* The high water mark must be low enough to fit one full frame
* (or the size used for early receive) above it in the Rx FIFO.
* Set it to the lower of:
* - 90% of the Rx FIFO size, and
* - the full Rx FIFO size minus the early receive size (for parts
* with ERT support assuming ERT set to E1000_ERT_2048), or
* - the full Rx FIFO size minus one full frame
*/
hwm = min(((pba << 10) * 9 / 10),
((pba << 10) - hw->max_frame_size));
hw->fc_high_water = hwm & 0xFFF8; /* 8-byte granularity */
hw->fc_low_water = hw->fc_high_water - 8;
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
hw->fc_pause_time = E1000_FC_PAUSE_TIME;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
hw->fc_send_xon = 1;
hw->fc = hw->original_fc;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
/* Allow time for pending master requests to run */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_reset_hw(hw);
if (hw->mac_type >= e1000_82544)
ew32(WUC, 0);
e1000: gather hardware bit tweaks. Several hardware bits were set all over the driver and have been consolidated into a single function. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:51 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (e1000_init_hw(hw))
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_dev_err("Hardware Error\n");
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
e1000_update_mng_vlan(adapter);
[PATCH] e1000: Fix Wake-on-Lan with forced gigabit speed If the user has forced gigabit speed, phy power management must be disabled; otherwise the NIC would try to negotiate to a linkspeed of 10/100 mbit on shutdown, which would lead to a total loss of link. This loss of link breaks Wake-on-Lan and IPMI. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:33:46 +00:00
/* if (adapter->hwflags & HWFLAGS_PHY_PWR_BIT) { */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (hw->mac_type >= e1000_82544 &&
hw->autoneg == 1 &&
hw->autoneg_advertised == ADVERTISE_1000_FULL) {
u32 ctrl = er32(CTRL);
[PATCH] e1000: Fix Wake-on-Lan with forced gigabit speed If the user has forced gigabit speed, phy power management must be disabled; otherwise the NIC would try to negotiate to a linkspeed of 10/100 mbit on shutdown, which would lead to a total loss of link. This loss of link breaks Wake-on-Lan and IPMI. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:33:46 +00:00
/* clear phy power management bit if we are in gig only mode,
* which if enabled will attempt negotiation to 100Mb, which
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* can cause a loss of link at power off or driver unload
*/
[PATCH] e1000: Fix Wake-on-Lan with forced gigabit speed If the user has forced gigabit speed, phy power management must be disabled; otherwise the NIC would try to negotiate to a linkspeed of 10/100 mbit on shutdown, which would lead to a total loss of link. This loss of link breaks Wake-on-Lan and IPMI. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:33:46 +00:00
ctrl &= ~E1000_CTRL_SWDPIN3;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(CTRL, ctrl);
[PATCH] e1000: Fix Wake-on-Lan with forced gigabit speed If the user has forced gigabit speed, phy power management must be disabled; otherwise the NIC would try to negotiate to a linkspeed of 10/100 mbit on shutdown, which would lead to a total loss of link. This loss of link breaks Wake-on-Lan and IPMI. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:33:46 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Enable h/w to recognize an 802.1Q VLAN Ethernet packet */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(VET, ETHERNET_IEEE_VLAN_TYPE);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_reset_adaptive(hw);
e1000_phy_get_info(hw, &adapter->phy_info);
e1000: add smart power down code Smart Power Down is a power saving feature in newer e1000 hardware. We disable it because it causes time to link to be long, but make it a user choice. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:45 +00:00
e1000: consolidate managability enabling/disabling Several bugs existed in how we handle manageability issues all over the driver. This patch consolidates all the managability release and init code in two single functions and call them from appropriate locations. This fixes several BMC packet redirect issues and powerup/down hiccups. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:56:10 +00:00
e1000_release_manageability(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
drivers/net/ethernet: Fix non-kernel-doc comments with kernel-doc start markers Convert doxygen (or similar) formatted comments to kernel-doc or unformatted comment. Delete a few that are content-free. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-10 10:56:59 +00:00
/* Dump the eeprom for users having checksum issues */
e1000: make e1000_dump_eeprom() static This patch makes the needlessly global e1000_dump_eeprom() static. Signed-off-by: Adrian Bunk <bunk@kernel.org> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-02-01 16:21:28 +00:00
static void e1000_dump_eeprom(struct e1000_adapter *adapter)
e1000: Dump the eeprom when a user encounters a bad checksum To help supporting users with a bad eeprom checksum, dump the eeprom info when such a situation is encountered by a user. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-17 21:50:23 +00:00
{
struct net_device *netdev = adapter->netdev;
struct ethtool_eeprom eeprom;
const struct ethtool_ops *ops = netdev->ethtool_ops;
u8 *data;
int i;
u16 csum_old, csum_new = 0;
eeprom.len = ops->get_eeprom_len(netdev);
eeprom.offset = 0;
data = kmalloc(eeprom.len, GFP_KERNEL);
drivers/net: Remove unnecessary k.alloc/v.alloc OOM messages alloc failures use dump_stack so emitting an additional out-of-memory message is an unnecessary duplication. Remove the allocation failure messages. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-01-29 12:56:23 +00:00
if (!data)
e1000: Dump the eeprom when a user encounters a bad checksum To help supporting users with a bad eeprom checksum, dump the eeprom info when such a situation is encountered by a user. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-17 21:50:23 +00:00
return;
ops->get_eeprom(netdev, &eeprom, data);
csum_old = (data[EEPROM_CHECKSUM_REG * 2]) +
(data[EEPROM_CHECKSUM_REG * 2 + 1] << 8);
for (i = 0; i < EEPROM_CHECKSUM_REG * 2; i += 2)
csum_new += data[i] + (data[i + 1] << 8);
csum_new = EEPROM_SUM - csum_new;
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
pr_err("/*********************/\n");
pr_err("Current EEPROM Checksum : 0x%04x\n", csum_old);
pr_err("Calculated : 0x%04x\n", csum_new);
e1000: Dump the eeprom when a user encounters a bad checksum To help supporting users with a bad eeprom checksum, dump the eeprom info when such a situation is encountered by a user. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-17 21:50:23 +00:00
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
pr_err("Offset Values\n");
pr_err("======== ======\n");
e1000: Dump the eeprom when a user encounters a bad checksum To help supporting users with a bad eeprom checksum, dump the eeprom info when such a situation is encountered by a user. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-17 21:50:23 +00:00
print_hex_dump(KERN_ERR, "", DUMP_PREFIX_OFFSET, 16, 1, data, 128, 0);
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
pr_err("Include this output when contacting your support provider.\n");
pr_err("This is not a software error! Something bad happened to\n");
pr_err("your hardware or EEPROM image. Ignoring this problem could\n");
pr_err("result in further problems, possibly loss of data,\n");
pr_err("corruption or system hangs!\n");
pr_err("The MAC Address will be reset to 00:00:00:00:00:00,\n");
pr_err("which is invalid and requires you to set the proper MAC\n");
pr_err("address manually before continuing to enable this network\n");
pr_err("device. Please inspect the EEPROM dump and report the\n");
pr_err("issue to your hardware vendor or Intel Customer Support.\n");
pr_err("/*********************/\n");
e1000: Dump the eeprom when a user encounters a bad checksum To help supporting users with a bad eeprom checksum, dump the eeprom info when such a situation is encountered by a user. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-17 21:50:23 +00:00
kfree(data);
}
e1000: make ioport free This patch makes e1000 driver ioport-free. This corrects behavior in probe function so as not to request ioport resources as long as they are not really needed. This is based on the ioport-free patch of e1000 driver from Auke Kok and Tomohiro Kusumi. Signed-off-by: Tomohiro Kusumi <kusumi.tomohiro@jp.fujitsu.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Taku Izumi <izumi.taku@jp.fujitsu.com> Signed-off-by: Jeff Kirsher<jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:44 +00:00
/**
* e1000_is_need_ioport - determine if an adapter needs ioport resources or not
* @pdev: PCI device information struct
*
* Return true if an adapter needs ioport resources
**/
static int e1000_is_need_ioport(struct pci_dev *pdev)
{
switch (pdev->device) {
case E1000_DEV_ID_82540EM:
case E1000_DEV_ID_82540EM_LOM:
case E1000_DEV_ID_82540EP:
case E1000_DEV_ID_82540EP_LOM:
case E1000_DEV_ID_82540EP_LP:
case E1000_DEV_ID_82541EI:
case E1000_DEV_ID_82541EI_MOBILE:
case E1000_DEV_ID_82541ER:
case E1000_DEV_ID_82541ER_LOM:
case E1000_DEV_ID_82541GI:
case E1000_DEV_ID_82541GI_LF:
case E1000_DEV_ID_82541GI_MOBILE:
case E1000_DEV_ID_82544EI_COPPER:
case E1000_DEV_ID_82544EI_FIBER:
case E1000_DEV_ID_82544GC_COPPER:
case E1000_DEV_ID_82544GC_LOM:
case E1000_DEV_ID_82545EM_COPPER:
case E1000_DEV_ID_82545EM_FIBER:
case E1000_DEV_ID_82546EB_COPPER:
case E1000_DEV_ID_82546EB_FIBER:
case E1000_DEV_ID_82546EB_QUAD_COPPER:
return true;
default:
return false;
}
}
net: introduce and use netdev_features_t for device features sets v2: add couple missing conversions in drivers split unexporting netdev_fix_features() implemented %pNF convert sock::sk_route_(no?)caps Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-11-15 15:29:55 +00:00
static netdev_features_t e1000_fix_features(struct net_device *netdev,
netdev_features_t features)
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
{
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* Since there is no support for separate Rx/Tx vlan accel
* enable/disable make sure Tx flag is always in same state as Rx.
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
*/
net: vlan: rename NETIF_F_HW_VLAN_* feature flags to NETIF_F_HW_VLAN_CTAG_* Rename the hardware VLAN acceleration features to include "CTAG" to indicate that they only support CTAGs. Follow up patches will introduce 802.1ad server provider tagging (STAGs) and require the distinction for hardware not supporting acclerating both. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-04-19 02:04:27 +00:00
if (features & NETIF_F_HW_VLAN_CTAG_RX)
features |= NETIF_F_HW_VLAN_CTAG_TX;
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
else
net: vlan: rename NETIF_F_HW_VLAN_* feature flags to NETIF_F_HW_VLAN_CTAG_* Rename the hardware VLAN acceleration features to include "CTAG" to indicate that they only support CTAGs. Follow up patches will introduce 802.1ad server provider tagging (STAGs) and require the distinction for hardware not supporting acclerating both. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-04-19 02:04:27 +00:00
features &= ~NETIF_F_HW_VLAN_CTAG_TX;
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
return features;
}
net: introduce and use netdev_features_t for device features sets v2: add couple missing conversions in drivers split unexporting netdev_fix_features() implemented %pNF convert sock::sk_route_(no?)caps Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-11-15 15:29:55 +00:00
static int e1000_set_features(struct net_device *netdev,
netdev_features_t features)
e1000: convert to ndo_fix_features Private rx_csum flags are now duplicate of netdev->features & NETIF_F_RXCSUM. Removing this needs deeper surgery. Things noticed: - RX csum disabled by default - HW VLAN acceleration probably can be toggled, but it's left as is - the resets on RX csum offload change can probably be avoided - there is A LOT of copy-and-pasted code here Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-06-08 08:36:42 +00:00
{
struct e1000_adapter *adapter = netdev_priv(netdev);
net: introduce and use netdev_features_t for device features sets v2: add couple missing conversions in drivers split unexporting netdev_fix_features() implemented %pNF convert sock::sk_route_(no?)caps Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-11-15 15:29:55 +00:00
netdev_features_t changed = features ^ netdev->features;
e1000: convert to ndo_fix_features Private rx_csum flags are now duplicate of netdev->features & NETIF_F_RXCSUM. Removing this needs deeper surgery. Things noticed: - RX csum disabled by default - HW VLAN acceleration probably can be toggled, but it's left as is - the resets on RX csum offload change can probably be avoided - there is A LOT of copy-and-pasted code here Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-06-08 08:36:42 +00:00
net: vlan: rename NETIF_F_HW_VLAN_* feature flags to NETIF_F_HW_VLAN_CTAG_* Rename the hardware VLAN acceleration features to include "CTAG" to indicate that they only support CTAGs. Follow up patches will introduce 802.1ad server provider tagging (STAGs) and require the distinction for hardware not supporting acclerating both. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-04-19 02:04:27 +00:00
if (changed & NETIF_F_HW_VLAN_CTAG_RX)
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
e1000_vlan_mode(netdev, features);
e1000: Support RX-ALL flag. This allows the NIC to receive errored frames (bad FCS, etc) and pass them up the stack. This can be useful when using sniffers. Signed-off-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-04-04 06:01:29 +00:00
if (!(changed & (NETIF_F_RXCSUM | NETIF_F_RXALL)))
e1000: convert to ndo_fix_features Private rx_csum flags are now duplicate of netdev->features & NETIF_F_RXCSUM. Removing this needs deeper surgery. Things noticed: - RX csum disabled by default - HW VLAN acceleration probably can be toggled, but it's left as is - the resets on RX csum offload change can probably be avoided - there is A LOT of copy-and-pasted code here Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-06-08 08:36:42 +00:00
return 0;
e1000: Support RX-ALL flag. This allows the NIC to receive errored frames (bad FCS, etc) and pass them up the stack. This can be useful when using sniffers. Signed-off-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-04-04 06:01:29 +00:00
netdev->features = features;
e1000: convert to ndo_fix_features Private rx_csum flags are now duplicate of netdev->features & NETIF_F_RXCSUM. Removing this needs deeper surgery. Things noticed: - RX csum disabled by default - HW VLAN acceleration probably can be toggled, but it's left as is - the resets on RX csum offload change can probably be avoided - there is A LOT of copy-and-pasted code here Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-06-08 08:36:42 +00:00
adapter->rx_csum = !!(features & NETIF_F_RXCSUM);
if (netif_running(netdev))
e1000_reinit_locked(adapter);
else
e1000_reset(adapter);
return 0;
}
e1000: convert to net_device_ops Convert to new network device ops interface. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-11-20 06:18:22 +00:00
static const struct net_device_ops e1000_netdev_ops = {
.ndo_open = e1000_open,
.ndo_stop = e1000_close,
netdev: add more functions to netdevice ops This patch moves neigh_setup and hard_start_xmit into the network device ops structure. For bisection, fix all the previously converted drivers as well. Bonding driver took the biggest hit on this. Added a prefetch of the hard_start_xmit in the fast path to try and reduce any impact this would have. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-11-21 04:14:53 +00:00
.ndo_start_xmit = e1000_xmit_frame,
e1000: convert to net_device_ops Convert to new network device ops interface. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-11-20 06:18:22 +00:00
.ndo_get_stats = e1000_get_stats,
.ndo_set_rx_mode = e1000_set_rx_mode,
.ndo_set_mac_address = e1000_set_mac,
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
.ndo_tx_timeout = e1000_tx_timeout,
e1000: convert to net_device_ops Convert to new network device ops interface. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-11-20 06:18:22 +00:00
.ndo_change_mtu = e1000_change_mtu,
.ndo_do_ioctl = e1000_ioctl,
.ndo_validate_addr = eth_validate_addr,
.ndo_vlan_rx_add_vid = e1000_vlan_rx_add_vid,
.ndo_vlan_rx_kill_vid = e1000_vlan_rx_kill_vid,
#ifdef CONFIG_NET_POLL_CONTROLLER
.ndo_poll_controller = e1000_netpoll,
#endif
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
.ndo_fix_features = e1000_fix_features,
.ndo_set_features = e1000_set_features,
e1000: convert to net_device_ops Convert to new network device ops interface. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-11-20 06:18:22 +00:00
};
e1000: fix Tx hangs by disabling 64-bit DMA Several users report issues with 32-bit adapters when plugged into PCI slots in machines with >= 4GB ram. In particular AMD systems with HyperTransport to PCI bridges seem to trigger the issue, but it isn't limited to only them. This issue is not easily reproducible here, yet still continues to occur in the field. For e1000 on PCI devices, just disable DMA addresses over the 4GB boundary when in PCI (not PCI-X) mode, to prevent the issue from continuing to pop up. The performance impact for this is negligible. The code was refactored to move the init of the hw struct to its own function. This allows the init to be called very early in probe, which then allows using hw-> members for this fix. A slight refactor to the DMA mask code was done for minor correctness based on the instructions in DMA-API-HOWTO. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-07 21:01:12 +00:00
/**
* e1000_init_hw_struct - initialize members of hw struct
* @adapter: board private struct
* @hw: structure used by e1000_hw.c
*
* Factors out initialization of the e1000_hw struct to its own function
* that can be called very early at init (just after struct allocation).
* Fields are initialized based on PCI device information and
* OS network device settings (MTU size).
* Returns negative error codes if MAC type setup fails.
*/
static int e1000_init_hw_struct(struct e1000_adapter *adapter,
struct e1000_hw *hw)
{
struct pci_dev *pdev = adapter->pdev;
/* PCI config space info */
hw->vendor_id = pdev->vendor;
hw->device_id = pdev->device;
hw->subsystem_vendor_id = pdev->subsystem_vendor;
hw->subsystem_id = pdev->subsystem_device;
hw->revision_id = pdev->revision;
pci_read_config_word(pdev, PCI_COMMAND, &hw->pci_cmd_word);
hw->max_frame_size = adapter->netdev->mtu +
ENET_HEADER_SIZE + ETHERNET_FCS_SIZE;
hw->min_frame_size = MINIMUM_ETHERNET_FRAME_SIZE;
/* identify the MAC */
if (e1000_set_mac_type(hw)) {
e_err(probe, "Unknown MAC Type\n");
return -EIO;
}
switch (hw->mac_type) {
default:
break;
case e1000_82541:
case e1000_82547:
case e1000_82541_rev_2:
case e1000_82547_rev_2:
hw->phy_init_script = 1;
break;
}
e1000_set_media_type(hw);
e1000_get_bus_info(hw);
hw->wait_autoneg_complete = false;
hw->tbi_compatibility_en = true;
hw->adaptive_ifs = true;
/* Copper options */
if (hw->media_type == e1000_media_type_copper) {
hw->mdix = AUTO_ALL_MODES;
hw->disable_polarity_correction = false;
hw->master_slave = E1000_MASTER_SLAVE;
}
return 0;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/**
* e1000_probe - Device Initialization Routine
* @pdev: PCI device information struct
* @ent: entry in e1000_pci_tbl
*
* Returns 0 on success, negative on failure
*
* e1000_probe initializes an adapter identified by a pci_dev structure.
* The OS initialization, configuring of the adapter private structure,
* and a hardware reset occur.
**/
drivers/net: fix up function prototypes after __dev* removals The __dev* removal patches for the network drivers ended up messing up the function prototypes for a bunch of drivers. This patch fixes all of them back up to be properly aligned. Bonus is that this almost removes 100 lines of code, always a nice surprise. Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-12-06 14:30:56 +00:00
static int e1000_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct net_device *netdev;
struct e1000_adapter *adapter;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw;
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
static int cards_found = 0;
e1000: unify WoL capability detection code WoL is constantly giving problems and needed a rewrite. Consolidates all WoL capabilities into a single function, and disables WoL for all other ports on the device except for port A. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-31 21:27:46 +00:00
static int global_quad_port_a = 0; /* global ksp3 port a indication */
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
int i, err, pci_using_dac;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u16 eeprom_data = 0;
e1000: Add support for the CE4100 reference platform This patch adds support for the gigabit phys present on the CE4100 reference platforms. Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-06 14:29:54 +00:00
u16 tmp = 0;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u16 eeprom_apme_mask = E1000_EEPROM_APME;
e1000: make ioport free This patch makes e1000 driver ioport-free. This corrects behavior in probe function so as not to request ioport resources as long as they are not really needed. This is based on the ioport-free patch of e1000 driver from Auke Kok and Tomohiro Kusumi. Signed-off-by: Tomohiro Kusumi <kusumi.tomohiro@jp.fujitsu.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Taku Izumi <izumi.taku@jp.fujitsu.com> Signed-off-by: Jeff Kirsher<jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:44 +00:00
int bars, need_ioport;
[NET]: Introduce and use print_mac() and DECLARE_MAC_BUF() This is nicer than the MAC_FMT stuff. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 00:59:30 +00:00
e1000: make ioport free This patch makes e1000 driver ioport-free. This corrects behavior in probe function so as not to request ioport resources as long as they are not really needed. This is based on the ioport-free patch of e1000 driver from Auke Kok and Tomohiro Kusumi. Signed-off-by: Tomohiro Kusumi <kusumi.tomohiro@jp.fujitsu.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Taku Izumi <izumi.taku@jp.fujitsu.com> Signed-off-by: Jeff Kirsher<jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:44 +00:00
/* do not allocate ioport bars when not needed */
need_ioport = e1000_is_need_ioport(pdev);
if (need_ioport) {
bars = pci_select_bars(pdev, IORESOURCE_MEM | IORESOURCE_IO);
err = pci_enable_device(pdev);
} else {
bars = pci_select_bars(pdev, IORESOURCE_MEM);
e1000: Fix PCI enable to honor the need_ioport flag On machine were no IO ports are assigned the call to pci_enable_device() will fail, even if need_ioport is false, we need to use pci_enable_device_mem() here. Signed-off-by: Karsten Keil <kkeil@suse.de> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-02-03 23:18:01 +00:00
err = pci_enable_device_mem(pdev);
e1000: make ioport free This patch makes e1000 driver ioport-free. This corrects behavior in probe function so as not to request ioport resources as long as they are not really needed. This is based on the ioport-free patch of e1000 driver from Auke Kok and Tomohiro Kusumi. Signed-off-by: Tomohiro Kusumi <kusumi.tomohiro@jp.fujitsu.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Taku Izumi <izumi.taku@jp.fujitsu.com> Signed-off-by: Jeff Kirsher<jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:44 +00:00
}
e1000: Move assignments in tests before test Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:28 +00:00
if (err)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return err;
e1000: make ioport free This patch makes e1000 driver ioport-free. This corrects behavior in probe function so as not to request ioport resources as long as they are not really needed. This is based on the ioport-free patch of e1000 driver from Auke Kok and Tomohiro Kusumi. Signed-off-by: Tomohiro Kusumi <kusumi.tomohiro@jp.fujitsu.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Taku Izumi <izumi.taku@jp.fujitsu.com> Signed-off-by: Jeff Kirsher<jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:44 +00:00
err = pci_request_selected_regions(pdev, bars, e1000_driver_name);
e1000: Move assignments in tests before test Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:28 +00:00
if (err)
e1000: e1000_probe resources cleanup Fix resources cleanup in e1000_probe() Signed-off-by: Vasily Averin <vvs@sw.ru> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-28 21:56:22 +00:00
goto err_pci_reg;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
pci_set_master(pdev);
e1000: call pci_save_state after pci_restore_state This patch adds a call to pci_save_state() immediately after the call to pci_restore_state(). Due to a change in the behavior of pci_restore_state() it is necessary to call pci_save_state() to keep the state_saved flag. This patch is based on a similar patch for ixgbe. Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-02-03 14:49:48 +00:00
err = pci_save_state(pdev);
if (err)
goto err_alloc_etherdev;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: e1000_probe resources cleanup Fix resources cleanup in e1000_probe() Signed-off-by: Vasily Averin <vvs@sw.ru> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-28 21:56:22 +00:00
err = -ENOMEM;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
netdev = alloc_etherdev(sizeof(struct e1000_adapter));
e1000: e1000_probe resources cleanup Fix resources cleanup in e1000_probe() Signed-off-by: Vasily Averin <vvs@sw.ru> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-28 21:56:22 +00:00
if (!netdev)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto err_alloc_etherdev;
SET_NETDEV_DEV(netdev, &pdev->dev);
pci_set_drvdata(pdev, netdev);
[PATCH] e1000: Use netdev_priv() to get to netdev->priv Use netdev_priv() to get to netdev->priv - from shemminger@osdl.org Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:41:45 +00:00
adapter = netdev_priv(netdev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
adapter->netdev = netdev;
adapter->pdev = pdev;
intel: make wired ethernet driver message level consistent (rev2) Dan Carpenter noticed that ixgbevf initial default was different than the rest. But the problem is broader than that, only one Intel driver (ixgb) was doing it almost right. The convention for default debug level should be consistent among Intel drivers and follow established convention. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-03-13 06:04:20 +00:00
adapter->msg_enable = netif_msg_init(debug, DEFAULT_MSG_ENABLE);
e1000: make ioport free This patch makes e1000 driver ioport-free. This corrects behavior in probe function so as not to request ioport resources as long as they are not really needed. This is based on the ioport-free patch of e1000 driver from Auke Kok and Tomohiro Kusumi. Signed-off-by: Tomohiro Kusumi <kusumi.tomohiro@jp.fujitsu.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Taku Izumi <izumi.taku@jp.fujitsu.com> Signed-off-by: Jeff Kirsher<jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:44 +00:00
adapter->bars = bars;
adapter->need_ioport = need_ioport;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
hw = &adapter->hw;
hw->back = adapter;
e1000: e1000_probe resources cleanup Fix resources cleanup in e1000_probe() Signed-off-by: Vasily Averin <vvs@sw.ru> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-28 21:56:22 +00:00
err = -EIO;
pci: use pci_ioremap_bar() in drivers/net Use the newly introduced pci_ioremap_bar() function in drivers/net. pci_ioremap_bar() just takes a pci device and a bar number, with the goal of making it really hard to get wrong, while also having a central place to stick sanity checks. Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-10-21 04:42:39 +00:00
hw->hw_addr = pci_ioremap_bar(pdev, BAR_0);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (!hw->hw_addr)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto err_ioremap;
e1000: make ioport free This patch makes e1000 driver ioport-free. This corrects behavior in probe function so as not to request ioport resources as long as they are not really needed. This is based on the ioport-free patch of e1000 driver from Auke Kok and Tomohiro Kusumi. Signed-off-by: Tomohiro Kusumi <kusumi.tomohiro@jp.fujitsu.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Taku Izumi <izumi.taku@jp.fujitsu.com> Signed-off-by: Jeff Kirsher<jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:44 +00:00
if (adapter->need_ioport) {
for (i = BAR_1; i <= BAR_5; i++) {
if (pci_resource_len(pdev, i) == 0)
continue;
if (pci_resource_flags(pdev, i) & IORESOURCE_IO) {
hw->io_base = pci_resource_start(pdev, i);
break;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
}
e1000: fix Tx hangs by disabling 64-bit DMA Several users report issues with 32-bit adapters when plugged into PCI slots in machines with >= 4GB ram. In particular AMD systems with HyperTransport to PCI bridges seem to trigger the issue, but it isn't limited to only them. This issue is not easily reproducible here, yet still continues to occur in the field. For e1000 on PCI devices, just disable DMA addresses over the 4GB boundary when in PCI (not PCI-X) mode, to prevent the issue from continuing to pop up. The performance impact for this is negligible. The code was refactored to move the init of the hw struct to its own function. This allows the init to be called very early in probe, which then allows using hw-> members for this fix. A slight refactor to the DMA mask code was done for minor correctness based on the instructions in DMA-API-HOWTO. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-07 21:01:12 +00:00
/* make ready for any if (hw->...) below */
err = e1000_init_hw_struct(adapter, hw);
if (err)
goto err_sw_init;
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* there is a workaround being applied below that limits
e1000: fix Tx hangs by disabling 64-bit DMA Several users report issues with 32-bit adapters when plugged into PCI slots in machines with >= 4GB ram. In particular AMD systems with HyperTransport to PCI bridges seem to trigger the issue, but it isn't limited to only them. This issue is not easily reproducible here, yet still continues to occur in the field. For e1000 on PCI devices, just disable DMA addresses over the 4GB boundary when in PCI (not PCI-X) mode, to prevent the issue from continuing to pop up. The performance impact for this is negligible. The code was refactored to move the init of the hw struct to its own function. This allows the init to be called very early in probe, which then allows using hw-> members for this fix. A slight refactor to the DMA mask code was done for minor correctness based on the instructions in DMA-API-HOWTO. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-07 21:01:12 +00:00
* 64-bit DMA addresses to 64-bit hardware. There are some
* 32-bit adapters that Tx hang when given 64-bit DMA addresses
*/
pci_using_dac = 0;
if ((hw->bus_type == e1000_bus_type_pcix) &&
DMA-API: net: intel/e1000: replace dma_set_mask()+dma_set_coherent_mask() with new helper Replace the following sequence: dma_set_mask(dev, mask); dma_set_coherent_mask(dev, mask); with a call to the new helper dma_set_mask_and_coherent(). Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2013-06-26 22:49:11 +00:00
!dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64))) {
e1000: fix Tx hangs by disabling 64-bit DMA Several users report issues with 32-bit adapters when plugged into PCI slots in machines with >= 4GB ram. In particular AMD systems with HyperTransport to PCI bridges seem to trigger the issue, but it isn't limited to only them. This issue is not easily reproducible here, yet still continues to occur in the field. For e1000 on PCI devices, just disable DMA addresses over the 4GB boundary when in PCI (not PCI-X) mode, to prevent the issue from continuing to pop up. The performance impact for this is negligible. The code was refactored to move the init of the hw struct to its own function. This allows the init to be called very early in probe, which then allows using hw-> members for this fix. A slight refactor to the DMA mask code was done for minor correctness based on the instructions in DMA-API-HOWTO. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-07 21:01:12 +00:00
pci_using_dac = 1;
} else {
DMA-API: net: intel/e1000: replace dma_set_mask()+dma_set_coherent_mask() with new helper Replace the following sequence: dma_set_mask(dev, mask); dma_set_coherent_mask(dev, mask); with a call to the new helper dma_set_mask_and_coherent(). Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2013-06-26 22:49:11 +00:00
err = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32));
e1000: fix return value not set on error Dean noticed that 'err' wasn't being set when the "goto err_dma" statement is executed in the following hunk from the commit. It's value will be zero as a result of a successful call to e1000_init_hw_struct(). This patch changes the error condition to be correctly propagated. CC: stable@kernel.org Signed-off-by: Dean Nelson <dnelson@redhat.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2010-11-11 05:50:25 +00:00
if (err) {
pr_err("No usable DMA config, aborting\n");
goto err_dma;
}
e1000: fix Tx hangs by disabling 64-bit DMA Several users report issues with 32-bit adapters when plugged into PCI slots in machines with >= 4GB ram. In particular AMD systems with HyperTransport to PCI bridges seem to trigger the issue, but it isn't limited to only them. This issue is not easily reproducible here, yet still continues to occur in the field. For e1000 on PCI devices, just disable DMA addresses over the 4GB boundary when in PCI (not PCI-X) mode, to prevent the issue from continuing to pop up. The performance impact for this is negligible. The code was refactored to move the init of the hw struct to its own function. This allows the init to be called very early in probe, which then allows using hw-> members for this fix. A slight refactor to the DMA mask code was done for minor correctness based on the instructions in DMA-API-HOWTO. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-07 21:01:12 +00:00
}
e1000: convert to net_device_ops Convert to new network device ops interface. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-11-20 06:18:22 +00:00
netdev->netdev_ops = &e1000_netdev_ops;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_set_ethtool_ops(netdev);
netdev->watchdog_timeo = 5 * HZ;
[NET]: Make NAPI polling independent of struct net_device objects. Several devices have multiple independant RX queues per net device, and some have a single interrupt doorbell for several queues. In either case, it's easier to support layouts like that if the structure representing the poll is independant from the net device itself. The signature of the ->poll() call back goes from: int foo_poll(struct net_device *dev, int *budget) to int foo_poll(struct napi_struct *napi, int budget) The caller is returned the number of RX packets processed (or the number of "NAPI credits" consumed if you want to get abstract). The callee no longer messes around bumping dev->quota, *budget, etc. because that is all handled in the caller upon return. The napi_struct is to be embedded in the device driver private data structures. Furthermore, it is the driver's responsibility to disable all NAPI instances in it's ->stop() device close handler. Since the napi_struct is privatized into the driver's private data structures, only the driver knows how to get at all of the napi_struct instances it may have per-device. With lots of help and suggestions from Rusty Russell, Roland Dreier, Michael Chan, Jeff Garzik, and Jamal Hadi Salim. Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra, Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan. [ Ported to current tree and all drivers converted. Integrated Stephen's follow-on kerneldoc additions, and restored poll_list handling to the old style to fix mutual exclusion issues. -DaveM ] Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-03 23:41:36 +00:00
netif_napi_add(netdev, &adapter->napi, e1000_clean, 64);
e1000: convert to net_device_ops Convert to new network device ops interface. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-11-20 06:18:22 +00:00
e100, e1000, ixgb: Fix an impossible memory overwrite bug We keep getting requests from people that think that this might be an exploitable hole where we would overwrite 4 bytes in the netdev struct if the pci name would exceed 15 characters. In reality this will never happen but we fix it anyway. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:17 +00:00
strncpy(netdev->name, pci_name(pdev), sizeof(netdev->name) - 1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
adapter->bd_number = cards_found;
/* setup the private structure */
e1000: Move assignments in tests before test Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:28 +00:00
err = e1000_sw_init(adapter);
if (err)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto err_sw_init;
e1000: e1000_probe resources cleanup Fix resources cleanup in e1000_probe() Signed-off-by: Vasily Averin <vvs@sw.ru> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-28 21:56:22 +00:00
err = -EIO;
e1000: Add support for the CE4100 reference platform This patch adds support for the gigabit phys present on the CE4100 reference platforms. Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-06 14:29:54 +00:00
if (hw->mac_type == e1000_ce4100) {
e1000: cleanup CE4100 MDIO registers access A global variable is currently used to hold the virtual address of the CE4100 MDIO base register address. Store the address in the e1000_hw structure and update macros accordingly. Signed-off-by: Florian Fainelli <ffainelli@freebox.fr> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-01-04 20:23:35 +00:00
hw->ce4100_gbe_mdio_base_virt =
ioremap(pci_resource_start(pdev, BAR_1),
e1000: Add support for the CE4100 reference platform This patch adds support for the gigabit phys present on the CE4100 reference platforms. Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-06 14:29:54 +00:00
pci_resource_len(pdev, BAR_1));
e1000: cleanup CE4100 MDIO registers access A global variable is currently used to hold the virtual address of the CE4100 MDIO base register address. Store the address in the e1000_hw structure and update macros accordingly. Signed-off-by: Florian Fainelli <ffainelli@freebox.fr> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-01-04 20:23:35 +00:00
if (!hw->ce4100_gbe_mdio_base_virt)
e1000: Add support for the CE4100 reference platform This patch adds support for the gigabit phys present on the CE4100 reference platforms. Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-06 14:29:54 +00:00
goto err_mdio_ioremap;
}
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (hw->mac_type >= e1000_82543) {
e1000: convert to ndo_fix_features Private rx_csum flags are now duplicate of netdev->features & NETIF_F_RXCSUM. Removing this needs deeper surgery. Things noticed: - RX csum disabled by default - HW VLAN acceleration probably can be toggled, but it's left as is - the resets on RX csum offload change can probably be avoided - there is A LOT of copy-and-pasted code here Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-06-08 08:36:42 +00:00
netdev->hw_features = NETIF_F_SG |
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
NETIF_F_HW_CSUM |
net: vlan: rename NETIF_F_HW_VLAN_* feature flags to NETIF_F_HW_VLAN_CTAG_* Rename the hardware VLAN acceleration features to include "CTAG" to indicate that they only support CTAGs. Follow up patches will introduce 802.1ad server provider tagging (STAGs) and require the distinction for hardware not supporting acclerating both. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-04-19 02:04:27 +00:00
NETIF_F_HW_VLAN_CTAG_RX;
netdev->features = NETIF_F_HW_VLAN_CTAG_TX |
NETIF_F_HW_VLAN_CTAG_FILTER;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if ((hw->mac_type >= e1000_82544) &&
(hw->mac_type != e1000_82547))
e1000: convert to ndo_fix_features Private rx_csum flags are now duplicate of netdev->features & NETIF_F_RXCSUM. Removing this needs deeper surgery. Things noticed: - RX csum disabled by default - HW VLAN acceleration probably can be toggled, but it's left as is - the resets on RX csum offload change can probably be avoided - there is A LOT of copy-and-pasted code here Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-06-08 08:36:42 +00:00
netdev->hw_features |= NETIF_F_TSO;
e1000: Support sending custom Ethernet CRC. Good for testing the RX logic for bad CRC handling. Signed-off-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-11 15:40:01 +00:00
netdev->priv_flags |= IFF_SUPP_NOFCS;
e1000: convert to ndo_fix_features Private rx_csum flags are now duplicate of netdev->features & NETIF_F_RXCSUM. Removing this needs deeper surgery. Things noticed: - RX csum disabled by default - HW VLAN acceleration probably can be toggled, but it's left as is - the resets on RX csum offload change can probably be avoided - there is A LOT of copy-and-pasted code here Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-06-08 08:36:42 +00:00
netdev->features |= netdev->hw_features;
e1000: Combining Bitwise OR in one expression. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-06-12 13:03:29 +00:00
netdev->hw_features |= (NETIF_F_RXCSUM |
NETIF_F_RXALL |
NETIF_F_RXFCS);
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
e1000/e1000e/igb/ixgb/ixgbe: set NETIF_F_HIGHDMA for VLAN feature flags If the netdev->features is set with NETIF_F_HIGHDMA, we should set the corresponding netdev->vlan_features as well to allow VLAN netdev created on top of the real netdev to be able to also benefit from HIGHDMA on 32bit system, reducing the performance hit that is caused by __skb_linearize(), particularly for large send. This is fixed in this patch for all Intel e1000, e1000e, igb, ixgbe, and ixgbe drivers since this should be beneficial to all devices supported by these drivers. Signed-off-by: Yi Zou <yi.zou@intel.com> Tested-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-22 17:57:58 +00:00
if (pci_using_dac) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
netdev->features |= NETIF_F_HIGHDMA;
e1000/e1000e/igb/ixgb/ixgbe: set NETIF_F_HIGHDMA for VLAN feature flags If the netdev->features is set with NETIF_F_HIGHDMA, we should set the corresponding netdev->vlan_features as well to allow VLAN netdev created on top of the real netdev to be able to also benefit from HIGHDMA on 32bit system, reducing the performance hit that is caused by __skb_linearize(), particularly for large send. This is fixed in this patch for all Intel e1000, e1000e, igb, ixgbe, and ixgbe drivers since this should be beneficial to all devices supported by these drivers. Signed-off-by: Yi Zou <yi.zou@intel.com> Tested-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-22 17:57:58 +00:00
netdev->vlan_features |= NETIF_F_HIGHDMA;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Combining Bitwise OR in one expression. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-06-12 13:03:29 +00:00
netdev->vlan_features |= (NETIF_F_TSO |
NETIF_F_HW_CSUM |
NETIF_F_SG);
e1000: allow VLAN devices to use TSO and CSUM offload This patch changes e1000 to set vlan_features so TSO and CSUM offload can be used by VLAN devices, similar as with the other Intel drivers. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-11 19:25:59 +00:00
net: introduce IFF_UNICAST_FLT private flag Use IFF_UNICAST_FTL to find out if driver handles unicast address filtering. In case it does not, promisc mode is entered. Patch also fixes following drivers: stmmac, niu: support uc filtering and yet it propagated ndo_set_multicast_list bna, benet, pxa168_eth, ks8851, ks8851_mll, ksz884x : has set ndo_set_rx_mode but do not support uc filtering Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-08-16 06:29:00 +00:00
netdev->priv_flags |= IFF_UNICAST_FLT;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
adapter->en_mng_pt = e1000_enable_mng_pass_thru(hw);
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
e1000: integrate ich8 support into driver This hooks up the ich8 structure into the driver itself. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:08:22 +00:00
/* initialize eeprom parameters */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (e1000_init_eeprom_params(hw)) {
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(probe, "EEPROM initialization failed\n");
e1000: e1000_probe resources cleanup Fix resources cleanup in e1000_probe() Signed-off-by: Vasily Averin <vvs@sw.ru> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-28 21:56:22 +00:00
goto err_eeprom;
e1000: integrate ich8 support into driver This hooks up the ich8 structure into the driver itself. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:08:22 +00:00
}
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
/* before reading the EEPROM, reset the controller to
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* put the device in a known good starting state
*/
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_reset_hw(hw);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* make sure the EEPROM is good */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (e1000_validate_eeprom_checksum(hw) < 0) {
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(probe, "The EEPROM Checksum Is Not Valid\n");
e1000: Dump the eeprom when a user encounters a bad checksum To help supporting users with a bad eeprom checksum, dump the eeprom info when such a situation is encountered by a user. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-17 21:50:23 +00:00
e1000_dump_eeprom(adapter);
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* set MAC address to all zeroes to invalidate and temporary
e1000: Dump the eeprom when a user encounters a bad checksum To help supporting users with a bad eeprom checksum, dump the eeprom info when such a situation is encountered by a user. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-17 21:50:23 +00:00
* disable this device for the user. This blocks regular
* traffic while still permitting ethtool ioctls from reaching
* the hardware as well as allowing the user to run the
* interface after manually setting a hw addr using
* `ip set address`
*/
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
memset(hw->mac_addr, 0, netdev->addr_len);
e1000: Dump the eeprom when a user encounters a bad checksum To help supporting users with a bad eeprom checksum, dump the eeprom info when such a situation is encountered by a user. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-17 21:50:23 +00:00
} else {
/* copy the MAC address out of the EEPROM */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (e1000_read_mac_addr(hw))
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(probe, "EEPROM Read Error\n");
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: Dump the eeprom when a user encounters a bad checksum To help supporting users with a bad eeprom checksum, dump the eeprom info when such a situation is encountered by a user. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-17 21:50:23 +00:00
/* don't block initalization here due to bad MAC address */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
memcpy(netdev->dev_addr, hw->mac_addr, netdev->addr_len);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
remove init of dev->perm_addr in drivers perm_addr is initialized correctly in register_netdevice() so to init it in drivers is no longer needed. Signed-off-by: Jiri Pirko <jiri@resnulli.us> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-01-08 01:38:26 +00:00
if (!is_valid_ether_addr(netdev->dev_addr))
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(probe, "Invalid MAC Address\n");
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
INIT_DELAYED_WORK(&adapter->watchdog_task, e1000_watchdog);
INIT_DELAYED_WORK(&adapter->fifo_stall_task,
e1000_82547_tx_fifo_stall_task);
INIT_DELAYED_WORK(&adapter->phy_info_task, e1000_update_phy_info_task);
WorkStruct: Pass the work_struct pointer instead of context data Pass the work_struct pointer to the work function rather than context data. The work function can use container_of() to work out the data. For the cases where the container of the work_struct may go away the moment the pending bit is cleared, it is made possible to defer the release of the structure by deferring the clearing of the pending bit. To make this work, an extra flag is introduced into the management side of the work_struct. This governs auto-release of the structure upon execution. Ordinarily, the work queue executor would release the work_struct for further scheduling or deallocation by clearing the pending bit prior to jumping to the work function. This means that, unless the driver makes some guarantee itself that the work_struct won't go away, the work function may not access anything else in the work_struct or its container lest they be deallocated.. This is a problem if the auxiliary data is taken away (as done by the last patch). However, if the pending bit is *not* cleared before jumping to the work function, then the work function *may* access the work_struct and its container with no problems. But then the work function must itself release the work_struct by calling work_release(). In most cases, automatic release is fine, so this is the default. Special initiators exist for the non-auto-release case (ending in _NAR). Signed-Off-By: David Howells <dhowells@redhat.com>
2006-11-22 14:55:48 +00:00
INIT_WORK(&adapter->reset_task, e1000_reset_task);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_check_options(adapter);
/* Initial Wake on LAN setting
* If APM wake is enabled in the EEPROM,
* enable the ACPI Magic Packet filter
*/
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
switch (hw->mac_type) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
case e1000_82542_rev2_0:
case e1000_82542_rev2_1:
case e1000_82543:
break;
case e1000_82544:
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_read_eeprom(hw,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
EEPROM_INIT_CONTROL2_REG, 1, &eeprom_data);
eeprom_apme_mask = E1000_EEPROM_82544_APM;
break;
case e1000_82546:
case e1000_82546_rev_3:
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (er32(STATUS) & E1000_STATUS_FUNC_1){
e1000_read_eeprom(hw,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
EEPROM_INIT_CONTROL3_PORT_B, 1, &eeprom_data);
break;
}
/* Fall Through */
default:
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_read_eeprom(hw,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
EEPROM_INIT_CONTROL3_PORT_A, 1, &eeprom_data);
break;
}
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (eeprom_data & eeprom_apme_mask)
e1000: unify WoL capability detection code WoL is constantly giving problems and needed a rewrite. Consolidates all WoL capabilities into a single function, and disables WoL for all other ports on the device except for port A. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-31 21:27:46 +00:00
adapter->eeprom_wol |= E1000_WUFC_MAG;
/* now that we have the eeprom settings, apply the special cases
* where the eeprom may be wrong or the board simply won't support
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* wake on lan on a particular port
*/
e1000: unify WoL capability detection code WoL is constantly giving problems and needed a rewrite. Consolidates all WoL capabilities into a single function, and disables WoL for all other ports on the device except for port A. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-31 21:27:46 +00:00
switch (pdev->device) {
case E1000_DEV_ID_82546GB_PCIE:
adapter->eeprom_wol = 0;
break;
case E1000_DEV_ID_82546EB_FIBER:
case E1000_DEV_ID_82546GB_FIBER:
/* Wake events only supported on port A for dual fiber
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* regardless of eeprom setting
*/
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (er32(STATUS) & E1000_STATUS_FUNC_1)
e1000: unify WoL capability detection code WoL is constantly giving problems and needed a rewrite. Consolidates all WoL capabilities into a single function, and disables WoL for all other ports on the device except for port A. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-31 21:27:46 +00:00
adapter->eeprom_wol = 0;
break;
case E1000_DEV_ID_82546GB_QUAD_COPPER_KSP3:
/* if quad port adapter, disable WoL on all but port A */
if (global_quad_port_a != 0)
adapter->eeprom_wol = 0;
else
net: fix assignment of 0/1 to bool variables. DaveM said: Please, this kind of stuff rots forever and not using bool properly drives me crazy. Joe Perches <joe@perches.com> gave me the spatch script: @@ bool b; @@ -b = 0 +b = false @@ bool b; @@ -b = 1 +b = true I merely installed coccinelle, read the documentation and took credit. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-12-19 13:56:45 +00:00
adapter->quad_port_a = true;
e1000: unify WoL capability detection code WoL is constantly giving problems and needed a rewrite. Consolidates all WoL capabilities into a single function, and disables WoL for all other ports on the device except for port A. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-31 21:27:46 +00:00
/* Reset for multiple quad port adapters */
if (++global_quad_port_a == 4)
global_quad_port_a = 0;
break;
}
/* initialize the wol settings based on the eeprom settings */
adapter->wol = adapter->eeprom_wol;
e1000: Use device_set_wakeup_enable Since dev->power.should_wakeup bit is used by the PCI core to decide whether the device should wake up the system from sleep states, set/unset this bit whenever WOL is enabled/disabled using e1000_set_wol(). Accordingly, use device_can_wakeup() for checking if wake-up is supported by the device. Signed-off-by: "Rafael J. Wysocki" <rjw@sisk.pl> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-11-07 20:30:19 +00:00
device_set_wakeup_enable(&adapter->pdev->dev, adapter->wol);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Add support for the CE4100 reference platform This patch adds support for the gigabit phys present on the CE4100 reference platforms. Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-06 14:29:54 +00:00
/* Auto detect PHY address */
if (hw->mac_type == e1000_ce4100) {
for (i = 0; i < 32; i++) {
hw->phy_addr = i;
e1000_read_phy_reg(hw, PHY_ID2, &tmp);
if (tmp == 0 || tmp == 0xFF) {
if (i == 31)
goto err_eeprom;
continue;
} else
break;
}
}
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
/* reset the hardware with the new settings */
e1000_reset(adapter);
strcpy(netdev->name, "eth%d");
err = register_netdev(netdev);
if (err)
goto err_register;
e1000: fix vlan processing regression This patch fixes a regression introduced by commit "e1000: do vlan cleanup (799d531)". Apparently some e1000 chips (not mine) are sensitive about the order of setting vlan filter and vlan stripping/inserting functionality. So this patch changes the order so it's the same as before vlan cleanup. Reported-by: Ben Greear <greearb@candelatech.com> Signed-off-by: Jiri Pirko <jpirko@redhat.com> Tested-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-03-20 18:10:01 +00:00
e1000_vlan_filter_on_off(adapter, false);
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
[PATCH] e1000: Added PCIe bus information This is two patches, the first is adding additional bus information for the 8257{1|2|3} controllers. The second patch was orginally a community patch to print bus type/speed/width, and enhanced by us. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:25 +00:00
/* print bus type/speed/width info */
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_info(probe, "(PCI%s:%dMHz:%d-bit) %pM\n",
e1000: Fix message logging defect commit 675ad47375c76a7c3be4ace9554d92cd55518ced removed the capability to use ethtool.set_msglevel to control the types of messages emitted by the driver. That commit should probably be reverted. If not, then this patch fixes a message logging defect introduced by converting a printk without KERN_<level> to e_info. This also reduces text by about 200 bytes. Signed-off-by: Joe Perches <joe@perches.com> Tested-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-06-11 12:51:49 +00:00
((hw->bus_type == e1000_bus_type_pcix) ? "-X" : ""),
((hw->bus_speed == e1000_bus_speed_133) ? 133 :
(hw->bus_speed == e1000_bus_speed_120) ? 120 :
(hw->bus_speed == e1000_bus_speed_100) ? 100 :
(hw->bus_speed == e1000_bus_speed_66) ? 66 : 33),
((hw->bus_width == e1000_bus_width_64) ? 64 : 32),
netdev->dev_addr);
e1000: driver state fixes (race fix) We were plagued by our interrupt handler posting a watchdog event which could occur when our adapter was going down in case a late packet arrived just before e1000_down() finished. This caused the watchdog timer to start after the NIC was down and keep rescheduling it every N seconds. Once the driver unloaded it would panic. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:54:02 +00:00
e1000: init link state correctly As reported by Andrew Lutomirski <amluto@gmail.com> All the intel wired ethernet drivers were calling netif_carrier_off and netif_stop_queue (or variants) before calling register_netdevice This is incorrect behavior as was pointed out by davem, and causes ifconfig and friends to report a strange state before first link after the driver was loaded. This apparently confused *some* versions of networkmanager. Andy tested this for e1000e and confirmed it was working for him. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Reported-by: Andrew Lutomirski <amluto@gmail.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-17 20:44:36 +00:00
/* carrier off reporting is important to ethtool even BEFORE open */
netif_carrier_off(netdev);
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_info(probe, "Intel(R) PRO/1000 Network Connection\n");
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
cards_found++;
return 0;
err_register:
e1000: e1000_probe resources cleanup Fix resources cleanup in e1000_probe() Signed-off-by: Vasily Averin <vvs@sw.ru> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-28 21:56:22 +00:00
err_eeprom:
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
e1000_phy_hw_reset(hw);
e1000: e1000_probe resources cleanup Fix resources cleanup in e1000_probe() Signed-off-by: Vasily Averin <vvs@sw.ru> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-28 21:56:22 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (hw->flash_address)
iounmap(hw->flash_address);
e1000: e1000_probe resources cleanup Fix resources cleanup in e1000_probe() Signed-off-by: Vasily Averin <vvs@sw.ru> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-28 21:56:22 +00:00
kfree(adapter->tx_ring);
kfree(adapter->rx_ring);
e1000: fix Tx hangs by disabling 64-bit DMA Several users report issues with 32-bit adapters when plugged into PCI slots in machines with >= 4GB ram. In particular AMD systems with HyperTransport to PCI bridges seem to trigger the issue, but it isn't limited to only them. This issue is not easily reproducible here, yet still continues to occur in the field. For e1000 on PCI devices, just disable DMA addresses over the 4GB boundary when in PCI (not PCI-X) mode, to prevent the issue from continuing to pop up. The performance impact for this is negligible. The code was refactored to move the init of the hw struct to its own function. This allows the init to be called very early in probe, which then allows using hw-> members for this fix. A slight refactor to the DMA mask code was done for minor correctness based on the instructions in DMA-API-HOWTO. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-07 21:01:12 +00:00
err_dma:
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
err_sw_init:
e1000: Add support for the CE4100 reference platform This patch adds support for the gigabit phys present on the CE4100 reference platforms. Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-06 14:29:54 +00:00
err_mdio_ioremap:
e1000: cleanup CE4100 MDIO registers access A global variable is currently used to hold the virtual address of the CE4100 MDIO base register address. Store the address in the e1000_hw structure and update macros accordingly. Signed-off-by: Florian Fainelli <ffainelli@freebox.fr> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-01-04 20:23:35 +00:00
iounmap(hw->ce4100_gbe_mdio_base_virt);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
iounmap(hw->hw_addr);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
err_ioremap:
free_netdev(netdev);
err_alloc_etherdev:
e1000: make ioport free This patch makes e1000 driver ioport-free. This corrects behavior in probe function so as not to request ioport resources as long as they are not really needed. This is based on the ioport-free patch of e1000 driver from Auke Kok and Tomohiro Kusumi. Signed-off-by: Tomohiro Kusumi <kusumi.tomohiro@jp.fujitsu.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Taku Izumi <izumi.taku@jp.fujitsu.com> Signed-off-by: Jeff Kirsher<jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:44 +00:00
pci_release_selected_regions(pdev, bars);
e1000: e1000_probe resources cleanup Fix resources cleanup in e1000_probe() Signed-off-by: Vasily Averin <vvs@sw.ru> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-28 21:56:22 +00:00
err_pci_reg:
pci_disable_device(pdev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return err;
}
/**
* e1000_remove - Device Removal Routine
* @pdev: PCI device information struct
*
* e1000_remove is called by the PCI subsystem to alert the driver
* that it should release a PCI device. The could be caused by a
* Hot-Plug event, or because the driver is going to be removed from
* memory.
**/
net/intel: remove __dev* attributes CONFIG_HOTPLUG is going away as an option. As result the __dev* markings will be going away. Remove use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit. Signed-off-by: Bill Pemberton <wfp5p@virginia.edu> Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Bruce Allan <bruce.w.allan@intel.com> Cc: Carolyn Wyborny <carolyn.wyborny@intel.com> Cc: Don Skidmore <donald.c.skidmore@intel.com> Cc: Greg Rose <gregory.v.rose@intel.com> Cc: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com> Cc: Alex Duyck <alexander.h.duyck@intel.com> Cc: John Ronciak <john.ronciak@intel.com> Cc: Tushar Dave <tushar.n.dave@intel.com> Cc: e1000-devel@lists.sourceforge.net Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-03 14:24:25 +00:00
static void e1000_remove(struct pci_dev *pdev)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct net_device *netdev = pci_get_drvdata(pdev);
[PATCH] e1000: Use netdev_priv() to get to netdev->priv Use netdev_priv() to get to netdev->priv - from shemminger@osdl.org Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:41:45 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
e1000_down_and_stop(adapter);
e1000: consolidate managability enabling/disabling Several bugs existed in how we handle manageability issues all over the driver. This patch consolidates all the managability release and init code in two single functions and call them from appropriate locations. This fixes several BMC packet redirect issues and powerup/down hiccups. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:56:10 +00:00
e1000_release_manageability(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[NET]: Make NAPI polling independent of struct net_device objects. Several devices have multiple independant RX queues per net device, and some have a single interrupt doorbell for several queues. In either case, it's easier to support layouts like that if the structure representing the poll is independant from the net device itself. The signature of the ->poll() call back goes from: int foo_poll(struct net_device *dev, int *budget) to int foo_poll(struct napi_struct *napi, int budget) The caller is returned the number of RX packets processed (or the number of "NAPI credits" consumed if you want to get abstract). The callee no longer messes around bumping dev->quota, *budget, etc. because that is all handled in the caller upon return. The napi_struct is to be embedded in the device driver private data structures. Furthermore, it is the driver's responsibility to disable all NAPI instances in it's ->stop() device close handler. Since the napi_struct is privatized into the driver's private data structures, only the driver knows how to get at all of the napi_struct instances it may have per-device. With lots of help and suggestions from Rusty Russell, Roland Dreier, Michael Chan, Jeff Garzik, and Jamal Hadi Salim. Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra, Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan. [ Ported to current tree and all drivers converted. Integrated Stephen's follow-on kerneldoc additions, and restored poll_list handling to the old style to fix mutual exclusion issues. -DaveM ] Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-03 23:41:36 +00:00
unregister_netdev(netdev);
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
e1000_phy_hw_reset(hw);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: implementation of the multi-queue feature Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:03:23 +00:00
kfree(adapter->tx_ring);
kfree(adapter->rx_ring);
e1000: unmap ce4100_gbe_mdio_base_virt in e1000_remove We are not unmapping ce4100_gbe_mdio_base_virt in exit path in case we are running on a CE4100 adapter, fix that. Signed-off-by: Florian Fainelli <ffainelli@freebox.fr> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-01-04 20:23:34 +00:00
if (hw->mac_type == e1000_ce4100)
e1000: cleanup CE4100 MDIO registers access A global variable is currently used to hold the virtual address of the CE4100 MDIO base register address. Store the address in the e1000_hw structure and update macros accordingly. Signed-off-by: Florian Fainelli <ffainelli@freebox.fr> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-01-04 20:23:35 +00:00
iounmap(hw->ce4100_gbe_mdio_base_virt);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
iounmap(hw->hw_addr);
if (hw->flash_address)
iounmap(hw->flash_address);
e1000: make ioport free This patch makes e1000 driver ioport-free. This corrects behavior in probe function so as not to request ioport resources as long as they are not really needed. This is based on the ioport-free patch of e1000 driver from Auke Kok and Tomohiro Kusumi. Signed-off-by: Tomohiro Kusumi <kusumi.tomohiro@jp.fujitsu.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Taku Izumi <izumi.taku@jp.fujitsu.com> Signed-off-by: Jeff Kirsher<jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:44 +00:00
pci_release_selected_regions(pdev, adapter->bars);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
free_netdev(netdev);
pci_disable_device(pdev);
}
/**
* e1000_sw_init - Initialize general software structures (struct e1000_adapter)
* @adapter: board private structure to initialize
*
* e1000_sw_init initializes the Adapter private data structure.
e1000: fix Tx hangs by disabling 64-bit DMA Several users report issues with 32-bit adapters when plugged into PCI slots in machines with >= 4GB ram. In particular AMD systems with HyperTransport to PCI bridges seem to trigger the issue, but it isn't limited to only them. This issue is not easily reproducible here, yet still continues to occur in the field. For e1000 on PCI devices, just disable DMA addresses over the 4GB boundary when in PCI (not PCI-X) mode, to prevent the issue from continuing to pop up. The performance impact for this is negligible. The code was refactored to move the init of the hw struct to its own function. This allows the init to be called very early in probe, which then allows using hw-> members for this fix. A slight refactor to the DMA mask code was done for minor correctness based on the instructions in DMA-API-HOWTO. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-07 21:01:12 +00:00
* e1000_init_hw_struct MUST be called before this function
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
**/
net/intel: remove __dev* attributes CONFIG_HOTPLUG is going away as an option. As result the __dev* markings will be going away. Remove use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit. Signed-off-by: Bill Pemberton <wfp5p@virginia.edu> Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Bruce Allan <bruce.w.allan@intel.com> Cc: Carolyn Wyborny <carolyn.wyborny@intel.com> Cc: Don Skidmore <donald.c.skidmore@intel.com> Cc: Greg Rose <gregory.v.rose@intel.com> Cc: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com> Cc: Alex Duyck <alexander.h.duyck@intel.com> Cc: John Ronciak <john.ronciak@intel.com> Cc: Tushar Dave <tushar.n.dave@intel.com> Cc: e1000-devel@lists.sourceforge.net Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-03 14:24:25 +00:00
static int e1000_sw_init(struct e1000_adapter *adapter)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: fix panic on large frame receive when mtu=default A panic was reported when receiving 1522 size packets if using the default MTU. we should set the initial rx buffer length to the value that e1000changemtu sets so that we can receive any packet that would not be dropped by LPE=0. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke.jan.h.kok@intel.com>
2006-07-14 23:14:48 +00:00
adapter->rx_buffer_len = MAXIMUM_ETHERNET_VLAN_SIZE;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[PATCH] e1000: Fix adapter structure and prepare for multique fix Fix adapter structure to handle multiple queues and prepping the driver for full multiple queue support, some changes are ifdef'd our unless you define CONFIG_E1000_MQ. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:39 +00:00
adapter->num_tx_queues = 1;
adapter->num_rx_queues = 1;
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
if (e1000_alloc_queues(adapter)) {
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(probe, "Unable to allocate memory for queues\n");
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
return -ENOMEM;
}
e1000: restore netif_poll_enable call but make sure IRQs are off This restores the previously removed netif_poll_enable call in e1000_open. It's needed on all but the first call to e1000_open for a NIC as e1000_close always calls netif_poll_disable. netif_poll_enable can only be called safely if no polls have been scheduled. This should be the case as long as we don't enter our IRQ handler. In order to guarantee this we explicitly disable IRQs as early as possible when we're probing the NIC. Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Cc: "Kok, Auke" <auke-jan.h.kok@intel.com> Cc: Jeff Garzik <jeff@garzik.org> Cc: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-05-29 22:07:31 +00:00
/* Explicitly disable IRQ since the NIC can be in any state. */
e1000_irq_disable(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
spin_lock_init(&adapter->stats_lock);
e1000: convert to private mutex from rtnl The e1000 driver when running with lockdep could run into some possible deadlocks between the work items acquiring rtnl and the rtnl lock being acquired before work items were cancelled. Use a private mutex to make sure lock ordering isn't violated. The private mutex is only used to protect areas not generally covered by the rtnl lock already. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:51 +00:00
mutex_init(&adapter->mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: driver state fixes (race fix) We were plagued by our interrupt handler posting a watchdog event which could occur when our adapter was going down in case a late packet arrived just before e1000_down() finished. This caused the watchdog timer to start after the NIC was down and keep rescheduling it every N seconds. Once the driver unloaded it would panic. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:54:02 +00:00
set_bit(__E1000_DOWN, &adapter->flags);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return 0;
}
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
/**
* e1000_alloc_queues - Allocate memory for all rings
* @adapter: board private structure to initialize
*
* We allocate one ring per queue at run-time since we don't know the
e1000: e1000_adapter->polling_netdev is useless Commit bea3348eef27e6044b6161fd04c3152215f96411 "[NET]: Make NAPI polling independent of struct net_device objects." made NAPI polling to be independent of net_device. So e1000_adapter->polling_netdev is no longer used. Kill it. Signed-off-by: Wang Chen <wangchen@cn.fujitsu.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-12-04 06:07:10 +00:00
* number of queues at compile-time.
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
**/
net/intel: remove __dev* attributes CONFIG_HOTPLUG is going away as an option. As result the __dev* markings will be going away. Remove use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit. Signed-off-by: Bill Pemberton <wfp5p@virginia.edu> Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Bruce Allan <bruce.w.allan@intel.com> Cc: Carolyn Wyborny <carolyn.wyborny@intel.com> Cc: Don Skidmore <donald.c.skidmore@intel.com> Cc: Greg Rose <gregory.v.rose@intel.com> Cc: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com> Cc: Alex Duyck <alexander.h.duyck@intel.com> Cc: John Ronciak <john.ronciak@intel.com> Cc: Tushar Dave <tushar.n.dave@intel.com> Cc: e1000-devel@lists.sourceforge.net Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-03 14:24:25 +00:00
static int e1000_alloc_queues(struct e1000_adapter *adapter)
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
{
e1000: Use kcalloc() Replace kmalloc+memsetout the driver. Slightly modified by Auke Kok. Signed-off-by: Yan Burman <burman.yan@gmail.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-03-06 16:58:04 +00:00
adapter->tx_ring = kcalloc(adapter->num_tx_queues,
sizeof(struct e1000_tx_ring), GFP_KERNEL);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
if (!adapter->tx_ring)
return -ENOMEM;
e1000: Use kcalloc() Replace kmalloc+memsetout the driver. Slightly modified by Auke Kok. Signed-off-by: Yan Burman <burman.yan@gmail.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-03-06 16:58:04 +00:00
adapter->rx_ring = kcalloc(adapter->num_rx_queues,
sizeof(struct e1000_rx_ring), GFP_KERNEL);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
if (!adapter->rx_ring) {
kfree(adapter->tx_ring);
return -ENOMEM;
}
return E1000_SUCCESS;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/**
* e1000_open - Called when a network interface is made active
* @netdev: network interface device structure
*
* Returns 0 on success, negative value on failure
*
* The open entry point is called when a network interface is made
* active by the system (IFF_UP). At this point all resources needed
* for transmit and receive operations are allocated, the interrupt
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
* handler is registered with the OS, the watchdog task is started,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* and the stack is notified that the interface is ready.
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static int e1000_open(struct net_device *netdev)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
[PATCH] e1000: Use netdev_priv() to get to netdev->priv Use netdev_priv() to get to netdev->priv - from shemminger@osdl.org Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:41:45 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
int err;
e1000: rework driver hardware reset locking After studying the driver mac reset code it was found that there were multiple race conditions possible to reset the unit twice or bring it e1000_up() double. This fixes all occurences where the driver needs to reset the mac. We also remove irq requesting/releasing into _open and _close so that while the device is _up we will never touch the irq's. This fixes the double free irq bug that people saw. To make sure that the watchdog task doesn't cause another race we let it run as a non-scheduled task. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:28 +00:00
/* disallow open during test */
e1000: driver state fixes (race fix) We were plagued by our interrupt handler posting a watchdog event which could occur when our adapter was going down in case a late packet arrived just before e1000_down() finished. This caused the watchdog timer to start after the NIC was down and keep rescheduling it every N seconds. Once the driver unloaded it would panic. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:54:02 +00:00
if (test_bit(__E1000_TESTING, &adapter->flags))
e1000: rework driver hardware reset locking After studying the driver mac reset code it was found that there were multiple race conditions possible to reset the unit twice or bring it e1000_up() double. This fixes all occurences where the driver needs to reset the mac. We also remove irq requesting/releasing into _open and _close so that while the device is _up we will never touch the irq's. This fixes the double free irq bug that people saw. To make sure that the watchdog task doesn't cause another race we let it run as a non-scheduled task. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:28 +00:00
return -EBUSY;
e1000: init link state correctly As reported by Andrew Lutomirski <amluto@gmail.com> All the intel wired ethernet drivers were calling netif_carrier_off and netif_stop_queue (or variants) before calling register_netdevice This is incorrect behavior as was pointed out by davem, and causes ifconfig and friends to report a strange state before first link after the driver was loaded. This apparently confused *some* versions of networkmanager. Andy tested this for e1000e and confirmed it was working for him. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Reported-by: Andrew Lutomirski <amluto@gmail.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-17 20:44:36 +00:00
netif_carrier_off(netdev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* allocate transmit descriptors */
e1000: FIX: be ready for incoming irq at pci_request_irq DEBUG_SHIRQ code exposed that e1000 was not ready for incoming interrupts after having called pci_request_irq. This obviously requires us to finish our software setup which assigns the irq handler before we request the irq. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-03-06 16:57:21 +00:00
err = e1000_setup_all_tx_resources(adapter);
if (err)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto err_setup_tx;
/* allocate receive descriptors */
e1000: FIX: be ready for incoming irq at pci_request_irq DEBUG_SHIRQ code exposed that e1000 was not ready for incoming interrupts after having called pci_request_irq. This obviously requires us to finish our software setup which assigns the irq handler before we request the irq. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-03-06 16:57:21 +00:00
err = e1000_setup_all_rx_resources(adapter);
Revert "e1000: fix shared interrupt warning message" This reverts commit d2ed16356ff4fb9de23fbc5e5d582ce580390106. As Thomas Gleixner reports: "e1000 is not working anymore. ifup fails permanentely. ADDRCONF(NETDEV_UP): eth0: link is not ready nothing else" The broken commit was identified with "git bisect". Auke Kok says: "I think we need to drop this now. The report that says that this *fixes* something might have been on regular interrupts only. I currently suspect that it breaks all MSI interrupts, which would make sense if I look a the code. Very bad indeed." Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Acked-by: Auke Kok <auke-jan.h.kok@intel.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Jeff Garzik <jeff@garzik.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-21 19:21:44 +00:00
if (err)
e1000: FIX: be ready for incoming irq at pci_request_irq DEBUG_SHIRQ code exposed that e1000 was not ready for incoming interrupts after having called pci_request_irq. This obviously requires us to finish our software setup which assigns the irq handler before we request the irq. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-03-06 16:57:21 +00:00
goto err_setup_rx;
Revert "e1000: fix shared interrupt warning message" This reverts commit d2ed16356ff4fb9de23fbc5e5d582ce580390106. As Thomas Gleixner reports: "e1000 is not working anymore. ifup fails permanentely. ADDRCONF(NETDEV_UP): eth0: link is not ready nothing else" The broken commit was identified with "git bisect". Auke Kok says: "I think we need to drop this now. The report that says that this *fixes* something might have been on regular interrupts only. I currently suspect that it breaks all MSI interrupts, which would make sense if I look a the code. Very bad indeed." Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Acked-by: Auke Kok <auke-jan.h.kok@intel.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Jeff Garzik <jeff@garzik.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-21 19:21:44 +00:00
e1000: Make PHY powerup/down a function In relation to the irq work done earlier we also move the PHY powerup and powerdown functions into separate functions and move the calls to _close and _open, making the PHY stay in it's power state as long as the device is _up. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:32 +00:00
e1000_power_up_phy(adapter);
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
adapter->mng_vlan_id = E1000_MNG_VLAN_NONE;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if ((hw->mng_cookie.status &
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
E1000_MNG_DHCP_COOKIE_STATUS_VLAN_SUPPORT)) {
e1000_update_mng_vlan(adapter);
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: FIX: be ready for incoming irq at pci_request_irq DEBUG_SHIRQ code exposed that e1000 was not ready for incoming interrupts after having called pci_request_irq. This obviously requires us to finish our software setup which assigns the irq handler before we request the irq. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-03-06 16:57:21 +00:00
/* before we allocate an interrupt, we must be ready to handle it.
* Setting DEBUG_SHIRQ in the kernel makes it fire an interrupt
* as soon as we call pci_request_irq, so we have to setup our
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* clean_rx handler before we do so.
*/
e1000: FIX: be ready for incoming irq at pci_request_irq DEBUG_SHIRQ code exposed that e1000 was not ready for incoming interrupts after having called pci_request_irq. This obviously requires us to finish our software setup which assigns the irq handler before we request the irq. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-03-06 16:57:21 +00:00
e1000_configure(adapter);
err = e1000_request_irq(adapter);
if (err)
goto err_req_irq;
/* From here on the code is the same as e1000_up() */
clear_bit(__E1000_DOWN, &adapter->flags);
[NET]: Make NAPI polling independent of struct net_device objects. Several devices have multiple independant RX queues per net device, and some have a single interrupt doorbell for several queues. In either case, it's easier to support layouts like that if the structure representing the poll is independant from the net device itself. The signature of the ->poll() call back goes from: int foo_poll(struct net_device *dev, int *budget) to int foo_poll(struct napi_struct *napi, int budget) The caller is returned the number of RX packets processed (or the number of "NAPI credits" consumed if you want to get abstract). The callee no longer messes around bumping dev->quota, *budget, etc. because that is all handled in the caller upon return. The napi_struct is to be embedded in the device driver private data structures. Furthermore, it is the driver's responsibility to disable all NAPI instances in it's ->stop() device close handler. Since the napi_struct is privatized into the driver's private data structures, only the driver knows how to get at all of the napi_struct instances it may have per-device. With lots of help and suggestions from Rusty Russell, Roland Dreier, Michael Chan, Jeff Garzik, and Jamal Hadi Salim. Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra, Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan. [ Ported to current tree and all drivers converted. Integrated Stephen's follow-on kerneldoc additions, and restored poll_list handling to the old style to fix mutual exclusion issues. -DaveM ] Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-03 23:41:36 +00:00
napi_enable(&adapter->napi);
e1000: restore netif_poll_enable call but make sure IRQs are off This restores the previously removed netif_poll_enable call in e1000_open. It's needed on all but the first call to e1000_open for a NIC as e1000_close always calls netif_poll_disable. netif_poll_enable can only be called safely if no polls have been scheduled. This should be the case as long as we don't enter our IRQ handler. In order to guarantee this we explicitly disable IRQs as early as possible when we're probing the NIC. Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Cc: "Kok, Auke" <auke-jan.h.kok@intel.com> Cc: Jeff Garzik <jeff@garzik.org> Cc: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-05-29 22:07:31 +00:00
e1000: FIX: be ready for incoming irq at pci_request_irq DEBUG_SHIRQ code exposed that e1000 was not ready for incoming interrupts after having called pci_request_irq. This obviously requires us to finish our software setup which assigns the irq handler before we request the irq. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-03-06 16:57:21 +00:00
e1000_irq_enable(adapter);
e1000: resolve tx multiqueue bug With the recent changes to tx mutiqueue, e1000 was not calling netif_start_queue() before calling netif_wake_queue(). This causes an oops during loading of the driver. (Based on commit d55b53fff0c2ddb639dca04c3f5a0854f292d982 ("igb/ixgbe/e1000e: resolve tx multiqueue bug").) Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-19 00:50:57 +00:00
netif_start_queue(netdev);
e1000: FIX: be ready for incoming irq at pci_request_irq DEBUG_SHIRQ code exposed that e1000 was not ready for incoming interrupts after having called pci_request_irq. This obviously requires us to finish our software setup which assigns the irq handler before we request the irq. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-03-06 16:57:21 +00:00
/* fire a link status change interrupt to start the watchdog */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(ICS, E1000_ICS_LSC);
e1000: FIX: be ready for incoming irq at pci_request_irq DEBUG_SHIRQ code exposed that e1000 was not ready for incoming interrupts after having called pci_request_irq. This obviously requires us to finish our software setup which assigns the irq handler before we request the irq. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-03-06 16:57:21 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return E1000_SUCCESS;
Revert "e1000: fix shared interrupt warning message" This reverts commit d2ed16356ff4fb9de23fbc5e5d582ce580390106. As Thomas Gleixner reports: "e1000 is not working anymore. ifup fails permanentely. ADDRCONF(NETDEV_UP): eth0: link is not ready nothing else" The broken commit was identified with "git bisect". Auke Kok says: "I think we need to drop this now. The report that says that this *fixes* something might have been on regular interrupts only. I currently suspect that it breaks all MSI interrupts, which would make sense if I look a the code. Very bad indeed." Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Acked-by: Auke Kok <auke-jan.h.kok@intel.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Jeff Garzik <jeff@garzik.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-21 19:21:44 +00:00
err_req_irq:
e1000: FIX: be ready for incoming irq at pci_request_irq DEBUG_SHIRQ code exposed that e1000 was not ready for incoming interrupts after having called pci_request_irq. This obviously requires us to finish our software setup which assigns the irq handler before we request the irq. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-03-06 16:57:21 +00:00
e1000_power_down_phy(adapter);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
e1000_free_all_rx_resources(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
err_setup_rx:
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
e1000_free_all_tx_resources(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
err_setup_tx:
e1000_reset(adapter);
return err;
}
/**
* e1000_close - Disables a network interface
* @netdev: network interface device structure
*
* Returns 0, this is not allowed to fail
*
* The close entry point is called when an interface is de-activated
* by the OS. The hardware is still under the drivers control, but
* needs to be disabled. A global MAC reset is issued to stop the
* hardware, and all transmit and receive resources are freed.
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static int e1000_close(struct net_device *netdev)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
[PATCH] e1000: Use netdev_priv() to get to netdev->priv Use netdev_priv() to get to netdev->priv - from shemminger@osdl.org Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:41:45 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: prevent oops when adapter is being closed and reset simultaneously This change is based on a similar change made to e1000e support in commit bb9e44d0d0f4 ("e1000e: prevent oops when adapter is being closed and reset simultaneously"). The same issue has also been observed on the older e1000 cards. Here, we have increased the RESET_COUNT value to 50 because there are too many accesses to e1000 nic on stress tests to e1000 nic, it is not enough to set RESET_COUT 25. Experimentation has shown that it is enough to set RESET_COUNT 50. Signed-off-by: yzhu1 <yanjun.zhu@windriver.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2013-11-23 07:07:40 +00:00
int count = E1000_CHECK_RESET_COUNT;
while (test_bit(__E1000_RESETTING, &adapter->flags) && count--)
usleep_range(10000, 20000);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: rework driver hardware reset locking After studying the driver mac reset code it was found that there were multiple race conditions possible to reset the unit twice or bring it e1000_up() double. This fixes all occurences where the driver needs to reset the mac. We also remove irq requesting/releasing into _open and _close so that while the device is _up we will never touch the irq's. This fixes the double free irq bug that people saw. To make sure that the watchdog task doesn't cause another race we let it run as a non-scheduled task. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:28 +00:00
WARN_ON(test_bit(__E1000_RESETTING, &adapter->flags));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_down(adapter);
e1000: Make PHY powerup/down a function In relation to the irq work done earlier we also move the PHY powerup and powerdown functions into separate functions and move the calls to _close and _open, making the PHY stay in it's power state as long as the device is _up. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:32 +00:00
e1000_power_down_phy(adapter);
e1000: rework driver hardware reset locking After studying the driver mac reset code it was found that there were multiple race conditions possible to reset the unit twice or bring it e1000_up() double. This fixes all occurences where the driver needs to reset the mac. We also remove irq requesting/releasing into _open and _close so that while the device is _up we will never touch the irq's. This fixes the double free irq bug that people saw. To make sure that the watchdog task doesn't cause another race we let it run as a non-scheduled task. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:28 +00:00
e1000_free_irq(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
e1000_free_all_tx_resources(adapter);
e1000_free_all_rx_resources(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: don't strip vlan ID if 8021q claims it Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:54:08 +00:00
/* kill manageability vlan ID if supported, but not if a vlan with
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* the same ID is registered on the host OS (let 8021q kill it)
*/
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if ((hw->mng_cookie.status &
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
E1000_MNG_DHCP_COOKIE_STATUS_VLAN_SUPPORT) &&
!test_bit(adapter->mng_vlan_id, adapter->active_vlans)) {
net: vlan: prepare for 802.1ad VLAN filtering offload Change the rx_{add,kill}_vid callbacks to take a protocol argument in preparation of 802.1ad support. The protocol argument used so far is always htons(ETH_P_8021Q). Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-04-19 02:04:28 +00:00
e1000_vlan_rx_kill_vid(netdev, htons(ETH_P_8021Q),
adapter->mng_vlan_id);
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
}
[PATCH] e1000: Fix ASF/AMT for 8257{1|2|3} controllers The 82573 controller required different logic than 82571|2 controllers. Corrected the reset logic for 8257{1|2|3} controllers. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:30 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return 0;
}
/**
* e1000_check_64k_bound - check that memory doesn't cross 64kB boundary
* @adapter: address of board private structure
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
* @start: address of beginning of memory
* @len: length of memory
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static bool e1000_check_64k_bound(struct e1000_adapter *adapter, void *start,
unsigned long len)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: Remove spaces after casts and function names Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:18 +00:00
unsigned long begin = (unsigned long)start;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
unsigned long end = begin + len;
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
/* First rev 82545 and 82546 need to not allow any memory
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* write location to cross 64k boundary due to errata 23
*/
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (hw->mac_type == e1000_82545 ||
e1000: Add support for the CE4100 reference platform This patch adds support for the gigabit phys present on the CE4100 reference platforms. Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-01-06 14:29:54 +00:00
hw->mac_type == e1000_ce4100 ||
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
hw->mac_type == e1000_82546) {
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
return ((begin ^ (end - 1)) >> 16) != 0 ? false : true;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
return true;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/**
* e1000_setup_tx_resources - allocate Tx resources (Descriptors)
* @adapter: board private structure
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
* @txdr: tx descriptor ring (for a specific queue) to setup
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*
* Return 0 on success, negative on failure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static int e1000_setup_tx_resources(struct e1000_adapter *adapter,
struct e1000_tx_ring *txdr)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct pci_dev *pdev = adapter->pdev;
int size;
size = sizeof(struct e1000_buffer) * txdr->count;
drivers/net: use vzalloc() Use vzalloc() and vzalloc_node() in net drivers Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Acked-by: Jon Mason <jon.mason@exar.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-11-22 00:15:06 +00:00
txdr->buffer_info = vzalloc(size);
drivers: net: Remove remaining alloc/OOM messages alloc failures already get standardized OOM messages and a dump_stack. For the affected mallocs around these OOM messages: Converted kmallocs with multiplies to kmalloc_array. Converted a kmalloc/memcpy to kmemdup. Removed now unused stack variables. Removed unnecessary parentheses. Neatened alignment. Signed-off-by: Joe Perches <joe@perches.com> Acked-by: Arend van Spriel <arend@broadcom.com> Acked-by: Marc Kleine-Budde <mkl@pengutronix.de> Acked-by: John W. Linville <linville@tuxdriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-02-07 11:46:27 +00:00
if (!txdr->buffer_info)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return -ENOMEM;
/* round up to nearest 4K */
txdr->size = txdr->count * sizeof(struct e1000_tx_desc);
e1000: ROUND_UP macro cleanup in drivers/net/e1000 E1000_ROUNDUP macro cleanup, use ALIGN Signed-off-by: Milind Arun Choudhary <milindchoudhary@gmail.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-04-27 20:55:29 +00:00
txdr->size = ALIGN(txdr->size, 4096);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
txdr->desc = dma_alloc_coherent(&pdev->dev, txdr->size, &txdr->dma,
GFP_KERNEL);
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (!txdr->desc) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
setup_tx_desc_die:
vfree(txdr->buffer_info);
return -ENOMEM;
}
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
/* Fix for errata 23, can't cross 64kB boundary */
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (!e1000_check_64k_bound(adapter, txdr->desc, txdr->size)) {
void *olddesc = txdr->desc;
dma_addr_t olddma = txdr->dma;
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(tx_err, "txdr align check failed: %u bytes at %p\n",
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
txdr->size, txdr->desc);
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
/* Try again, without freeing the previous */
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
txdr->desc = dma_alloc_coherent(&pdev->dev, txdr->size,
&txdr->dma, GFP_KERNEL);
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
/* Failed allocation, critical failure */
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (!txdr->desc) {
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
dma_free_coherent(&pdev->dev, txdr->size, olddesc,
olddma);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto setup_tx_desc_die;
}
if (!e1000_check_64k_bound(adapter, txdr->desc, txdr->size)) {
/* give up */
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
dma_free_coherent(&pdev->dev, txdr->size, txdr->desc,
txdr->dma);
dma_free_coherent(&pdev->dev, txdr->size, olddesc,
olddma);
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(probe, "Unable to allocate aligned memory "
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
"for the transmit descriptor ring\n");
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
vfree(txdr->buffer_info);
return -ENOMEM;
} else {
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
/* Free old allocation, new allocation was successful */
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
dma_free_coherent(&pdev->dev, txdr->size, olddesc,
olddma);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
}
memset(txdr->desc, 0, txdr->size);
txdr->next_to_use = 0;
txdr->next_to_clean = 0;
return 0;
}
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
/**
* e1000_setup_all_tx_resources - wrapper to allocate Tx resources
* (Descriptors) for all queues
* @adapter: board private structure
*
* Return 0 on success, negative on failure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
int e1000_setup_all_tx_resources(struct e1000_adapter *adapter)
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
{
int i, err = 0;
[PATCH] e1000: Fix adapter structure and prepare for multique fix Fix adapter structure to handle multiple queues and prepping the driver for full multiple queue support, some changes are ifdef'd our unless you define CONFIG_E1000_MQ. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:39 +00:00
for (i = 0; i < adapter->num_tx_queues; i++) {
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
err = e1000_setup_tx_resources(adapter, &adapter->tx_ring[i]);
if (err) {
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(probe, "Allocation for Tx Queue %u failed\n", i);
e1000: ring buffers resources cleanup Memory leak was found in 2.6.18-rc4 and e1000 7.2.7 from sourceforge: We should free resources allocated for previous rings if following allocation fails. Signed-off-by: Vasily Averin <vvs@sw.ru> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-28 21:56:24 +00:00
for (i-- ; i >= 0; i--)
e1000_free_tx_resources(adapter,
&adapter->tx_ring[i]);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
break;
}
}
return err;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/**
* e1000_configure_tx - Configure 8254x Transmit Unit after Reset
* @adapter: board private structure
*
* Configure the Tx unit of the MAC after a reset.
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_configure_tx(struct e1000_adapter *adapter)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u64 tdba;
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
u32 tdlen, tctl, tipg;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u32 ipgr1, ipgr2;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Setup the HW Tx Head and Tail descriptor pointers */
[PATCH] e1000: Fix adapter structure and prepare for multique fix Fix adapter structure to handle multiple queues and prepping the driver for full multiple queue support, some changes are ifdef'd our unless you define CONFIG_E1000_MQ. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:39 +00:00
switch (adapter->num_tx_queues) {
e1000: implementation of the multi-queue feature Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:03:23 +00:00
case 1:
default:
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
tdba = adapter->tx_ring[0].dma;
tdlen = adapter->tx_ring[0].count *
sizeof(struct e1000_tx_desc);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(TDLEN, tdlen);
ew32(TDBAH, (tdba >> 32));
ew32(TDBAL, (tdba & 0x00000000ffffffffULL));
ew32(TDT, 0);
ew32(TDH, 0);
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
adapter->tx_ring[0].tdh = ((hw->mac_type >= e1000_82543) ?
E1000_TDH : E1000_82542_TDH);
adapter->tx_ring[0].tdt = ((hw->mac_type >= e1000_82543) ?
E1000_TDT : E1000_82542_TDT);
e1000: implementation of the multi-queue feature Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:03:23 +00:00
break;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Set the default values for the Tx Inter Packet Gap timer */
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
if ((hw->media_type == e1000_media_type_fiber ||
[PATCH] e1000: narrow down the scope of the tipg timer tweak the driver has (ancient) code for messing with TIPG from the 82542 days. Unfortunately this code was running on our current adapters and setting TIPG for fiber to be +1 over the copper value. This caused 1.45Mpps to be sent instead of 1.487Mpps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:38:32 +00:00
hw->media_type == e1000_media_type_internal_serdes))
[PATCH] e1000: Fix collision distance Fixed the collision distance for 82543 controllers and newer. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:05 +00:00
tipg = DEFAULT_82543_TIPG_IPGT_FIBER;
else
tipg = DEFAULT_82543_TIPG_IPGT_COPPER;
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
switch (hw->mac_type) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
case e1000_82542_rev2_0:
case e1000_82542_rev2_1:
tipg = DEFAULT_82542_TIPG_IPGT;
[PATCH] e1000: Fix collision distance Fixed the collision distance for 82543 controllers and newer. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:05 +00:00
ipgr1 = DEFAULT_82542_TIPG_IPGR1;
ipgr2 = DEFAULT_82542_TIPG_IPGR2;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
default:
[PATCH] e1000: Fix collision distance Fixed the collision distance for 82543 controllers and newer. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:05 +00:00
ipgr1 = DEFAULT_82543_TIPG_IPGR1;
ipgr2 = DEFAULT_82543_TIPG_IPGR2;
break;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
[PATCH] e1000: Fix collision distance Fixed the collision distance for 82543 controllers and newer. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:05 +00:00
tipg |= ipgr1 << E1000_TIPG_IPGR1_SHIFT;
tipg |= ipgr2 << E1000_TIPG_IPGR2_SHIFT;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(TIPG, tipg);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Set the Tx Interrupt Delay register */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(TIDV, adapter->tx_int_delay);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
if (hw->mac_type >= e1000_82540)
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(TADV, adapter->tx_abs_int_delay);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Program the Transmit Control Register */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
tctl = er32(TCTL);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
tctl &= ~E1000_TCTL_CT;
e1000: Fix network problems when forced at 100Mb/s and to fix TSO when forced at 100Mb/s Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:19:30 +00:00
tctl |= E1000_TCTL_PSP | E1000_TCTL_RTLC |
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
(E1000_COLLISION_THRESHOLD << E1000_CT_SHIFT);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
e1000_config_collision_dist(hw);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Setup Transmit Descriptor Settings for eop descriptor */
e1000: Only set IDE for tx when we are using TIDV/TADV Spec fix: don't set IDE unless we are actually setting the tx int delay time. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2006-11-01 16:48:04 +00:00
adapter->txd_cmd = E1000_TXD_CMD_EOP | E1000_TXD_CMD_IFCS;
/* only set IDE if we are delaying interrupts using the timers */
if (adapter->tx_int_delay)
adapter->txd_cmd |= E1000_TXD_CMD_IDE;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
if (hw->mac_type < e1000_82543)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
adapter->txd_cmd |= E1000_TXD_CMD_RPS;
else
adapter->txd_cmd |= E1000_TXD_CMD_RS;
/* Cache if we're 82544 running in PCI-X because we'll
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* need this to apply a workaround later in the send path.
*/
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
if (hw->mac_type == e1000_82544 &&
hw->bus_type == e1000_bus_type_pcix)
net: fix assignment of 0/1 to bool variables. DaveM said: Please, this kind of stuff rots forever and not using bool properly drives me crazy. Joe Perches <joe@perches.com> gave me the spatch script: @@ bool b; @@ -b = 0 +b = false @@ bool b; @@ -b = 1 +b = true I merely installed coccinelle, read the documentation and took credit. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-12-19 13:56:45 +00:00
adapter->pcix_82544 = true;
e1000: Fix network problems when forced at 100Mb/s and to fix TSO when forced at 100Mb/s Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:19:30 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(TCTL, tctl);
e1000: Fix network problems when forced at 100Mb/s and to fix TSO when forced at 100Mb/s Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:19:30 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/**
* e1000_setup_rx_resources - allocate Rx resources (Descriptors)
* @adapter: board private structure
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
* @rxdr: rx descriptor ring (for a specific queue) to setup
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*
* Returns 0 on success, negative on failure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static int e1000_setup_rx_resources(struct e1000_adapter *adapter,
struct e1000_rx_ring *rxdr)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct pci_dev *pdev = adapter->pdev;
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
int size, desc_len;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
size = sizeof(struct e1000_buffer) * rxdr->count;
drivers/net: use vzalloc() Use vzalloc() and vzalloc_node() in net drivers Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Acked-by: Jon Mason <jon.mason@exar.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-11-22 00:15:06 +00:00
rxdr->buffer_info = vzalloc(size);
drivers: net: Remove remaining alloc/OOM messages alloc failures already get standardized OOM messages and a dump_stack. For the affected mallocs around these OOM messages: Converted kmallocs with multiplies to kmalloc_array. Converted a kmalloc/memcpy to kmemdup. Removed now unused stack variables. Removed unnecessary parentheses. Neatened alignment. Signed-off-by: Joe Perches <joe@perches.com> Acked-by: Arend van Spriel <arend@broadcom.com> Acked-by: Marc Kleine-Budde <mkl@pengutronix.de> Acked-by: John W. Linville <linville@tuxdriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-02-07 11:46:27 +00:00
if (!rxdr->buffer_info)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return -ENOMEM;
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
desc_len = sizeof(struct e1000_rx_desc);
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Round up to nearest 4K */
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
rxdr->size = rxdr->count * desc_len;
e1000: ROUND_UP macro cleanup in drivers/net/e1000 E1000_ROUNDUP macro cleanup, use ALIGN Signed-off-by: Milind Arun Choudhary <milindchoudhary@gmail.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-04-27 20:55:29 +00:00
rxdr->size = ALIGN(rxdr->size, 4096);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
rxdr->desc = dma_alloc_coherent(&pdev->dev, rxdr->size, &rxdr->dma,
GFP_KERNEL);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
if (!rxdr->desc) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
setup_rx_desc_die:
vfree(rxdr->buffer_info);
return -ENOMEM;
}
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
/* Fix for errata 23, can't cross 64kB boundary */
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (!e1000_check_64k_bound(adapter, rxdr->desc, rxdr->size)) {
void *olddesc = rxdr->desc;
dma_addr_t olddma = rxdr->dma;
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(rx_err, "rxdr align check failed: %u bytes at %p\n",
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
rxdr->size, rxdr->desc);
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
/* Try again, without freeing the previous */
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
rxdr->desc = dma_alloc_coherent(&pdev->dev, rxdr->size,
&rxdr->dma, GFP_KERNEL);
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
/* Failed allocation, critical failure */
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
if (!rxdr->desc) {
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
dma_free_coherent(&pdev->dev, rxdr->size, olddesc,
olddma);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto setup_rx_desc_die;
}
if (!e1000_check_64k_bound(adapter, rxdr->desc, rxdr->size)) {
/* give up */
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
dma_free_coherent(&pdev->dev, rxdr->size, rxdr->desc,
rxdr->dma);
dma_free_coherent(&pdev->dev, rxdr->size, olddesc,
olddma);
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(probe, "Unable to allocate aligned memory for "
"the Rx descriptor ring\n");
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
goto setup_rx_desc_die;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
} else {
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
/* Free old allocation, new allocation was successful */
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
dma_free_coherent(&pdev->dev, rxdr->size, olddesc,
olddma);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
}
memset(rxdr->desc, 0, rxdr->size);
rxdr->next_to_clean = 0;
rxdr->next_to_use = 0;
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
rxdr->rx_skb_top = NULL;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return 0;
}
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
/**
* e1000_setup_all_rx_resources - wrapper to allocate Rx resources
* (Descriptors) for all queues
* @adapter: board private structure
*
* Return 0 on success, negative on failure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
int e1000_setup_all_rx_resources(struct e1000_adapter *adapter)
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
{
int i, err = 0;
[PATCH] e1000: Fix adapter structure and prepare for multique fix Fix adapter structure to handle multiple queues and prepping the driver for full multiple queue support, some changes are ifdef'd our unless you define CONFIG_E1000_MQ. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:39 +00:00
for (i = 0; i < adapter->num_rx_queues; i++) {
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
err = e1000_setup_rx_resources(adapter, &adapter->rx_ring[i]);
if (err) {
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(probe, "Allocation for Rx Queue %u failed\n", i);
e1000: ring buffers resources cleanup Memory leak was found in 2.6.18-rc4 and e1000 7.2.7 from sourceforge: We should free resources allocated for previous rings if following allocation fails. Signed-off-by: Vasily Averin <vvs@sw.ru> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-28 21:56:24 +00:00
for (i-- ; i >= 0; i--)
e1000_free_rx_resources(adapter,
&adapter->rx_ring[i]);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
break;
}
}
return err;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/**
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
* e1000_setup_rctl - configure the receive control registers
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* @adapter: Board private structure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_setup_rctl(struct e1000_adapter *adapter)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: remove unused Kconfig option for disabling packet split Since the e1000/e1000e split, no hardware supported by e1000 supports packet split, just remove the Kconfig option and associated code from the driver. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-09-16 20:01:28 +00:00
u32 rctl;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
rctl = er32(RCTL);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
rctl &= ~(3 << E1000_RCTL_MO_SHIFT);
e1000: don't enable dma receives until after dma address has been setup Doing an 'ifconfig ethN down' followed by an 'ifconfig ethN up' on a qemu-kvm guest system configured with two e1000 NICs can result in an 'unable to handle kernel paging request at 0000000100000000' or 'bad page map in process ...' or something similar. These result from a 4096-byte page being corrupted with the following two-word pattern (16-bytes) repeated throughout the entire page: 0x0000000000000000 0x0000000100000000 There can be other bits set as well. What is a constant is that the 2nd word has the 32nd bit set. So one could see: : 0x0000000000000000 0x0000000100000000 0x0000000000000000 0x0000000172adc067 <<< bad pte 0x800000006ec60067 0x0000000700000040 0x0000000000000000 0x0000000100000000 : Which came from from a process' page table I dumped out when the marked line was seen as bad by print_bad_pte(). The repeating pattern represents the e1000's two-word receive descriptor: struct e1000_rx_desc { __le64 buffer_addr; /* Address of the descriptor's data buffer */ __le16 length; /* Length of data DMAed into data buffer */ __le16 csum; /* Packet checksum */ u8 status; /* Descriptor status */ u8 errors; /* Descriptor Errors */ __le16 special; }; And the 32nd bit of the 2nd word maps to the 'u8 status' member, and corresponds to E1000_RXD_STAT_DD which indicates the descriptor is done. The corruption appears to result from the following... . An 'ifconfig ethN down' gets us into e1000_close(), which through a number of subfunctions results in: 1. E1000_RCTL_EN being cleared in RCTL register. [e1000_down()] 2. dma_free_coherent() being called. [e1000_free_rx_resources()] . An 'ifconfig ethN up' gets us into e1000_open(), which through a number of subfunctions results in: 1. dma_alloc_coherent() being called. [e1000_setup_rx_resources()] 2. E1000_RCTL_EN being set in RCTL register. [e1000_setup_rctl()] 3. E1000_RCTL_EN being cleared in RCTL register. [e1000_configure_rx()] 4. RDLEN, RDBAH and RDBAL registers being set to reflect the dma page allocated in step 1. [e1000_configure_rx()] 5. E1000_RCTL_EN being set in RCTL register. [e1000_configure_rx()] During the 'ifconfig ethN up' there is a window opened, starting in step 2 where the receives are enabled up until they are disabled in step 3, in which the address of the receive descriptor dma page known by the NIC is still the previous one which was freed during the 'ifconfig ethN down'. If this memory has been reallocated for some other use and the NIC feels so inclined, it will write to that former dma page with predictably unpleasant results. I realize that in the guest, we're dealing with an e1000 NIC that is software emulated by qemu-kvm. The problem doesn't appear to occur on bare-metal. Andy suspects that this is because in the emulator link-up is essentially instant and traffic can start flowing immediately. Whereas on bare-metal, link-up usually seems to take at least a few milliseconds. And this might be enough to prevent traffic from flowing into the device inside the window where E1000_RCTL_EN is set. So perhaps a modification needs to be made to the qemu-kvm e1000 NIC emulator to delay the link-up. But in defense of the emulator, it seems like a bad idea to enable dma operations before the address of the memory to be involved has been made known. The following patch no longer enables receives in e1000_setup_rctl() but leaves them however they were. It only enables receives in e1000_configure_rx(), and only after the dma address has been made known to the hardware. There are two places where e1000_setup_rctl() gets called. The one in e1000_configure() is followed immediately by a call to e1000_configure_rx(), so there's really no change functionally (except for the removal of the problem window. The other is in __e1000_shutdown() and is not followed by a call to e1000_configure_rx(), so there is a change functionally. But consider... . An 'ifconfig ethN down' (just as described above). . A 'suspend' of the system, which (I'm assuming) will find its way into e1000_suspend() which calls __e1000_shutdown() resulting in: 1. E1000_RCTL_EN being set in RCTL register. [e1000_setup_rctl()] And again we've re-opened the problem window for some unknown amount of time. Signed-off-by: Andy Gospodarek <andy@greyhouse.net> Signed-off-by: Dean Nelson <dnelson@redhat.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-09-16 16:52:54 +00:00
rctl |= E1000_RCTL_BAM | E1000_RCTL_LBM_NO |
E1000_RCTL_RDMTS_HALF |
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
(hw->mc_filter_type << E1000_RCTL_MO_SHIFT);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (hw->tbi_compatibility_on == 1)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
rctl |= E1000_RCTL_SBP;
else
rctl &= ~E1000_RCTL_SBP;
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
if (adapter->netdev->mtu <= ETH_DATA_LEN)
rctl &= ~E1000_RCTL_LPE;
else
rctl |= E1000_RCTL_LPE;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Setup buffer sizes */
e1000: Buffer optimizations for small MTU Remove multi-descriptor support from legacy recieve path Add memory usage efficiency by using more correct size descriptors for small MTU sizes and optimize using LPE for <= 1522 byte frame sizes An extra performance fix that effected our TCP window size growth as a receiver. Set our initial buffer to be 128 bytes instead of 256 to prevent over-socket charge when truesize is computed in the stack. old way: truesize = 256 + l1 = 256 + 1460 = 1716 new way: truesize = 128 + l1 = 128 + 1460 = 1588 The magic value that we can't cross is 1648. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-04-15 02:05:18 +00:00
rctl &= ~E1000_RCTL_SZ_4096;
rctl |= E1000_RCTL_BSEX;
switch (adapter->rx_buffer_len) {
[PATCH] e1000: revert to single descriptor for legacy receive path A recent patch attempted to enable more efficient memory usage by using only 2kB descriptors for jumbo frames. The method used to implement this has since been commented upon as "illegal" and in recent kernels even causes a BUG when receiving ip fragments while using jumbo frames. This patch simply goes back to the way things were. We expect some complaints due to order 3 allocations failing to come back due to this change. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2006-03-01 04:24:07 +00:00
case E1000_RXBUFFER_2048:
default:
rctl |= E1000_RCTL_SZ_2048;
rctl &= ~E1000_RCTL_BSEX;
break;
case E1000_RXBUFFER_4096:
rctl |= E1000_RCTL_SZ_4096;
break;
case E1000_RXBUFFER_8192:
rctl |= E1000_RCTL_SZ_8192;
break;
case E1000_RXBUFFER_16384:
rctl |= E1000_RCTL_SZ_16384;
break;
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
}
e1000: Support RX-ALL flag. This allows the NIC to receive errored frames (bad FCS, etc) and pass them up the stack. This can be useful when using sniffers. Signed-off-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-04-04 06:01:29 +00:00
/* This is useful for sniffing bad packets. */
if (adapter->netdev->features & NETIF_F_RXALL) {
/* UPE and MPE will be handled by normal PROMISC logic
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* in e1000e_set_rx_mode
*/
e1000: Support RX-ALL flag. This allows the NIC to receive errored frames (bad FCS, etc) and pass them up the stack. This can be useful when using sniffers. Signed-off-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-04-04 06:01:29 +00:00
rctl |= (E1000_RCTL_SBP | /* Receive bad packets */
E1000_RCTL_BAM | /* RX All Bcast Pkts */
E1000_RCTL_PMCF); /* RX All MAC Ctrl Pkts */
rctl &= ~(E1000_RCTL_VFE | /* Disable VLAN filter */
E1000_RCTL_DPF | /* Allow filtered pause */
E1000_RCTL_CFIEN); /* Dis VLAN CFIEN Filter */
/* Do not mess with E1000_CTRL_VME, it affects transmit as well,
* and that breaks VLANs.
*/
}
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(RCTL, rctl);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/**
* e1000_configure_rx - Configure 8254x Receive Unit after Reset
* @adapter: board private structure
*
* Configure the Rx unit of the MAC after a reset.
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_configure_rx(struct e1000_adapter *adapter)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u64 rdba;
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
u32 rdlen, rctl, rxcsum;
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
if (adapter->netdev->mtu > ETH_DATA_LEN) {
rdlen = adapter->rx_ring[0].count *
sizeof(struct e1000_rx_desc);
adapter->clean_rx = e1000_clean_jumbo_rx_irq;
adapter->alloc_rx_buf = e1000_alloc_jumbo_rx_buffers;
} else {
rdlen = adapter->rx_ring[0].count *
sizeof(struct e1000_rx_desc);
adapter->clean_rx = e1000_clean_rx_irq;
adapter->alloc_rx_buf = e1000_alloc_rx_buffers;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* disable receives while setting up the descriptors */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
rctl = er32(RCTL);
ew32(RCTL, rctl & ~E1000_RCTL_EN);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* set the Receive Delay Timer Register */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(RDTR, adapter->rx_int_delay);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
if (hw->mac_type >= e1000_82540) {
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(RADV, adapter->rx_abs_int_delay);
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
if (adapter->itr_setting != 0)
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(ITR, 1000000000 / (adapter->itr * 256));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
/* Setup the HW Rx Head and Tail Descriptor Pointers and
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* the Base and Length of the Rx Descriptor Ring
*/
[PATCH] e1000: Fix adapter structure and prepare for multique fix Fix adapter structure to handle multiple queues and prepping the driver for full multiple queue support, some changes are ifdef'd our unless you define CONFIG_E1000_MQ. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:39 +00:00
switch (adapter->num_rx_queues) {
e1000: implementation of the multi-queue feature Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:03:23 +00:00
case 1:
default:
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
rdba = adapter->rx_ring[0].dma;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(RDLEN, rdlen);
ew32(RDBAH, (rdba >> 32));
ew32(RDBAL, (rdba & 0x00000000ffffffffULL));
ew32(RDT, 0);
ew32(RDH, 0);
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
adapter->rx_ring[0].rdh = ((hw->mac_type >= e1000_82543) ?
E1000_RDH : E1000_82542_RDH);
adapter->rx_ring[0].rdt = ((hw->mac_type >= e1000_82543) ?
E1000_RDT : E1000_82542_RDT);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
break;
e1000: implementation of the multi-queue feature Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:03:23 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Enable 82543 Receive Checksum Offload for TCP and UDP */
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
if (hw->mac_type >= e1000_82543) {
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
rxcsum = er32(RXCSUM);
e1000: remove unused Kconfig option for disabling packet split Since the e1000/e1000e split, no hardware supported by e1000 supports packet split, just remove the Kconfig option and associated code from the driver. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-09-16 20:01:28 +00:00
if (adapter->rx_csum)
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
rxcsum |= E1000_RXCSUM_TUOFL;
e1000: remove unused Kconfig option for disabling packet split Since the e1000/e1000e split, no hardware supported by e1000 supports packet split, just remove the Kconfig option and associated code from the driver. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-09-16 20:01:28 +00:00
else
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
/* don't need to clear IPPCSE as it defaults to 0 */
e1000: remove unused Kconfig option for disabling packet split Since the e1000/e1000e split, no hardware supported by e1000 supports packet split, just remove the Kconfig option and associated code from the driver. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-09-16 20:01:28 +00:00
rxcsum &= ~E1000_RXCSUM_TUOFL;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(RXCSUM, rxcsum);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/* Enable Receives */
e1000: don't enable dma receives until after dma address has been setup Doing an 'ifconfig ethN down' followed by an 'ifconfig ethN up' on a qemu-kvm guest system configured with two e1000 NICs can result in an 'unable to handle kernel paging request at 0000000100000000' or 'bad page map in process ...' or something similar. These result from a 4096-byte page being corrupted with the following two-word pattern (16-bytes) repeated throughout the entire page: 0x0000000000000000 0x0000000100000000 There can be other bits set as well. What is a constant is that the 2nd word has the 32nd bit set. So one could see: : 0x0000000000000000 0x0000000100000000 0x0000000000000000 0x0000000172adc067 <<< bad pte 0x800000006ec60067 0x0000000700000040 0x0000000000000000 0x0000000100000000 : Which came from from a process' page table I dumped out when the marked line was seen as bad by print_bad_pte(). The repeating pattern represents the e1000's two-word receive descriptor: struct e1000_rx_desc { __le64 buffer_addr; /* Address of the descriptor's data buffer */ __le16 length; /* Length of data DMAed into data buffer */ __le16 csum; /* Packet checksum */ u8 status; /* Descriptor status */ u8 errors; /* Descriptor Errors */ __le16 special; }; And the 32nd bit of the 2nd word maps to the 'u8 status' member, and corresponds to E1000_RXD_STAT_DD which indicates the descriptor is done. The corruption appears to result from the following... . An 'ifconfig ethN down' gets us into e1000_close(), which through a number of subfunctions results in: 1. E1000_RCTL_EN being cleared in RCTL register. [e1000_down()] 2. dma_free_coherent() being called. [e1000_free_rx_resources()] . An 'ifconfig ethN up' gets us into e1000_open(), which through a number of subfunctions results in: 1. dma_alloc_coherent() being called. [e1000_setup_rx_resources()] 2. E1000_RCTL_EN being set in RCTL register. [e1000_setup_rctl()] 3. E1000_RCTL_EN being cleared in RCTL register. [e1000_configure_rx()] 4. RDLEN, RDBAH and RDBAL registers being set to reflect the dma page allocated in step 1. [e1000_configure_rx()] 5. E1000_RCTL_EN being set in RCTL register. [e1000_configure_rx()] During the 'ifconfig ethN up' there is a window opened, starting in step 2 where the receives are enabled up until they are disabled in step 3, in which the address of the receive descriptor dma page known by the NIC is still the previous one which was freed during the 'ifconfig ethN down'. If this memory has been reallocated for some other use and the NIC feels so inclined, it will write to that former dma page with predictably unpleasant results. I realize that in the guest, we're dealing with an e1000 NIC that is software emulated by qemu-kvm. The problem doesn't appear to occur on bare-metal. Andy suspects that this is because in the emulator link-up is essentially instant and traffic can start flowing immediately. Whereas on bare-metal, link-up usually seems to take at least a few milliseconds. And this might be enough to prevent traffic from flowing into the device inside the window where E1000_RCTL_EN is set. So perhaps a modification needs to be made to the qemu-kvm e1000 NIC emulator to delay the link-up. But in defense of the emulator, it seems like a bad idea to enable dma operations before the address of the memory to be involved has been made known. The following patch no longer enables receives in e1000_setup_rctl() but leaves them however they were. It only enables receives in e1000_configure_rx(), and only after the dma address has been made known to the hardware. There are two places where e1000_setup_rctl() gets called. The one in e1000_configure() is followed immediately by a call to e1000_configure_rx(), so there's really no change functionally (except for the removal of the problem window. The other is in __e1000_shutdown() and is not followed by a call to e1000_configure_rx(), so there is a change functionally. But consider... . An 'ifconfig ethN down' (just as described above). . A 'suspend' of the system, which (I'm assuming) will find its way into e1000_suspend() which calls __e1000_shutdown() resulting in: 1. E1000_RCTL_EN being set in RCTL register. [e1000_setup_rctl()] And again we've re-opened the problem window for some unknown amount of time. Signed-off-by: Andy Gospodarek <andy@greyhouse.net> Signed-off-by: Dean Nelson <dnelson@redhat.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-09-16 16:52:54 +00:00
ew32(RCTL, rctl | E1000_RCTL_EN);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/**
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
* e1000_free_tx_resources - Free Tx Resources per Queue
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* @adapter: board private structure
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
* @tx_ring: Tx descriptor ring for a specific queue
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*
* Free all transmit software resources
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_free_tx_resources(struct e1000_adapter *adapter,
struct e1000_tx_ring *tx_ring)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct pci_dev *pdev = adapter->pdev;
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
e1000_clean_tx_ring(adapter, tx_ring);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
vfree(tx_ring->buffer_info);
tx_ring->buffer_info = NULL;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
dma_free_coherent(&pdev->dev, tx_ring->size, tx_ring->desc,
tx_ring->dma);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
tx_ring->desc = NULL;
}
/**
* e1000_free_all_tx_resources - Free Tx Resources for All Queues
* @adapter: board private structure
*
* Free all transmit software resources
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
void e1000_free_all_tx_resources(struct e1000_adapter *adapter)
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
{
int i;
[PATCH] e1000: Fix adapter structure and prepare for multique fix Fix adapter structure to handle multiple queues and prepping the driver for full multiple queue support, some changes are ifdef'd our unless you define CONFIG_E1000_MQ. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:39 +00:00
for (i = 0; i < adapter->num_tx_queues; i++)
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
e1000_free_tx_resources(adapter, &adapter->tx_ring[i]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_unmap_and_free_tx_resource(struct e1000_adapter *adapter,
struct e1000_buffer *buffer_info)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: remove use of skb_dma_map from e1000 driver Remove the use of skb_dma_map from the e1000 driver in order to avoid issues when HW iommu are in use. Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-02 16:46:00 +00:00
if (buffer_info->dma) {
if (buffer_info->mapped_as_page)
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
dma_unmap_page(&adapter->pdev->dev, buffer_info->dma,
buffer_info->length, DMA_TO_DEVICE);
e1000: remove use of skb_dma_map from e1000 driver Remove the use of skb_dma_map from the e1000 driver in order to avoid issues when HW iommu are in use. Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-02 16:46:00 +00:00
else
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
dma_unmap_single(&adapter->pdev->dev, buffer_info->dma,
e1000: remove use of skb_dma_map from e1000 driver Remove the use of skb_dma_map from the e1000 driver in order to avoid issues when HW iommu are in use. Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-02 16:46:00 +00:00
buffer_info->length,
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
DMA_TO_DEVICE);
e1000: remove use of skb_dma_map from e1000 driver Remove the use of skb_dma_map from the e1000 driver in order to avoid issues when HW iommu are in use. Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-02 16:46:00 +00:00
buffer_info->dma = 0;
}
e1000: Remove unneeded and unwanted memsets This memsetting was added in a paranoid rage debugging TX hangs, but are no longer of importance. We can beef up the performance quite a bit removing them. Make sure to fill in next_to_watch to allow this. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2006-11-01 16:47:53 +00:00
if (buffer_info->skb) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
dev_kfree_skb_any(buffer_info->skb);
e1000: Remove unneeded and unwanted memsets This memsetting was added in a paranoid rage debugging TX hangs, but are no longer of importance. We can beef up the performance quite a bit removing them. Make sure to fill in next_to_watch to allow this. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2006-11-01 16:47:53 +00:00
buffer_info->skb = NULL;
}
e1000: fix tx hang detect logic and address dma mapping issues This patch changes the dma mapping to better support skb_dma_map/skb_dma_unmap and addresses and redefines the tx hang logic to be based off of time stamp instead of if the dma field is populated Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:58:45 +00:00
buffer_info->time_stamp = 0;
e1000: Remove unneeded and unwanted memsets This memsetting was added in a paranoid rage debugging TX hangs, but are no longer of importance. We can beef up the performance quite a bit removing them. Make sure to fill in next_to_watch to allow this. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2006-11-01 16:47:53 +00:00
/* buffer_info must be completely set up in the transmit path */
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/**
* e1000_clean_tx_ring - Free Tx Buffers
* @adapter: board private structure
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
* @tx_ring: ring to be cleaned
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_clean_tx_ring(struct e1000_adapter *adapter,
struct e1000_tx_ring *tx_ring)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct e1000_buffer *buffer_info;
unsigned long size;
unsigned int i;
/* Free all the Tx ring sk_buffs */
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
for (i = 0; i < tx_ring->count; i++) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
buffer_info = &tx_ring->buffer_info[i];
e1000_unmap_and_free_tx_resource(adapter, buffer_info);
}
e1000: add byte queue limits Signed-off-by: Otto Estuardo Solares Cabrera <solca@galileo.edu> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-08-24 08:04:58 +00:00
netdev_reset_queue(adapter->netdev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
size = sizeof(struct e1000_buffer) * tx_ring->count;
memset(tx_ring->buffer_info, 0, size);
/* Zero out the descriptor ring */
memset(tx_ring->desc, 0, tx_ring->size);
tx_ring->next_to_use = 0;
tx_ring->next_to_clean = 0;
net: fix assignment of 0/1 to bool variables. DaveM said: Please, this kind of stuff rots forever and not using bool properly drives me crazy. Joe Perches <joe@perches.com> gave me the spatch script: @@ bool b; @@ -b = 0 +b = false @@ bool b; @@ -b = 1 +b = true I merely installed coccinelle, read the documentation and took credit. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-12-19 13:56:45 +00:00
tx_ring->last_tx_tso = false;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
writel(0, hw->hw_addr + tx_ring->tdh);
writel(0, hw->hw_addr + tx_ring->tdt);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
}
/**
* e1000_clean_all_tx_rings - Free Tx Buffers for all queues
* @adapter: board private structure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_clean_all_tx_rings(struct e1000_adapter *adapter)
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
{
int i;
[PATCH] e1000: Fix adapter structure and prepare for multique fix Fix adapter structure to handle multiple queues and prepping the driver for full multiple queue support, some changes are ifdef'd our unless you define CONFIG_E1000_MQ. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:39 +00:00
for (i = 0; i < adapter->num_tx_queues; i++)
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
e1000_clean_tx_ring(adapter, &adapter->tx_ring[i]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/**
* e1000_free_rx_resources - Free Rx Resources
* @adapter: board private structure
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
* @rx_ring: ring to clean the resources from
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*
* Free all receive software resources
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_free_rx_resources(struct e1000_adapter *adapter,
struct e1000_rx_ring *rx_ring)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct pci_dev *pdev = adapter->pdev;
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
e1000_clean_rx_ring(adapter, rx_ring);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
vfree(rx_ring->buffer_info);
rx_ring->buffer_info = NULL;
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
dma_free_coherent(&pdev->dev, rx_ring->size, rx_ring->desc,
rx_ring->dma);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
rx_ring->desc = NULL;
}
/**
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
* e1000_free_all_rx_resources - Free Rx Resources for All Queues
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* @adapter: board private structure
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
*
* Free all receive software resources
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
void e1000_free_all_rx_resources(struct e1000_adapter *adapter)
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
{
int i;
[PATCH] e1000: Fix adapter structure and prepare for multique fix Fix adapter structure to handle multiple queues and prepping the driver for full multiple queue support, some changes are ifdef'd our unless you define CONFIG_E1000_MQ. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:39 +00:00
for (i = 0; i < adapter->num_rx_queues; i++)
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
e1000_free_rx_resources(adapter, &adapter->rx_ring[i]);
}
/**
* e1000_clean_rx_ring - Free Rx Buffers per Queue
* @adapter: board private structure
* @rx_ring: ring to free buffers from
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_clean_rx_ring(struct e1000_adapter *adapter,
struct e1000_rx_ring *rx_ring)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct e1000_buffer *buffer_info;
struct pci_dev *pdev = adapter->pdev;
unsigned long size;
e1000: remove unused Kconfig option for disabling packet split Since the e1000/e1000e split, no hardware supported by e1000 supports packet split, just remove the Kconfig option and associated code from the driver. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-09-16 20:01:28 +00:00
unsigned int i;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Free all the Rx ring sk_buffs */
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
for (i = 0; i < rx_ring->count; i++) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
buffer_info = &rx_ring->buffer_info[i];
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
if (buffer_info->dma &&
adapter->clean_rx == e1000_clean_rx_irq) {
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
dma_unmap_single(&pdev->dev, buffer_info->dma,
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
buffer_info->length,
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
DMA_FROM_DEVICE);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
} else if (buffer_info->dma &&
adapter->clean_rx == e1000_clean_jumbo_rx_irq) {
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
dma_unmap_page(&pdev->dev, buffer_info->dma,
buffer_info->length,
DMA_FROM_DEVICE);
e1000: fix unmap bug as reported by kerneloops.org [ 121.781161] ------------[ cut here ]------------ [ 121.781171] WARNING: at lib/dma-debug.c:793 check_unmap+0x14e/0x577() [ 121.781173] Hardware name: S5520HC [ 121.781177] e1000 0000:0a:00.0: DMA-API: device driver tries to free DMA memory it has not allocated [device address=0x00000001d688b0fa] [size=1522 bytes] [ 121.781180] Modules linked in: e1000 mdio dca [last unloaded: ixgbe] [ 121.781187] Pid: 4793, comm: bash Tainted: P 2.6.30-master-06161113 #3 [ 121.781190] Call Trace: [ 121.781195] [<ffffffff8123056f>] ? check_unmap+0x14e/0x577 [ 121.781201] [<ffffffff81057a19>] warn_slowpath_common+0x77/0x8f [ 121.781205] [<ffffffff81057ae1>] warn_slowpath_fmt+0x9f/0xa1 [ 121.781212] [<ffffffff81477ce2>] ? _spin_lock_irqsave+0x3f/0x49 [ 121.781216] [<ffffffff8122fa97>] ? get_hash_bucket+0x28/0x33 [ 121.781220] [<ffffffff8123056f>] check_unmap+0x14e/0x577 [ 121.781225] [<ffffffff810e4f48>] ? check_bytes_and_report+0x38/0xcb [ 121.781230] [<ffffffff81230bbf>] debug_dma_unmap_page+0x80/0x92 [ 121.781234] [<ffffffff8122e549>] ? unmap_single+0x1a/0x4e [ 121.781239] [<ffffffff813901e1>] ? __kfree_skb+0x74/0x78 [ 121.781250] [<ffffffffa00662ef>] pci_unmap_single+0x64/0x6d [e1000] [ 121.781259] [<ffffffffa0066344>] e1000_clean_rx_ring+0x4c/0xbf [e1000] [ 121.781268] [<ffffffffa00663df>] e1000_clean_all_rx_rings+0x28/0x36 [e1000] [ 121.781277] [<ffffffffa0067464>] e1000_down+0x138/0x141 [e1000] [ 121.781286] [<ffffffffa00681c2>] __e1000_shutdown+0x6b/0x198 [e1000] [ 121.781296] [<ffffffffa0068405>] e1000_suspend+0x17/0x50 [e1000] [ 121.781301] [<ffffffff81237665>] pci_legacy_suspend+0x3b/0xbe [ 121.781305] [<ffffffff81237bc6>] pci_pm_suspend+0x3e/0xf1 [ 121.781310] [<ffffffff812eaf1c>] pm_op+0x57/0xde [ 121.781314] [<ffffffff812eb444>] dpm_suspend_start+0x31e/0x470 [ 121.781319] [<ffffffff810877da>] suspend_devices_and_enter+0x3e/0x1a2 [ 121.781323] [<ffffffff81087a0f>] enter_state+0xd1/0x127 [ 121.781327] [<ffffffff8108717a>] state_store+0xa7/0xc9 [ 121.781332] [<ffffffff81221843>] kobj_attr_store+0x17/0x19 [ 121.781336] [<ffffffff8113c01e>] sysfs_write_file+0xe5/0x121 [ 121.781341] [<ffffffff810ed165>] vfs_write+0xab/0x105 [ 121.781344] [<ffffffff810ed279>] sys_write+0x47/0x6d [ 121.781349] [<ffffffff81027aab>] system_call_fastpath+0x16/0x1b [ 121.781352] ---[ end trace 97bacaaac2ed7786 ]--- Fix is to correctly zero out internal ->dma value when unmapping and make sure never to unmap unless there specifically was a mapping done. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-06-30 12:45:34 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: fix unmap bug as reported by kerneloops.org [ 121.781161] ------------[ cut here ]------------ [ 121.781171] WARNING: at lib/dma-debug.c:793 check_unmap+0x14e/0x577() [ 121.781173] Hardware name: S5520HC [ 121.781177] e1000 0000:0a:00.0: DMA-API: device driver tries to free DMA memory it has not allocated [device address=0x00000001d688b0fa] [size=1522 bytes] [ 121.781180] Modules linked in: e1000 mdio dca [last unloaded: ixgbe] [ 121.781187] Pid: 4793, comm: bash Tainted: P 2.6.30-master-06161113 #3 [ 121.781190] Call Trace: [ 121.781195] [<ffffffff8123056f>] ? check_unmap+0x14e/0x577 [ 121.781201] [<ffffffff81057a19>] warn_slowpath_common+0x77/0x8f [ 121.781205] [<ffffffff81057ae1>] warn_slowpath_fmt+0x9f/0xa1 [ 121.781212] [<ffffffff81477ce2>] ? _spin_lock_irqsave+0x3f/0x49 [ 121.781216] [<ffffffff8122fa97>] ? get_hash_bucket+0x28/0x33 [ 121.781220] [<ffffffff8123056f>] check_unmap+0x14e/0x577 [ 121.781225] [<ffffffff810e4f48>] ? check_bytes_and_report+0x38/0xcb [ 121.781230] [<ffffffff81230bbf>] debug_dma_unmap_page+0x80/0x92 [ 121.781234] [<ffffffff8122e549>] ? unmap_single+0x1a/0x4e [ 121.781239] [<ffffffff813901e1>] ? __kfree_skb+0x74/0x78 [ 121.781250] [<ffffffffa00662ef>] pci_unmap_single+0x64/0x6d [e1000] [ 121.781259] [<ffffffffa0066344>] e1000_clean_rx_ring+0x4c/0xbf [e1000] [ 121.781268] [<ffffffffa00663df>] e1000_clean_all_rx_rings+0x28/0x36 [e1000] [ 121.781277] [<ffffffffa0067464>] e1000_down+0x138/0x141 [e1000] [ 121.781286] [<ffffffffa00681c2>] __e1000_shutdown+0x6b/0x198 [e1000] [ 121.781296] [<ffffffffa0068405>] e1000_suspend+0x17/0x50 [e1000] [ 121.781301] [<ffffffff81237665>] pci_legacy_suspend+0x3b/0xbe [ 121.781305] [<ffffffff81237bc6>] pci_pm_suspend+0x3e/0xf1 [ 121.781310] [<ffffffff812eaf1c>] pm_op+0x57/0xde [ 121.781314] [<ffffffff812eb444>] dpm_suspend_start+0x31e/0x470 [ 121.781319] [<ffffffff810877da>] suspend_devices_and_enter+0x3e/0x1a2 [ 121.781323] [<ffffffff81087a0f>] enter_state+0xd1/0x127 [ 121.781327] [<ffffffff8108717a>] state_store+0xa7/0xc9 [ 121.781332] [<ffffffff81221843>] kobj_attr_store+0x17/0x19 [ 121.781336] [<ffffffff8113c01e>] sysfs_write_file+0xe5/0x121 [ 121.781341] [<ffffffff810ed165>] vfs_write+0xab/0x105 [ 121.781344] [<ffffffff810ed279>] sys_write+0x47/0x6d [ 121.781349] [<ffffffff81027aab>] system_call_fastpath+0x16/0x1b [ 121.781352] ---[ end trace 97bacaaac2ed7786 ]--- Fix is to correctly zero out internal ->dma value when unmapping and make sure never to unmap unless there specifically was a mapping done. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-06-30 12:45:34 +00:00
buffer_info->dma = 0;
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
if (buffer_info->page) {
put_page(buffer_info->page);
buffer_info->page = NULL;
}
e1000: fix unmap bug as reported by kerneloops.org [ 121.781161] ------------[ cut here ]------------ [ 121.781171] WARNING: at lib/dma-debug.c:793 check_unmap+0x14e/0x577() [ 121.781173] Hardware name: S5520HC [ 121.781177] e1000 0000:0a:00.0: DMA-API: device driver tries to free DMA memory it has not allocated [device address=0x00000001d688b0fa] [size=1522 bytes] [ 121.781180] Modules linked in: e1000 mdio dca [last unloaded: ixgbe] [ 121.781187] Pid: 4793, comm: bash Tainted: P 2.6.30-master-06161113 #3 [ 121.781190] Call Trace: [ 121.781195] [<ffffffff8123056f>] ? check_unmap+0x14e/0x577 [ 121.781201] [<ffffffff81057a19>] warn_slowpath_common+0x77/0x8f [ 121.781205] [<ffffffff81057ae1>] warn_slowpath_fmt+0x9f/0xa1 [ 121.781212] [<ffffffff81477ce2>] ? _spin_lock_irqsave+0x3f/0x49 [ 121.781216] [<ffffffff8122fa97>] ? get_hash_bucket+0x28/0x33 [ 121.781220] [<ffffffff8123056f>] check_unmap+0x14e/0x577 [ 121.781225] [<ffffffff810e4f48>] ? check_bytes_and_report+0x38/0xcb [ 121.781230] [<ffffffff81230bbf>] debug_dma_unmap_page+0x80/0x92 [ 121.781234] [<ffffffff8122e549>] ? unmap_single+0x1a/0x4e [ 121.781239] [<ffffffff813901e1>] ? __kfree_skb+0x74/0x78 [ 121.781250] [<ffffffffa00662ef>] pci_unmap_single+0x64/0x6d [e1000] [ 121.781259] [<ffffffffa0066344>] e1000_clean_rx_ring+0x4c/0xbf [e1000] [ 121.781268] [<ffffffffa00663df>] e1000_clean_all_rx_rings+0x28/0x36 [e1000] [ 121.781277] [<ffffffffa0067464>] e1000_down+0x138/0x141 [e1000] [ 121.781286] [<ffffffffa00681c2>] __e1000_shutdown+0x6b/0x198 [e1000] [ 121.781296] [<ffffffffa0068405>] e1000_suspend+0x17/0x50 [e1000] [ 121.781301] [<ffffffff81237665>] pci_legacy_suspend+0x3b/0xbe [ 121.781305] [<ffffffff81237bc6>] pci_pm_suspend+0x3e/0xf1 [ 121.781310] [<ffffffff812eaf1c>] pm_op+0x57/0xde [ 121.781314] [<ffffffff812eb444>] dpm_suspend_start+0x31e/0x470 [ 121.781319] [<ffffffff810877da>] suspend_devices_and_enter+0x3e/0x1a2 [ 121.781323] [<ffffffff81087a0f>] enter_state+0xd1/0x127 [ 121.781327] [<ffffffff8108717a>] state_store+0xa7/0xc9 [ 121.781332] [<ffffffff81221843>] kobj_attr_store+0x17/0x19 [ 121.781336] [<ffffffff8113c01e>] sysfs_write_file+0xe5/0x121 [ 121.781341] [<ffffffff810ed165>] vfs_write+0xab/0x105 [ 121.781344] [<ffffffff810ed279>] sys_write+0x47/0x6d [ 121.781349] [<ffffffff81027aab>] system_call_fastpath+0x16/0x1b [ 121.781352] ---[ end trace 97bacaaac2ed7786 ]--- Fix is to correctly zero out internal ->dma value when unmapping and make sure never to unmap unless there specifically was a mapping done. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-06-30 12:45:34 +00:00
if (buffer_info->skb) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
dev_kfree_skb(buffer_info->skb);
buffer_info->skb = NULL;
[PATCH] e1000: Fix Desc. Rings and Jumbo Frames This patch contains two fixes. The first fix is to the tx and rx descriptor rings clean up process. The second fix is to jumbo frames, which cleans up the code logic and removes most of the fifo related limitations on jumbo frames. This is because the driver code now supports splitting a packet across multiple descriptors. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:55 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
/* there also may be some cached data from a chained receive */
if (rx_ring->rx_skb_top) {
dev_kfree_skb(rx_ring->rx_skb_top);
rx_ring->rx_skb_top = NULL;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
size = sizeof(struct e1000_buffer) * rx_ring->count;
memset(rx_ring->buffer_info, 0, size);
/* Zero out the descriptor ring */
memset(rx_ring->desc, 0, rx_ring->size);
rx_ring->next_to_clean = 0;
rx_ring->next_to_use = 0;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
writel(0, hw->hw_addr + rx_ring->rdh);
writel(0, hw->hw_addr + rx_ring->rdt);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
}
/**
* e1000_clean_all_rx_rings - Free Rx Buffers for all queues
* @adapter: board private structure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_clean_all_rx_rings(struct e1000_adapter *adapter)
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
{
int i;
[PATCH] e1000: Fix adapter structure and prepare for multique fix Fix adapter structure to handle multiple queues and prepping the driver for full multiple queue support, some changes are ifdef'd our unless you define CONFIG_E1000_MQ. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:39 +00:00
for (i = 0; i < adapter->num_rx_queues; i++)
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
e1000_clean_rx_ring(adapter, &adapter->rx_ring[i]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/* The 82542 2.0 (revision 2) needs to have the receive unit in reset
* and memory write and invalidate disabled for certain operations
*/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_enter_82542_rst(struct e1000_adapter *adapter)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct net_device *netdev = adapter->netdev;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u32 rctl;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_pci_clear_mwi(hw);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
rctl = er32(RCTL);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
rctl |= E1000_RCTL_RST;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(RCTL, rctl);
E1000_WRITE_FLUSH();
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
mdelay(5);
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (netif_running(netdev))
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
e1000_clean_all_rx_rings(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_leave_82542_rst(struct e1000_adapter *adapter)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct net_device *netdev = adapter->netdev;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u32 rctl;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
rctl = er32(RCTL);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
rctl &= ~E1000_RCTL_RST;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(RCTL, rctl);
E1000_WRITE_FLUSH();
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
mdelay(5);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (hw->pci_cmd_word & PCI_COMMAND_INVALIDATE)
e1000_pci_set_mwi(hw);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (netif_running(netdev)) {
[PATCH] e1000: Added cleaned_count to RX buffer allocation Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:19 +00:00
/* No need to loop, because 82542 supports only 1 queue */
struct e1000_rx_ring *ring = &adapter->rx_ring[0];
[PATCH] e1000: fix compile warning Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:45 +00:00
e1000_configure_rx(adapter);
[PATCH] e1000: Added cleaned_count to RX buffer allocation Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:19 +00:00
adapter->alloc_rx_buf(adapter, ring, E1000_DESC_UNUSED(ring));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
}
/**
* e1000_set_mac - Change the Ethernet Address of the NIC
* @netdev: network interface device structure
* @p: pointer to an address structure
*
* Returns 0 on success, negative on failure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static int e1000_set_mac(struct net_device *netdev, void *p)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
[PATCH] e1000: Use netdev_priv() to get to netdev->priv Use netdev_priv() to get to netdev->priv - from shemminger@osdl.org Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:41:45 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct sockaddr *addr = p;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (!is_valid_ether_addr(addr->sa_data))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return -EADDRNOTAVAIL;
/* 82542 2.0 needs to be in reset to write receive address registers */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (hw->mac_type == e1000_82542_rev2_0)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_enter_82542_rst(adapter);
memcpy(netdev->dev_addr, addr->sa_data, netdev->addr_len);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
memcpy(hw->mac_addr, addr->sa_data, netdev->addr_len);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_rar_set(hw, hw->mac_addr, 0);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (hw->mac_type == e1000_82542_rev2_0)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_leave_82542_rst(adapter);
return 0;
}
/**
[E1000]: Secondary unicast address support Add support for configuring secondary unicast addresses. Unicast addresses take precendece over multicast addresses when filling the exact address filters to avoid going to promiscous mode. When more unicast addresses are present than filter slots, unicast filtering is disabled and all slots can be used for multicast addresses. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-11-14 04:54:59 +00:00
* e1000_set_rx_mode - Secondary Unicast, Multicast and Promiscuous mode set
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* @netdev: network interface device structure
*
[E1000]: Secondary unicast address support Add support for configuring secondary unicast addresses. Unicast addresses take precendece over multicast addresses when filling the exact address filters to avoid going to promiscous mode. When more unicast addresses are present than filter slots, unicast filtering is disabled and all slots can be used for multicast addresses. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-11-14 04:54:59 +00:00
* The set_rx_mode entry point is called whenever the unicast or multicast
* address lists or the network interface flags are updated. This routine is
* responsible for configuring the hardware for proper unicast, multicast,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* promiscuous mode, and all-multi behavior.
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_set_rx_mode(struct net_device *netdev)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
[PATCH] e1000: Use netdev_priv() to get to netdev->priv Use netdev_priv() to get to netdev->priv - from shemminger@osdl.org Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:41:45 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct e1000_hw *hw = &adapter->hw;
net: convert unicast addr list This patch converts unicast address list to standard list_head using previously introduced struct netdev_hw_addr. It also relaxes the locking. Original spinlock (still used for multicast addresses) is not needed and is no longer used for a protection of this list. All reading and writing takes place under rtnl (with no changes). I also removed a possibility to specify the length of the address while adding or deleting unicast address. It's always dev->addr_len. The convertion touched especially e1000 and ixgbe codes when the change is not so trivial. Signed-off-by: Jiri Pirko <jpirko@redhat.com> drivers/net/bnx2.c | 13 +-- drivers/net/e1000/e1000_main.c | 24 +++-- drivers/net/ixgbe/ixgbe_common.c | 14 ++-- drivers/net/ixgbe/ixgbe_common.h | 4 +- drivers/net/ixgbe/ixgbe_main.c | 6 +- drivers/net/ixgbe/ixgbe_type.h | 4 +- drivers/net/macvlan.c | 11 +- drivers/net/mv643xx_eth.c | 11 +- drivers/net/niu.c | 7 +- drivers/net/virtio_net.c | 7 +- drivers/s390/net/qeth_l2_main.c | 6 +- drivers/scsi/fcoe/fcoe.c | 16 ++-- include/linux/netdevice.h | 18 ++-- net/8021q/vlan.c | 4 +- net/8021q/vlan_dev.c | 10 +- net/core/dev.c | 195 +++++++++++++++++++++++++++----------- net/dsa/slave.c | 10 +- net/packet/af_packet.c | 4 +- 18 files changed, 227 insertions(+), 137 deletions(-) Signed-off-by: David S. Miller <davem@davemloft.net>
2009-05-22 23:22:17 +00:00
struct netdev_hw_addr *ha;
bool use_uc = false;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u32 rctl;
u32 hash_value;
e1000: Support for 82571 and 82572 controllers Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 10:58:59 +00:00
int i, rar_entries = E1000_RAR_ENTRIES;
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
int mta_reg_count = E1000_NUM_MTA_REGISTERS;
e1000: fix loss of multicast packets e1000 (and e1000e, igb, ixgbe, ixgb) all do a series of operations each time a multicast address is added. The flow goes something like 1) stack adds one multicast address 2) stack passes whole current list of unicast and multicast addresses to driver 3) driver clears entire list in hardware 4) driver programs each multicast address using iomem in a loop This was causing multicast packets to be lost during the reprogramming process. reference with test program: http://kerneltrap.org/mailarchive/linux-netdev/2009/3/14/5160514/thread Thanks to Dave Boutcher for his report and test program. This driver fix prepares an array all at once in memory and programs it in one shot to the hardware, not requiring an "erase" cycle. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-04 23:36:53 +00:00
u32 *mcarray = kcalloc(mta_reg_count, sizeof(u32), GFP_ATOMIC);
drivers: net: Remove remaining alloc/OOM messages alloc failures already get standardized OOM messages and a dump_stack. For the affected mallocs around these OOM messages: Converted kmallocs with multiplies to kmalloc_array. Converted a kmalloc/memcpy to kmemdup. Removed now unused stack variables. Removed unnecessary parentheses. Neatened alignment. Signed-off-by: Joe Perches <joe@perches.com> Acked-by: Arend van Spriel <arend@broadcom.com> Acked-by: Marc Kleine-Budde <mkl@pengutronix.de> Acked-by: John W. Linville <linville@tuxdriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-02-07 11:46:27 +00:00
if (!mcarray)
e1000: fix loss of multicast packets e1000 (and e1000e, igb, ixgbe, ixgb) all do a series of operations each time a multicast address is added. The flow goes something like 1) stack adds one multicast address 2) stack passes whole current list of unicast and multicast addresses to driver 3) driver clears entire list in hardware 4) driver programs each multicast address using iomem in a loop This was causing multicast packets to be lost during the reprogramming process. reference with test program: http://kerneltrap.org/mailarchive/linux-netdev/2009/3/14/5160514/thread Thanks to Dave Boutcher for his report and test program. This driver fix prepares an array all at once in memory and programs it in one shot to the hardware, not requiring an "erase" cycle. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-04 23:36:53 +00:00
return;
e1000: integrate ich8 support into driver This hooks up the ich8 structure into the driver itself. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:08:22 +00:00
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
/* Check for Promiscuous and All Multicast modes */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
rctl = er32(RCTL);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (netdev->flags & IFF_PROMISC) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
rctl |= (E1000_RCTL_UPE | E1000_RCTL_MPE);
netdrv intel: disable VLAN filtering in promiscous mode As discussed in this thread: http://www.mail-archive.com/netdev@vger.kernel.org/msg53976.html promiscous mode means to disable *all* filters. Currently only unicast and multicast filtering is disabled. This patch changes all Intel drivers to also disable VLAN filtering. Signed-off-by: Patrick McHardy <kaber@trash.net> Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Acked-by: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-17 03:15:45 +00:00
rctl &= ~E1000_RCTL_VFE;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
} else {
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
if (netdev->flags & IFF_ALLMULTI)
netdrv intel: disable VLAN filtering in promiscous mode As discussed in this thread: http://www.mail-archive.com/netdev@vger.kernel.org/msg53976.html promiscous mode means to disable *all* filters. Currently only unicast and multicast filtering is disabled. This patch changes all Intel drivers to also disable VLAN filtering. Signed-off-by: Patrick McHardy <kaber@trash.net> Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Acked-by: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-17 03:15:45 +00:00
rctl |= E1000_RCTL_MPE;
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
else
netdrv intel: disable VLAN filtering in promiscous mode As discussed in this thread: http://www.mail-archive.com/netdev@vger.kernel.org/msg53976.html promiscous mode means to disable *all* filters. Currently only unicast and multicast filtering is disabled. This patch changes all Intel drivers to also disable VLAN filtering. Signed-off-by: Patrick McHardy <kaber@trash.net> Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Acked-by: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-17 03:15:45 +00:00
rctl &= ~E1000_RCTL_MPE;
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
/* Enable VLAN filter if there is a VLAN */
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
if (e1000_vlan_used(adapter))
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
rctl |= E1000_RCTL_VFE;
[E1000]: Secondary unicast address support Add support for configuring secondary unicast addresses. Unicast addresses take precendece over multicast addresses when filling the exact address filters to avoid going to promiscous mode. When more unicast addresses are present than filter slots, unicast filtering is disabled and all slots can be used for multicast addresses. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-11-14 04:54:59 +00:00
}
net: use helpers to access uc list V2 This patch introduces three macros to work with uc list from net drivers. Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-01-25 21:36:10 +00:00
if (netdev_uc_count(netdev) > rar_entries - 1) {
[E1000]: Secondary unicast address support Add support for configuring secondary unicast addresses. Unicast addresses take precendece over multicast addresses when filling the exact address filters to avoid going to promiscous mode. When more unicast addresses are present than filter slots, unicast filtering is disabled and all slots can be used for multicast addresses. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-11-14 04:54:59 +00:00
rctl |= E1000_RCTL_UPE;
} else if (!(netdev->flags & IFF_PROMISC)) {
rctl &= ~E1000_RCTL_UPE;
net: convert unicast addr list This patch converts unicast address list to standard list_head using previously introduced struct netdev_hw_addr. It also relaxes the locking. Original spinlock (still used for multicast addresses) is not needed and is no longer used for a protection of this list. All reading and writing takes place under rtnl (with no changes). I also removed a possibility to specify the length of the address while adding or deleting unicast address. It's always dev->addr_len. The convertion touched especially e1000 and ixgbe codes when the change is not so trivial. Signed-off-by: Jiri Pirko <jpirko@redhat.com> drivers/net/bnx2.c | 13 +-- drivers/net/e1000/e1000_main.c | 24 +++-- drivers/net/ixgbe/ixgbe_common.c | 14 ++-- drivers/net/ixgbe/ixgbe_common.h | 4 +- drivers/net/ixgbe/ixgbe_main.c | 6 +- drivers/net/ixgbe/ixgbe_type.h | 4 +- drivers/net/macvlan.c | 11 +- drivers/net/mv643xx_eth.c | 11 +- drivers/net/niu.c | 7 +- drivers/net/virtio_net.c | 7 +- drivers/s390/net/qeth_l2_main.c | 6 +- drivers/scsi/fcoe/fcoe.c | 16 ++-- include/linux/netdevice.h | 18 ++-- net/8021q/vlan.c | 4 +- net/8021q/vlan_dev.c | 10 +- net/core/dev.c | 195 +++++++++++++++++++++++++++----------- net/dsa/slave.c | 10 +- net/packet/af_packet.c | 4 +- 18 files changed, 227 insertions(+), 137 deletions(-) Signed-off-by: David S. Miller <davem@davemloft.net>
2009-05-22 23:22:17 +00:00
use_uc = true;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(RCTL, rctl);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* 82542 2.0 needs to be in reset to write receive address registers */
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (hw->mac_type == e1000_82542_rev2_0)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_enter_82542_rst(adapter);
[E1000]: Secondary unicast address support Add support for configuring secondary unicast addresses. Unicast addresses take precendece over multicast addresses when filling the exact address filters to avoid going to promiscous mode. When more unicast addresses are present than filter slots, unicast filtering is disabled and all slots can be used for multicast addresses. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-11-14 04:54:59 +00:00
/* load the first 14 addresses into the exact filters 1-14. Unicast
* addresses take precedence to avoid disabling unicast filtering
* when possible.
*
tree-wide: fix comment/printk typos "gadget", "through", "command", "maintain", "maintain", "controller", "address", "between", "initiali[zs]e", "instead", "function", "select", "already", "equal", "access", "management", "hierarchy", "registration", "interest", "relative", "memory", "offset", "already", Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2010-11-01 19:38:34 +00:00
* RAR 0 is used for the station MAC address
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* if there are not 14 addresses, go ahead and clear the filters
*/
net: convert unicast addr list This patch converts unicast address list to standard list_head using previously introduced struct netdev_hw_addr. It also relaxes the locking. Original spinlock (still used for multicast addresses) is not needed and is no longer used for a protection of this list. All reading and writing takes place under rtnl (with no changes). I also removed a possibility to specify the length of the address while adding or deleting unicast address. It's always dev->addr_len. The convertion touched especially e1000 and ixgbe codes when the change is not so trivial. Signed-off-by: Jiri Pirko <jpirko@redhat.com> drivers/net/bnx2.c | 13 +-- drivers/net/e1000/e1000_main.c | 24 +++-- drivers/net/ixgbe/ixgbe_common.c | 14 ++-- drivers/net/ixgbe/ixgbe_common.h | 4 +- drivers/net/ixgbe/ixgbe_main.c | 6 +- drivers/net/ixgbe/ixgbe_type.h | 4 +- drivers/net/macvlan.c | 11 +- drivers/net/mv643xx_eth.c | 11 +- drivers/net/niu.c | 7 +- drivers/net/virtio_net.c | 7 +- drivers/s390/net/qeth_l2_main.c | 6 +- drivers/scsi/fcoe/fcoe.c | 16 ++-- include/linux/netdevice.h | 18 ++-- net/8021q/vlan.c | 4 +- net/8021q/vlan_dev.c | 10 +- net/core/dev.c | 195 +++++++++++++++++++++++++++----------- net/dsa/slave.c | 10 +- net/packet/af_packet.c | 4 +- 18 files changed, 227 insertions(+), 137 deletions(-) Signed-off-by: David S. Miller <davem@davemloft.net>
2009-05-22 23:22:17 +00:00
i = 1;
if (use_uc)
net: use helpers to access uc list V2 This patch introduces three macros to work with uc list from net drivers. Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-01-25 21:36:10 +00:00
netdev_for_each_uc_addr(ha, netdev) {
net: convert unicast addr list This patch converts unicast address list to standard list_head using previously introduced struct netdev_hw_addr. It also relaxes the locking. Original spinlock (still used for multicast addresses) is not needed and is no longer used for a protection of this list. All reading and writing takes place under rtnl (with no changes). I also removed a possibility to specify the length of the address while adding or deleting unicast address. It's always dev->addr_len. The convertion touched especially e1000 and ixgbe codes when the change is not so trivial. Signed-off-by: Jiri Pirko <jpirko@redhat.com> drivers/net/bnx2.c | 13 +-- drivers/net/e1000/e1000_main.c | 24 +++-- drivers/net/ixgbe/ixgbe_common.c | 14 ++-- drivers/net/ixgbe/ixgbe_common.h | 4 +- drivers/net/ixgbe/ixgbe_main.c | 6 +- drivers/net/ixgbe/ixgbe_type.h | 4 +- drivers/net/macvlan.c | 11 +- drivers/net/mv643xx_eth.c | 11 +- drivers/net/niu.c | 7 +- drivers/net/virtio_net.c | 7 +- drivers/s390/net/qeth_l2_main.c | 6 +- drivers/scsi/fcoe/fcoe.c | 16 ++-- include/linux/netdevice.h | 18 ++-- net/8021q/vlan.c | 4 +- net/8021q/vlan_dev.c | 10 +- net/core/dev.c | 195 +++++++++++++++++++++++++++----------- net/dsa/slave.c | 10 +- net/packet/af_packet.c | 4 +- 18 files changed, 227 insertions(+), 137 deletions(-) Signed-off-by: David S. Miller <davem@davemloft.net>
2009-05-22 23:22:17 +00:00
if (i == rar_entries)
break;
e1000_rar_set(hw, ha->addr, i++);
}
net: convert multicast list to list_head Converts the list and the core manipulating with it to be the same as uc_list. +uses two functions for adding/removing mc address (normal and "global" variant) instead of a function parameter. +removes dev_mcast.c completely. +exposes netdev_hw_addr_list_* macros along with __hw_addr_* functions for manipulation with lists on a sandbox (used in bonding and 80211 drivers) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-01 21:22:57 +00:00
netdev_for_each_mc_addr(ha, netdev) {
e1000: convert to use netdev_for_each_mc_addr Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-02-22 09:10:44 +00:00
if (i == rar_entries) {
/* load any remaining addresses into the hash table */
u32 hash_reg, hash_bit, mta;
net: convert multicast list to list_head Converts the list and the core manipulating with it to be the same as uc_list. +uses two functions for adding/removing mc address (normal and "global" variant) instead of a function parameter. +removes dev_mcast.c completely. +exposes netdev_hw_addr_list_* macros along with __hw_addr_* functions for manipulation with lists on a sandbox (used in bonding and 80211 drivers) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-01 21:22:57 +00:00
hash_value = e1000_hash_mc_addr(hw, ha->addr);
e1000: convert to use netdev_for_each_mc_addr Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-02-22 09:10:44 +00:00
hash_reg = (hash_value >> 5) & 0x7F;
hash_bit = hash_value & 0x1F;
mta = (1 << hash_bit);
mcarray[hash_reg] |= mta;
e1000: correct wrong coding style for "else" Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-02-23 09:19:22 +00:00
} else {
net: convert multicast list to list_head Converts the list and the core manipulating with it to be the same as uc_list. +uses two functions for adding/removing mc address (normal and "global" variant) instead of a function parameter. +removes dev_mcast.c completely. +exposes netdev_hw_addr_list_* macros along with __hw_addr_* functions for manipulation with lists on a sandbox (used in bonding and 80211 drivers) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-01 21:22:57 +00:00
e1000_rar_set(hw, ha->addr, i++);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
}
e1000: convert to use netdev_for_each_mc_addr Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-02-22 09:10:44 +00:00
for (; i < rar_entries; i++) {
E1000_WRITE_REG_ARRAY(hw, RA, i << 1, 0);
E1000_WRITE_FLUSH();
E1000_WRITE_REG_ARRAY(hw, RA, (i << 1) + 1, 0);
E1000_WRITE_FLUSH();
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: fix loss of multicast packets e1000 (and e1000e, igb, ixgbe, ixgb) all do a series of operations each time a multicast address is added. The flow goes something like 1) stack adds one multicast address 2) stack passes whole current list of unicast and multicast addresses to driver 3) driver clears entire list in hardware 4) driver programs each multicast address using iomem in a loop This was causing multicast packets to be lost during the reprogramming process. reference with test program: http://kerneltrap.org/mailarchive/linux-netdev/2009/3/14/5160514/thread Thanks to Dave Boutcher for his report and test program. This driver fix prepares an array all at once in memory and programs it in one shot to the hardware, not requiring an "erase" cycle. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-04 23:36:53 +00:00
/* write the hash table completely, write from bottom to avoid
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* both stupid write combining chipsets, and flushing each write
*/
e1000: fix loss of multicast packets e1000 (and e1000e, igb, ixgbe, ixgb) all do a series of operations each time a multicast address is added. The flow goes something like 1) stack adds one multicast address 2) stack passes whole current list of unicast and multicast addresses to driver 3) driver clears entire list in hardware 4) driver programs each multicast address using iomem in a loop This was causing multicast packets to be lost during the reprogramming process. reference with test program: http://kerneltrap.org/mailarchive/linux-netdev/2009/3/14/5160514/thread Thanks to Dave Boutcher for his report and test program. This driver fix prepares an array all at once in memory and programs it in one shot to the hardware, not requiring an "erase" cycle. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-04 23:36:53 +00:00
for (i = mta_reg_count - 1; i >= 0 ; i--) {
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* If we are on an 82544 has an errata where writing odd
e1000: fix loss of multicast packets e1000 (and e1000e, igb, ixgbe, ixgb) all do a series of operations each time a multicast address is added. The flow goes something like 1) stack adds one multicast address 2) stack passes whole current list of unicast and multicast addresses to driver 3) driver clears entire list in hardware 4) driver programs each multicast address using iomem in a loop This was causing multicast packets to be lost during the reprogramming process. reference with test program: http://kerneltrap.org/mailarchive/linux-netdev/2009/3/14/5160514/thread Thanks to Dave Boutcher for his report and test program. This driver fix prepares an array all at once in memory and programs it in one shot to the hardware, not requiring an "erase" cycle. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-04 23:36:53 +00:00
* offsets overwrites the previous even offset, but writing
* backwards over the range solves the issue by always
* writing the odd offset first
*/
E1000_WRITE_REG_ARRAY(hw, MTA, i, mcarray[i]);
}
E1000_WRITE_FLUSH();
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (hw->mac_type == e1000_82542_rev2_0)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_leave_82542_rst(adapter);
e1000: fix loss of multicast packets e1000 (and e1000e, igb, ixgbe, ixgb) all do a series of operations each time a multicast address is added. The flow goes something like 1) stack adds one multicast address 2) stack passes whole current list of unicast and multicast addresses to driver 3) driver clears entire list in hardware 4) driver programs each multicast address using iomem in a loop This was causing multicast packets to be lost during the reprogramming process. reference with test program: http://kerneltrap.org/mailarchive/linux-netdev/2009/3/14/5160514/thread Thanks to Dave Boutcher for his report and test program. This driver fix prepares an array all at once in memory and programs it in one shot to the hardware, not requiring an "erase" cycle. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-04 23:36:53 +00:00
kfree(mcarray);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
/**
* e1000_update_phy_info_task - get phy info
* @work: work struct contained inside adapter struct
*
* Need to wait a few seconds after link up to get diagnostic information from
* the phy
*/
e1000: use work queues E1000 is using several timers that in a follow on patch will need to acquire the rtnl_lock in order to be safe. This patch moves the timer bodies into work queues which will allow the next patch to add rtnl_lock. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-22 18:22:17 +00:00
static void e1000_update_phy_info_task(struct work_struct *work)
{
struct e1000_adapter *adapter = container_of(work,
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
struct e1000_adapter,
phy_info_task.work);
e1000: convert to private mutex from rtnl The e1000 driver when running with lockdep could run into some possible deadlocks between the work items acquiring rtnl and the rtnl lock being acquired before work items were cancelled. Use a private mutex to make sure lock ordering isn't violated. The private mutex is only used to protect areas not generally covered by the rtnl lock already. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:51 +00:00
if (test_bit(__E1000_DOWN, &adapter->flags))
return;
mutex_lock(&adapter->mutex);
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
e1000_phy_get_info(&adapter->hw, &adapter->phy_info);
e1000: convert to private mutex from rtnl The e1000 driver when running with lockdep could run into some possible deadlocks between the work items acquiring rtnl and the rtnl lock being acquired before work items were cancelled. Use a private mutex to make sure lock ordering isn't violated. The private mutex is only used to protect areas not generally covered by the rtnl lock already. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:51 +00:00
mutex_unlock(&adapter->mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: use work queues E1000 is using several timers that in a follow on patch will need to acquire the rtnl_lock in order to be safe. This patch moves the timer bodies into work queues which will allow the next patch to add rtnl_lock. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-22 18:22:17 +00:00
/**
* e1000_82547_tx_fifo_stall_task - task to complete work
* @work: work struct contained inside adapter struct
**/
static void e1000_82547_tx_fifo_stall_task(struct work_struct *work)
{
struct e1000_adapter *adapter = container_of(work,
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
struct e1000_adapter,
fifo_stall_task.work);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct net_device *netdev = adapter->netdev;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u32 tctl;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: convert to private mutex from rtnl The e1000 driver when running with lockdep could run into some possible deadlocks between the work items acquiring rtnl and the rtnl lock being acquired before work items were cancelled. Use a private mutex to make sure lock ordering isn't violated. The private mutex is only used to protect areas not generally covered by the rtnl lock already. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:51 +00:00
if (test_bit(__E1000_DOWN, &adapter->flags))
return;
mutex_lock(&adapter->mutex);
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (atomic_read(&adapter->tx_fifo_stall)) {
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if ((er32(TDT) == er32(TDH)) &&
(er32(TDFT) == er32(TDFH)) &&
(er32(TDFTS) == er32(TDFHS))) {
tctl = er32(TCTL);
ew32(TCTL, tctl & ~E1000_TCTL_EN);
ew32(TDFT, adapter->tx_head_addr);
ew32(TDFH, adapter->tx_head_addr);
ew32(TDFTS, adapter->tx_head_addr);
ew32(TDFHS, adapter->tx_head_addr);
ew32(TCTL, tctl);
E1000_WRITE_FLUSH();
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
adapter->tx_fifo_head = 0;
atomic_set(&adapter->tx_fifo_stall, 0);
netif_wake_queue(netdev);
e1000: stop timers at appropriate times there were some hotplug cases that made timers still run after the driver had been removed, make sure to stop all the timers and not allow racy reschedules. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:17:23 +00:00
} else if (!test_bit(__E1000_DOWN, &adapter->flags)) {
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
schedule_delayed_work(&adapter->fifo_stall_task, 1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
}
e1000: convert to private mutex from rtnl The e1000 driver when running with lockdep could run into some possible deadlocks between the work items acquiring rtnl and the rtnl lock being acquired before work items were cancelled. Use a private mutex to make sure lock ordering isn't violated. The private mutex is only used to protect areas not generally covered by the rtnl lock already. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:51 +00:00
mutex_unlock(&adapter->mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: Report link status in ethtool when interface is down With this change ethtool will correctly report link status when the interface is down. Currently ethtool reports the link as not detected when the interface is down. Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-02-03 14:49:28 +00:00
bool e1000_has_link(struct e1000_adapter *adapter)
e1000: test link state conclusively e1000 was using one particular way to detect link, but with the advent of some of the newer hardware designs using SERDES connections, tests for link must completely cover all cases. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:17:44 +00:00
{
struct e1000_hw *hw = &adapter->hw;
bool link_active = false;
e1000: always call e1000_check_for_link() on e1000_ce4100 MACs. Interrupts about link lost or rx sequence errors are not reported by the ce4100 hardware, leading to transitions from link UP to link DOWN never being reported. Signed-off-by: Nicolas Schichan <nschichan@freebox.fr> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-07-09 00:24:18 +00:00
/* get_link_status is set on LSC (link status) interrupt or rx
* sequence error interrupt (except on intel ce4100).
* get_link_status will stay false until the
* e1000_check_for_link establishes link for copper adapters
* ONLY
e1000: test link state conclusively e1000 was using one particular way to detect link, but with the advent of some of the newer hardware designs using SERDES connections, tests for link must completely cover all cases. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:17:44 +00:00
*/
switch (hw->media_type) {
case e1000_media_type_copper:
e1000: always call e1000_check_for_link() on e1000_ce4100 MACs. Interrupts about link lost or rx sequence errors are not reported by the ce4100 hardware, leading to transitions from link UP to link DOWN never being reported. Signed-off-by: Nicolas Schichan <nschichan@freebox.fr> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-07-09 00:24:18 +00:00
if (hw->mac_type == e1000_ce4100)
hw->get_link_status = 1;
e1000: test link state conclusively e1000 was using one particular way to detect link, but with the advent of some of the newer hardware designs using SERDES connections, tests for link must completely cover all cases. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:17:44 +00:00
if (hw->get_link_status) {
e1000: updated whitespace and comments A large whitespace change to e1000_hw.[ch] in order to update it to kernel coding style (by running lindent). Updated function header comments into kdoc style. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 22:19:46 +00:00
e1000_check_for_link(hw);
e1000: test link state conclusively e1000 was using one particular way to detect link, but with the advent of some of the newer hardware designs using SERDES connections, tests for link must completely cover all cases. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:17:44 +00:00
link_active = !hw->get_link_status;
} else {
link_active = true;
}
break;
case e1000_media_type_fiber:
e1000: updated whitespace and comments A large whitespace change to e1000_hw.[ch] in order to update it to kernel coding style (by running lindent). Updated function header comments into kdoc style. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 22:19:46 +00:00
e1000_check_for_link(hw);
e1000: test link state conclusively e1000 was using one particular way to detect link, but with the advent of some of the newer hardware designs using SERDES connections, tests for link must completely cover all cases. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:17:44 +00:00
link_active = !!(er32(STATUS) & E1000_STATUS_LU);
break;
case e1000_media_type_internal_serdes:
e1000: updated whitespace and comments A large whitespace change to e1000_hw.[ch] in order to update it to kernel coding style (by running lindent). Updated function header comments into kdoc style. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 22:19:46 +00:00
e1000_check_for_link(hw);
e1000: test link state conclusively e1000 was using one particular way to detect link, but with the advent of some of the newer hardware designs using SERDES connections, tests for link must completely cover all cases. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:17:44 +00:00
link_active = hw->serdes_has_link;
break;
default:
break;
}
return link_active;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/**
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
* e1000_watchdog - work function
* @work: work struct contained inside adapter struct
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
**/
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
static void e1000_watchdog(struct work_struct *work)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
struct e1000_adapter *adapter = container_of(work,
struct e1000_adapter,
watchdog_task.work);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct net_device *netdev = adapter->netdev;
[PATCH] e1000: General Fixes These fixes update the TX and RX ring structures. Prepare driver for up-coming fixes. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:25 +00:00
struct e1000_tx_ring *txdr = adapter->tx_ring;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u32 link, tctl;
e1000: whitespace changes, comments, typos Small whitespace changes, comment changes, typo fixes. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:47:30 +00:00
e1000: convert to private mutex from rtnl The e1000 driver when running with lockdep could run into some possible deadlocks between the work items acquiring rtnl and the rtnl lock being acquired before work items were cancelled. Use a private mutex to make sure lock ordering isn't violated. The private mutex is only used to protect areas not generally covered by the rtnl lock already. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:51 +00:00
if (test_bit(__E1000_DOWN, &adapter->flags))
return;
mutex_lock(&adapter->mutex);
e1000: test link state conclusively e1000 was using one particular way to detect link, but with the advent of some of the newer hardware designs using SERDES connections, tests for link must completely cover all cases. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:17:44 +00:00
link = e1000_has_link(adapter);
if ((netif_carrier_ok(netdev)) && link)
goto link_up;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (link) {
if (!netif_carrier_ok(netdev)) {
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u32 ctrl;
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
bool txb2b = true;
e1000: test link state conclusively e1000 was using one particular way to detect link, but with the advent of some of the newer hardware designs using SERDES connections, tests for link must completely cover all cases. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:17:44 +00:00
/* update snapshot of PHY registers on LSC */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_get_speed_and_duplex(hw,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
&adapter->link_speed,
&adapter->link_duplex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ctrl = er32(CTRL);
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
pr_info("%s NIC Link is Up %d Mbps %s, "
"Flow Control: %s\n",
netdev->name,
adapter->link_speed,
adapter->link_duplex == FULL_DUPLEX ?
"Full Duplex" : "Half Duplex",
((ctrl & E1000_CTRL_TFCE) && (ctrl &
E1000_CTRL_RFCE)) ? "RX/TX" : ((ctrl &
E1000_CTRL_RFCE) ? "RX" : ((ctrl &
E1000_CTRL_TFCE) ? "TX" : "None")));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: do not modify tx_queue_len on link speed change Previously the driver tweaked txqueuelen to avoid false Tx hang reports seen at half duplex. This had the effect of overriding user set values on link change/reset. Testing shows that adjusting only the timeout factor is sufficient to prevent Tx hang reports at half duplex. This patch removes all instances of tx_queue_len in the driver. Based on e1000e patch by Franco Fichtner <franco@lastsummer.de> CC: Franco Fichtner <franco@lastsummer.de> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-03-26 11:25:58 +00:00
/* adjust timeout factor according to speed/duplex */
[PATCH] e1000: Fix TX queue length based on link speed 10/100 speeds seem to have some problems reporting false tx timeouts especially at half duplex. Fixed by using a timeout factor to attempt to mitigate the false timeouts. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:53 +00:00
adapter->tx_timeout_factor = 1;
e1000: Fix network problems when forced at 100Mb/s and to fix TSO when forced at 100Mb/s Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:19:30 +00:00
switch (adapter->link_speed) {
case SPEED_10:
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
txb2b = false;
e1000: test link state conclusively e1000 was using one particular way to detect link, but with the advent of some of the newer hardware designs using SERDES connections, tests for link must completely cover all cases. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:17:44 +00:00
adapter->tx_timeout_factor = 16;
e1000: Fix network problems when forced at 100Mb/s and to fix TSO when forced at 100Mb/s Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:19:30 +00:00
break;
case SPEED_100:
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
txb2b = false;
e1000: Fix network problems when forced at 100Mb/s and to fix TSO when forced at 100Mb/s Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:19:30 +00:00
/* maybe add some timeout factor ? */
break;
}
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
/* enable transmits in the hardware */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
tctl = er32(TCTL);
e1000: Fix network problems when forced at 100Mb/s and to fix TSO when forced at 100Mb/s Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:19:30 +00:00
tctl |= E1000_TCTL_EN;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(TCTL, tctl);
[PATCH] e1000: Fix TX queue length based on link speed 10/100 speeds seem to have some problems reporting false tx timeouts especially at half duplex. Fixed by using a timeout factor to attempt to mitigate the false timeouts. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:53 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
netif_carrier_on(netdev);
e1000: stop timers at appropriate times there were some hotplug cases that made timers still run after the driver had been removed, make sure to stop all the timers and not allow racy reschedules. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:17:23 +00:00
if (!test_bit(__E1000_DOWN, &adapter->flags))
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
schedule_delayed_work(&adapter->phy_info_task,
2 * HZ);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
adapter->smartspeed = 0;
}
} else {
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (netif_carrier_ok(netdev)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
adapter->link_speed = 0;
adapter->link_duplex = 0;
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
pr_info("%s NIC Link is Down\n",
netdev->name);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
netif_carrier_off(netdev);
e1000: stop timers at appropriate times there were some hotplug cases that made timers still run after the driver had been removed, make sure to stop all the timers and not allow racy reschedules. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:17:23 +00:00
if (!test_bit(__E1000_DOWN, &adapter->flags))
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
schedule_delayed_work(&adapter->phy_info_task,
2 * HZ);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000_smartspeed(adapter);
}
e1000: test link state conclusively e1000 was using one particular way to detect link, but with the advent of some of the newer hardware designs using SERDES connections, tests for link must completely cover all cases. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:17:44 +00:00
link_up:
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_update_stats(adapter);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
hw->tx_packet_delta = adapter->stats.tpt - adapter->tpt_old;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
adapter->tpt_old = adapter->stats.tpt;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
hw->collision_delta = adapter->stats.colc - adapter->colc_old;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
adapter->colc_old = adapter->stats.colc;
adapter->gorcl = adapter->stats.gorcl - adapter->gorcl_old;
adapter->gorcl_old = adapter->stats.gorcl;
adapter->gotcl = adapter->stats.gotcl - adapter->gotcl_old;
adapter->gotcl_old = adapter->stats.gotcl;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_update_adaptive(hw);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[PATCH] e1000: Fix adapter structure and prepare for multique fix Fix adapter structure to handle multiple queues and prepping the driver for full multiple queue support, some changes are ifdef'd our unless you define CONFIG_E1000_MQ. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:39 +00:00
if (!netif_carrier_ok(netdev)) {
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
if (E1000_DESC_UNUSED(txdr) + 1 < txdr->count) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* We've lost link, so the controller stops DMA,
* but we've got queued Tx work that's never going
* to get done, so reset controller to flush Tx.
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* (Do the reset outside of interrupt context).
*/
e1000: Fixed the following issues with ESB2 (requires ESB2 support): - Add restriction for ESB2 to MTU size <=9216 - Removed FIFO errors which were not being used - Fixed issues with loopback - Power management change for saving state and config space - WA to disable recieves and reset device on link loss. Reset needed to be done outside the interrupt context - modified existing tx_timeout_task Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:21:24 +00:00
adapter->tx_timeout_count++;
schedule_work(&adapter->reset_task);
e1000: convert to private mutex from rtnl The e1000 driver when running with lockdep could run into some possible deadlocks between the work items acquiring rtnl and the rtnl lock being acquired before work items were cancelled. Use a private mutex to make sure lock ordering isn't violated. The private mutex is only used to protect areas not generally covered by the rtnl lock already. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:51 +00:00
/* exit immediately since reset is imminent */
goto unlock;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
}
e1000/e1000e: implement a simple interrupt moderation Back before e1000-7.3.20, the e1000 driver had a simple algorithm that managed interrupt moderation. The driver was updated in 7.3.20 to have the new "adaptive" interrupt moderation but we have customer requests to redeploy the old way as an option. This patch adds the old functionality back. The new functionality can be enabled via module parameter or at runtime via ethtool. Module parameter: (InterruptThrottleRate=4) to use this new moderation method. Ethtool method: ethtool -C ethX rx-usecs 4 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-05-04 22:26:03 +00:00
/* Simple mode for Interrupt Throttle Rate (ITR) */
if (hw->mac_type >= e1000_82540 && adapter->itr_setting == 4) {
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* Symmetric Tx/Rx gets a reduced ITR=2000;
e1000/e1000e: implement a simple interrupt moderation Back before e1000-7.3.20, the e1000 driver had a simple algorithm that managed interrupt moderation. The driver was updated in 7.3.20 to have the new "adaptive" interrupt moderation but we have customer requests to redeploy the old way as an option. This patch adds the old functionality back. The new functionality can be enabled via module parameter or at runtime via ethtool. Module parameter: (InterruptThrottleRate=4) to use this new moderation method. Ethtool method: ethtool -C ethX rx-usecs 4 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-05-04 22:26:03 +00:00
* Total asymmetrical Tx or Rx gets ITR=8000;
* everyone else is between 2000-8000.
*/
u32 goc = (adapter->gotcl + adapter->gorcl) / 10000;
u32 dif = (adapter->gotcl > adapter->gorcl ?
adapter->gotcl - adapter->gorcl :
adapter->gorcl - adapter->gotcl) / 10000;
u32 itr = goc > 0 ? (dif * 6000 / goc + 2000) : 8000;
ew32(ITR, 1000000000 / (itr * 256));
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Cause software interrupt to ensure rx ring is cleaned */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(ICS, E1000_ICS_RXDMT0);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
/* Force detection of hung controller every watchdog period */
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
adapter->detect_tx_hung = true;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
/* Reschedule the task */
e1000: stop timers at appropriate times there were some hotplug cases that made timers still run after the driver had been removed, make sure to stop all the timers and not allow racy reschedules. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:17:23 +00:00
if (!test_bit(__E1000_DOWN, &adapter->flags))
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
schedule_delayed_work(&adapter->watchdog_task, 2 * HZ);
e1000: convert to private mutex from rtnl The e1000 driver when running with lockdep could run into some possible deadlocks between the work items acquiring rtnl and the rtnl lock being acquired before work items were cancelled. Use a private mutex to make sure lock ordering isn't violated. The private mutex is only used to protect areas not generally covered by the rtnl lock already. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:51 +00:00
unlock:
mutex_unlock(&adapter->mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
enum latency_range {
lowest_latency = 0,
low_latency = 1,
bulk_latency = 2,
latency_invalid = 255
};
/**
* e1000_update_itr - update the dynamic ITR value based on statistics
e1000: two workarounds were incomplete, fix them 1) 82544 does not need last_tx_tso workaround, it interferes with the 82544 workaround too 2) 82544 hang workaround was using the address of the page struct instead of the physical address as its "workaround decider" not sure how that ever worked Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:18:41 +00:00
* @adapter: pointer to adapter
* @itr_setting: current adapter->itr
* @packets: the number of packets during this measurement interval
* @bytes: the number of bytes during this measurement interval
*
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
* Stores a new ITR value based on packets and byte
* counts during the last interrupt. The advantage of per interrupt
* computation is faster updates and more accurate ITR for the current
* traffic pattern. Constants in this function were computed
* based on theoretical maximum wire speed and thresholds were set based
* on testing data as well as attempting to minimize response time
* while increasing bulk throughput.
* this functionality is controlled by the InterruptThrottleRate module
* parameter (see e1000_param.c)
**/
static unsigned int e1000_update_itr(struct e1000_adapter *adapter,
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
u16 itr_setting, int packets, int bytes)
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
{
unsigned int retval = itr_setting;
struct e1000_hw *hw = &adapter->hw;
if (unlikely(hw->mac_type < e1000_82540))
goto update_itr_done;
if (packets == 0)
goto update_itr_done;
switch (itr_setting) {
case lowest_latency:
[PATCH] e1000: dynamic itr: take TSO and jumbo into account The dynamic interrupt rate control patches omitted proper counting for jumbo's and TSO resulting in suboptimal interrupt mitigation strategies. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:30:44 +00:00
/* jumbo frames get bulk treatment*/
if (bytes/packets > 8000)
retval = bulk_latency;
else if ((packets < 5) && (bytes > 512))
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
retval = low_latency;
break;
case low_latency: /* 50 usec aka 20000 ints/s */
if (bytes > 10000) {
[PATCH] e1000: dynamic itr: take TSO and jumbo into account The dynamic interrupt rate control patches omitted proper counting for jumbo's and TSO resulting in suboptimal interrupt mitigation strategies. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:30:44 +00:00
/* jumbo frames need bulk latency setting */
if (bytes/packets > 8000)
retval = bulk_latency;
else if ((packets < 10) || ((bytes/packets) > 1200))
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
retval = bulk_latency;
else if ((packets > 35))
retval = lowest_latency;
[PATCH] e1000: dynamic itr: take TSO and jumbo into account The dynamic interrupt rate control patches omitted proper counting for jumbo's and TSO resulting in suboptimal interrupt mitigation strategies. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:30:44 +00:00
} else if (bytes/packets > 2000)
retval = bulk_latency;
else if (packets <= 2 && bytes < 512)
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
retval = lowest_latency;
break;
case bulk_latency: /* 250 usec aka 4000 ints/s */
if (bytes > 25000) {
if (packets > 35)
retval = low_latency;
[PATCH] e1000: dynamic itr: take TSO and jumbo into account The dynamic interrupt rate control patches omitted proper counting for jumbo's and TSO resulting in suboptimal interrupt mitigation strategies. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:30:44 +00:00
} else if (bytes < 6000) {
retval = low_latency;
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
}
break;
}
update_itr_done:
return retval;
}
static void e1000_set_itr(struct e1000_adapter *adapter)
{
struct e1000_hw *hw = &adapter->hw;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u16 current_itr;
u32 new_itr = adapter->itr;
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
if (unlikely(hw->mac_type < e1000_82540))
return;
/* for non-gigabit speeds, just fix the interrupt rate at 4000 */
if (unlikely(adapter->link_speed != SPEED_1000)) {
current_itr = 0;
new_itr = 4000;
goto set_itr_now;
}
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
adapter->tx_itr = e1000_update_itr(adapter, adapter->tx_itr,
adapter->total_tx_packets,
adapter->total_tx_bytes);
[PATCH] e1000: dynamic itr: take TSO and jumbo into account The dynamic interrupt rate control patches omitted proper counting for jumbo's and TSO resulting in suboptimal interrupt mitigation strategies. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:30:44 +00:00
/* conservative mode (itr 3) eliminates the lowest_latency setting */
if (adapter->itr_setting == 3 && adapter->tx_itr == lowest_latency)
adapter->tx_itr = low_latency;
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
adapter->rx_itr = e1000_update_itr(adapter, adapter->rx_itr,
adapter->total_rx_packets,
adapter->total_rx_bytes);
[PATCH] e1000: dynamic itr: take TSO and jumbo into account The dynamic interrupt rate control patches omitted proper counting for jumbo's and TSO resulting in suboptimal interrupt mitigation strategies. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 09:30:44 +00:00
/* conservative mode (itr 3) eliminates the lowest_latency setting */
if (adapter->itr_setting == 3 && adapter->rx_itr == lowest_latency)
adapter->rx_itr = low_latency;
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
current_itr = max(adapter->rx_itr, adapter->tx_itr);
switch (current_itr) {
/* counts and packets in update_itr are dependent on these numbers */
case lowest_latency:
new_itr = 70000;
break;
case low_latency:
new_itr = 20000; /* aka hwitr = ~200 */
break;
case bulk_latency:
new_itr = 4000;
break;
default:
break;
}
set_itr_now:
if (new_itr != adapter->itr) {
/* this attempts to bias the interrupt rate towards Bulk
* by adding intermediate steps when interrupt rate is
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* increasing
*/
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
new_itr = new_itr > adapter->itr ?
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
min(adapter->itr + (new_itr >> 2), new_itr) :
new_itr;
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
adapter->itr = new_itr;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(ITR, 1000000000 / (new_itr * 256));
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
}
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#define E1000_TX_FLAGS_CSUM 0x00000001
#define E1000_TX_FLAGS_VLAN 0x00000002
#define E1000_TX_FLAGS_TSO 0x00000004
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
#define E1000_TX_FLAGS_IPV4 0x00000008
e1000: Support sending custom Ethernet CRC. Good for testing the RX logic for bad CRC handling. Signed-off-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-11 15:40:01 +00:00
#define E1000_TX_FLAGS_NO_FCS 0x00000010
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#define E1000_TX_FLAGS_VLAN_MASK 0xffff0000
#define E1000_TX_FLAGS_VLAN_SHIFT 16
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static int e1000_tso(struct e1000_adapter *adapter,
struct e1000_tx_ring *tx_ring, struct sk_buff *skb)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct e1000_context_desc *context_desc;
[PATCH] e1000: General Fixes These fixes update the TX and RX ring structures. Prepare driver for up-coming fixes. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:25 +00:00
struct e1000_buffer *buffer_info;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
unsigned int i;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u32 cmd_length = 0;
u16 ipcse = 0, tucse, mss;
u8 ipcss, ipcso, tucss, tucso, hdr_len;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
int err;
[NET] gso: Add skb_is_gso This patch adds the wrapper function skb_is_gso which can be used instead of directly testing skb_shinfo(skb)->gso_size. This makes things a little nicer and allows us to change the primary key for indicating whether an skb is GSO (if we ever want to do that). Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-07-08 20:34:32 +00:00
if (skb_is_gso(skb)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (skb_header_cloned(skb)) {
err = pskb_expand_head(skb, 0, 0, GFP_ATOMIC);
if (err)
return err;
}
[TCP]: Introduce tcp_hdrlen() and tcp_optlen() The ip_hdrlen() buddy, created to reduce the number of skb->h.th-> uses and to avoid the longer, open coded equivalent. Ditched a no-op in bnx2 in the process. I wonder if we should have a BUG_ON(skb->h.th->doff < 5) in tcp_optlen()... Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-19 00:43:48 +00:00
hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb);
[NET]: Merge TSO/UFO fields in sk_buff Having separate fields in sk_buff for TSO/UFO (tso_size/ufo_size) is not going to scale if we add any more segmentation methods (e.g., DCCP). So let's merge them. They were used to tell the protocol of a packet. This function has been subsumed by the new gso_type field. This is essentially a set of netdev feature bits (shifted by 16 bits) that are required to process a specific skb. As such it's easy to tell whether a given device can process a GSO skb: you just have to and the gso_type field and the netdev's features field. I've made gso_type a conjunction. The idea is that you have a base type (e.g., SKB_GSO_TCPV4) that can be modified further to support new features. For example, if we add a hardware TSO type that supports ECN, they would declare NETIF_F_TSO | NETIF_F_TSO_ECN. All TSO packets with CWR set would have a gso_type of SKB_GSO_TCPV4 | SKB_GSO_TCPV4_ECN while all other TSO packets would be SKB_GSO_TCPV4. This means that only the CWR packets need to be emulated in software. Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-22 09:40:14 +00:00
mss = skb_shinfo(skb)->gso_size;
[PATCH] e1000: endian fixes Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> Cc: "Ronciak, John" <john.ronciak@intel.com> Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Acked-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-05-23 21:52:21 +00:00
if (skb->protocol == htons(ETH_P_IP)) {
[SK_BUFF]: Introduce ip_hdr(), remove skb->nh.iph Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-21 05:47:35 +00:00
struct iphdr *iph = ip_hdr(skb);
iph->tot_len = 0;
iph->check = 0;
[SK_BUFF]: Introduce tcp_hdr(), remove skb->h.th Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-11 04:04:22 +00:00
tcp_hdr(skb)->check = ~csum_tcpudp_magic(iph->saddr,
iph->daddr, 0,
IPPROTO_TCP,
0);
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
cmd_length = E1000_TXD_CMD_IP;
[SK_BUFF]: Introduce skb_transport_offset() For the quite common 'skb->h.raw - skb->data' sequence. Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-26 00:55:53 +00:00
ipcse = skb_transport_offset(skb) - 1;
e1000: Same cosmetic fix as earlier sent out for IPV4. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-16 18:28:45 +00:00
} else if (skb->protocol == htons(ETH_P_IPV6)) {
[SK_BUFF]: Introduce ipv6_hdr(), remove skb->nh.ipv6h Now the skb->nh union has just one member, .raw, i.e. it is just like the skb->mac union, strange, no? I'm just leaving it like that till the transport layer is done with, when we'll rename skb->mac.raw to skb->mac_header (or ->mac_header_offset?), ditto for ->{h,nh}. Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-26 00:54:47 +00:00
ipv6_hdr(skb)->payload_len = 0;
[SK_BUFF]: Introduce tcp_hdr(), remove skb->h.th Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-11 04:04:22 +00:00
tcp_hdr(skb)->check =
[SK_BUFF]: Introduce ipv6_hdr(), remove skb->nh.ipv6h Now the skb->nh union has just one member, .raw, i.e. it is just like the skb->mac union, strange, no? I'm just leaving it like that till the transport layer is done with, when we'll rename skb->mac.raw to skb->mac_header (or ->mac_header_offset?), ditto for ->{h,nh}. Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-26 00:54:47 +00:00
~csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
&ipv6_hdr(skb)->daddr,
0, IPPROTO_TCP, 0);
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
ipcse = 0;
}
[SK_BUFF]: Introduce skb_network_offset() For the quite common 'skb->nh.raw - skb->data' sequence. Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-11 01:16:10 +00:00
ipcss = skb_network_offset(skb);
[SK_BUFF]: Introduce ip_hdr(), remove skb->nh.iph Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-21 05:47:35 +00:00
ipcso = (void *)&(ip_hdr(skb)->check) - (void *)skb->data;
[SK_BUFF]: Introduce skb_transport_offset() For the quite common 'skb->h.raw - skb->data' sequence. Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-26 00:55:53 +00:00
tucss = skb_transport_offset(skb);
[SK_BUFF]: Introduce tcp_hdr(), remove skb->h.th Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-11 04:04:22 +00:00
tucso = (void *)&(tcp_hdr(skb)->check) - (void *)skb->data;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
tucse = 0;
cmd_length |= (E1000_TXD_CMD_DEXT | E1000_TXD_CMD_TSE |
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
E1000_TXD_CMD_TCP | (skb->len - (hdr_len)));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
i = tx_ring->next_to_use;
context_desc = E1000_CONTEXT_DESC(*tx_ring, i);
[PATCH] e1000: General Fixes These fixes update the TX and RX ring structures. Prepare driver for up-coming fixes. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:25 +00:00
buffer_info = &tx_ring->buffer_info[i];
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
context_desc->lower_setup.ip_fields.ipcss = ipcss;
context_desc->lower_setup.ip_fields.ipcso = ipcso;
context_desc->lower_setup.ip_fields.ipcse = cpu_to_le16(ipcse);
context_desc->upper_setup.tcp_fields.tucss = tucss;
context_desc->upper_setup.tcp_fields.tucso = tucso;
context_desc->upper_setup.tcp_fields.tucse = cpu_to_le16(tucse);
context_desc->tcp_seg_setup.fields.mss = cpu_to_le16(mss);
context_desc->tcp_seg_setup.fields.hdr_len = hdr_len;
context_desc->cmd_and_length = cpu_to_le32(cmd_length);
[PATCH] e1000: General Fixes These fixes update the TX and RX ring structures. Prepare driver for up-coming fixes. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:25 +00:00
buffer_info->time_stamp = jiffies;
e1000: Remove unneeded and unwanted memsets This memsetting was added in a paranoid rage debugging TX hangs, but are no longer of importance. We can beef up the performance quite a bit removing them. Make sure to fill in next_to_watch to allow this. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2006-11-01 16:47:53 +00:00
buffer_info->next_to_watch = i;
[PATCH] e1000: General Fixes These fixes update the TX and RX ring structures. Prepare driver for up-coming fixes. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:25 +00:00
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
if (++i == tx_ring->count) i = 0;
tx_ring->next_to_use = i;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
return true;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
return false;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static bool e1000_tx_csum(struct e1000_adapter *adapter,
struct e1000_tx_ring *tx_ring, struct sk_buff *skb)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct e1000_context_desc *context_desc;
[PATCH] e1000: General Fixes These fixes update the TX and RX ring structures. Prepare driver for up-coming fixes. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:25 +00:00
struct e1000_buffer *buffer_info;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
unsigned int i;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u8 css;
e1000: don't generate bad checksums for tcp packets with 0 csum When offloading transmit checksums only, the driver was not correctly configuring the hardware to handle the case of a zero checksum. For UDP the correct behavior is to leave it alone, but for tcp the checksum must be changed from 0x0000 to 0xFFFF. The hardware takes care of this case but only if it is told the packet is tcp. same patch as e1000e Signed-off-by: Dave Graham <david.graham@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-09 21:29:26 +00:00
u32 cmd_len = E1000_TXD_CMD_DEXT;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: don't generate bad checksums for tcp packets with 0 csum When offloading transmit checksums only, the driver was not correctly configuring the hardware to handle the case of a zero checksum. For UDP the correct behavior is to leave it alone, but for tcp the checksum must be changed from 0x0000 to 0xFFFF. The hardware takes care of this case but only if it is told the packet is tcp. same patch as e1000e Signed-off-by: Dave Graham <david.graham@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-09 21:29:26 +00:00
if (skb->ip_summed != CHECKSUM_PARTIAL)
return false;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: don't generate bad checksums for tcp packets with 0 csum When offloading transmit checksums only, the driver was not correctly configuring the hardware to handle the case of a zero checksum. For UDP the correct behavior is to leave it alone, but for tcp the checksum must be changed from 0x0000 to 0xFFFF. The hardware takes care of this case but only if it is told the packet is tcp. same patch as e1000e Signed-off-by: Dave Graham <david.graham@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-09 21:29:26 +00:00
switch (skb->protocol) {
net: replace uses of __constant_{endian} Base versions handle constant folding now. Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-02-01 08:45:17 +00:00
case cpu_to_be16(ETH_P_IP):
e1000: don't generate bad checksums for tcp packets with 0 csum When offloading transmit checksums only, the driver was not correctly configuring the hardware to handle the case of a zero checksum. For UDP the correct behavior is to leave it alone, but for tcp the checksum must be changed from 0x0000 to 0xFFFF. The hardware takes care of this case but only if it is told the packet is tcp. same patch as e1000e Signed-off-by: Dave Graham <david.graham@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-09 21:29:26 +00:00
if (ip_hdr(skb)->protocol == IPPROTO_TCP)
cmd_len |= E1000_TXD_CMD_TCP;
break;
net: replace uses of __constant_{endian} Base versions handle constant folding now. Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-02-01 08:45:17 +00:00
case cpu_to_be16(ETH_P_IPV6):
e1000: don't generate bad checksums for tcp packets with 0 csum When offloading transmit checksums only, the driver was not correctly configuring the hardware to handle the case of a zero checksum. For UDP the correct behavior is to leave it alone, but for tcp the checksum must be changed from 0x0000 to 0xFFFF. The hardware takes care of this case but only if it is told the packet is tcp. same patch as e1000e Signed-off-by: Dave Graham <david.graham@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-09 21:29:26 +00:00
/* XXX not handling all IPV6 headers */
if (ipv6_hdr(skb)->nexthdr == IPPROTO_TCP)
cmd_len |= E1000_TXD_CMD_TCP;
break;
default:
if (unlikely(net_ratelimit()))
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_warn(drv, "checksum_partial proto=%x!\n",
skb->protocol);
e1000: don't generate bad checksums for tcp packets with 0 csum When offloading transmit checksums only, the driver was not correctly configuring the hardware to handle the case of a zero checksum. For UDP the correct behavior is to leave it alone, but for tcp the checksum must be changed from 0x0000 to 0xFFFF. The hardware takes care of this case but only if it is told the packet is tcp. same patch as e1000e Signed-off-by: Dave Graham <david.graham@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-09 21:29:26 +00:00
break;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
net: Fix drivers advertising HW_CSUM feature to use csum_start Some drivers are using skb_transport_offset(skb) instead of skb->csum_start for NETIF_F_HW_CSUM offload. This does not matter now, but if someone implements checksumming of encapsulated packets then this will break silently. TSO output paths are left as they are, since they are for IP+TCP only (might be worth converting though). Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-12-14 15:24:08 +00:00
css = skb_checksum_start_offset(skb);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: don't generate bad checksums for tcp packets with 0 csum When offloading transmit checksums only, the driver was not correctly configuring the hardware to handle the case of a zero checksum. For UDP the correct behavior is to leave it alone, but for tcp the checksum must be changed from 0x0000 to 0xFFFF. The hardware takes care of this case but only if it is told the packet is tcp. same patch as e1000e Signed-off-by: Dave Graham <david.graham@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-09 21:29:26 +00:00
i = tx_ring->next_to_use;
buffer_info = &tx_ring->buffer_info[i];
context_desc = E1000_CONTEXT_DESC(*tx_ring, i);
[PATCH] e1000: General Fixes These fixes update the TX and RX ring structures. Prepare driver for up-coming fixes. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:25 +00:00
e1000: don't generate bad checksums for tcp packets with 0 csum When offloading transmit checksums only, the driver was not correctly configuring the hardware to handle the case of a zero checksum. For UDP the correct behavior is to leave it alone, but for tcp the checksum must be changed from 0x0000 to 0xFFFF. The hardware takes care of this case but only if it is told the packet is tcp. same patch as e1000e Signed-off-by: Dave Graham <david.graham@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-09 21:29:26 +00:00
context_desc->lower_setup.ip_config = 0;
context_desc->upper_setup.tcp_fields.tucss = css;
context_desc->upper_setup.tcp_fields.tucso =
css + skb->csum_offset;
context_desc->upper_setup.tcp_fields.tucse = 0;
context_desc->tcp_seg_setup.data = 0;
context_desc->cmd_and_length = cpu_to_le32(cmd_len);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: don't generate bad checksums for tcp packets with 0 csum When offloading transmit checksums only, the driver was not correctly configuring the hardware to handle the case of a zero checksum. For UDP the correct behavior is to leave it alone, but for tcp the checksum must be changed from 0x0000 to 0xFFFF. The hardware takes care of this case but only if it is told the packet is tcp. same patch as e1000e Signed-off-by: Dave Graham <david.graham@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-09 21:29:26 +00:00
buffer_info->time_stamp = jiffies;
buffer_info->next_to_watch = i;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: don't generate bad checksums for tcp packets with 0 csum When offloading transmit checksums only, the driver was not correctly configuring the hardware to handle the case of a zero checksum. For UDP the correct behavior is to leave it alone, but for tcp the checksum must be changed from 0x0000 to 0xFFFF. The hardware takes care of this case but only if it is told the packet is tcp. same patch as e1000e Signed-off-by: Dave Graham <david.graham@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-09 21:29:26 +00:00
if (unlikely(++i == tx_ring->count)) i = 0;
tx_ring->next_to_use = i;
return true;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
#define E1000_MAX_TXD_PWR 12
#define E1000_MAX_DATA_PER_TXD (1<<E1000_MAX_TXD_PWR)
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static int e1000_tx_map(struct e1000_adapter *adapter,
struct e1000_tx_ring *tx_ring,
struct sk_buff *skb, unsigned int first,
unsigned int max_per_txd, unsigned int nr_frags,
unsigned int mss)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: remove use of skb_dma_map from e1000 driver Remove the use of skb_dma_map from the e1000 driver in order to avoid issues when HW iommu are in use. Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-02 16:46:00 +00:00
struct pci_dev *pdev = adapter->pdev;
e1000: fix tx hang detect logic and address dma mapping issues This patch changes the dma mapping to better support skb_dma_map/skb_dma_unmap and addresses and redefines the tx hang logic to be based off of time stamp instead of if the dma field is populated Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:58:45 +00:00
struct e1000_buffer *buffer_info;
e1000: fix unmap bug This is in reference to the issue shown in kerneloops (search e1000 unmap) The e1000 transmit code was calling pci_unmap_page on dma handles that it might have called pci_map_single on. Same bug as e1000e Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Acked-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-03 00:03:21 +00:00
unsigned int len = skb_headlen(skb);
e1000: remove use of skb_dma_map from e1000 driver Remove the use of skb_dma_map from the e1000 driver in order to avoid issues when HW iommu are in use. Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-02 16:46:00 +00:00
unsigned int offset = 0, size, count = 0, i;
e1000: save skb counts in TX to avoid cache misses Virtual Machines with emulated e1000 network adapter running on Parallels' server were seeing kernel panics due to the e1000 driver dereferencing an unexpected NULL pointer retrieved from buffer_info->skb. The problem has been addressed for the e1000e driver, but not for the e1000. Since the two drivers share similar code in the affected area, a port of the following e1000e driver commit solves the issue for the e1000 driver: commit 9ed318d546a29d7a591dbe648fd1a2efe3be1180 Author: Tom Herbert <therbert@google.com> Date: Wed May 5 14:02:27 2010 +0000 e1000e: save skb counts in TX to avoid cache misses In e1000_tx_map, precompute number of segements and bytecounts which are derived from fields in skb; these are stored in buffer_info. When cleaning tx in e1000_clean_tx_irq use the values in the associated buffer_info for statistics counting, this eliminates cache misses on skb fields. Signed-off-by: Dean Nelson <dnelson@redhat.com> Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-08-25 14:39:24 +00:00
unsigned int f, bytecount, segs;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
i = tx_ring->next_to_use;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
while (len) {
e1000: fix tx hang detect logic and address dma mapping issues This patch changes the dma mapping to better support skb_dma_map/skb_dma_unmap and addresses and redefines the tx hang logic to be based off of time stamp instead of if the dma field is populated Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:58:45 +00:00
buffer_info = &tx_ring->buffer_info[i];
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
size = min(len, max_per_txd);
e1000: Fixes for 8357x - TSO workaround - Fixes eeprom version reporting - Fix loopback test - Fix for WOL Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2005-12-13 05:06:22 +00:00
/* Workaround for Controller erratum --
* descriptor for non-tso packet in a linear SKB that follows a
* tso gets written back prematurely before the data is fully
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* DMA'd to the controller
*/
e1000: Fixes for 8357x - TSO workaround - Fixes eeprom version reporting - Fix loopback test - Fix for WOL Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2005-12-13 05:06:22 +00:00
if (!skb->data_len && tx_ring->last_tx_tso &&
[NET] gso: Add skb_is_gso This patch adds the wrapper function skb_is_gso which can be used instead of directly testing skb_shinfo(skb)->gso_size. This makes things a little nicer and allows us to change the primary key for indicating whether an skb is GSO (if we ever want to do that). Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-07-08 20:34:32 +00:00
!skb_is_gso(skb)) {
net: fix assignment of 0/1 to bool variables. DaveM said: Please, this kind of stuff rots forever and not using bool properly drives me crazy. Joe Perches <joe@perches.com> gave me the spatch script: @@ bool b; @@ -b = 0 +b = false @@ bool b; @@ -b = 1 +b = true I merely installed coccinelle, read the documentation and took credit. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-12-19 13:56:45 +00:00
tx_ring->last_tx_tso = false;
e1000: Fixes for 8357x - TSO workaround - Fixes eeprom version reporting - Fix loopback test - Fix for WOL Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2005-12-13 05:06:22 +00:00
size -= 4;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Workaround for premature desc write-backs
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* in TSO mode. Append 4-byte sentinel desc
*/
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (unlikely(mss && !nr_frags && size == len && size > 8))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
size -= 4;
[PATCH] e1000: Implement a workaround for 82546 errata 10 Implement 82546 errata 10 -- first Tx descriptor cannot have more than 2015 byte of data in it or it could hang the transmitter. Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_main.c net-drivers-2.6/drivers/net/e1000.new/e1000_main.c
2005-04-29 02:41:46 +00:00
/* work-around for errata 10 and it applies
* to all controllers in PCI-X mode
* The fix is to make sure that the first descriptor of a
* packet is smaller than 2048 - 16 - 16 (or 2016) bytes
*/
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (unlikely((hw->bus_type == e1000_bus_type_pcix) &&
[PATCH] e1000: Implement a workaround for 82546 errata 10 Implement 82546 errata 10 -- first Tx descriptor cannot have more than 2015 byte of data in it or it could hang the transmitter. Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_main.c net-drivers-2.6/drivers/net/e1000.new/e1000_main.c
2005-04-29 02:41:46 +00:00
(size > 2015) && count == 0))
size = 2015;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Workaround for potential 82544 hang in PCI-X. Avoid
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* terminating buffers within evenly-aligned dwords.
*/
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (unlikely(adapter->pcix_82544 &&
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
!((unsigned long)(skb->data + offset + size - 1) & 4) &&
size > 4))
size -= 4;
buffer_info->length = size;
e1000: fix tx waking queue after queue stopped during shutdown This fix closes a race where the adapter can be shutting down while hard_start_xmit is being called and interrupts are being handled. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:18:07 +00:00
/* set time_stamp *before* dma to help avoid a possible race */
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
buffer_info->time_stamp = jiffies;
e1000: remove use of skb_dma_map from e1000 driver Remove the use of skb_dma_map from the e1000 driver in order to avoid issues when HW iommu are in use. Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-02 16:46:00 +00:00
buffer_info->mapped_as_page = false;
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
buffer_info->dma = dma_map_single(&pdev->dev,
skb->data + offset,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
size, DMA_TO_DEVICE);
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
if (dma_mapping_error(&pdev->dev, buffer_info->dma))
e1000: remove use of skb_dma_map from e1000 driver Remove the use of skb_dma_map from the e1000 driver in order to avoid issues when HW iommu are in use. Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-02 16:46:00 +00:00
goto dma_error;
e1000: Remove unneeded and unwanted memsets This memsetting was added in a paranoid rage debugging TX hangs, but are no longer of importance. We can beef up the performance quite a bit removing them. Make sure to fill in next_to_watch to allow this. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2006-11-01 16:47:53 +00:00
buffer_info->next_to_watch = i;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
len -= size;
offset += size;
count++;
e1000: fix tx hang detect logic and address dma mapping issues This patch changes the dma mapping to better support skb_dma_map/skb_dma_unmap and addresses and redefines the tx hang logic to be based off of time stamp instead of if the dma field is populated Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:58:45 +00:00
if (len) {
i++;
if (unlikely(i == tx_ring->count))
i = 0;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
for (f = 0; f < nr_frags; f++) {
net: add skb frag size accessors To ease skb->truesize sanitization, its better to be able to localize all references to skb frags size. Define accessors : skb_frag_size() to fetch frag size, and skb_frag_size_{set|add|sub}() to manipulate it. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-10-18 21:00:24 +00:00
const struct skb_frag_struct *frag;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
frag = &skb_shinfo(skb)->frags[f];
net: add skb frag size accessors To ease skb->truesize sanitization, its better to be able to localize all references to skb frags size. Define accessors : skb_frag_size() to fetch frag size, and skb_frag_size_{set|add|sub}() to manipulate it. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-10-18 21:00:24 +00:00
len = skb_frag_size(frag);
intel: convert to SKB paged frag API. Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Bruce Allan <bruce.w.allan@intel.com> Cc: Carolyn Wyborny <carolyn.wyborny@intel.com> Cc: Don Skidmore <donald.c.skidmore@intel.com> Cc: Greg Rose <gregory.v.rose@intel.com> Cc: PJ Waskiewicz <peter.p.waskiewicz.jr@intel.com> Cc: Alex Duyck <alexander.h.duyck@intel.com> Cc: John Ronciak <john.ronciak@intel.com> Cc: e1000-devel@lists.sourceforge.net Cc: netdev@vger.kernel.org Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-08-29 23:18:26 +00:00
offset = 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
while (len) {
intel: convert to SKB paged frag API. Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Bruce Allan <bruce.w.allan@intel.com> Cc: Carolyn Wyborny <carolyn.wyborny@intel.com> Cc: Don Skidmore <donald.c.skidmore@intel.com> Cc: Greg Rose <gregory.v.rose@intel.com> Cc: PJ Waskiewicz <peter.p.waskiewicz.jr@intel.com> Cc: Alex Duyck <alexander.h.duyck@intel.com> Cc: John Ronciak <john.ronciak@intel.com> Cc: e1000-devel@lists.sourceforge.net Cc: netdev@vger.kernel.org Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-08-29 23:18:26 +00:00
unsigned long bufend;
e1000: fix tx hang detect logic and address dma mapping issues This patch changes the dma mapping to better support skb_dma_map/skb_dma_unmap and addresses and redefines the tx hang logic to be based off of time stamp instead of if the dma field is populated Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:58:45 +00:00
i++;
if (unlikely(i == tx_ring->count))
i = 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
buffer_info = &tx_ring->buffer_info[i];
size = min(len, max_per_txd);
/* Workaround for premature desc write-backs
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* in TSO mode. Append 4-byte sentinel desc
*/
if (unlikely(mss && f == (nr_frags-1) &&
size == len && size > 8))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
size -= 4;
/* Workaround for potential 82544 hang in PCI-X.
* Avoid terminating buffers within evenly-aligned
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* dwords.
*/
intel: convert to SKB paged frag API. Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Bruce Allan <bruce.w.allan@intel.com> Cc: Carolyn Wyborny <carolyn.wyborny@intel.com> Cc: Don Skidmore <donald.c.skidmore@intel.com> Cc: Greg Rose <gregory.v.rose@intel.com> Cc: PJ Waskiewicz <peter.p.waskiewicz.jr@intel.com> Cc: Alex Duyck <alexander.h.duyck@intel.com> Cc: John Ronciak <john.ronciak@intel.com> Cc: e1000-devel@lists.sourceforge.net Cc: netdev@vger.kernel.org Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-08-29 23:18:26 +00:00
bufend = (unsigned long)
page_to_phys(skb_frag_page(frag));
bufend += offset + size - 1;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (unlikely(adapter->pcix_82544 &&
intel: convert to SKB paged frag API. Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Bruce Allan <bruce.w.allan@intel.com> Cc: Carolyn Wyborny <carolyn.wyborny@intel.com> Cc: Don Skidmore <donald.c.skidmore@intel.com> Cc: Greg Rose <gregory.v.rose@intel.com> Cc: PJ Waskiewicz <peter.p.waskiewicz.jr@intel.com> Cc: Alex Duyck <alexander.h.duyck@intel.com> Cc: John Ronciak <john.ronciak@intel.com> Cc: e1000-devel@lists.sourceforge.net Cc: netdev@vger.kernel.org Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-08-29 23:18:26 +00:00
!(bufend & 4) &&
size > 4))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
size -= 4;
buffer_info->length = size;
buffer_info->time_stamp = jiffies;
e1000: remove use of skb_dma_map from e1000 driver Remove the use of skb_dma_map from the e1000 driver in order to avoid issues when HW iommu are in use. Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-02 16:46:00 +00:00
buffer_info->mapped_as_page = true;
intel: convert to SKB paged frag API. Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Bruce Allan <bruce.w.allan@intel.com> Cc: Carolyn Wyborny <carolyn.wyborny@intel.com> Cc: Don Skidmore <donald.c.skidmore@intel.com> Cc: Greg Rose <gregory.v.rose@intel.com> Cc: PJ Waskiewicz <peter.p.waskiewicz.jr@intel.com> Cc: Alex Duyck <alexander.h.duyck@intel.com> Cc: John Ronciak <john.ronciak@intel.com> Cc: e1000-devel@lists.sourceforge.net Cc: netdev@vger.kernel.org Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-08-29 23:18:26 +00:00
buffer_info->dma = skb_frag_dma_map(&pdev->dev, frag,
offset, size, DMA_TO_DEVICE);
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
if (dma_mapping_error(&pdev->dev, buffer_info->dma))
e1000: remove use of skb_dma_map from e1000 driver Remove the use of skb_dma_map from the e1000 driver in order to avoid issues when HW iommu are in use. Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-02 16:46:00 +00:00
goto dma_error;
e1000: Remove unneeded and unwanted memsets This memsetting was added in a paranoid rage debugging TX hangs, but are no longer of importance. We can beef up the performance quite a bit removing them. Make sure to fill in next_to_watch to allow this. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2006-11-01 16:47:53 +00:00
buffer_info->next_to_watch = i;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
len -= size;
offset += size;
count++;
}
}
e1000: save skb counts in TX to avoid cache misses Virtual Machines with emulated e1000 network adapter running on Parallels' server were seeing kernel panics due to the e1000 driver dereferencing an unexpected NULL pointer retrieved from buffer_info->skb. The problem has been addressed for the e1000e driver, but not for the e1000. Since the two drivers share similar code in the affected area, a port of the following e1000e driver commit solves the issue for the e1000 driver: commit 9ed318d546a29d7a591dbe648fd1a2efe3be1180 Author: Tom Herbert <therbert@google.com> Date: Wed May 5 14:02:27 2010 +0000 e1000e: save skb counts in TX to avoid cache misses In e1000_tx_map, precompute number of segements and bytecounts which are derived from fields in skb; these are stored in buffer_info. When cleaning tx in e1000_clean_tx_irq use the values in the associated buffer_info for statistics counting, this eliminates cache misses on skb fields. Signed-off-by: Dean Nelson <dnelson@redhat.com> Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-08-25 14:39:24 +00:00
segs = skb_shinfo(skb)->gso_segs ?: 1;
/* multiply data chunks by size of headers */
bytecount = ((segs - 1) * skb_headlen(skb)) + skb->len;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
tx_ring->buffer_info[i].skb = skb;
e1000: save skb counts in TX to avoid cache misses Virtual Machines with emulated e1000 network adapter running on Parallels' server were seeing kernel panics due to the e1000 driver dereferencing an unexpected NULL pointer retrieved from buffer_info->skb. The problem has been addressed for the e1000e driver, but not for the e1000. Since the two drivers share similar code in the affected area, a port of the following e1000e driver commit solves the issue for the e1000 driver: commit 9ed318d546a29d7a591dbe648fd1a2efe3be1180 Author: Tom Herbert <therbert@google.com> Date: Wed May 5 14:02:27 2010 +0000 e1000e: save skb counts in TX to avoid cache misses In e1000_tx_map, precompute number of segements and bytecounts which are derived from fields in skb; these are stored in buffer_info. When cleaning tx in e1000_clean_tx_irq use the values in the associated buffer_info for statistics counting, this eliminates cache misses on skb fields. Signed-off-by: Dean Nelson <dnelson@redhat.com> Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-08-25 14:39:24 +00:00
tx_ring->buffer_info[i].segs = segs;
tx_ring->buffer_info[i].bytecount = bytecount;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
tx_ring->buffer_info[first].next_to_watch = i;
return count;
e1000: remove use of skb_dma_map from e1000 driver Remove the use of skb_dma_map from the e1000 driver in order to avoid issues when HW iommu are in use. Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-02 16:46:00 +00:00
dma_error:
dev_err(&pdev->dev, "TX DMA map failed\n");
buffer_info->dma = 0;
e1000/e1000e/igb/igbvf/ixgb/ixgbe: Fix tests of unsigned in *_tx_map() The variable count and i are unsigned so the (<|>=)0 tests do not work. Signed-off-by: Roel Kluin <roel.kluin@gmail.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-01-19 14:21:45 +00:00
if (count)
e1000: remove use of skb_dma_map from e1000 driver Remove the use of skb_dma_map from the e1000 driver in order to avoid issues when HW iommu are in use. Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-02 16:46:00 +00:00
count--;
e1000/e1000e/igb/igbvf/ixgb/ixgbe: Fix tests of unsigned in *_tx_map() The variable count and i are unsigned so the (<|>=)0 tests do not work. Signed-off-by: Roel Kluin <roel.kluin@gmail.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-01-19 14:21:45 +00:00
while (count--) {
if (i==0)
e1000: remove use of skb_dma_map from e1000 driver Remove the use of skb_dma_map from the e1000 driver in order to avoid issues when HW iommu are in use. Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-02 16:46:00 +00:00
i += tx_ring->count;
e1000/e1000e/igb/igbvf/ixgb/ixgbe: Fix tests of unsigned in *_tx_map() The variable count and i are unsigned so the (<|>=)0 tests do not work. Signed-off-by: Roel Kluin <roel.kluin@gmail.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-01-19 14:21:45 +00:00
i--;
e1000: remove use of skb_dma_map from e1000 driver Remove the use of skb_dma_map from the e1000 driver in order to avoid issues when HW iommu are in use. Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-02 16:46:00 +00:00
buffer_info = &tx_ring->buffer_info[i];
e1000_unmap_and_free_tx_resource(adapter, buffer_info);
}
return 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_tx_queue(struct e1000_adapter *adapter,
struct e1000_tx_ring *tx_ring, int tx_flags,
int count)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct e1000_tx_desc *tx_desc = NULL;
struct e1000_buffer *buffer_info;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u32 txd_upper = 0, txd_lower = E1000_TXD_CMD_IFCS;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
unsigned int i;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (likely(tx_flags & E1000_TX_FLAGS_TSO)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
txd_lower |= E1000_TXD_CMD_DEXT | E1000_TXD_DTYP_D |
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
E1000_TXD_CMD_TSE;
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
txd_upper |= E1000_TXD_POPTS_TXSM << 8;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (likely(tx_flags & E1000_TX_FLAGS_IPV4))
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
txd_upper |= E1000_TXD_POPTS_IXSM << 8;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (likely(tx_flags & E1000_TX_FLAGS_CSUM)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
txd_lower |= E1000_TXD_CMD_DEXT | E1000_TXD_DTYP_D;
txd_upper |= E1000_TXD_POPTS_TXSM << 8;
}
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (unlikely(tx_flags & E1000_TX_FLAGS_VLAN)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
txd_lower |= E1000_TXD_CMD_VLE;
txd_upper |= (tx_flags & E1000_TX_FLAGS_VLAN_MASK);
}
e1000: Support sending custom Ethernet CRC. Good for testing the RX logic for bad CRC handling. Signed-off-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-11 15:40:01 +00:00
if (unlikely(tx_flags & E1000_TX_FLAGS_NO_FCS))
txd_lower &= ~(E1000_TXD_CMD_IFCS);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
i = tx_ring->next_to_use;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
while (count--) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
buffer_info = &tx_ring->buffer_info[i];
tx_desc = E1000_TX_DESC(*tx_ring, i);
tx_desc->buffer_addr = cpu_to_le64(buffer_info->dma);
tx_desc->lower.data =
cpu_to_le32(txd_lower | buffer_info->length);
tx_desc->upper.data = cpu_to_le32(txd_upper);
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (unlikely(++i == tx_ring->count)) i = 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
tx_desc->lower.data |= cpu_to_le32(adapter->txd_cmd);
e1000: Support sending custom Ethernet CRC. Good for testing the RX logic for bad CRC handling. Signed-off-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-11 15:40:01 +00:00
/* txd_cmd re-enables FCS, so we'll re-disable it here as desired. */
if (unlikely(tx_flags & E1000_TX_FLAGS_NO_FCS))
tx_desc->lower.data &= ~(cpu_to_le32(E1000_TXD_CMD_IFCS));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Force memory writes to complete before letting h/w
* know there are new descriptors to fetch. (Only
* applicable for weak-ordered memory model archs,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* such as IA-64).
*/
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
wmb();
tx_ring->next_to_use = i;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
writel(i, hw->hw_addr + tx_ring->tdt);
e1000: add mmiowb() for IA64 to sync tail writes IA64 SMP systems were seeing TX issues with multiple cpu's attempting to write tail registers unordered. This mmiowb() fixes the issue. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2006-11-01 16:47:42 +00:00
/* we need this if more than one processor can write to our tail
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* at a time, it synchronizes IO on IA64/Altix systems
*/
e1000: add mmiowb() for IA64 to sync tail writes IA64 SMP systems were seeing TX issues with multiple cpu's attempting to write tail registers unordered. This mmiowb() fixes the issue. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2006-11-01 16:47:42 +00:00
mmiowb();
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
drivers/net/ethernet: Fix non-kernel-doc comments with kernel-doc start markers Convert doxygen (or similar) formatted comments to kernel-doc or unformatted comment. Delete a few that are content-free. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-10 10:56:59 +00:00
/* 82547 workaround to avoid controller hang in half-duplex environment.
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* The workaround is to avoid queuing a large packet that would span
* the internal Tx FIFO ring boundary by notifying the stack to resend
* the packet at a later time. This gives the Tx FIFO an opportunity to
* flush all packets. When that occurs, we reset the Tx FIFO pointers
* to the beginning of the Tx FIFO.
drivers/net/ethernet: Fix non-kernel-doc comments with kernel-doc start markers Convert doxygen (or similar) formatted comments to kernel-doc or unformatted comment. Delete a few that are content-free. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-10 10:56:59 +00:00
*/
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#define E1000_FIFO_HDR 0x10
#define E1000_82547_PAD_LEN 0x3E0
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static int e1000_82547_fifo_workaround(struct e1000_adapter *adapter,
struct sk_buff *skb)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u32 fifo_space = adapter->tx_fifo_size - adapter->tx_fifo_head;
u32 skb_fifo_len = skb->len + E1000_FIFO_HDR;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: ROUND_UP macro cleanup in drivers/net/e1000 E1000_ROUNDUP macro cleanup, use ALIGN Signed-off-by: Milind Arun Choudhary <milindchoudhary@gmail.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-04-27 20:55:29 +00:00
skb_fifo_len = ALIGN(skb_fifo_len, E1000_FIFO_HDR);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (adapter->link_duplex != HALF_DUPLEX)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto no_fifo_stall_required;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (atomic_read(&adapter->tx_fifo_stall))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return 1;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (skb_fifo_len >= (E1000_82547_PAD_LEN + fifo_space)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
atomic_set(&adapter->tx_fifo_stall, 1);
return 1;
}
no_fifo_stall_required:
adapter->tx_fifo_head += skb_fifo_len;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (adapter->tx_fifo_head >= adapter->tx_fifo_size)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
adapter->tx_fifo_head -= adapter->tx_fifo_size;
return 0;
}
e1000: Maybe stop TX if not enough free descriptors Cc: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:48 +00:00
static int __e1000_maybe_stop_tx(struct net_device *netdev, int size)
{
struct e1000_adapter *adapter = netdev_priv(netdev);
struct e1000_tx_ring *tx_ring = adapter->tx_ring;
netif_stop_queue(netdev);
/* Herbert's original patch had:
* smp_mb__after_netif_stop_queue();
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* but since that doesn't exist yet, just open code it.
*/
e1000: Maybe stop TX if not enough free descriptors Cc: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:48 +00:00
smp_mb();
/* We need to check again in a case another CPU has just
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* made room available.
*/
e1000: Maybe stop TX if not enough free descriptors Cc: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:48 +00:00
if (likely(E1000_DESC_UNUSED(tx_ring) < size))
return -EBUSY;
/* A reprieve! */
netif_start_queue(netdev);
e1000: add queue restart counter Add a netif_wake/start_queue counter to the ethtool statistics to indicated to the user that their transmit ring could be too small for their workload. Signed-off-by: Jesse brandeburg <jesse.brandeburg@intel.com> Cc: Jamal Hadi <hadi@cyberus.ca> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:47:59 +00:00
++adapter->restart_queue;
e1000: Maybe stop TX if not enough free descriptors Cc: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:48 +00:00
return 0;
}
static int e1000_maybe_stop_tx(struct net_device *netdev,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
struct e1000_tx_ring *tx_ring, int size)
e1000: Maybe stop TX if not enough free descriptors Cc: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:48 +00:00
{
if (likely(E1000_DESC_UNUSED(tx_ring) >= size))
return 0;
return __e1000_maybe_stop_tx(netdev, size);
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#define TXD_USE_COUNT(S, X) (((S) >> (X)) + 1 )
intel: convert drivers to netdev_tx_t Get rid of some bogus return wrapping as well. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-08-31 19:50:55 +00:00
static netdev_tx_t e1000_xmit_frame(struct sk_buff *skb,
struct net_device *netdev)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
[PATCH] e1000: Use netdev_priv() to get to netdev->priv Use netdev_priv() to get to netdev->priv - from shemminger@osdl.org Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:41:45 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
struct e1000_tx_ring *tx_ring;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
unsigned int first, max_per_txd = E1000_MAX_DATA_PER_TXD;
unsigned int max_txd_pwr = E1000_MAX_TXD_PWR;
unsigned int tx_flags = 0;
drivers: net: use skb_headlen() replaces (skb->len - skb->data_len) occurrences by skb_headlen(skb) Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-14 22:59:40 +00:00
unsigned int len = skb_headlen(skb);
e1000: Simple optimizations in e1000_xmit_frame Some simple optimizations in e1000_xmit_frame. Signed-off-by: Krishna Kumar <krkumar2@in.ibm.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-10-05 21:15:16 +00:00
unsigned int nr_frags;
unsigned int mss;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
int count = 0;
e1000: remove leading and trailing whitespace. Some leading and trailing whitespace made it into the driver code here. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-05-23 20:36:06 +00:00
int tso;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
unsigned int f;
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* This goes back to the question of how to logically map a Tx queue
e1000: Maybe stop TX if not enough free descriptors Cc: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:48 +00:00
* to a flow. Right now, performance is impacted slightly negatively
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* if using multiple Tx queues. If the stack breaks away from a
* single qdisc implementation, we can look at this again.
*/
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
tx_ring = adapter->tx_ring;
e1000: implementation of the multi-queue feature Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:03:23 +00:00
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
if (unlikely(skb->len <= 0)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
dev_kfree_skb_any(skb);
return NETDEV_TX_OK;
}
e1000: Small packets may get corrupted during padding by HW On PCI/PCI-X HW, if packet size is less than ETH_ZLEN, packets may get corrupted during padding by HW. To WA this issue, pad all small packets manually. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-09-15 10:16:57 +00:00
/* On PCI/PCI-X HW, if packet size is less than ETH_ZLEN,
* packets may get corrupted during padding by HW.
* To WA this issue, pad all small packets manually.
*/
if (skb->len < ETH_ZLEN) {
if (skb_pad(skb, ETH_ZLEN - skb->len))
return NETDEV_TX_OK;
skb->len = ETH_ZLEN;
skb_set_tail_pointer(skb, ETH_ZLEN);
}
[NET]: Merge TSO/UFO fields in sk_buff Having separate fields in sk_buff for TSO/UFO (tso_size/ufo_size) is not going to scale if we add any more segmentation methods (e.g., DCCP). So let's merge them. They were used to tell the protocol of a packet. This function has been subsumed by the new gso_type field. This is essentially a set of netdev feature bits (shifted by 16 bits) that are required to process a specific skb. As such it's easy to tell whether a given device can process a GSO skb: you just have to and the gso_type field and the netdev's features field. I've made gso_type a conjunction. The idea is that you have a base type (e.g., SKB_GSO_TCPV4) that can be modified further to support new features. For example, if we add a hardware TSO type that supports ECN, they would declare NETIF_F_TSO | NETIF_F_TSO_ECN. All TSO packets with CWR set would have a gso_type of SKB_GSO_TCPV4 | SKB_GSO_TCPV4_ECN while all other TSO packets would be SKB_GSO_TCPV4. This means that only the CWR packets need to be emulated in software. Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-22 09:40:14 +00:00
mss = skb_shinfo(skb)->gso_size;
e1000: remove leading and trailing whitespace. Some leading and trailing whitespace made it into the driver code here. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-05-23 20:36:06 +00:00
/* The controller does a simple calculation to
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* make sure there is enough room in the FIFO before
* initiating the DMA for each buffer. The calc is:
* 4 = ceil(buffer len/mss). To make sure we don't
* overrun the FIFO, adjust the max buffer len if mss
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* drops.
*/
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (mss) {
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u8 hdr_len;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
max_per_txd = min(mss << 2, max_per_txd);
max_txd_pwr = fls(max_per_txd) - 1;
[PATCH] e1000: Fix TSO Fixed the TSO workaround for 82571/2 controllers. Fixed TSO issue where a non-tso packet in a linear SKB which followed a TSO packet would get written back prematurely. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:23 +00:00
[TCP]: Introduce tcp_hdrlen() and tcp_optlen() The ip_hdrlen() buddy, created to reduce the number of skb->h.th-> uses and to avoid the longer, open coded equivalent. Ditched a no-op in bnx2 in the process. I wonder if we should have a BUG_ON(skb->h.th->doff < 5) in tcp_optlen()... Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-19 00:43:48 +00:00
hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb);
e1000: Simple optimizations in e1000_xmit_frame Some simple optimizations in e1000_xmit_frame. Signed-off-by: Krishna Kumar <krkumar2@in.ibm.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-10-05 21:15:16 +00:00
if (skb->data_len && hdr_len == len) {
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
switch (hw->mac_type) {
e1000: Add 82573 controller support to TSO fix Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:20:17 +00:00
unsigned int pull_size;
e1000: Do not truncate TSO TCP header with 82544 workaround The e1000 driver has a workaround for 82544 on PCI-X where if the terminating byte of a buffer is at addresses 0-3 mod 8, then 4 bytes are shaved off it and defered to a new segment. This is due to an erratum that could otherwise cause TX hangs. Unfortunately this breaks TSO because it may cause the TCP header to be split over two segments which itself causes TX hangs. The solution is to pull 4 bytes of data up from the next segment rather than pushing 4 bytes off. This ensures the TCP header remains in one piece and works around the PCI-X hang. This patch is based on one from Jesse Brandeburg. This bug has been trigered by both CONFIG_DEBUG_SLAB as well as Xen. Note that the only reason we don't see this normally is because the TCP stack starts writing from the end, i.e., it writes the TCP header first then slaps on the IP header, etc. So the end of the TCP header (skb->tail - 1 here) is always aligned correctly. Had we made the start of the IP header (e.g., IPv6) 8-byte aligned instead, this would happen for normal TCP traffic as well. Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> -- Visit Openswan at http://www.openswan.org/ Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt -- Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-16 01:04:33 +00:00
case e1000_82544:
/* Make sure we have room to chop off 4 bytes,
* and that the end alignment will work out to
* this hardware's requirements
* NOTE: this is a TSO only workaround
* if end byte alignment not correct move us
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* into the next dword
*/
if ((unsigned long)(skb_tail_pointer(skb) - 1)
& 4)
e1000: Do not truncate TSO TCP header with 82544 workaround The e1000 driver has a workaround for 82544 on PCI-X where if the terminating byte of a buffer is at addresses 0-3 mod 8, then 4 bytes are shaved off it and defered to a new segment. This is due to an erratum that could otherwise cause TX hangs. Unfortunately this breaks TSO because it may cause the TCP header to be split over two segments which itself causes TX hangs. The solution is to pull 4 bytes of data up from the next segment rather than pushing 4 bytes off. This ensures the TCP header remains in one piece and works around the PCI-X hang. This patch is based on one from Jesse Brandeburg. This bug has been trigered by both CONFIG_DEBUG_SLAB as well as Xen. Note that the only reason we don't see this normally is because the TCP stack starts writing from the end, i.e., it writes the TCP header first then slaps on the IP header, etc. So the end of the TCP header (skb->tail - 1 here) is always aligned correctly. Had we made the start of the IP header (e.g., IPv6) 8-byte aligned instead, this would happen for normal TCP traffic as well. Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> -- Visit Openswan at http://www.openswan.org/ Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt -- Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-16 01:04:33 +00:00
break;
/* fall through */
e1000: Add 82573 controller support to TSO fix Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:20:17 +00:00
pull_size = min((unsigned int)4, skb->data_len);
if (!__pskb_pull_tail(skb, pull_size)) {
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(drv, "__pskb_pull_tail "
"failed.\n");
e1000: Add 82573 controller support to TSO fix Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:20:17 +00:00
dev_kfree_skb_any(skb);
Merge branch 'upstream-fixes'
2006-03-11 18:35:31 +00:00
return NETDEV_TX_OK;
e1000: Add 82573 controller support to TSO fix Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:20:17 +00:00
}
drivers: net: use skb_headlen() replaces (skb->len - skb->data_len) occurrences by skb_headlen(skb) Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-14 22:59:40 +00:00
len = skb_headlen(skb);
e1000: Add 82573 controller support to TSO fix Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:20:17 +00:00
break;
default:
/* do nothing */
break;
[PATCH] e1000: Fix __pskb_pull_tail Fixed by moving code to correct location (for 82572 and 82571 controllers). Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:07 +00:00
}
[PATCH] e1000: Fix TSO Fixed the TSO workaround for 82571/2 controllers. Fixed TSO issue where a non-tso packet in a linear SKB which followed a TSO packet would get written back prematurely. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:23 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
[PATCH] e1000: Fix TSO Fixed the TSO workaround for 82571/2 controllers. Fixed TSO issue where a non-tso packet in a linear SKB which followed a TSO packet would get written back prematurely. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:23 +00:00
/* reserve a descriptor for the offload context */
[NET]: Replace CHECKSUM_HW by CHECKSUM_PARTIAL/CHECKSUM_COMPLETE Replace CHECKSUM_HW by CHECKSUM_PARTIAL (for outgoing packets, whose checksum still needs to be completed) and CHECKSUM_COMPLETE (for incoming packets, device supplied full checksum). Patch originally from Herbert Xu, updated by myself for 2.6.18-rc3. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-08-29 23:44:56 +00:00
if ((mss) || (skb->ip_summed == CHECKSUM_PARTIAL))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
count++;
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
count++;
e1000: Fixes for 8357x - TSO workaround - Fixes eeprom version reporting - Fix loopback test - Fix for WOL Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2005-12-13 05:06:22 +00:00
/* Controller Erratum workaround */
[NET] gso: Add skb_is_gso This patch adds the wrapper function skb_is_gso which can be used instead of directly testing skb_shinfo(skb)->gso_size. This makes things a little nicer and allows us to change the primary key for indicating whether an skb is GSO (if we ever want to do that). Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-07-08 20:34:32 +00:00
if (!skb->data_len && tx_ring->last_tx_tso && !skb_is_gso(skb))
e1000: Fixes for 8357x - TSO workaround - Fixes eeprom version reporting - Fix loopback test - Fix for WOL Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2005-12-13 05:06:22 +00:00
count++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
count += TXD_USE_COUNT(len, max_txd_pwr);
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (adapter->pcix_82544)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
count++;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
/* work-around for errata 10 and it applies to all controllers
[PATCH] e1000: Implement a workaround for 82546 errata 10 Implement 82546 errata 10 -- first Tx descriptor cannot have more than 2015 byte of data in it or it could hang the transmitter. Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_main.c net-drivers-2.6/drivers/net/e1000.new/e1000_main.c
2005-04-29 02:41:46 +00:00
* in PCI-X mode, so add one more descriptor to the count
*/
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (unlikely((hw->bus_type == e1000_bus_type_pcix) &&
[PATCH] e1000: Implement a workaround for 82546 errata 10 Implement 82546 errata 10 -- first Tx descriptor cannot have more than 2015 byte of data in it or it could hang the transmitter. Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_main.c net-drivers-2.6/drivers/net/e1000.new/e1000_main.c
2005-04-29 02:41:46 +00:00
(len > 2015)))
count++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
nr_frags = skb_shinfo(skb)->nr_frags;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
for (f = 0; f < nr_frags; f++)
net: add skb frag size accessors To ease skb->truesize sanitization, its better to be able to localize all references to skb frags size. Define accessors : skb_frag_size() to fetch frag size, and skb_frag_size_{set|add|sub}() to manipulate it. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-10-18 21:00:24 +00:00
count += TXD_USE_COUNT(skb_frag_size(&skb_shinfo(skb)->frags[f]),
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
max_txd_pwr);
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (adapter->pcix_82544)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
count += nr_frags;
/* need: count + 2 desc gap to keep tail from touching
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* head, otherwise try next time
*/
e1000: drop lltx, remove unnecessary lock LLTX is deprecated, don't use it. This completes the removal of LLTX from the Intel Network drivers. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-01-21 22:42:47 +00:00
if (unlikely(e1000_maybe_stop_tx(netdev, tx_ring, count + 2)))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return NETDEV_TX_BUSY;
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
if (unlikely((hw->mac_type == e1000_82547) &&
(e1000_82547_fifo_workaround(adapter, skb)))) {
netif_stop_queue(netdev);
if (!test_bit(__E1000_DOWN, &adapter->flags))
schedule_delayed_work(&adapter->fifo_stall_task, 1);
return NETDEV_TX_BUSY;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
if (vlan_tx_tag_present(skb)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
tx_flags |= E1000_TX_FLAGS_VLAN;
tx_flags |= (vlan_tx_tag_get(skb) << E1000_TX_FLAGS_VLAN_SHIFT);
}
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
first = tx_ring->next_to_use;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
tso = e1000_tso(adapter, tx_ring, skb);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (tso < 0) {
dev_kfree_skb_any(skb);
return NETDEV_TX_OK;
}
e1000: Fixes for 8357x - TSO workaround - Fixes eeprom version reporting - Fix loopback test - Fix for WOL Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2005-12-13 05:06:22 +00:00
if (likely(tso)) {
e1000: two workarounds were incomplete, fix them 1) 82544 does not need last_tx_tso workaround, it interferes with the 82544 workaround too 2) 82544 hang workaround was using the address of the page struct instead of the physical address as its "workaround decider" not sure how that ever worked Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:18:41 +00:00
if (likely(hw->mac_type != e1000_82544))
net: fix assignment of 0/1 to bool variables. DaveM said: Please, this kind of stuff rots forever and not using bool properly drives me crazy. Joe Perches <joe@perches.com> gave me the spatch script: @@ bool b; @@ -b = 0 +b = false @@ bool b; @@ -b = 1 +b = true I merely installed coccinelle, read the documentation and took credit. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-12-19 13:56:45 +00:00
tx_ring->last_tx_tso = true;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
tx_flags |= E1000_TX_FLAGS_TSO;
e1000: Fixes for 8357x - TSO workaround - Fixes eeprom version reporting - Fix loopback test - Fix for WOL Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2005-12-13 05:06:22 +00:00
} else if (likely(e1000_tx_csum(adapter, tx_ring, skb)))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
tx_flags |= E1000_TX_FLAGS_CSUM;
[PATCH] e1000: endian fixes Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> Cc: "Ronciak, John" <john.ronciak@intel.com> Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Acked-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-05-23 21:52:21 +00:00
if (likely(skb->protocol == htons(ETH_P_IP)))
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
tx_flags |= E1000_TX_FLAGS_IPV4;
e1000: Support sending custom Ethernet CRC. Good for testing the RX logic for bad CRC handling. Signed-off-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-11 15:40:01 +00:00
if (unlikely(skb->no_fcs))
tx_flags |= E1000_TX_FLAGS_NO_FCS;
e1000: fix tx hang detect logic and address dma mapping issues This patch changes the dma mapping to better support skb_dma_map/skb_dma_unmap and addresses and redefines the tx hang logic to be based off of time stamp instead of if the dma field is populated Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:58:45 +00:00
count = e1000_tx_map(adapter, tx_ring, skb, first, max_per_txd,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
nr_frags, mss);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: fix tx hang detect logic and address dma mapping issues This patch changes the dma mapping to better support skb_dma_map/skb_dma_unmap and addresses and redefines the tx hang logic to be based off of time stamp instead of if the dma field is populated Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:58:45 +00:00
if (count) {
e1000: add byte queue limits Signed-off-by: Otto Estuardo Solares Cabrera <solca@galileo.edu> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-08-24 08:04:58 +00:00
netdev_sent_queue(netdev, skb->len);
e1000: add transmit timestamping support Signed-off-by: Willem de Bruijn <willemb@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-04-27 09:04:04 +00:00
skb_tx_timestamp(skb);
e1000: fix tx hang detect logic and address dma mapping issues This patch changes the dma mapping to better support skb_dma_map/skb_dma_unmap and addresses and redefines the tx hang logic to be based off of time stamp instead of if the dma field is populated Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:58:45 +00:00
e1000_tx_queue(adapter, tx_ring, tx_flags, count);
/* Make sure there is space in the ring for the next send. */
e1000_maybe_stop_tx(netdev, tx_ring, MAX_SKB_FRAGS + 2);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: fix tx hang detect logic and address dma mapping issues This patch changes the dma mapping to better support skb_dma_map/skb_dma_unmap and addresses and redefines the tx hang logic to be based off of time stamp instead of if the dma field is populated Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:58:45 +00:00
} else {
dev_kfree_skb_any(skb);
tx_ring->buffer_info[first].time_stamp = 0;
tx_ring->next_to_use = first;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return NETDEV_TX_OK;
}
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
#define NUM_REGS 38 /* 1 based count */
static void e1000_regdump(struct e1000_adapter *adapter)
{
struct e1000_hw *hw = &adapter->hw;
u32 regs[NUM_REGS];
u32 *regs_buff = regs;
int i = 0;
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
static const char * const reg_name[] = {
"CTRL", "STATUS",
"RCTL", "RDLEN", "RDH", "RDT", "RDTR",
"TCTL", "TDBAL", "TDBAH", "TDLEN", "TDH", "TDT",
"TIDV", "TXDCTL", "TADV", "TARC0",
"TDBAL1", "TDBAH1", "TDLEN1", "TDH1", "TDT1",
"TXDCTL1", "TARC1",
"CTRL_EXT", "ERT", "RDBAL", "RDBAH",
"TDFH", "TDFT", "TDFHS", "TDFTS", "TDFPC",
"RDFH", "RDFT", "RDFHS", "RDFTS", "RDFPC"
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
};
regs_buff[0] = er32(CTRL);
regs_buff[1] = er32(STATUS);
regs_buff[2] = er32(RCTL);
regs_buff[3] = er32(RDLEN);
regs_buff[4] = er32(RDH);
regs_buff[5] = er32(RDT);
regs_buff[6] = er32(RDTR);
regs_buff[7] = er32(TCTL);
regs_buff[8] = er32(TDBAL);
regs_buff[9] = er32(TDBAH);
regs_buff[10] = er32(TDLEN);
regs_buff[11] = er32(TDH);
regs_buff[12] = er32(TDT);
regs_buff[13] = er32(TIDV);
regs_buff[14] = er32(TXDCTL);
regs_buff[15] = er32(TADV);
regs_buff[16] = er32(TARC0);
regs_buff[17] = er32(TDBAL1);
regs_buff[18] = er32(TDBAH1);
regs_buff[19] = er32(TDLEN1);
regs_buff[20] = er32(TDH1);
regs_buff[21] = er32(TDT1);
regs_buff[22] = er32(TXDCTL1);
regs_buff[23] = er32(TARC1);
regs_buff[24] = er32(CTRL_EXT);
regs_buff[25] = er32(ERT);
regs_buff[26] = er32(RDBAL0);
regs_buff[27] = er32(RDBAH0);
regs_buff[28] = er32(TDFH);
regs_buff[29] = er32(TDFT);
regs_buff[30] = er32(TDFHS);
regs_buff[31] = er32(TDFTS);
regs_buff[32] = er32(TDFPC);
regs_buff[33] = er32(RDFH);
regs_buff[34] = er32(RDFT);
regs_buff[35] = er32(RDFHS);
regs_buff[36] = er32(RDFTS);
regs_buff[37] = er32(RDFPC);
pr_info("Register dump\n");
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
for (i = 0; i < NUM_REGS; i++)
pr_info("%-15s %08x\n", reg_name[i], regs_buff[i]);
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
}
/*
* e1000_dump: Print registers, tx ring and rx ring
*/
static void e1000_dump(struct e1000_adapter *adapter)
{
/* this code doesn't handle multiple rings */
struct e1000_tx_ring *tx_ring = adapter->tx_ring;
struct e1000_rx_ring *rx_ring = adapter->rx_ring;
int i;
if (!netif_msg_hw(adapter))
return;
/* Print Registers */
e1000_regdump(adapter);
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* transmit dump */
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
pr_info("TX Desc ring0 dump\n");
/* Transmit Descriptor Formats - DEXT[29] is 0 (Legacy) or 1 (Extended)
*
* Legacy Transmit Descriptor
* +--------------------------------------------------------------+
* 0 | Buffer Address [63:0] (Reserved on Write Back) |
* +--------------------------------------------------------------+
* 8 | Special | CSS | Status | CMD | CSO | Length |
* +--------------------------------------------------------------+
* 63 48 47 36 35 32 31 24 23 16 15 0
*
* Extended Context Descriptor (DTYP=0x0) for TSO or checksum offload
* 63 48 47 40 39 32 31 16 15 8 7 0
* +----------------------------------------------------------------+
* 0 | TUCSE | TUCS0 | TUCSS | IPCSE | IPCS0 | IPCSS |
* +----------------------------------------------------------------+
* 8 | MSS | HDRLEN | RSV | STA | TUCMD | DTYP | PAYLEN |
* +----------------------------------------------------------------+
* 63 48 47 40 39 36 35 32 31 24 23 20 19 0
*
* Extended Data Descriptor (DTYP=0x1)
* +----------------------------------------------------------------+
* 0 | Buffer Address [63:0] |
* +----------------------------------------------------------------+
* 8 | VLAN tag | POPTS | Rsvd | Status | Command | DTYP | DTALEN |
* +----------------------------------------------------------------+
* 63 48 47 40 39 36 35 32 31 24 23 20 19 0
*/
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
pr_info("Tc[desc] [Ce CoCsIpceCoS] [MssHlRSCm0Plen] [bi->dma ] leng ntw timestmp bi->skb\n");
pr_info("Td[desc] [address 63:0 ] [VlaPoRSCm1Dlen] [bi->dma ] leng ntw timestmp bi->skb\n");
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
if (!netif_msg_tx_done(adapter))
goto rx_ring_summary;
for (i = 0; tx_ring->desc && (i < tx_ring->count); i++) {
struct e1000_tx_desc *tx_desc = E1000_TX_DESC(*tx_ring, i);
struct e1000_buffer *buffer_info = &tx_ring->buffer_info[i];
e1000: Silence sparse warnings by correcting type Silence sparse warnings shown below: ... drivers/net/ethernet/intel/e1000/e1000_main.c:3435:17: warning: cast to restricted __le64 drivers/net/ethernet/intel/e1000/e1000_main.c:3435:17: warning: cast to restricted __le64 ... Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-03-25 17:49:25 +00:00
struct my_u { __le64 a; __le64 b; };
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
struct my_u *u = (struct my_u *)tx_desc;
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
const char *type;
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
if (i == tx_ring->next_to_use && i == tx_ring->next_to_clean)
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
type = "NTC/U";
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
else if (i == tx_ring->next_to_use)
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
type = "NTU";
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
else if (i == tx_ring->next_to_clean)
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
type = "NTC";
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
else
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
type = "";
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
pr_info("T%c[0x%03X] %016llX %016llX %016llX %04X %3X %016llX %p %s\n",
((le64_to_cpu(u->b) & (1<<20)) ? 'd' : 'c'), i,
le64_to_cpu(u->a), le64_to_cpu(u->b),
(u64)buffer_info->dma, buffer_info->length,
buffer_info->next_to_watch,
(u64)buffer_info->time_stamp, buffer_info->skb, type);
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
}
rx_ring_summary:
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* receive dump */
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
pr_info("\nRX Desc ring dump\n");
/* Legacy Receive Descriptor Format
*
* +-----------------------------------------------------+
* | Buffer Address [63:0] |
* +-----------------------------------------------------+
* | VLAN Tag | Errors | Status 0 | Packet csum | Length |
* +-----------------------------------------------------+
* 63 48 47 40 39 32 31 16 15 0
*/
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
pr_info("R[desc] [address 63:0 ] [vl er S cks ln] [bi->dma ] [bi->skb]\n");
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
if (!netif_msg_rx_status(adapter))
goto exit;
for (i = 0; rx_ring->desc && (i < rx_ring->count); i++) {
struct e1000_rx_desc *rx_desc = E1000_RX_DESC(*rx_ring, i);
struct e1000_buffer *buffer_info = &rx_ring->buffer_info[i];
e1000: Silence sparse warnings by correcting type Silence sparse warnings shown below: ... drivers/net/ethernet/intel/e1000/e1000_main.c:3435:17: warning: cast to restricted __le64 drivers/net/ethernet/intel/e1000/e1000_main.c:3435:17: warning: cast to restricted __le64 ... Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-03-25 17:49:25 +00:00
struct my_u { __le64 a; __le64 b; };
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
struct my_u *u = (struct my_u *)rx_desc;
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
const char *type;
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
if (i == rx_ring->next_to_use)
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
type = "NTU";
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
else if (i == rx_ring->next_to_clean)
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
type = "NTC";
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
else
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
type = "";
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
pr_info("R[0x%03X] %016llX %016llX %016llX %p %s\n",
i, le64_to_cpu(u->a), le64_to_cpu(u->b),
(u64)buffer_info->dma, buffer_info->skb, type);
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
} /* for */
/* dump the descriptor caches */
/* rx */
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
pr_info("Rx descriptor cache in 64bit format\n");
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
for (i = 0x6000; i <= 0x63FF ; i += 0x10) {
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
pr_info("R%04X: %08X|%08X %08X|%08X\n",
i,
readl(adapter->hw.hw_addr + i+4),
readl(adapter->hw.hw_addr + i),
readl(adapter->hw.hw_addr + i+12),
readl(adapter->hw.hw_addr + i+8));
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
}
/* tx */
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
pr_info("Tx descriptor cache in 64bit format\n");
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
for (i = 0x7000; i <= 0x73FF ; i += 0x10) {
v2 e1000: Neaten e1000_dump function Use pr_<level> for printk Use temporary instead of multiple pr_conts Coalesce formats. Save a few bytes of object code too: $ size drivers/net/ethernet/intel/e1000/e1000_main.o* text data bss dec hex filename 60507 369 14120 74996 124f4 drivers/net/ethernet/intel/e1000/e1000_main.o.new 60717 369 14176 75262 125fe drivers/net/ethernet/intel/e1000/e1000_main.o.old Removed printing of pktdata. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-10 08:06:36 +00:00
pr_info("T%04X: %08X|%08X %08X|%08X\n",
i,
readl(adapter->hw.hw_addr + i+4),
readl(adapter->hw.hw_addr + i),
readl(adapter->hw.hw_addr + i+12),
readl(adapter->hw.hw_addr + i+8));
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
}
exit:
return;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/**
* e1000_tx_timeout - Respond to a Tx Hang
* @netdev: network interface device structure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_tx_timeout(struct net_device *netdev)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
[PATCH] e1000: Use netdev_priv() to get to netdev->priv Use netdev_priv() to get to netdev->priv - from shemminger@osdl.org Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:41:45 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Do the reset outside of interrupt context */
e1000: Fixed the following issues with ESB2 (requires ESB2 support): - Add restriction for ESB2 to MTU size <=9216 - Removed FIFO errors which were not being used - Fixed issues with loopback - Power management change for saving state and config space - WA to disable recieves and reset device on link loss. Reset needed to be done outside the interrupt context - modified existing tx_timeout_task Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:21:24 +00:00
adapter->tx_timeout_count++;
schedule_work(&adapter->reset_task);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_reset_task(struct work_struct *work)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
WorkStruct: Pass the work_struct pointer instead of context data Pass the work_struct pointer to the work function rather than context data. The work function can use container_of() to work out the data. For the cases where the container of the work_struct may go away the moment the pending bit is cleared, it is made possible to defer the release of the structure by deferring the clearing of the pending bit. To make this work, an extra flag is introduced into the management side of the work_struct. This governs auto-release of the structure upon execution. Ordinarily, the work queue executor would release the work_struct for further scheduling or deallocation by clearing the pending bit prior to jumping to the work function. This means that, unless the driver makes some guarantee itself that the work_struct won't go away, the work function may not access anything else in the work_struct or its container lest they be deallocated.. This is a problem if the auxiliary data is taken away (as done by the last patch). However, if the pending bit is *not* cleared before jumping to the work function, then the work function *may* access the work_struct and its container with no problems. But then the work function must itself release the work_struct by calling work_release(). In most cases, automatic release is fine, so this is the default. Special initiators exist for the non-auto-release case (ending in _NAR). Signed-Off-By: David Howells <dhowells@redhat.com>
2006-11-22 14:55:48 +00:00
struct e1000_adapter *adapter =
container_of(work, struct e1000_adapter, reset_task);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: convert to private mutex from rtnl The e1000 driver when running with lockdep could run into some possible deadlocks between the work items acquiring rtnl and the rtnl lock being acquired before work items were cancelled. Use a private mutex to make sure lock ordering isn't violated. The private mutex is only used to protect areas not generally covered by the rtnl lock already. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:51 +00:00
if (test_bit(__E1000_DOWN, &adapter->flags))
return;
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
e_err(drv, "Reset adapter\n");
e1000: fix occasional panic on unload Net drivers in general have an issue where timers fired by mod_timer or work threads with schedule_work are running outside of the rtnl_lock. With no other lock protection these routines are vulnerable to races with driver unload or reset paths. The longer term solution to this might be a redesign with safer locks being taken in the driver to guarantee no reentrance, but for now a safe and effective fix is to take the rtnl_lock in these routines. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-22 18:22:42 +00:00
e1000_reinit_safe(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/**
* e1000_get_stats - Get System Network Statistics
* @netdev: network interface device structure
*
* Returns the address of the device statistics structure.
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
* The statistics are actually updated from the watchdog.
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static struct net_device_stats *e1000_get_stats(struct net_device *netdev)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
[PATCH] Fix e1000 stats Updated the e1000_stats structure and removed mpx for rx_errors and rx_dropped. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:35 +00:00
/* only return the current stats */
e1000: Use the instance of net_device_stats from net_device. Since net_device has an instance of net_device_stats, we can remove the instance of this from the adapter structure. Signed-off-by: Ajit Khaparde <ajitk@serverengines.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-07 02:42:23 +00:00
return &netdev->stats;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/**
* e1000_change_mtu - Change the Maximum Transfer Unit
* @netdev: network interface device structure
* @new_mtu: new value for maximum frame size
*
* Returns 0 on success, negative on failure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static int e1000_change_mtu(struct net_device *netdev, int new_mtu)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
[PATCH] e1000: Use netdev_priv() to get to netdev->priv Use netdev_priv() to get to netdev->priv - from shemminger@osdl.org Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:41:45 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
int max_frame = new_mtu + ENET_HEADER_SIZE + ETHERNET_FCS_SIZE;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if ((max_frame < MINIMUM_ETHERNET_FRAME_SIZE) ||
(max_frame > MAX_JUMBO_FRAME_SIZE)) {
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(probe, "Invalid MTU setting\n");
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return -EINVAL;
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[PATCH] e1000: Fix Desc. Rings and Jumbo Frames This patch contains two fixes. The first fix is to the tx and rx descriptor rings clean up process. The second fix is to jumbo frames, which cleans up the code logic and removes most of the fifo related limitations on jumbo frames. This is because the driver code now supports splitting a packet across multiple descriptors. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:55 +00:00
/* Adapter-specific max frame size limits. */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
switch (hw->mac_type) {
e1000: Buffer optimizations for small MTU Remove multi-descriptor support from legacy recieve path Add memory usage efficiency by using more correct size descriptors for small MTU sizes and optimize using LPE for <= 1522 byte frame sizes An extra performance fix that effected our TCP window size growth as a receiver. Set our initial buffer to be 128 bytes instead of 256 to prevent over-socket charge when truesize is computed in the stack. old way: truesize = 256 + l1 = 256 + 1460 = 1716 new way: truesize = 128 + l1 = 128 + 1460 = 1588 The magic value that we can't cross is 1648. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-04-15 02:05:18 +00:00
case e1000_undefined ... e1000_82542_rev2_1:
e1000: fix flow control thresholds when testing the jumbo frames with pages patch, the stats would show rx_missed errors (dropped packets) even when connected to a link partner with flow control enabled. this indicates that for this MTU (9000) the flow control thresholds are not adjusting correctly. In fact, before this change, the FCRTH (xoff threshold) is 36864 when the fifo size is only 40000, with 9000 byte MTU. fix it so that we at least have room for one frame after we send the xoff. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:45:01 +00:00
if (max_frame > (ETH_FRAME_LEN + ETH_FCS_LEN)) {
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(probe, "Jumbo Frames not supported.\n");
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
return -EINVAL;
}
[PATCH] e1000: Fix Desc. Rings and Jumbo Frames This patch contains two fixes. The first fix is to the tx and rx descriptor rings clean up process. The second fix is to jumbo frames, which cleans up the code logic and removes most of the fifo related limitations on jumbo frames. This is because the driver code now supports splitting a packet across multiple descriptors. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:55 +00:00
break;
default:
/* Capable of supporting up to MAX_JUMBO_FRAME_SIZE limit. */
break;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: remove races when changing mtu this patch fixes a bug that occurs when routing packets and simultaneously changing the mtu. the rx_buffer_len variable is used during the rx cleanup and if that changes on the fly without stopping traffic bad things happen Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:19:02 +00:00
while (test_and_set_bit(__E1000_RESETTING, &adapter->flags))
msleep(1);
/* e1000_down has a dependency on max_frame_size */
hw->max_frame_size = max_frame;
if (netif_running(netdev))
e1000_down(adapter);
[E1000]: Convert to netdev_alloc_skb Signed-off-by: David S. Miller <davem@davemloft.net>
2006-08-01 05:39:40 +00:00
/* NOTE: netdev_alloc_skb reserves 16 bytes, and typically NET_IP_ALIGN
e1000: Buffer optimizations for small MTU Remove multi-descriptor support from legacy recieve path Add memory usage efficiency by using more correct size descriptors for small MTU sizes and optimize using LPE for <= 1522 byte frame sizes An extra performance fix that effected our TCP window size growth as a receiver. Set our initial buffer to be 128 bytes instead of 256 to prevent over-socket charge when truesize is computed in the stack. old way: truesize = 256 + l1 = 256 + 1460 = 1716 new way: truesize = 128 + l1 = 128 + 1460 = 1588 The magic value that we can't cross is 1648. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-04-15 02:05:18 +00:00
* means we reserve 2 more, this pushes us to allocate from the next
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
* larger slab size.
* i.e. RXBUFFER_2048 --> size-4096 slab
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* however with the new *_jumbo_rx* routines, jumbo receives will use
* fragmented skbs
*/
e1000: Buffer optimizations for small MTU Remove multi-descriptor support from legacy recieve path Add memory usage efficiency by using more correct size descriptors for small MTU sizes and optimize using LPE for <= 1522 byte frame sizes An extra performance fix that effected our TCP window size growth as a receiver. Set our initial buffer to be 128 bytes instead of 256 to prevent over-socket charge when truesize is computed in the stack. old way: truesize = 256 + l1 = 256 + 1460 = 1716 new way: truesize = 128 + l1 = 128 + 1460 = 1588 The magic value that we can't cross is 1648. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-04-15 02:05:18 +00:00
e1000/e1000e: don't use small hardware rx buffers When testing the "e1000: enhance frame fragment detection" (and e1000e) patches we found some bugs with reducing the MTU size. The 1024 byte descriptor used with the 1000 mtu test also (re) introduced the (originally) reported bug, and causes us to need the e1000_clean_tx_irq "enhance frame fragment detection" fix. So what has occured here is that 2.6.32 is only vulnerable for mtu < 1500 due to the jumbo specific routines in both e1000 and e1000e. So, 2.6.32 needs the 2kB buffer len fix for those smaller MTUs, but is not vulnerable to the original issue reported. It has been pointed out that this vulnerability needs to be patched in older kernels that don't have the e1000 jumbo routine. Without the jumbo routines, we need the "enhance frame fragment detection" fix the e1000, old e1000e is only vulnerable for < 1500 mtu, and needs a similar fix. We split the patches up to provide easy backport paths. There is only a slight bit of extra code when this fix and the original "enhance frame fragment detection" fixes are applied, so please apply both, even though it is a bit of overkill. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-01-22 22:56:16 +00:00
if (max_frame <= E1000_RXBUFFER_2048)
e1000: Buffer optimizations for small MTU Remove multi-descriptor support from legacy recieve path Add memory usage efficiency by using more correct size descriptors for small MTU sizes and optimize using LPE for <= 1522 byte frame sizes An extra performance fix that effected our TCP window size growth as a receiver. Set our initial buffer to be 128 bytes instead of 256 to prevent over-socket charge when truesize is computed in the stack. old way: truesize = 256 + l1 = 256 + 1460 = 1716 new way: truesize = 128 + l1 = 128 + 1460 = 1588 The magic value that we can't cross is 1648. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-04-15 02:05:18 +00:00
adapter->rx_buffer_len = E1000_RXBUFFER_2048;
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
else
#if (PAGE_SIZE >= E1000_RXBUFFER_16384)
e1000: Buffer optimizations for small MTU Remove multi-descriptor support from legacy recieve path Add memory usage efficiency by using more correct size descriptors for small MTU sizes and optimize using LPE for <= 1522 byte frame sizes An extra performance fix that effected our TCP window size growth as a receiver. Set our initial buffer to be 128 bytes instead of 256 to prevent over-socket charge when truesize is computed in the stack. old way: truesize = 256 + l1 = 256 + 1460 = 1716 new way: truesize = 128 + l1 = 128 + 1460 = 1588 The magic value that we can't cross is 1648. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-04-15 02:05:18 +00:00
adapter->rx_buffer_len = E1000_RXBUFFER_16384;
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
#elif (PAGE_SIZE >= E1000_RXBUFFER_4096)
adapter->rx_buffer_len = PAGE_SIZE;
#endif
e1000: Buffer optimizations for small MTU Remove multi-descriptor support from legacy recieve path Add memory usage efficiency by using more correct size descriptors for small MTU sizes and optimize using LPE for <= 1522 byte frame sizes An extra performance fix that effected our TCP window size growth as a receiver. Set our initial buffer to be 128 bytes instead of 256 to prevent over-socket charge when truesize is computed in the stack. old way: truesize = 256 + l1 = 256 + 1460 = 1716 new way: truesize = 128 + l1 = 128 + 1460 = 1588 The magic value that we can't cross is 1648. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-04-15 02:05:18 +00:00
/* adjust allocation if LPE protects us, and we aren't using SBP */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (!hw->tbi_compatibility_on &&
e1000: fix flow control thresholds when testing the jumbo frames with pages patch, the stats would show rx_missed errors (dropped packets) even when connected to a link partner with flow control enabled. this indicates that for this MTU (9000) the flow control thresholds are not adjusting correctly. In fact, before this change, the FCRTH (xoff threshold) is 36864 when the fifo size is only 40000, with 9000 byte MTU. fix it so that we at least have room for one frame after we send the xoff. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:45:01 +00:00
((max_frame == (ETH_FRAME_LEN + ETH_FCS_LEN)) ||
e1000: Buffer optimizations for small MTU Remove multi-descriptor support from legacy recieve path Add memory usage efficiency by using more correct size descriptors for small MTU sizes and optimize using LPE for <= 1522 byte frame sizes An extra performance fix that effected our TCP window size growth as a receiver. Set our initial buffer to be 128 bytes instead of 256 to prevent over-socket charge when truesize is computed in the stack. old way: truesize = 256 + l1 = 256 + 1460 = 1716 new way: truesize = 128 + l1 = 128 + 1460 = 1588 The magic value that we can't cross is 1648. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-04-15 02:05:18 +00:00
(max_frame == MAXIMUM_ETHERNET_VLAN_SIZE)))
adapter->rx_buffer_len = MAXIMUM_ETHERNET_VLAN_SIZE;
[PATCH] e1000: Fix Desc. Rings and Jumbo Frames This patch contains two fixes. The first fix is to the tx and rx descriptor rings clean up process. The second fix is to jumbo frames, which cleans up the code logic and removes most of the fifo related limitations on jumbo frames. This is because the driver code now supports splitting a packet across multiple descriptors. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:55 +00:00
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
pr_info("%s changing MTU from %d to %d\n",
netdev->name, netdev->mtu, new_mtu);
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
netdev->mtu = new_mtu;
e1000: rework driver hardware reset locking After studying the driver mac reset code it was found that there were multiple race conditions possible to reset the unit twice or bring it e1000_up() double. This fixes all occurences where the driver needs to reset the mac. We also remove irq requesting/releasing into _open and _close so that while the device is _up we will never touch the irq's. This fixes the double free irq bug that people saw. To make sure that the watchdog task doesn't cause another race we let it run as a non-scheduled task. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:28 +00:00
if (netif_running(netdev))
e1000: remove races when changing mtu this patch fixes a bug that occurs when routing packets and simultaneously changing the mtu. the rx_buffer_len variable is used during the rx cleanup and if that changes on the fly without stopping traffic bad things happen Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:19:02 +00:00
e1000_up(adapter);
else
e1000_reset(adapter);
clear_bit(__E1000_RESETTING, &adapter->flags);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return 0;
}
/**
* e1000_update_stats - Update the board statistics counters
* @adapter: board private structure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
void e1000_update_stats(struct e1000_adapter *adapter)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use the instance of net_device_stats from net_device. Since net_device has an instance of net_device_stats, we can remove the instance of this from the adapter structure. Signed-off-by: Ajit Khaparde <ajitk@serverengines.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-07 02:42:23 +00:00
struct net_device *netdev = adapter->netdev;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct e1000_hw *hw = &adapter->hw;
[PATCH] e1000: prevent statistics from getting garbled during reset If a PCI bus error/fault triggers a PCI bus reset, attempts to get the ethernet packet count statistics from the hardware will fail, returning garbage data upstream. This patch skips statistics data collection if the PCI device is not on the bus. This patch presumes that an earlier patch, [PATCH] PCI Error Recovery: e1000 network device driver has already been applied. Signed-off-by: Linas Vepstas <linas@austin.ibm.com> Cc: John Ronciak <john.ronciak@intel.com> Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Cc: Jeff Garzik <jeff@garzik.org> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-06-09 05:19:44 +00:00
struct pci_dev *pdev = adapter->pdev;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
unsigned long flags;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u16 phy_tmp;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#define PHY_IDLE_ERROR_COUNT_MASK 0x00FF
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* Prevent stats update while adapter is being reset, or if the pci
[PATCH] e1000: prevent statistics from getting garbled during reset If a PCI bus error/fault triggers a PCI bus reset, attempts to get the ethernet packet count statistics from the hardware will fail, returning garbage data upstream. This patch skips statistics data collection if the PCI device is not on the bus. This patch presumes that an earlier patch, [PATCH] PCI Error Recovery: e1000 network device driver has already been applied. Signed-off-by: Linas Vepstas <linas@austin.ibm.com> Cc: John Ronciak <john.ronciak@intel.com> Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Cc: Jeff Garzik <jeff@garzik.org> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-06-09 05:19:44 +00:00
* connection is down.
*/
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
if (adapter->link_speed == 0)
[PATCH] e1000: prevent statistics from getting garbled during reset If a PCI bus error/fault triggers a PCI bus reset, attempts to get the ethernet packet count statistics from the hardware will fail, returning garbage data upstream. This patch skips statistics data collection if the PCI device is not on the bus. This patch presumes that an earlier patch, [PATCH] PCI Error Recovery: e1000 network device driver has already been applied. Signed-off-by: Linas Vepstas <linas@austin.ibm.com> Cc: John Ronciak <john.ronciak@intel.com> Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Cc: Jeff Garzik <jeff@garzik.org> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-06-09 05:19:44 +00:00
return;
PCI: Use newly defined PCI channel offline routine Use newly minted routine to access the PCI channel state. Signed-off-by: Linas Vepstas <linas@linas.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-12-13 00:29:15 +00:00
if (pci_channel_offline(pdev))
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
return;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
spin_lock_irqsave(&adapter->stats_lock, flags);
fix typo about TBI in e1000 comment Signed-off-by: Masatake YAMATO <jet@gyve.org> Signed-off-by: Adrian Bunk <bunk@kernel.org>
2007-10-20 01:06:37 +00:00
/* these counters are modified from e1000_tbi_adjust_stats,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* called from the interrupt context, so they must only
* be written while holding adapter->stats_lock
*/
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
adapter->stats.crcerrs += er32(CRCERRS);
adapter->stats.gprc += er32(GPRC);
adapter->stats.gorcl += er32(GORCL);
adapter->stats.gorch += er32(GORCH);
adapter->stats.bprc += er32(BPRC);
adapter->stats.mprc += er32(MPRC);
adapter->stats.roc += er32(ROC);
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
adapter->stats.prc64 += er32(PRC64);
adapter->stats.prc127 += er32(PRC127);
adapter->stats.prc255 += er32(PRC255);
adapter->stats.prc511 += er32(PRC511);
adapter->stats.prc1023 += er32(PRC1023);
adapter->stats.prc1522 += er32(PRC1522);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
adapter->stats.symerrs += er32(SYMERRS);
adapter->stats.mpc += er32(MPC);
adapter->stats.scc += er32(SCC);
adapter->stats.ecol += er32(ECOL);
adapter->stats.mcc += er32(MCC);
adapter->stats.latecol += er32(LATECOL);
adapter->stats.dc += er32(DC);
adapter->stats.sec += er32(SEC);
adapter->stats.rlec += er32(RLEC);
adapter->stats.xonrxc += er32(XONRXC);
adapter->stats.xontxc += er32(XONTXC);
adapter->stats.xoffrxc += er32(XOFFRXC);
adapter->stats.xofftxc += er32(XOFFTXC);
adapter->stats.fcruc += er32(FCRUC);
adapter->stats.gptc += er32(GPTC);
adapter->stats.gotcl += er32(GOTCL);
adapter->stats.gotch += er32(GOTCH);
adapter->stats.rnbc += er32(RNBC);
adapter->stats.ruc += er32(RUC);
adapter->stats.rfc += er32(RFC);
adapter->stats.rjc += er32(RJC);
adapter->stats.torl += er32(TORL);
adapter->stats.torh += er32(TORH);
adapter->stats.totl += er32(TOTL);
adapter->stats.toth += er32(TOTH);
adapter->stats.tpr += er32(TPR);
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
adapter->stats.ptc64 += er32(PTC64);
adapter->stats.ptc127 += er32(PTC127);
adapter->stats.ptc255 += er32(PTC255);
adapter->stats.ptc511 += er32(PTC511);
adapter->stats.ptc1023 += er32(PTC1023);
adapter->stats.ptc1522 += er32(PTC1522);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
adapter->stats.mptc += er32(MPTC);
adapter->stats.bptc += er32(BPTC);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* used for adaptive IFS */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
hw->tx_packet_delta = er32(TPT);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
adapter->stats.tpt += hw->tx_packet_delta;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
hw->collision_delta = er32(COLC);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
adapter->stats.colc += hw->collision_delta;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (hw->mac_type >= e1000_82543) {
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
adapter->stats.algnerrc += er32(ALGNERRC);
adapter->stats.rxerrc += er32(RXERRC);
adapter->stats.tncrs += er32(TNCRS);
adapter->stats.cexterr += er32(CEXTERR);
adapter->stats.tsctc += er32(TSCTC);
adapter->stats.tsctfc += er32(TSCTFC);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/* Fill out the OS statistics structure */
e1000: Use the instance of net_device_stats from net_device. Since net_device has an instance of net_device_stats, we can remove the instance of this from the adapter structure. Signed-off-by: Ajit Khaparde <ajitk@serverengines.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-07 02:42:23 +00:00
netdev->stats.multicast = adapter->stats.mprc;
netdev->stats.collisions = adapter->stats.colc;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Rx Errors */
e1000: Fixed the following issues with ESB2 (requires ESB2 support): - Add restriction for ESB2 to MTU size <=9216 - Removed FIFO errors which were not being used - Fixed issues with loopback - Power management change for saving state and config space - WA to disable recieves and reset device on link loss. Reset needed to be done outside the interrupt context - modified existing tx_timeout_task Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:21:24 +00:00
/* RLEC on some newer hardware can be incorrect so build
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* our own version based on RUC and ROC
*/
e1000: Use the instance of net_device_stats from net_device. Since net_device has an instance of net_device_stats, we can remove the instance of this from the adapter structure. Signed-off-by: Ajit Khaparde <ajitk@serverengines.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-07 02:42:23 +00:00
netdev->stats.rx_errors = adapter->stats.rxerrc +
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
adapter->stats.crcerrs + adapter->stats.algnerrc +
e1000: Fixed the following issues with ESB2 (requires ESB2 support): - Add restriction for ESB2 to MTU size <=9216 - Removed FIFO errors which were not being used - Fixed issues with loopback - Power management change for saving state and config space - WA to disable recieves and reset device on link loss. Reset needed to be done outside the interrupt context - modified existing tx_timeout_task Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:21:24 +00:00
adapter->stats.ruc + adapter->stats.roc +
adapter->stats.cexterr;
e1000: add multicast stats counters Add 4 multicast and broadcast hardware counters (rx/tx), and eliminate as many non-hardware counters as possible. Signed-off-by: Mitch Williams <mitch.a.williams@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:37 +00:00
adapter->stats.rlerrc = adapter->stats.ruc + adapter->stats.roc;
e1000: Use the instance of net_device_stats from net_device. Since net_device has an instance of net_device_stats, we can remove the instance of this from the adapter structure. Signed-off-by: Ajit Khaparde <ajitk@serverengines.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-07 02:42:23 +00:00
netdev->stats.rx_length_errors = adapter->stats.rlerrc;
netdev->stats.rx_crc_errors = adapter->stats.crcerrs;
netdev->stats.rx_frame_errors = adapter->stats.algnerrc;
netdev->stats.rx_missed_errors = adapter->stats.mpc;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Tx Errors */
e1000: add multicast stats counters Add 4 multicast and broadcast hardware counters (rx/tx), and eliminate as many non-hardware counters as possible. Signed-off-by: Mitch Williams <mitch.a.williams@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:37 +00:00
adapter->stats.txerrc = adapter->stats.ecol + adapter->stats.latecol;
e1000: Use the instance of net_device_stats from net_device. Since net_device has an instance of net_device_stats, we can remove the instance of this from the adapter structure. Signed-off-by: Ajit Khaparde <ajitk@serverengines.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-07 02:42:23 +00:00
netdev->stats.tx_errors = adapter->stats.txerrc;
netdev->stats.tx_aborted_errors = adapter->stats.ecol;
netdev->stats.tx_window_errors = adapter->stats.latecol;
netdev->stats.tx_carrier_errors = adapter->stats.tncrs;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (hw->bad_tx_carr_stats_fd &&
e1000: omit stats for broken counter in 82543 The 82543 chip does not count tx_carrier_errors properly in FD mode; report zeros instead of garbage. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:41:15 +00:00
adapter->link_duplex == FULL_DUPLEX) {
e1000: Use the instance of net_device_stats from net_device. Since net_device has an instance of net_device_stats, we can remove the instance of this from the adapter structure. Signed-off-by: Ajit Khaparde <ajitk@serverengines.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-07 02:42:23 +00:00
netdev->stats.tx_carrier_errors = 0;
e1000: omit stats for broken counter in 82543 The 82543 chip does not count tx_carrier_errors properly in FD mode; report zeros instead of garbage. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:41:15 +00:00
adapter->stats.tncrs = 0;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Tx Dropped needs to be maintained elsewhere */
/* Phy Stats */
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (hw->media_type == e1000_media_type_copper) {
if ((adapter->link_speed == SPEED_1000) &&
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
(!e1000_read_phy_reg(hw, PHY_1000T_STATUS, &phy_tmp))) {
phy_tmp &= PHY_IDLE_ERROR_COUNT_MASK;
adapter->phy_stats.idle_errors += phy_tmp;
}
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if ((hw->mac_type <= e1000_82546) &&
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
(hw->phy_type == e1000_phy_m88) &&
!e1000_read_phy_reg(hw, M88E1000_RX_ERR_CNTR, &phy_tmp))
adapter->phy_stats.receive_errors += phy_tmp;
}
e1000: 3 new driver stats for managability testing Add 3 extra packet redirect counters for tracking purposes to make sure we can test that all packets arrive properly. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 16:16:33 +00:00
/* Management Stats */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (hw->has_smbus) {
adapter->stats.mgptc += er32(MGTPTC);
adapter->stats.mgprc += er32(MGTPRC);
adapter->stats.mgpdc += er32(MGTPDC);
e1000: 3 new driver stats for managability testing Add 3 extra packet redirect counters for tracking purposes to make sure we can test that all packets arrive properly. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 16:16:33 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
spin_unlock_irqrestore(&adapter->stats_lock, flags);
}
e1000: add dynamic generic MSI interrupt routine Add a generic MSI interrupt routine that is IO read-free, speeding up MSI interrupt handling. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:10 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/**
* e1000_intr - Interrupt Handler
* @irq: interrupt number
* @data: pointer to a network interface device structure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static irqreturn_t e1000_intr(int irq, void *data)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct net_device *netdev = data;
[PATCH] e1000: Use netdev_priv() to get to netdev->priv Use netdev_priv() to get to netdev->priv - from shemminger@osdl.org Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:41:45 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
u32 icr = er32(ICR);
e1000: delete non NAPI code from the driver Compile-tested only. Signed-off-by: Francois Romieu <romieu@fr.zoreil.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:38 +00:00
e1000: Avoid unhandled IRQ If hardware asserted an interrupt and driver is down, then there is nothing to do so return IRQ_HANDLED instead of IRQ_NONE. Returning IRQ_NONE in above situation causes screaming IRQ on virtual machines. CC: Andy Gospodarek <gospo@redhat.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-01-13 07:48:13 +00:00
if (unlikely((!icr)))
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
return IRQ_NONE; /* Not our interrupt */
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* we might have caused the interrupt, but the above
e1000: Avoid unhandled IRQ If hardware asserted an interrupt and driver is down, then there is nothing to do so return IRQ_HANDLED instead of IRQ_NONE. Returning IRQ_NONE in above situation causes screaming IRQ on virtual machines. CC: Andy Gospodarek <gospo@redhat.com> Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-01-13 07:48:13 +00:00
* read cleared it, and just in case the driver is
* down there is nothing to do so return handled
*/
if (unlikely(test_bit(__E1000_DOWN, &adapter->flags)))
return IRQ_HANDLED;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (unlikely(icr & (E1000_ICR_RXSEQ | E1000_ICR_LSC))) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
hw->get_link_status = 1;
e1000: driver state fixes (race fix) We were plagued by our interrupt handler posting a watchdog event which could occur when our adapter was going down in case a late packet arrived just before e1000_down() finished. This caused the watchdog timer to start after the NIC was down and keep rescheduling it every N seconds. Once the driver unloaded it would panic. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:54:02 +00:00
/* guard against interrupt when we're going down */
if (!test_bit(__E1000_DOWN, &adapter->flags))
e1000: convert hardware management from timers to threads Thomas Gleixner (tglx) reported that e1000 was delaying for many milliseconds (using mdelay) from inside timer/interrupt context. None of these paths are performance critical and can be moved into threads/work items. This patch implements the work items and the next patch changes the mdelays to msleeps. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Thomas Gleixner <tglx@linutronix.de> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2011-10-05 07:24:41 +00:00
schedule_delayed_work(&adapter->watchdog_task, 1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
/* disable interrupts, without the synchronize_irq bit */
ew32(IMC, ~0);
E1000_WRITE_FLUSH();
net: Remove redundant NAPI functions Following the removal of the unused struct net_device * parameter from the NAPI functions named *netif_rx_* in commit 908a7a1, they are exactly equivalent to the corresponding *napi_* functions and are therefore redundant. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-01-20 00:43:59 +00:00
if (likely(napi_schedule_prep(&adapter->napi))) {
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
adapter->total_tx_bytes = 0;
adapter->total_tx_packets = 0;
adapter->total_rx_bytes = 0;
adapter->total_rx_packets = 0;
net: Remove redundant NAPI functions Following the removal of the unused struct net_device * parameter from the NAPI functions named *netif_rx_* in commit 908a7a1, they are exactly equivalent to the corresponding *napi_* functions and are therefore redundant. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-01-20 00:43:59 +00:00
__napi_schedule(&adapter->napi);
e1000: fix close race with interrupt this is in regards to http://bugzilla.kernel.org/show_bug.cgi?id=12876 where it appears that e1000 can leave its interrupt enabled after exiting the driver. Fix the bug by making the interrupt enable paths more aware of the driver exiting. Thanks to Alan Cox for the poke and initial investigation. CC: Alan Cox <alan@lxorguk.ukuu.org.uk> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:59:22 +00:00
} else {
e1000: whitespace changes, comments, typos Small whitespace changes, comment changes, typo fixes. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:47:30 +00:00
/* this really should not happen! if it does it is basically a
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* bug, but not a hard error, so enable ints and continue
*/
e1000: fix close race with interrupt this is in regards to http://bugzilla.kernel.org/show_bug.cgi?id=12876 where it appears that e1000 can leave its interrupt enabled after exiting the driver. Fix the bug by making the interrupt enable paths more aware of the driver exiting. Thanks to Alan Cox for the poke and initial investigation. CC: Alan Cox <alan@lxorguk.ukuu.org.uk> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:59:22 +00:00
if (!test_bit(__E1000_DOWN, &adapter->flags))
e1000_irq_enable(adapter);
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return IRQ_HANDLED;
}
/**
* e1000_clean - NAPI Rx polling callback
* @adapter: board private structure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static int e1000_clean(struct napi_struct *napi, int budget)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
struct e1000_adapter *adapter = container_of(napi, struct e1000_adapter,
napi);
e1000: drop redunant line of code, cleanup adapter was being assigned twice, also clarified variable name and unwrapped line. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:19:23 +00:00
int tx_clean_complete = 0, work_done = 0;
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
e1000: drop redunant line of code, cleanup adapter was being assigned twice, also clarified variable name and unwrapped line. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:19:23 +00:00
tx_clean_complete = e1000_clean_tx_irq(adapter, &adapter->tx_ring[0]);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
e1000: drop redunant line of code, cleanup adapter was being assigned twice, also clarified variable name and unwrapped line. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:19:23 +00:00
adapter->clean_rx(adapter, &adapter->rx_ring[0], &work_done, budget);
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
e1000: drop redunant line of code, cleanup adapter was being assigned twice, also clarified variable name and unwrapped line. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:19:23 +00:00
if (!tx_clean_complete)
[NET]: Fix TX timeout regression in Intel drivers. This fixes a regression added by changeset 53e52c729cc169db82a6105fac7a166e10c2ec36 ("[NET]: Make ->poll() breakout consistent in Intel ethernet drivers.") As pointed out by Jesse Brandeburg, for three of the drivers edited above there is breakout logic in the *_clean_tx_irq() code to prevent running TX reclaim forever. If this occurs, we have to elide NAPI poll completion or else those TX events will never be serviced. Signed-off-by: David S. Miller <davem@davemloft.net> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2008-01-16 06:43:24 +00:00
work_done = budget;
[NET]: Make ->poll() breakout consistent in Intel ethernet drivers. This makes the ->poll() routines of the E100, E1000, E1000E, IXGB, and IXGBE drivers complete ->poll() consistently. Now they will all break out when the amount of RX work done is less than 'budget'. At a later time, we may want put back code to include the TX work as well (as at least one other NAPI driver does, but by in large NAPI drivers do not do this). But if so, it should be done consistently across the board to all of these drivers. Signed-off-by: David S. Miller <davem@davemloft.net> Acked-by: Auke Kok <auke-jan.h.kok@intel.com>
2008-01-08 05:06:12 +00:00
/* If budget not fully consumed, exit the polling mode */
if (work_done < budget) {
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
if (likely(adapter->itr_setting & 3))
e1000_set_itr(adapter);
net: Remove redundant NAPI functions Following the removal of the unused struct net_device * parameter from the NAPI functions named *netif_rx_* in commit 908a7a1, they are exactly equivalent to the corresponding *napi_* functions and are therefore redundant. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-01-20 00:43:59 +00:00
napi_complete(napi);
e1000: fix close race with interrupt this is in regards to http://bugzilla.kernel.org/show_bug.cgi?id=12876 where it appears that e1000 can leave its interrupt enabled after exiting the driver. Fix the bug by making the interrupt enable paths more aware of the driver exiting. Thanks to Alan Cox for the poke and initial investigation. CC: Alan Cox <alan@lxorguk.ukuu.org.uk> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:59:22 +00:00
if (!test_bit(__E1000_DOWN, &adapter->flags))
e1000_irq_enable(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
[NET]: Make NAPI polling independent of struct net_device objects. Several devices have multiple independant RX queues per net device, and some have a single interrupt doorbell for several queues. In either case, it's easier to support layouts like that if the structure representing the poll is independant from the net device itself. The signature of the ->poll() call back goes from: int foo_poll(struct net_device *dev, int *budget) to int foo_poll(struct napi_struct *napi, int budget) The caller is returned the number of RX packets processed (or the number of "NAPI credits" consumed if you want to get abstract). The callee no longer messes around bumping dev->quota, *budget, etc. because that is all handled in the caller upon return. The napi_struct is to be embedded in the device driver private data structures. Furthermore, it is the driver's responsibility to disable all NAPI instances in it's ->stop() device close handler. Since the napi_struct is privatized into the driver's private data structures, only the driver knows how to get at all of the napi_struct instances it may have per-device. With lots of help and suggestions from Rusty Russell, Roland Dreier, Michael Chan, Jeff Garzik, and Jamal Hadi Salim. Bug fixes from Thomas Graf, Roland Dreier, Peter Zijlstra, Joseph Fannin, Scott Wood, Hans J. Koch, and Michael Chan. [ Ported to current tree and all drivers converted. Integrated Stephen's follow-on kerneldoc additions, and restored poll_list handling to the old style to fix mutual exclusion issues. -DaveM ] Signed-off-by: Stephen Hemminger <shemminger@linux-foundation.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-03 23:41:36 +00:00
return work_done;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/**
* e1000_clean_tx_irq - Reclaim resources after transmit completes
* @adapter: board private structure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static bool e1000_clean_tx_irq(struct e1000_adapter *adapter,
struct e1000_tx_ring *tx_ring)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct net_device *netdev = adapter->netdev;
struct e1000_tx_desc *tx_desc, *eop_desc;
struct e1000_buffer *buffer_info;
unsigned int i, eop;
e1000: Add performance enahancement by balancing TX and RX Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:20:43 +00:00
unsigned int count = 0;
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
unsigned int total_tx_bytes=0, total_tx_packets=0;
e1000: add byte queue limits Signed-off-by: Otto Estuardo Solares Cabrera <solca@galileo.edu> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-08-24 08:04:58 +00:00
unsigned int bytes_compl = 0, pkts_compl = 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
i = tx_ring->next_to_clean;
eop = tx_ring->buffer_info[i].next_to_watch;
eop_desc = E1000_TX_DESC(*tx_ring, eop);
e1000: cleanup clean_tx_irq routine so that it completely cleans ring The tx cleanup routine was stopping after 64 packets and this was causing issues resulting in the ring not being completely cleaned. This change updates the driver to clean the entire ring and if it doesn't it then will retry on the next pass. Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-25 21:59:04 +00:00
while ((eop_desc->upper.data & cpu_to_le32(E1000_TXD_STAT_DD)) &&
(count < tx_ring->count)) {
e1000: fix transmit routine exit bug If the e1000 transmit cleanup inner loop exited early, then cleaned might not be true. This could cause tx hangs or other badness. Use count to track the total number of descriptors cleaned instead of basing a tx queue restart off of a temporary working state variable. This code now makes the flow the same for e1000/e1000e/igb/ixgbe Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-16 16:59:47 +00:00
bool cleaned = false;
e100/e1000*/igb*/ixgb*: Add missing read memory barrier Based on patches from Sonny Rao and Milton Miller... Combined the patches to fix up clean_tx_irq and clean_rx_irq. The PowerPC architecture does not require loads to independent bytes to be ordered without adding an explicit barrier. In ixgbe_clean_rx_irq we load the status bit then load the packet data. With packet split disabled if these loads go out of order we get a stale packet, but we will notice the bad sequence numbers and drop it. The problem occurs with packet split enabled where the TCP/IP header and data are in different descriptors. If the reads go out of order we may have data that doesn't match the TCP/IP header. Since we use hardware checksumming this bad data is never verified and it makes it all the way to the application. This bug was found during stress testing and adding this barrier has been shown to fix it. The bug can manifest as a data integrity issue (bad payload data) or as a BUG in skb_pull(). This was a nasty bug to hunt down, if people agree with the fix I think it's a candidate for stable. Previously Submitted to e1000-devel only for ixgbe http://marc.info/?l=e1000-devel&m=126593062701537&w=3 We've now seen this problem hit with other device drivers (e1000e mostly) So I'm resubmitting with fixes for other Intel Device Drivers with similar issues. CC: Milton Miller <miltonm@bga.com> CC: Anton Blanchard <anton@samba.org> CC: Sonny Rao <sonnyrao@us.ibm.com> CC: stable <stable@kernel.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-08-08 16:02:31 +00:00
rmb(); /* read buffer_info after eop_desc */
e1000: fix transmit routine exit bug If the e1000 transmit cleanup inner loop exited early, then cleaned might not be true. This could cause tx hangs or other badness. Use count to track the total number of descriptors cleaned instead of basing a tx queue restart off of a temporary working state variable. This code now makes the flow the same for e1000/e1000e/igb/ixgbe Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-16 16:59:47 +00:00
for ( ; !cleaned; count++) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
tx_desc = E1000_TX_DESC(*tx_ring, i);
buffer_info = &tx_ring->buffer_info[i];
cleaned = (i == eop);
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
if (cleaned) {
e1000: save skb counts in TX to avoid cache misses Virtual Machines with emulated e1000 network adapter running on Parallels' server were seeing kernel panics due to the e1000 driver dereferencing an unexpected NULL pointer retrieved from buffer_info->skb. The problem has been addressed for the e1000e driver, but not for the e1000. Since the two drivers share similar code in the affected area, a port of the following e1000e driver commit solves the issue for the e1000 driver: commit 9ed318d546a29d7a591dbe648fd1a2efe3be1180 Author: Tom Herbert <therbert@google.com> Date: Wed May 5 14:02:27 2010 +0000 e1000e: save skb counts in TX to avoid cache misses In e1000_tx_map, precompute number of segements and bytecounts which are derived from fields in skb; these are stored in buffer_info. When cleaning tx in e1000_clean_tx_irq use the values in the associated buffer_info for statistics counting, this eliminates cache misses on skb fields. Signed-off-by: Dean Nelson <dnelson@redhat.com> Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-08-25 14:39:24 +00:00
total_tx_packets += buffer_info->segs;
total_tx_bytes += buffer_info->bytecount;
e1000: add byte queue limits Signed-off-by: Otto Estuardo Solares Cabrera <solca@galileo.edu> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-08-24 08:04:58 +00:00
if (buffer_info->skb) {
bytes_compl += buffer_info->skb->len;
pkts_compl++;
}
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
}
e1000: Fixes for 8357x - TSO workaround - Fixes eeprom version reporting - Fix loopback test - Fix for WOL Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2005-12-13 05:06:22 +00:00
e1000_unmap_and_free_tx_resource(adapter, buffer_info);
e1000: Remove unneeded and unwanted memsets This memsetting was added in a paranoid rage debugging TX hangs, but are no longer of importance. We can beef up the performance quite a bit removing them. Make sure to fill in next_to_watch to allow this. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2006-11-01 16:47:53 +00:00
tx_desc->upper.data = 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (unlikely(++i == tx_ring->count)) i = 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
eop = tx_ring->buffer_info[i].next_to_watch;
eop_desc = E1000_TX_DESC(*tx_ring, eop);
}
tx_ring->next_to_clean = i;
e1000: add byte queue limits Signed-off-by: Otto Estuardo Solares Cabrera <solca@galileo.edu> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-08-24 08:04:58 +00:00
netdev_completed_queue(netdev, pkts_compl, bytes_compl);
e1000: implement more efficient tx queue locking Implement more efficient locking (avoid the lock) when checking for a stopped queue. Also don't wake the queue unless the threshold is reached to avoid queue on/off thrash. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-04-15 02:05:25 +00:00
#define TX_WAKE_THRESHOLD 32
e1000: fix transmit routine exit bug If the e1000 transmit cleanup inner loop exited early, then cleaned might not be true. This could cause tx hangs or other badness. Use count to track the total number of descriptors cleaned instead of basing a tx queue restart off of a temporary working state variable. This code now makes the flow the same for e1000/e1000e/igb/ixgbe Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-16 16:59:47 +00:00
if (unlikely(count && netif_carrier_ok(netdev) &&
e1000: Maybe stop TX if not enough free descriptors Cc: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:53:48 +00:00
E1000_DESC_UNUSED(tx_ring) >= TX_WAKE_THRESHOLD)) {
/* Make sure that anybody stopping the queue after this
* sees the new next_to_clean.
*/
smp_mb();
e1000: fix tx waking queue after queue stopped during shutdown This fix closes a race where the adapter can be shutting down while hard_start_xmit is being called and interrupts are being handled. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:18:07 +00:00
if (netif_queue_stopped(netdev) &&
!(test_bit(__E1000_DOWN, &adapter->flags))) {
e1000: implement more efficient tx queue locking Implement more efficient locking (avoid the lock) when checking for a stopped queue. Also don't wake the queue unless the threshold is reached to avoid queue on/off thrash. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-04-15 02:05:25 +00:00
netif_wake_queue(netdev);
e1000: add queue restart counter Add a netif_wake/start_queue counter to the ethtool statistics to indicated to the user that their transmit ring could be too small for their workload. Signed-off-by: Jesse brandeburg <jesse.brandeburg@intel.com> Cc: Jamal Hadi <hadi@cyberus.ca> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:47:59 +00:00
++adapter->restart_queue;
}
e1000: implement more efficient tx queue locking Implement more efficient locking (avoid the lock) when checking for a stopped queue. Also don't wake the queue unless the threshold is reached to avoid queue on/off thrash. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-04-15 02:05:25 +00:00
}
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
e1000: multi-queue defines/modification to data structures defines/modifies data structures, function prototypes and changes to the driver rendering it capable of handling <n> tx/rx queues Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-04 11:01:55 +00:00
if (adapter->detect_tx_hung) {
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
/* Detect a transmit hang in hardware, this serializes the
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* check with the clearing of time_stamp and movement of i
*/
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
adapter->detect_tx_hung = false;
e1000: fix tx waking queue after queue stopped during shutdown This fix closes a race where the adapter can be shutting down while hard_start_xmit is being called and interrupts are being handled. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:18:07 +00:00
if (tx_ring->buffer_info[eop].time_stamp &&
time_after(jiffies, tx_ring->buffer_info[eop].time_stamp +
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
(adapter->tx_timeout_factor * HZ)) &&
drivers/net: Move && and || to end of previous line Only files where David Miller is the primary git-signer. wireless, wimax, ixgbe, etc are not modified. Compile tested x86 allyesconfig only Not all files compiled (not x86 compatible) Added a few > 80 column lines, which I ignored. Existing checkpatch complaints ignored. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-03 07:58:21 +00:00
!(er32(STATUS) & E1000_STATUS_TXOFF)) {
[PATCH] e1000: Dump information on Tx ring Dump information on Tx ring when 'NETDEV: Watchdog' condition is reached Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_main.c net-drivers-2.6/drivers/net/e1000.new/e1000_main.c
2005-04-29 02:40:40 +00:00
/* detected Tx unit hang */
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(drv, "Detected Tx Unit Hang\n"
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
" Tx Queue <%lu>\n"
" TDH <%x>\n"
" TDT <%x>\n"
" next_to_use <%x>\n"
" next_to_clean <%x>\n"
"buffer_info[next_to_clean]\n"
" time_stamp <%lx>\n"
" next_to_watch <%x>\n"
" jiffies <%lx>\n"
" next_to_watch.status <%x>\n",
e1000: fix wrong queue idx calculation tx_ring and adapter->tx_ring are already of type "struct e1000_tx_ring *" Signed-off-by: Hong Zhiguo <zhiguohong@tencent.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2013-10-22 18:32:56 +00:00
(unsigned long)(tx_ring - adapter->tx_ring),
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
readl(hw->hw_addr + tx_ring->tdh),
readl(hw->hw_addr + tx_ring->tdt),
[PATCH] e1000: Dump information on Tx ring Dump information on Tx ring when 'NETDEV: Watchdog' condition is reached Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_main.c net-drivers-2.6/drivers/net/e1000.new/e1000_main.c
2005-04-29 02:40:40 +00:00
tx_ring->next_to_use,
[PATCH] e1000: Fix TX timeout logic Fixed the TX timeout logic to use "end of packet" rather than "next to clean". Updated message log. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:50:57 +00:00
tx_ring->next_to_clean,
e1000: fix tx waking queue after queue stopped during shutdown This fix closes a race where the adapter can be shutting down while hard_start_xmit is being called and interrupts are being handled. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:18:07 +00:00
tx_ring->buffer_info[eop].time_stamp,
[PATCH] e1000: Dump information on Tx ring Dump information on Tx ring when 'NETDEV: Watchdog' condition is reached Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_main.c net-drivers-2.6/drivers/net/e1000.new/e1000_main.c
2005-04-29 02:40:40 +00:00
eop,
jiffies,
eop_desc->upper.fields.status);
e1000: Adding e1000_dump function When TX hang occurs e1000_dump prints TX ring, RX ring and Device registers. Signed-off-by: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-27 09:00:46 +00:00
e1000_dump(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
netif_stop_queue(netdev);
[PATCH] e1000: Dump information on Tx ring Dump information on Tx ring when 'NETDEV: Watchdog' condition is reached Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_main.c net-drivers-2.6/drivers/net/e1000.new/e1000_main.c
2005-04-29 02:40:40 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
adapter->total_tx_bytes += total_tx_bytes;
adapter->total_tx_packets += total_tx_packets;
e1000: Use the instance of net_device_stats from net_device. Since net_device has an instance of net_device_stats, we can remove the instance of this from the adapter structure. Signed-off-by: Ajit Khaparde <ajitk@serverengines.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-07 02:42:23 +00:00
netdev->stats.tx_bytes += total_tx_bytes;
netdev->stats.tx_packets += total_tx_packets;
drivers/net: return operator cleanup Change "return (EXPR);" to "return EXPR;" return is not a function, parentheses are not required. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-23 05:40:09 +00:00
return count < tx_ring->count;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/**
* e1000_rx_checksum - Receive Checksum Offload for 82543
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
* @adapter: board private structure
* @status_err: receive descriptor status and error fields
* @csum: receive descriptor csum field
* @sk_buff: socket buffer with received data
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_rx_checksum(struct e1000_adapter *adapter, u32 status_err,
u32 csum, struct sk_buff *skb)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u16 status = (u16)status_err;
u8 errors = (u8)(status_err >> 24);
drivers/net: avoid some skb->ip_summed initializations fresh skbs have ip_summed set to CHECKSUM_NONE (0) We can avoid setting again skb->ip_summed to CHECKSUM_NONE in drivers. Introduce skb_checksum_none_assert() helper so that we keep this assertion documented in driver sources. Change most occurrences of : skb->ip_summed = CHECKSUM_NONE; by : skb_checksum_none_assert(skb); Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-02 20:07:41 +00:00
skb_checksum_none_assert(skb);
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* 82543 or newer only */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (unlikely(hw->mac_type < e1000_82543)) return;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Ignore Checksum bit is set */
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (unlikely(status & E1000_RXD_STAT_IXSM)) return;
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
/* TCP/UDP checksum error bit is set */
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (unlikely(errors & E1000_RXD_ERR_TCPE)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* let the stack verify checksum errors */
adapter->hw_csum_err++;
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
return;
}
/* TCP/UDP Checksum has not been calculated */
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
if (!(status & E1000_RXD_STAT_TCPCS))
return;
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
/* It must be a TCP or UDP packet with a valid checksum */
if (likely(status & E1000_RXD_STAT_TCPCS)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* TCP checksum is good */
skb->ip_summed = CHECKSUM_UNNECESSARY;
}
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
adapter->hw_csum_good++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
/**
* e1000_consume_page - helper function
**/
static void e1000_consume_page(struct e1000_buffer *bi, struct sk_buff *skb,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
u16 length)
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
{
bi->page = NULL;
skb->len += length;
skb->data_len += length;
e1000: fix skb truesize underestimation e1000 allocates a full page per skb fragment. We must account PAGE_SIZE increments on skb->truesize, not the actual frag length. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-10-13 07:53:42 +00:00
skb->truesize += PAGE_SIZE;
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
}
/**
* e1000_receive_skb - helper function to handle rx indications
* @adapter: board private structure
* @status: descriptor status field as written by hardware
* @vlan: descriptor vlan field as written by hardware (no le/be conversion)
* @skb: pointer to sk_buff to be indicated to stack
*/
static void e1000_receive_skb(struct e1000_adapter *adapter, u8 status,
__le16 vlan, struct sk_buff *skb)
{
e1000: use GRO for receive E1000 can benefit from calling the GRO receive functions. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-22 18:23:05 +00:00
skb->protocol = eth_type_trans(skb, adapter->netdev);
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
if (status & E1000_RXD_STAT_VP) {
u16 vid = le16_to_cpu(vlan) & E1000_RXD_SPC_VLAN_MASK;
net: vlan: add protocol argument to packet tagging functions Add a protocol argument to the VLAN packet tagging functions. In case of HW tagging, we need that protocol available in the ndo_start_xmit functions, so it is stored in a new field in the skb. The new field fits into a hole (on 64 bit) and doesn't increase the sks's size. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-04-19 02:04:30 +00:00
__vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), vid);
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
}
napi_gro_receive(&adapter->napi, skb);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
}
/**
* e1000_clean_jumbo_rx_irq - Send received data up the network stack; legacy
* @adapter: board private structure
* @rx_ring: ring to clean
* @work_done: amount of napi work completed this call
* @work_to_do: max amount of work allowed for this call to do
*
* the return value indicates whether actual cleaning was done, there
* is no guarantee that everything was cleaned
*/
static bool e1000_clean_jumbo_rx_irq(struct e1000_adapter *adapter,
struct e1000_rx_ring *rx_ring,
int *work_done, int work_to_do)
{
struct e1000_hw *hw = &adapter->hw;
struct net_device *netdev = adapter->netdev;
struct pci_dev *pdev = adapter->pdev;
struct e1000_rx_desc *rx_desc, *next_rxd;
struct e1000_buffer *buffer_info, *next_buffer;
unsigned long irq_flags;
u32 length;
unsigned int i;
int cleaned_count = 0;
bool cleaned = false;
unsigned int total_rx_bytes=0, total_rx_packets=0;
i = rx_ring->next_to_clean;
rx_desc = E1000_RX_DESC(*rx_ring, i);
buffer_info = &rx_ring->buffer_info[i];
while (rx_desc->status & E1000_RXD_STAT_DD) {
struct sk_buff *skb;
u8 status;
if (*work_done >= work_to_do)
break;
(*work_done)++;
e100/e1000*/igb*/ixgb*: Add missing read memory barrier Based on patches from Sonny Rao and Milton Miller... Combined the patches to fix up clean_tx_irq and clean_rx_irq. The PowerPC architecture does not require loads to independent bytes to be ordered without adding an explicit barrier. In ixgbe_clean_rx_irq we load the status bit then load the packet data. With packet split disabled if these loads go out of order we get a stale packet, but we will notice the bad sequence numbers and drop it. The problem occurs with packet split enabled where the TCP/IP header and data are in different descriptors. If the reads go out of order we may have data that doesn't match the TCP/IP header. Since we use hardware checksumming this bad data is never verified and it makes it all the way to the application. This bug was found during stress testing and adding this barrier has been shown to fix it. The bug can manifest as a data integrity issue (bad payload data) or as a BUG in skb_pull(). This was a nasty bug to hunt down, if people agree with the fix I think it's a candidate for stable. Previously Submitted to e1000-devel only for ixgbe http://marc.info/?l=e1000-devel&m=126593062701537&w=3 We've now seen this problem hit with other device drivers (e1000e mostly) So I'm resubmitting with fixes for other Intel Device Drivers with similar issues. CC: Milton Miller <miltonm@bga.com> CC: Anton Blanchard <anton@samba.org> CC: Sonny Rao <sonnyrao@us.ibm.com> CC: stable <stable@kernel.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-08-08 16:02:31 +00:00
rmb(); /* read descriptor and rx_buffer_info after status DD */
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
status = rx_desc->status;
skb = buffer_info->skb;
buffer_info->skb = NULL;
if (++i == rx_ring->count) i = 0;
next_rxd = E1000_RX_DESC(*rx_ring, i);
prefetch(next_rxd);
next_buffer = &rx_ring->buffer_info[i];
cleaned = true;
cleaned_count++;
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
dma_unmap_page(&pdev->dev, buffer_info->dma,
buffer_info->length, DMA_FROM_DEVICE);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
buffer_info->dma = 0;
length = le16_to_cpu(rx_desc->length);
/* errors is only valid for DD + EOP descriptors */
if (unlikely((status & E1000_RXD_STAT_EOP) &&
(rx_desc->errors & E1000_RXD_ERR_FRAME_ERR_MASK))) {
e1000: look in the page and not in skb->data for the last byte The code seems to want to look at the last byte where the HW puts some information. Since the skb->data area is never seen by the HW I guess it does not work as expected. We pass the page address to the HW so I *think* in order to get to the last byte where the information might be one should use the page buffer and take a look. This is of course not more than just compile tested. Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-05-11 16:30:46 +00:00
u8 *mapped;
u8 last_byte;
mapped = page_address(buffer_info->page);
last_byte = *(mapped + length - 1);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
if (TBI_ACCEPT(hw, status, rx_desc->errors, length,
last_byte)) {
spin_lock_irqsave(&adapter->stats_lock,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
irq_flags);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
e1000_tbi_adjust_stats(hw, &adapter->stats,
e1000: look into the page instead of skb->data for e1000_tbi_adjust_stats() This is another fixup where the data is not transfered into buffer addressed by skb->data but into a page. Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-05-15 09:18:55 +00:00
length, mapped);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
spin_unlock_irqrestore(&adapter->stats_lock,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
irq_flags);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
length--;
} else {
e1000: Support RX-ALL flag. This allows the NIC to receive errored frames (bad FCS, etc) and pass them up the stack. This can be useful when using sniffers. Signed-off-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-04-04 06:01:29 +00:00
if (netdev->features & NETIF_F_RXALL)
goto process_skb;
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
/* recycle both page and skb */
buffer_info->skb = skb;
/* an error means any chain goes out the window
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* too
*/
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
if (rx_ring->rx_skb_top)
dev_kfree_skb(rx_ring->rx_skb_top);
rx_ring->rx_skb_top = NULL;
goto next_desc;
}
}
#define rxtop rx_ring->rx_skb_top
e1000: Support RX-ALL flag. This allows the NIC to receive errored frames (bad FCS, etc) and pass them up the stack. This can be useful when using sniffers. Signed-off-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-04-04 06:01:29 +00:00
process_skb:
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
if (!(status & E1000_RXD_STAT_EOP)) {
/* this descriptor is only the beginning (or middle) */
if (!rxtop) {
/* this is the beginning of a chain */
rxtop = skb;
skb_fill_page_desc(rxtop, 0, buffer_info->page,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
0, length);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
} else {
/* this is the middle of a chain */
skb_fill_page_desc(rxtop,
skb_shinfo(rxtop)->nr_frags,
buffer_info->page, 0, length);
/* re-use the skb, only consumed the page */
buffer_info->skb = skb;
}
e1000_consume_page(buffer_info, rxtop, length);
goto next_desc;
} else {
if (rxtop) {
/* end of the chain */
skb_fill_page_desc(rxtop,
skb_shinfo(rxtop)->nr_frags,
buffer_info->page, 0, length);
/* re-use the current skb, we only consumed the
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* page
*/
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
buffer_info->skb = skb;
skb = rxtop;
rxtop = NULL;
e1000_consume_page(buffer_info, skb, length);
} else {
/* no chain, got EOP, this buf is the packet
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* copybreak to save the put_page/alloc_page
*/
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
if (length <= copybreak &&
skb_tailroom(skb) >= length) {
u8 *vaddr;
net: remove the second argument of k[un]map_atomic() Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: Cong Wang <amwang@redhat.com>
2011-11-25 15:14:23 +00:00
vaddr = kmap_atomic(buffer_info->page);
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
memcpy(skb_tail_pointer(skb), vaddr,
length);
net: remove the second argument of k[un]map_atomic() Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: Cong Wang <amwang@redhat.com>
2011-11-25 15:14:23 +00:00
kunmap_atomic(vaddr);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
/* re-use the page, so don't erase
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* buffer_info->page
*/
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
skb_put(skb, length);
} else {
skb_fill_page_desc(skb, 0,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
buffer_info->page, 0,
length);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
e1000_consume_page(buffer_info, skb,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
length);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
}
}
}
/* Receive Checksum Offload XXX recompute due to CRC strip? */
e1000_rx_checksum(adapter,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
(u32)(status) |
((u32)(rx_desc->errors) << 24),
le16_to_cpu(rx_desc->csum), skb);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
e1000: Support RX-FCS flag. This allows the NIC to pass the Ethernet Frame Checksum (FCS) up the stack. Useful when sniffing packets. Signed-off-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-11 15:40:11 +00:00
total_rx_bytes += (skb->len - 4); /* don't count FCS */
if (likely(!(netdev->features & NETIF_F_RXFCS)))
pskb_trim(skb, skb->len - 4);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
total_rx_packets++;
/* eth type trans needs skb->data to point to something */
if (!pskb_may_pull(skb, ETH_HLEN)) {
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(drv, "pskb_may_pull failed.\n");
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
dev_kfree_skb(skb);
goto next_desc;
}
e1000_receive_skb(adapter, status, rx_desc->special, skb);
next_desc:
rx_desc->status = 0;
/* return some buffers to hardware, one at a time is too slow */
if (unlikely(cleaned_count >= E1000_RX_BUFFER_WRITE)) {
adapter->alloc_rx_buf(adapter, rx_ring, cleaned_count);
cleaned_count = 0;
}
/* use prefetched values */
rx_desc = next_rxd;
buffer_info = next_buffer;
}
rx_ring->next_to_clean = i;
cleaned_count = E1000_DESC_UNUSED(rx_ring);
if (cleaned_count)
adapter->alloc_rx_buf(adapter, rx_ring, cleaned_count);
adapter->total_rx_packets += total_rx_packets;
adapter->total_rx_bytes += total_rx_bytes;
e1000: Use the instance of net_device_stats from net_device. Since net_device has an instance of net_device_stats, we can remove the instance of this from the adapter structure. Signed-off-by: Ajit Khaparde <ajitk@serverengines.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-07 02:42:23 +00:00
netdev->stats.rx_bytes += total_rx_bytes;
netdev->stats.rx_packets += total_rx_packets;
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
return cleaned;
}
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* this should improve performance for small packets with large amounts
ixgb and e1000: Use new function for copybreak tests There appears to be an off-by-1 defect in the maximum packet size copied when copybreak is speified in these modules. The copybreak module params are specified as: "Maximum size of packet that is copied to a new buffer on receive" The tests are changed from "< copybreak" to "<= copybreak" and moved into new static functions for readability. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-05-13 15:26:17 +00:00
* of reassembly being done in the stack
*/
static void e1000_check_copybreak(struct net_device *netdev,
struct e1000_buffer *buffer_info,
u32 length, struct sk_buff **skb)
{
struct sk_buff *new_skb;
if (length > copybreak)
return;
new_skb = netdev_alloc_skb_ip_align(netdev, length);
if (!new_skb)
return;
skb_copy_to_linear_data_offset(new_skb, -NET_IP_ALIGN,
(*skb)->data - NET_IP_ALIGN,
length + NET_IP_ALIGN);
/* save the skb in buffer_info as good */
buffer_info->skb = *skb;
*skb = new_skb;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/**
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
* e1000_clean_rx_irq - Send received data up the network stack; legacy
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* @adapter: board private structure
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
* @rx_ring: ring to clean
* @work_done: amount of napi work completed this call
* @work_to_do: max amount of work allowed for this call to do
*/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static bool e1000_clean_rx_irq(struct e1000_adapter *adapter,
struct e1000_rx_ring *rx_ring,
int *work_done, int work_to_do)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct net_device *netdev = adapter->netdev;
struct pci_dev *pdev = adapter->pdev;
[PATCH] e1000: fix receive breakage in attempting to not send the "prefetch" patch, we broke the receive code, this patch fixes that issue. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:43 +00:00
struct e1000_rx_desc *rx_desc, *next_rxd;
struct e1000_buffer *buffer_info, *next_buffer;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
unsigned long flags;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u32 length;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
unsigned int i;
[PATCH] e1000: Added cleaned_count to RX buffer allocation Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:19 +00:00
int cleaned_count = 0;
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
bool cleaned = false;
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
unsigned int total_rx_bytes=0, total_rx_packets=0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
i = rx_ring->next_to_clean;
rx_desc = E1000_RX_DESC(*rx_ring, i);
[PATCH] e1000: Added RX buffer enhancements Align the prefetches to a dword to help speed them up. Recycle skb's and early replenish. Force memory writes to complete before fetching more descriptors. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:32 +00:00
buffer_info = &rx_ring->buffer_info[i];
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[PATCH] e1000: Added RX buffer enhancements Align the prefetches to a dword to help speed them up. Recycle skb's and early replenish. Force memory writes to complete before fetching more descriptors. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:32 +00:00
while (rx_desc->status & E1000_RXD_STAT_DD) {
e1000: remove risky prefetch on next_skb->data It was brought to our attention that the prefetches break e1000 traffic on xscale/arm architectures. Remove them for now. We'll let them stay in mm for a while, or find a better solution to enable. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:28:47 +00:00
struct sk_buff *skb;
[PATCH] e1000: Added copy break code Improves small packet performance with large amounts of reassembly being done in the stack. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:30 +00:00
u8 status;
e1000: whitespace changes, comments, typos Small whitespace changes, comment changes, typo fixes. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:47:30 +00:00
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (*work_done >= work_to_do)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
(*work_done)++;
e100/e1000*/igb*/ixgb*: Add missing read memory barrier Based on patches from Sonny Rao and Milton Miller... Combined the patches to fix up clean_tx_irq and clean_rx_irq. The PowerPC architecture does not require loads to independent bytes to be ordered without adding an explicit barrier. In ixgbe_clean_rx_irq we load the status bit then load the packet data. With packet split disabled if these loads go out of order we get a stale packet, but we will notice the bad sequence numbers and drop it. The problem occurs with packet split enabled where the TCP/IP header and data are in different descriptors. If the reads go out of order we may have data that doesn't match the TCP/IP header. Since we use hardware checksumming this bad data is never verified and it makes it all the way to the application. This bug was found during stress testing and adding this barrier has been shown to fix it. The bug can manifest as a data integrity issue (bad payload data) or as a BUG in skb_pull(). This was a nasty bug to hunt down, if people agree with the fix I think it's a candidate for stable. Previously Submitted to e1000-devel only for ixgbe http://marc.info/?l=e1000-devel&m=126593062701537&w=3 We've now seen this problem hit with other device drivers (e1000e mostly) So I'm resubmitting with fixes for other Intel Device Drivers with similar issues. CC: Milton Miller <miltonm@bga.com> CC: Anton Blanchard <anton@samba.org> CC: Sonny Rao <sonnyrao@us.ibm.com> CC: stable <stable@kernel.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-08-08 16:02:31 +00:00
rmb(); /* read descriptor and rx_buffer_info after status DD */
e1000: delete non NAPI code from the driver Compile-tested only. Signed-off-by: Francois Romieu <romieu@fr.zoreil.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:38 +00:00
[PATCH] e1000: Added copy break code Improves small packet performance with large amounts of reassembly being done in the stack. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:30 +00:00
status = rx_desc->status;
[PATCH] e1000: Added RX buffer enhancements Align the prefetches to a dword to help speed them up. Recycle skb's and early replenish. Force memory writes to complete before fetching more descriptors. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:32 +00:00
skb = buffer_info->skb;
[PATCH] e1000: fix receive breakage in attempting to not send the "prefetch" patch, we broke the receive code, this patch fixes that issue. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:43 +00:00
buffer_info->skb = NULL;
e1000: Added a performance enhancement - prefetch - this implementation of prefetch was tested on new and old hardware Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:21:57 +00:00
prefetch(skb->data - NET_IP_ALIGN);
[PATCH] e1000: fix receive breakage in attempting to not send the "prefetch" patch, we broke the receive code, this patch fixes that issue. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:43 +00:00
if (++i == rx_ring->count) i = 0;
next_rxd = E1000_RX_DESC(*rx_ring, i);
e1000: Added a performance enhancement - prefetch - this implementation of prefetch was tested on new and old hardware Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:21:57 +00:00
prefetch(next_rxd);
[PATCH] e1000: fix receive breakage in attempting to not send the "prefetch" patch, we broke the receive code, this patch fixes that issue. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:43 +00:00
next_buffer = &rx_ring->buffer_info[i];
e1000: Convert boolean_t to bool On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote: > send me a patch for e1000 and for ixgb and I'll happily apply those :) boolean_t to bool TRUE to true FALSE to false comment typo ahread to ahead Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:25 +00:00
cleaned = true;
[PATCH] e1000: Added cleaned_count to RX buffer allocation Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:19 +00:00
cleaned_count++;
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
dma_unmap_single(&pdev->dev, buffer_info->dma,
buffer_info->length, DMA_FROM_DEVICE);
e1000: fix unmap bug as reported by kerneloops.org [ 121.781161] ------------[ cut here ]------------ [ 121.781171] WARNING: at lib/dma-debug.c:793 check_unmap+0x14e/0x577() [ 121.781173] Hardware name: S5520HC [ 121.781177] e1000 0000:0a:00.0: DMA-API: device driver tries to free DMA memory it has not allocated [device address=0x00000001d688b0fa] [size=1522 bytes] [ 121.781180] Modules linked in: e1000 mdio dca [last unloaded: ixgbe] [ 121.781187] Pid: 4793, comm: bash Tainted: P 2.6.30-master-06161113 #3 [ 121.781190] Call Trace: [ 121.781195] [<ffffffff8123056f>] ? check_unmap+0x14e/0x577 [ 121.781201] [<ffffffff81057a19>] warn_slowpath_common+0x77/0x8f [ 121.781205] [<ffffffff81057ae1>] warn_slowpath_fmt+0x9f/0xa1 [ 121.781212] [<ffffffff81477ce2>] ? _spin_lock_irqsave+0x3f/0x49 [ 121.781216] [<ffffffff8122fa97>] ? get_hash_bucket+0x28/0x33 [ 121.781220] [<ffffffff8123056f>] check_unmap+0x14e/0x577 [ 121.781225] [<ffffffff810e4f48>] ? check_bytes_and_report+0x38/0xcb [ 121.781230] [<ffffffff81230bbf>] debug_dma_unmap_page+0x80/0x92 [ 121.781234] [<ffffffff8122e549>] ? unmap_single+0x1a/0x4e [ 121.781239] [<ffffffff813901e1>] ? __kfree_skb+0x74/0x78 [ 121.781250] [<ffffffffa00662ef>] pci_unmap_single+0x64/0x6d [e1000] [ 121.781259] [<ffffffffa0066344>] e1000_clean_rx_ring+0x4c/0xbf [e1000] [ 121.781268] [<ffffffffa00663df>] e1000_clean_all_rx_rings+0x28/0x36 [e1000] [ 121.781277] [<ffffffffa0067464>] e1000_down+0x138/0x141 [e1000] [ 121.781286] [<ffffffffa00681c2>] __e1000_shutdown+0x6b/0x198 [e1000] [ 121.781296] [<ffffffffa0068405>] e1000_suspend+0x17/0x50 [e1000] [ 121.781301] [<ffffffff81237665>] pci_legacy_suspend+0x3b/0xbe [ 121.781305] [<ffffffff81237bc6>] pci_pm_suspend+0x3e/0xf1 [ 121.781310] [<ffffffff812eaf1c>] pm_op+0x57/0xde [ 121.781314] [<ffffffff812eb444>] dpm_suspend_start+0x31e/0x470 [ 121.781319] [<ffffffff810877da>] suspend_devices_and_enter+0x3e/0x1a2 [ 121.781323] [<ffffffff81087a0f>] enter_state+0xd1/0x127 [ 121.781327] [<ffffffff8108717a>] state_store+0xa7/0xc9 [ 121.781332] [<ffffffff81221843>] kobj_attr_store+0x17/0x19 [ 121.781336] [<ffffffff8113c01e>] sysfs_write_file+0xe5/0x121 [ 121.781341] [<ffffffff810ed165>] vfs_write+0xab/0x105 [ 121.781344] [<ffffffff810ed279>] sys_write+0x47/0x6d [ 121.781349] [<ffffffff81027aab>] system_call_fastpath+0x16/0x1b [ 121.781352] ---[ end trace 97bacaaac2ed7786 ]--- Fix is to correctly zero out internal ->dma value when unmapping and make sure never to unmap unless there specifically was a mapping done. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-06-30 12:45:34 +00:00
buffer_info->dma = 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
length = le16_to_cpu(rx_desc->length);
e1000: add missing length check to e1000 receive routine Patch to fix bad length checking in e1000. E1000 by default does two things: 1) Spans rx descriptors for packets that don't fit into 1 skb on recieve 2) Strips the crc from a frame by subtracting 4 bytes from the length prior to doing an skb_put Since the e1000 driver isn't written to support receiving packets that span multiple rx buffers, it checks the End of Packet bit of every frame, and discards it if its not set. This places us in a situation where, if we have a spanning packet, the first part is discarded, but the second part is not (since it is the end of packet, and it passes the EOP bit test). If the second part of the frame is small (4 bytes or less), we subtract 4 from it to remove its crc, underflow the length, and wind up in skb_over_panic, when we try to skb_put a huge number of bytes into the skb. This amounts to a remote DOS attack through careful selection of frame size in relation to interface MTU. The fix for this is already in the e1000e driver, as well as the e1000 sourceforge driver, but no one ever pushed it to e1000. This is lifted straight from e1000e, and prevents small frames from causing the underflow described above Signed-off-by: Neil Horman <nhorman@tuxdriver.com> Tested-by: Andy Gospodarek <andy@greyhouse.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-06-02 08:29:58 +00:00
/* !EOP means multiple descriptors were used to store a single
e1000: enhance frame fragment detection Originally From: Neil Horman <nhorman@tuxdriver.com> Modified by: Jesse Brandeburg <jesse.brandeburg@intel.com> Hey all- A security discussion was recently given: http://events.ccc.de/congress/2009/Fahrplan//events/3596.en.html And a patch that I submitted awhile back was brought up. Apparently some of their testing revealed that they were able to force a buffer fragment in e1000 in which the trailing fragment was greater than 4 bytes. As a result the fragment check I introduced failed to detect the fragement and a partial invalid frame was passed up into the network stack. I've written this patch to correct it. I'm in the process of testing it now, but it makes good logical sense to me. Effectively it maintains a per-adapter state variable which detects a non-EOP frame, and discards it and subsequent non-EOP frames leading up to _and_ _including_ the next positive-EOP frame (as it is by definition the last fragment). This should prevent any and all partial frames from entering the network stack from e1000. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-01-19 14:15:38 +00:00
* packet, if thats the case we need to toss it. In fact, we
* to toss every packet with the EOP bit clear and the next
* frame that _does_ have the EOP bit set, as it is by
* definition only a frame fragment
*/
if (unlikely(!(status & E1000_RXD_STAT_EOP)))
adapter->discarding = true;
if (adapter->discarding) {
[PATCH] e1000: revert to single descriptor for legacy receive path A recent patch attempted to enable more efficient memory usage by using only 2kB descriptors for jumbo frames. The method used to implement this has since been commented upon as "illegal" and in recent kernels even causes a BUG when receiving ip fragments while using jumbo frames. This patch simply goes back to the way things were. We expect some complaints due to order 3 allocations failing to come back due to this change. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
2006-03-01 04:24:07 +00:00
/* All receives must fit into a single buffer */
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_dbg("Receive packet consumed multiple buffers\n");
e1000: recycle skb Recycle an skb to improve performance a bit. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:53 +00:00
/* recycle */
e1000: Whitespace cleanup, cosmetic changes Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-28 21:56:16 +00:00
buffer_info->skb = skb;
e1000: enhance frame fragment detection Originally From: Neil Horman <nhorman@tuxdriver.com> Modified by: Jesse Brandeburg <jesse.brandeburg@intel.com> Hey all- A security discussion was recently given: http://events.ccc.de/congress/2009/Fahrplan//events/3596.en.html And a patch that I submitted awhile back was brought up. Apparently some of their testing revealed that they were able to force a buffer fragment in e1000 in which the trailing fragment was greater than 4 bytes. As a result the fragment check I introduced failed to detect the fragement and a partial invalid frame was passed up into the network stack. I've written this patch to correct it. I'm in the process of testing it now, but it makes good logical sense to me. Effectively it maintains a per-adapter state variable which detects a non-EOP frame, and discards it and subsequent non-EOP frames leading up to _and_ _including_ the next positive-EOP frame (as it is by definition the last fragment). This should prevent any and all partial frames from entering the network stack from e1000. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-01-19 14:15:38 +00:00
if (status & E1000_RXD_STAT_EOP)
adapter->discarding = false;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto next_desc;
}
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (unlikely(rx_desc->errors & E1000_RXD_ERR_FRAME_ERR_MASK)) {
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
u8 last_byte = *(skb->data + length - 1);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (TBI_ACCEPT(hw, status, rx_desc->errors, length,
last_byte)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
spin_lock_irqsave(&adapter->stats_lock, flags);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_tbi_adjust_stats(hw, &adapter->stats,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
length, skb->data);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
spin_unlock_irqrestore(&adapter->stats_lock,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
flags);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
length--;
} else {
e1000: Support RX-ALL flag. This allows the NIC to receive errored frames (bad FCS, etc) and pass them up the stack. This can be useful when using sniffers. Signed-off-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-04-04 06:01:29 +00:00
if (netdev->features & NETIF_F_RXALL)
goto process_skb;
e1000: Buffer optimizations for small MTU Remove multi-descriptor support from legacy recieve path Add memory usage efficiency by using more correct size descriptors for small MTU sizes and optimize using LPE for <= 1522 byte frame sizes An extra performance fix that effected our TCP window size growth as a receiver. Set our initial buffer to be 128 bytes instead of 256 to prevent over-socket charge when truesize is computed in the stack. old way: truesize = 256 + l1 = 256 + 1460 = 1716 new way: truesize = 128 + l1 = 128 + 1460 = 1588 The magic value that we can't cross is 1648. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-04-15 02:05:18 +00:00
/* recycle */
buffer_info->skb = skb;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto next_desc;
}
e1000: fix mismerge skb_put. Seems there was a bit of a fix needed to due a bad merge in the legacy receive path. Fixes a panic due to skb_over_panic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-04-18 19:31:04 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Support RX-ALL flag. This allows the NIC to receive errored frames (bad FCS, etc) and pass them up the stack. This can be useful when using sniffers. Signed-off-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-04-04 06:01:29 +00:00
process_skb:
e1000: Support RX-FCS flag. This allows the NIC to pass the Ethernet Frame Checksum (FCS) up the stack. Useful when sniffing packets. Signed-off-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-11 15:40:11 +00:00
total_rx_bytes += (length - 4); /* don't count FCS */
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
total_rx_packets++;
e1000: Support RX-FCS flag. This allows the NIC to pass the Ethernet Frame Checksum (FCS) up the stack. Useful when sniffing packets. Signed-off-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-02-11 15:40:11 +00:00
if (likely(!(netdev->features & NETIF_F_RXFCS)))
/* adjust length to remove Ethernet CRC, this must be
* done after the TBI_ACCEPT workaround above
*/
length -= 4;
ixgb and e1000: Use new function for copybreak tests There appears to be an off-by-1 defect in the maximum packet size copied when copybreak is speified in these modules. The copybreak module params are specified as: "Maximum size of packet that is copied to a new buffer on receive" The tests are changed from "< copybreak" to "<= copybreak" and moved into new static functions for readability. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-05-13 15:26:17 +00:00
e1000_check_copybreak(netdev, buffer_info, length, &skb);
e1000: simplify skb_put call. Simplify two calls to skb_put by removing one call to it. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:47:50 +00:00
skb_put(skb, length);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Receive Checksum Offload */
[PATCH] e1000: Added copy break code Improves small packet performance with large amounts of reassembly being done in the stack. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:30 +00:00
e1000_rx_checksum(adapter,
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
(u32)(status) |
((u32)(rx_desc->errors) << 24),
[PATCH] e1000 endianness bugs return -E_NO_BIG_ENDIAN_TESTING; [E1000]: Fix 4 missed endianness conversions on RX descriptor fields. Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-03-15 22:26:28 +00:00
le16_to_cpu(rx_desc->csum), skb);
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
e1000_receive_skb(adapter, status, rx_desc->special, skb);
e1000: delete non NAPI code from the driver Compile-tested only. Signed-off-by: Francois Romieu <romieu@fr.zoreil.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:38 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
next_desc:
rx_desc->status = 0;
[PATCH] e1000: Added cleaned_count to RX buffer allocation Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:19 +00:00
/* return some buffers to hardware, one at a time is too slow */
if (unlikely(cleaned_count >= E1000_RX_BUFFER_WRITE)) {
adapter->alloc_rx_buf(adapter, rx_ring, cleaned_count);
cleaned_count = 0;
}
e1000: Added a performance enhancement - prefetch - this implementation of prefetch was tested on new and old hardware Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-03-03 02:21:57 +00:00
/* use prefetched values */
[PATCH] e1000: fix receive breakage in attempting to not send the "prefetch" patch, we broke the receive code, this patch fixes that issue. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:43 +00:00
rx_desc = next_rxd;
buffer_info = next_buffer;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
rx_ring->next_to_clean = i;
[PATCH] e1000: Added cleaned_count to RX buffer allocation Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:19 +00:00
cleaned_count = E1000_DESC_UNUSED(rx_ring);
if (cleaned_count)
adapter->alloc_rx_buf(adapter, rx_ring, cleaned_count);
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
e1000: add dynamic itr modes Add a new dynamic itr algorithm, with 2 modes, and make it the default operation mode. This greatly reduces latency and increases small packet performance, at the "cost" of some CPU utilization. Bulk traffic throughput is unaffected. The driver can limit the amount of interrupts per second that the adapter will generate for incoming packets. It does this by writing a value to the adapter that is based on the maximum amount of interrupts that the adapter will generate per second. Setting InterruptThrottleRate to a value greater or equal to 100 will program the adapter to send out a maximum of that many interrupts per second, even if more packets have come in. This reduces interrupt load on the system and can lower CPU utilization under heavy load, but will increase latency as packets are not processed as quickly. The default behaviour of the driver previously assumed a static InterruptThrottleRate value of 8000, providing a good fallback value for all traffic types,but lacking in small packet performance and latency. The hardware can handle many more small packets per second however, and for this reason an adaptive interrupt moderation algorithm was implemented. Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which it dynamically adjusts the InterruptThrottleRate value based on the traffic that it receives. After determining the type of incoming traffic in the last timeframe, it will adjust the InterruptThrottleRate to an appropriate value for that traffic. The algorithm classifies the incoming traffic every interval into classes. Once the class is determined, the InterruptThrottleRate value is adjusted to suit that traffic type the best. There are three classes defined: "Bulk traffic", for large amounts of packets of normal size; "Low latency", for small amounts of traffic and/or a significant percentage of small packets; and "Lowest latency", for almost completely small packets or minimal traffic. In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 for traffic that falls in class "Bulk traffic". If traffic falls in the "Low latency" or "Lowest latency" class, the InterruptThrottleRate is increased stepwise to 20000. This default mode is suitable for most applications. For situations where low latency is vital such as cluster or grid computing, the algorithm can reduce latency even more when InterruptThrottleRate is set to mode 1. In this mode, which operates the same as mode 3, the InterruptThrottleRate will be increased stepwise to 70000 for traffic in class "Lowest latency". Setting InterruptThrottleRate to 0 turns off any interrupt moderation and may improve small packet latency, but is generally not suitable for bulk throughput traffic. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-11-01 16:48:13 +00:00
adapter->total_rx_packets += total_rx_packets;
adapter->total_rx_bytes += total_rx_bytes;
e1000: Use the instance of net_device_stats from net_device. Since net_device has an instance of net_device_stats, we can remove the instance of this from the adapter structure. Signed-off-by: Ajit Khaparde <ajitk@serverengines.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-07 02:42:23 +00:00
netdev->stats.rx_bytes += total_rx_bytes;
netdev->stats.rx_packets += total_rx_packets;
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
return cleaned;
}
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
/**
* e1000_alloc_jumbo_rx_buffers - Replace used jumbo receive buffers
* @adapter: address of board private structure
* @rx_ring: pointer to receive ring structure
* @cleaned_count: number of buffers to allocate this pass
**/
static void
e1000_alloc_jumbo_rx_buffers(struct e1000_adapter *adapter,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
struct e1000_rx_ring *rx_ring, int cleaned_count)
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
{
struct net_device *netdev = adapter->netdev;
struct pci_dev *pdev = adapter->pdev;
struct e1000_rx_desc *rx_desc;
struct e1000_buffer *buffer_info;
struct sk_buff *skb;
unsigned int i;
net: Use netdev_alloc_skb_ip_align() Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-13 05:34:20 +00:00
unsigned int bufsz = 256 - 16 /*for skb_reserve */ ;
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
i = rx_ring->next_to_use;
buffer_info = &rx_ring->buffer_info[i];
while (cleaned_count--) {
skb = buffer_info->skb;
if (skb) {
skb_trim(skb, 0);
goto check_page;
}
net: Use netdev_alloc_skb_ip_align() Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-13 05:34:20 +00:00
skb = netdev_alloc_skb_ip_align(netdev, bufsz);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
if (unlikely(!skb)) {
/* Better luck next round */
adapter->alloc_rx_buff_failed++;
break;
}
buffer_info->skb = skb;
buffer_info->length = adapter->rx_buffer_len;
check_page:
/* allocate a new page if necessary */
if (!buffer_info->page) {
buffer_info->page = alloc_page(GFP_ATOMIC);
if (unlikely(!buffer_info->page)) {
adapter->alloc_rx_buff_failed++;
break;
}
}
e1000: Fix DMA mapping error handling on RX Check for error return from pci_map_single/pci_map_page and clean up. With this and the previous patch the driver was able to handle a significant percentage of errors (I set the fault injection rate to 10% and could still download large files at a reasonable speed). Signed-off-by: Anton Blanchard <anton@samba.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-02-19 17:54:53 +00:00
if (!buffer_info->dma) {
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
buffer_info->dma = dma_map_page(&pdev->dev,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
buffer_info->page, 0,
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
buffer_info->length,
DMA_FROM_DEVICE);
if (dma_mapping_error(&pdev->dev, buffer_info->dma)) {
e1000: Fix DMA mapping error handling on RX Check for error return from pci_map_single/pci_map_page and clean up. With this and the previous patch the driver was able to handle a significant percentage of errors (I set the fault injection rate to 10% and could still download large files at a reasonable speed). Signed-off-by: Anton Blanchard <anton@samba.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-02-19 17:54:53 +00:00
put_page(buffer_info->page);
dev_kfree_skb(skb);
buffer_info->page = NULL;
buffer_info->skb = NULL;
buffer_info->dma = 0;
adapter->alloc_rx_buff_failed++;
break; /* while !buffer_info->skb */
}
}
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
rx_desc = E1000_RX_DESC(*rx_ring, i);
rx_desc->buffer_addr = cpu_to_le64(buffer_info->dma);
if (unlikely(++i == rx_ring->count))
i = 0;
buffer_info = &rx_ring->buffer_info[i];
}
if (likely(rx_ring->next_to_use != i)) {
rx_ring->next_to_use = i;
if (unlikely(i-- == 0))
i = (rx_ring->count - 1);
/* Force memory writes to complete before letting h/w
* know there are new descriptors to fetch. (Only
* applicable for weak-ordered memory model archs,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* such as IA-64).
*/
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
wmb();
writel(i, adapter->hw.hw_addr + rx_ring->rdt);
}
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/**
[PATCH] e1000:82573 specific code & packet split code 82573 specific code & packet split code Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000.h net-drivers-2.6/drivers/net/e1000.new/e1000.h
2005-04-29 02:43:52 +00:00
* e1000_alloc_rx_buffers - Replace used receive buffers; legacy & extended
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* @adapter: address of board private structure
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_alloc_rx_buffers(struct e1000_adapter *adapter,
struct e1000_rx_ring *rx_ring,
int cleaned_count)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct net_device *netdev = adapter->netdev;
struct pci_dev *pdev = adapter->pdev;
struct e1000_rx_desc *rx_desc;
struct e1000_buffer *buffer_info;
struct sk_buff *skb;
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
unsigned int i;
net: Use netdev_alloc_skb_ip_align() Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-13 05:34:20 +00:00
unsigned int bufsz = adapter->rx_buffer_len;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
i = rx_ring->next_to_use;
buffer_info = &rx_ring->buffer_info[i];
[PATCH] e1000: Added copy break code Improves small packet performance with large amounts of reassembly being done in the stack. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:30 +00:00
while (cleaned_count--) {
e1000: clean up skb allocation code Signed-off-by: Christoph Hellwig <hch@lst.de> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-31 21:27:47 +00:00
skb = buffer_info->skb;
if (skb) {
[PATCH] e1000: Added copy break code Improves small packet performance with large amounts of reassembly being done in the stack. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:30 +00:00
skb_trim(skb, 0);
goto map_skb;
}
net: Use netdev_alloc_skb_ip_align() Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-13 05:34:20 +00:00
skb = netdev_alloc_skb_ip_align(netdev, bufsz);
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (unlikely(!skb)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Better luck next round */
[PATCH] e1000: Added cleaned_count to RX buffer allocation Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:19 +00:00
adapter->alloc_rx_buff_failed++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
}
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
/* Fix for errata 23, can't cross 64kB boundary */
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (!e1000_check_64k_bound(adapter, skb->data, bufsz)) {
struct sk_buff *oldskb = skb;
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(rx_err, "skb align check failed: %u bytes at "
"%p\n", bufsz, skb->data);
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
/* Try again, without freeing the previous */
net: Use netdev_alloc_skb_ip_align() Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-13 05:34:20 +00:00
skb = netdev_alloc_skb_ip_align(netdev, bufsz);
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
/* Failed allocation, critical failure */
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (!skb) {
dev_kfree_skb(oldskb);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
adapter->alloc_rx_buff_failed++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
}
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (!e1000_check_64k_bound(adapter, skb->data, bufsz)) {
/* give up */
dev_kfree_skb(skb);
dev_kfree_skb(oldskb);
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
adapter->alloc_rx_buff_failed++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break; /* while !buffer_info->skb */
}
e1000: clean up skb allocation code Signed-off-by: Christoph Hellwig <hch@lst.de> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-31 21:27:47 +00:00
/* Use new allocation */
dev_kfree_skb(oldskb);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
buffer_info->skb = skb;
buffer_info->length = adapter->rx_buffer_len;
[PATCH] e1000: Added copy break code Improves small packet performance with large amounts of reassembly being done in the stack. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:30 +00:00
map_skb:
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
buffer_info->dma = dma_map_single(&pdev->dev,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
skb->data,
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
buffer_info->length,
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
DMA_FROM_DEVICE);
if (dma_mapping_error(&pdev->dev, buffer_info->dma)) {
e1000: Fix DMA mapping error handling on RX Check for error return from pci_map_single/pci_map_page and clean up. With this and the previous patch the driver was able to handle a significant percentage of errors (I set the fault injection rate to 10% and could still download large files at a reasonable speed). Signed-off-by: Anton Blanchard <anton@samba.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-02-19 17:54:53 +00:00
dev_kfree_skb(skb);
buffer_info->skb = NULL;
buffer_info->dma = 0;
adapter->alloc_rx_buff_failed++;
break; /* while !buffer_info->skb */
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* XXX if it was allocated cleanly it will never map to a
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
* boundary crossing
*/
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
/* Fix for errata 23, can't cross 64kB boundary */
if (!e1000_check_64k_bound(adapter,
(void *)(unsigned long)buffer_info->dma,
adapter->rx_buffer_len)) {
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(rx_err, "dma align check failed: %u bytes at "
"%p\n", adapter->rx_buffer_len,
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
(void *)(unsigned long)buffer_info->dma);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
dev_kfree_skb(skb);
buffer_info->skb = NULL;
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
dma_unmap_single(&pdev->dev, buffer_info->dma,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
adapter->rx_buffer_len,
e1000: use DMA API instead of PCI DMA functions Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 13:08:45 +00:00
DMA_FROM_DEVICE);
e1000: fix unmap bug as reported by kerneloops.org [ 121.781161] ------------[ cut here ]------------ [ 121.781171] WARNING: at lib/dma-debug.c:793 check_unmap+0x14e/0x577() [ 121.781173] Hardware name: S5520HC [ 121.781177] e1000 0000:0a:00.0: DMA-API: device driver tries to free DMA memory it has not allocated [device address=0x00000001d688b0fa] [size=1522 bytes] [ 121.781180] Modules linked in: e1000 mdio dca [last unloaded: ixgbe] [ 121.781187] Pid: 4793, comm: bash Tainted: P 2.6.30-master-06161113 #3 [ 121.781190] Call Trace: [ 121.781195] [<ffffffff8123056f>] ? check_unmap+0x14e/0x577 [ 121.781201] [<ffffffff81057a19>] warn_slowpath_common+0x77/0x8f [ 121.781205] [<ffffffff81057ae1>] warn_slowpath_fmt+0x9f/0xa1 [ 121.781212] [<ffffffff81477ce2>] ? _spin_lock_irqsave+0x3f/0x49 [ 121.781216] [<ffffffff8122fa97>] ? get_hash_bucket+0x28/0x33 [ 121.781220] [<ffffffff8123056f>] check_unmap+0x14e/0x577 [ 121.781225] [<ffffffff810e4f48>] ? check_bytes_and_report+0x38/0xcb [ 121.781230] [<ffffffff81230bbf>] debug_dma_unmap_page+0x80/0x92 [ 121.781234] [<ffffffff8122e549>] ? unmap_single+0x1a/0x4e [ 121.781239] [<ffffffff813901e1>] ? __kfree_skb+0x74/0x78 [ 121.781250] [<ffffffffa00662ef>] pci_unmap_single+0x64/0x6d [e1000] [ 121.781259] [<ffffffffa0066344>] e1000_clean_rx_ring+0x4c/0xbf [e1000] [ 121.781268] [<ffffffffa00663df>] e1000_clean_all_rx_rings+0x28/0x36 [e1000] [ 121.781277] [<ffffffffa0067464>] e1000_down+0x138/0x141 [e1000] [ 121.781286] [<ffffffffa00681c2>] __e1000_shutdown+0x6b/0x198 [e1000] [ 121.781296] [<ffffffffa0068405>] e1000_suspend+0x17/0x50 [e1000] [ 121.781301] [<ffffffff81237665>] pci_legacy_suspend+0x3b/0xbe [ 121.781305] [<ffffffff81237bc6>] pci_pm_suspend+0x3e/0xf1 [ 121.781310] [<ffffffff812eaf1c>] pm_op+0x57/0xde [ 121.781314] [<ffffffff812eb444>] dpm_suspend_start+0x31e/0x470 [ 121.781319] [<ffffffff810877da>] suspend_devices_and_enter+0x3e/0x1a2 [ 121.781323] [<ffffffff81087a0f>] enter_state+0xd1/0x127 [ 121.781327] [<ffffffff8108717a>] state_store+0xa7/0xc9 [ 121.781332] [<ffffffff81221843>] kobj_attr_store+0x17/0x19 [ 121.781336] [<ffffffff8113c01e>] sysfs_write_file+0xe5/0x121 [ 121.781341] [<ffffffff810ed165>] vfs_write+0xab/0x105 [ 121.781344] [<ffffffff810ed279>] sys_write+0x47/0x6d [ 121.781349] [<ffffffff81027aab>] system_call_fastpath+0x16/0x1b [ 121.781352] ---[ end trace 97bacaaac2ed7786 ]--- Fix is to correctly zero out internal ->dma value when unmapping and make sure never to unmap unless there specifically was a mapping done. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-06-30 12:45:34 +00:00
buffer_info->dma = 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: implement jumbo receive with partial descriptors This is code extremely similar to what is committed in e1000e already. e1000 will no longer request 32kB slab buffers to support jumbo frames on PCI/PCI-X adapters. This will significantly reduce the likelyhood of order:3 allocation failures. This new code adds support for using pages as receive buffers, and the driver will chain multiple pages together to build a jumbo frame for OS consumption. The hardware takes a power of two buffer size and will dump as much data as it can receive into 1 or more buffers. The benefits of applying this are 1) stop akpm's dissing :-) of this lame e1000 behavior [1] 2) more efficient memory allocation (half) when using jumbo frames, which will also allow for much better socket utilization with jumbos since the socket is charged for the full allocation of each receive buffer, regardless of how much is used. 3) this was a feature request by a customer 4) copybreak for small packets < 256 bytes still applies [1] http://lkml.org/lkml/2008/7/10/68 http://article.gmane.org/gmane.linux.network/130986 Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 10:44:39 +00:00
adapter->alloc_rx_buff_failed++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break; /* while !buffer_info->skb */
}
rx_desc = E1000_RX_DESC(*rx_ring, i);
rx_desc->buffer_addr = cpu_to_le64(buffer_info->dma);
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (unlikely(++i == rx_ring->count))
i = 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
buffer_info = &rx_ring->buffer_info[i];
}
[PATCH] e1000: Added RX buffer enhancements Align the prefetches to a dword to help speed them up. Recycle skb's and early replenish. Force memory writes to complete before fetching more descriptors. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:32 +00:00
if (likely(rx_ring->next_to_use != i)) {
rx_ring->next_to_use = i;
if (unlikely(i-- == 0))
i = (rx_ring->count - 1);
/* Force memory writes to complete before letting h/w
* know there are new descriptors to fetch. (Only
* applicable for weak-ordered memory model archs,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* such as IA-64).
*/
[PATCH] e1000: Added RX buffer enhancements Align the prefetches to a dword to help speed them up. Recycle skb's and early replenish. Force memory writes to complete before fetching more descriptors. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:32 +00:00
wmb();
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
writel(i, hw->hw_addr + rx_ring->rdt);
[PATCH] e1000: Added RX buffer enhancements Align the prefetches to a dword to help speed them up. Recycle skb's and early replenish. Force memory writes to complete before fetching more descriptors. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:32 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/**
* e1000_smartspeed - Workaround for SmartSpeed on 82541 and 82547 controllers.
* @adapter:
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_smartspeed(struct e1000_adapter *adapter)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u16 phy_status;
u16 phy_ctrl;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if ((hw->phy_type != e1000_phy_igp) || !hw->autoneg ||
!(hw->autoneg_advertised & ADVERTISE_1000_FULL))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (adapter->smartspeed == 0) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* If Master/Slave config fault is asserted twice,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* we assume back-to-back
*/
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_read_phy_reg(hw, PHY_1000T_STATUS, &phy_status);
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (!(phy_status & SR_1000T_MS_CONFIG_FAULT)) return;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_read_phy_reg(hw, PHY_1000T_STATUS, &phy_status);
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (!(phy_status & SR_1000T_MS_CONFIG_FAULT)) return;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_read_phy_reg(hw, PHY_1000T_CTRL, &phy_ctrl);
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (phy_ctrl & CR_1000T_MS_ENABLE) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
phy_ctrl &= ~CR_1000T_MS_ENABLE;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_write_phy_reg(hw, PHY_1000T_CTRL,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
phy_ctrl);
adapter->smartspeed++;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (!e1000_phy_setup_autoneg(hw) &&
!e1000_read_phy_reg(hw, PHY_CTRL,
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
&phy_ctrl)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
phy_ctrl |= (MII_CR_AUTO_NEG_EN |
MII_CR_RESTART_AUTO_NEG);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_write_phy_reg(hw, PHY_CTRL,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
phy_ctrl);
}
}
return;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
} else if (adapter->smartspeed == E1000_SMARTSPEED_DOWNSHIFT) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* If still no link, perhaps using 2/3 pair cable */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_read_phy_reg(hw, PHY_1000T_CTRL, &phy_ctrl);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
phy_ctrl |= CR_1000T_MS_ENABLE;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_write_phy_reg(hw, PHY_1000T_CTRL, phy_ctrl);
if (!e1000_phy_setup_autoneg(hw) &&
!e1000_read_phy_reg(hw, PHY_CTRL, &phy_ctrl)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
phy_ctrl |= (MII_CR_AUTO_NEG_EN |
MII_CR_RESTART_AUTO_NEG);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_write_phy_reg(hw, PHY_CTRL, phy_ctrl);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
}
/* Restart process after E1000_SMARTSPEED_MAX iterations */
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (adapter->smartspeed++ == E1000_SMARTSPEED_MAX)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
adapter->smartspeed = 0;
}
/**
* e1000_ioctl -
* @netdev:
* @ifreq:
* @cmd:
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static int e1000_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
switch (cmd) {
case SIOCGMIIPHY:
case SIOCGMIIREG:
case SIOCSMIIREG:
return e1000_mii_ioctl(netdev, ifr, cmd);
default:
return -EOPNOTSUPP;
}
}
/**
* e1000_mii_ioctl -
* @netdev:
* @ifreq:
* @cmd:
**/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static int e1000_mii_ioctl(struct net_device *netdev, struct ifreq *ifr,
int cmd)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
[PATCH] e1000: Use netdev_priv() to get to netdev->priv Use netdev_priv() to get to netdev->priv - from shemminger@osdl.org Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:41:45 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct mii_ioctl_data *data = if_mii(ifr);
int retval;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u16 mii_reg;
[PATCH] e1000: Synchronize phy access between stats update and MII_IOCTL Synchronize phy access between stats update and MII_IOCTL Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:40:19 +00:00
unsigned long flags;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (hw->media_type != e1000_media_type_copper)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return -EOPNOTSUPP;
switch (cmd) {
case SIOCGMIIPHY:
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
data->phy_id = hw->phy_addr;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
case SIOCGMIIREG:
[PATCH] e1000: Synchronize phy access between stats update and MII_IOCTL Synchronize phy access between stats update and MII_IOCTL Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:40:19 +00:00
spin_lock_irqsave(&adapter->stats_lock, flags);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (e1000_read_phy_reg(hw, data->reg_num & 0x1F,
[PATCH] e1000: Synchronize phy access between stats update and MII_IOCTL Synchronize phy access between stats update and MII_IOCTL Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:40:19 +00:00
&data->val_out)) {
spin_unlock_irqrestore(&adapter->stats_lock, flags);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return -EIO;
[PATCH] e1000: Synchronize phy access between stats update and MII_IOCTL Synchronize phy access between stats update and MII_IOCTL Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:40:19 +00:00
}
spin_unlock_irqrestore(&adapter->stats_lock, flags);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
case SIOCSMIIREG:
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (data->reg_num & ~(0x1F))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return -EFAULT;
mii_reg = data->val_in;
[PATCH] e1000: Synchronize phy access between stats update and MII_IOCTL Synchronize phy access between stats update and MII_IOCTL Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:40:19 +00:00
spin_lock_irqsave(&adapter->stats_lock, flags);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (e1000_write_phy_reg(hw, data->reg_num,
[PATCH] e1000: Synchronize phy access between stats update and MII_IOCTL Synchronize phy access between stats update and MII_IOCTL Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:40:19 +00:00
mii_reg)) {
spin_unlock_irqrestore(&adapter->stats_lock, flags);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return -EIO;
[PATCH] e1000: Synchronize phy access between stats update and MII_IOCTL Synchronize phy access between stats update and MII_IOCTL Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:40:19 +00:00
}
[E1000]: Fix schedule while atomic when called from mii-tool. mii-tool can cause the driver to call msleep during nway reset, bugzilla.kernel.org bug 8430. Fix by simply calling reinit_locked outside of the spinlock, which is safe from ethtool, so it should be safe from here. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-11-14 05:00:09 +00:00
spin_unlock_irqrestore(&adapter->stats_lock, flags);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (hw->media_type == e1000_media_type_copper) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
switch (data->reg_num) {
case PHY_CTRL:
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (mii_reg & MII_CR_POWER_DOWN)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (mii_reg & MII_CR_AUTO_NEG_EN) {
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
hw->autoneg = 1;
hw->autoneg_advertised = 0x2F;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
} else {
net/igb/e1000/e1000e: more robust ethtool duplex/speed configuration This makes sure that one cannot request a 99Mbps full-duplex and get a 100Mbps half-duplex configuration in return due to the way the speed/duplex parameters are handled internally. Tested: e1000 works Signed-off-by: David Decotigny <decot@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-04-27 18:32:43 +00:00
u32 speed;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (mii_reg & 0x40)
net/igb/e1000/e1000e: more robust ethtool duplex/speed configuration This makes sure that one cannot request a 99Mbps full-duplex and get a 100Mbps half-duplex configuration in return due to the way the speed/duplex parameters are handled internally. Tested: e1000 works Signed-off-by: David Decotigny <decot@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-04-27 18:32:43 +00:00
speed = SPEED_1000;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
else if (mii_reg & 0x2000)
net/igb/e1000/e1000e: more robust ethtool duplex/speed configuration This makes sure that one cannot request a 99Mbps full-duplex and get a 100Mbps half-duplex configuration in return due to the way the speed/duplex parameters are handled internally. Tested: e1000 works Signed-off-by: David Decotigny <decot@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-04-27 18:32:43 +00:00
speed = SPEED_100;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
else
net/igb/e1000/e1000e: more robust ethtool duplex/speed configuration This makes sure that one cannot request a 99Mbps full-duplex and get a 100Mbps half-duplex configuration in return due to the way the speed/duplex parameters are handled internally. Tested: e1000 works Signed-off-by: David Decotigny <decot@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-04-27 18:32:43 +00:00
speed = SPEED_10;
retval = e1000_set_spd_dplx(
adapter, speed,
((mii_reg & 0x100)
? DUPLEX_FULL :
DUPLEX_HALF));
[E1000]: Fix schedule while atomic when called from mii-tool. mii-tool can cause the driver to call msleep during nway reset, bugzilla.kernel.org bug 8430. Fix by simply calling reinit_locked outside of the spinlock, which is safe from ethtool, so it should be safe from here. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-11-14 05:00:09 +00:00
if (retval)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return retval;
}
e1000: rework driver hardware reset locking After studying the driver mac reset code it was found that there were multiple race conditions possible to reset the unit twice or bring it e1000_up() double. This fixes all occurences where the driver needs to reset the mac. We also remove irq requesting/releasing into _open and _close so that while the device is _up we will never touch the irq's. This fixes the double free irq bug that people saw. To make sure that the watchdog task doesn't cause another race we let it run as a non-scheduled task. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:28 +00:00
if (netif_running(adapter->netdev))
e1000_reinit_locked(adapter);
else
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_reset(adapter);
break;
case M88E1000_PHY_SPEC_CTRL:
case M88E1000_EXT_PHY_SPEC_CTRL:
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (e1000_phy_reset(hw))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return -EIO;
break;
}
} else {
switch (data->reg_num) {
case PHY_CTRL:
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (mii_reg & MII_CR_POWER_DOWN)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
e1000: rework driver hardware reset locking After studying the driver mac reset code it was found that there were multiple race conditions possible to reset the unit twice or bring it e1000_up() double. This fixes all occurences where the driver needs to reset the mac. We also remove irq requesting/releasing into _open and _close so that while the device is _up we will never touch the irq's. This fixes the double free irq bug that people saw. To make sure that the watchdog task doesn't cause another race we let it run as a non-scheduled task. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:28 +00:00
if (netif_running(adapter->netdev))
e1000_reinit_locked(adapter);
else
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_reset(adapter);
break;
}
}
break;
default:
return -EOPNOTSUPP;
}
return E1000_SUCCESS;
}
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
void e1000_pci_set_mwi(struct e1000_hw *hw)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct e1000_adapter *adapter = hw->back;
[PATCH] e1000:Driver version,white space,comments,device id Driver version, white space, comments, device id & other Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_ethtool.c net-drivers-2.6/drivers/net/e1000.new/e1000_ethtool.c
2005-04-29 02:44:46 +00:00
int ret_val = pci_set_mwi(adapter->pdev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (ret_val)
e1000: use netif_<level> instead of netdev_<level> This patch restores the ability to set msglvl through ethtool. The issue was introduced by: commit 675ad47375c76a7c3be4ace9554d92cd55518ced Reported-by: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-07-27 06:37:21 +00:00
e_err(probe, "Error in setting MWI\n");
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
void e1000_pci_clear_mwi(struct e1000_hw *hw)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct e1000_adapter *adapter = hw->back;
pci_clear_mwi(adapter->pdev);
}
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
int e1000_pcix_get_mmrbc(struct e1000_hw *hw)
PCI-X/PCI-Express read control interfaces: use them in e1000 These driver changes incorporate the proposed PCI-X / PCI-Express read byte count interface. Reading and setting those valuse doesn't take place "manually", instead wrapping functions are called to allow quirks for some PCI bridges. Signed-off by: Peter Oruba <peter.oruba@amd.com> Based on work by Stephen Hemminger <shemminger@linux-foundation.org> Acked-by: Auke Kok <auke-jan.h.kok@intel.com> Cc: Jeff Garzik <jeff@garzik.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-09-29 05:42:06 +00:00
{
struct e1000_adapter *adapter = hw->back;
return pcix_get_mmrbc(adapter->pdev);
}
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
void e1000_pcix_set_mmrbc(struct e1000_hw *hw, int mmrbc)
PCI-X/PCI-Express read control interfaces: use them in e1000 These driver changes incorporate the proposed PCI-X / PCI-Express read byte count interface. Reading and setting those valuse doesn't take place "manually", instead wrapping functions are called to allow quirks for some PCI bridges. Signed-off by: Peter Oruba <peter.oruba@amd.com> Based on work by Stephen Hemminger <shemminger@linux-foundation.org> Acked-by: Auke Kok <auke-jan.h.kok@intel.com> Cc: Jeff Garzik <jeff@garzik.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-09-29 05:42:06 +00:00
{
struct e1000_adapter *adapter = hw->back;
pcix_set_mmrbc(adapter->pdev, mmrbc);
}
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
void e1000_io_write(struct e1000_hw *hw, unsigned long port, u32 value)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
outl(value, port);
}
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
static bool e1000_vlan_used(struct e1000_adapter *adapter)
{
u16 vid;
for_each_set_bit(vid, adapter->active_vlans, VLAN_N_VID)
return true;
return false;
}
e1000: fix vlan processing regression This patch fixes a regression introduced by commit "e1000: do vlan cleanup (799d531)". Apparently some e1000 chips (not mine) are sensitive about the order of setting vlan filter and vlan stripping/inserting functionality. So this patch changes the order so it's the same as before vlan cleanup. Reported-by: Ben Greear <greearb@candelatech.com> Signed-off-by: Jiri Pirko <jpirko@redhat.com> Tested-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-03-20 18:10:01 +00:00
static void __e1000_vlan_mode(struct e1000_adapter *adapter,
netdev_features_t features)
{
struct e1000_hw *hw = &adapter->hw;
u32 ctrl;
ctrl = er32(CTRL);
net: vlan: rename NETIF_F_HW_VLAN_* feature flags to NETIF_F_HW_VLAN_CTAG_* Rename the hardware VLAN acceleration features to include "CTAG" to indicate that they only support CTAGs. Follow up patches will introduce 802.1ad server provider tagging (STAGs) and require the distinction for hardware not supporting acclerating both. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-04-19 02:04:27 +00:00
if (features & NETIF_F_HW_VLAN_CTAG_RX) {
e1000: fix vlan processing regression This patch fixes a regression introduced by commit "e1000: do vlan cleanup (799d531)". Apparently some e1000 chips (not mine) are sensitive about the order of setting vlan filter and vlan stripping/inserting functionality. So this patch changes the order so it's the same as before vlan cleanup. Reported-by: Ben Greear <greearb@candelatech.com> Signed-off-by: Jiri Pirko <jpirko@redhat.com> Tested-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-03-20 18:10:01 +00:00
/* enable VLAN tag insert/strip */
ctrl |= E1000_CTRL_VME;
} else {
/* disable VLAN tag insert/strip */
ctrl &= ~E1000_CTRL_VME;
}
ew32(CTRL, ctrl);
}
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
static void e1000_vlan_filter_on_off(struct e1000_adapter *adapter,
bool filter_on)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
u32 rctl;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: remove irq_sem irq_sem was just a hack to prevent interrupts from being enabled unexpectedly in deep call paths. Simply finding those call paths and fixing them by hand results in a driver that behaves as we expect and doesn't need the atomic at all. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:58 +00:00
if (!test_bit(__E1000_DOWN, &adapter->flags))
e1000_irq_disable(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: fix vlan processing regression This patch fixes a regression introduced by commit "e1000: do vlan cleanup (799d531)". Apparently some e1000 chips (not mine) are sensitive about the order of setting vlan filter and vlan stripping/inserting functionality. So this patch changes the order so it's the same as before vlan cleanup. Reported-by: Ben Greear <greearb@candelatech.com> Signed-off-by: Jiri Pirko <jpirko@redhat.com> Tested-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-03-20 18:10:01 +00:00
__e1000_vlan_mode(adapter, adapter->netdev->features);
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
if (filter_on) {
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
/* enable VLAN receive filtering */
rctl = er32(RCTL);
rctl &= ~E1000_RCTL_CFIEN;
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
if (!(adapter->netdev->flags & IFF_PROMISC))
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
rctl |= E1000_RCTL_VFE;
ew32(RCTL, rctl);
e1000_update_mng_vlan(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
} else {
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
/* disable VLAN receive filtering */
rctl = er32(RCTL);
rctl &= ~E1000_RCTL_VFE;
ew32(RCTL, rctl);
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
}
e1000: Fix for e1000 kills IPMI on a tagged vlan. Enabling VLAN filters (VFE) when the primary interface is brought up (per commit 78ed11a) has caused problems for some users who manage their systems using IPMI over a VLAN. This is because when the driver enables the VLAN filter, this same filter table is enabled for the management channel, and the table is initially empty, which means that the IPMI/VLAN packets are filtered out and not received by the BMC. This is a problem only on e1000 class adapters, as it is only on e1000 that the filter table is common to the management and host streams. With this change, filtering is only enabled when one or more host VLANs exist, and is disabled when the last host VLAN is removed. VLAN filtering is always disabled when the primary interface is in promiscuous mode, and will be (re)enabled if VLANs exist when the interface exits promiscuous mode. Note that this does not completely resolve the issue for those using VLAN management, because if the host adds a VLAN, then the above problem occurs when that VLAN is enabled. However, it does mean the there is no problem for configurations where management is on a VLAN and the host is not. A complete solution to this issue would require further driver changes. The driver would need to discover if (and which) management VLANs are active before enabling VLAN filtering, so that it could ensure that the managed VLANs are included in the VLAN filter table. This discovery requires that the BMC identifies its VLAN in registers accessible to the driver, and at least on Dell PE2850 systems the BMC does not identify its VLAN to allow such discovery. Intel is pursuing this issue with the BMC vendor. Signed-off-by: Dave Graham <david.graham@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Krzysztof Piotr Oledzki <ole@ans.pl> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-08-31 14:12:51 +00:00
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
if (!test_bit(__E1000_DOWN, &adapter->flags))
e1000_irq_enable(adapter);
}
net: introduce and use netdev_features_t for device features sets v2: add couple missing conversions in drivers split unexporting netdev_fix_features() implemented %pNF convert sock::sk_route_(no?)caps Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-11-15 15:29:55 +00:00
static void e1000_vlan_mode(struct net_device *netdev,
e1000: fix vlan processing regression This patch fixes a regression introduced by commit "e1000: do vlan cleanup (799d531)". Apparently some e1000 chips (not mine) are sensitive about the order of setting vlan filter and vlan stripping/inserting functionality. So this patch changes the order so it's the same as before vlan cleanup. Reported-by: Ben Greear <greearb@candelatech.com> Signed-off-by: Jiri Pirko <jpirko@redhat.com> Tested-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-03-20 18:10:01 +00:00
netdev_features_t features)
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
{
struct e1000_adapter *adapter = netdev_priv(netdev);
if (!test_bit(__E1000_DOWN, &adapter->flags))
e1000_irq_disable(adapter);
e1000: fix vlan processing regression This patch fixes a regression introduced by commit "e1000: do vlan cleanup (799d531)". Apparently some e1000 chips (not mine) are sensitive about the order of setting vlan filter and vlan stripping/inserting functionality. So this patch changes the order so it's the same as before vlan cleanup. Reported-by: Ben Greear <greearb@candelatech.com> Signed-off-by: Jiri Pirko <jpirko@redhat.com> Tested-by: Ben Greear <greearb@candelatech.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-03-20 18:10:01 +00:00
__e1000_vlan_mode(adapter, features);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: remove irq_sem irq_sem was just a hack to prevent interrupts from being enabled unexpectedly in deep call paths. Simply finding those call paths and fixing them by hand results in a driver that behaves as we expect and doesn't need the atomic at all. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:58 +00:00
if (!test_bit(__E1000_DOWN, &adapter->flags))
e1000_irq_enable(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
net: vlan: prepare for 802.1ad VLAN filtering offload Change the rx_{add,kill}_vid callbacks to take a protocol argument in preparation of 802.1ad support. The protocol argument used so far is always htons(ETH_P_8021Q). Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-04-19 02:04:28 +00:00
static int e1000_vlan_rx_add_vid(struct net_device *netdev,
__be16 proto, u16 vid)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
[PATCH] e1000: Use netdev_priv() to get to netdev->priv Use netdev_priv() to get to netdev->priv - from shemminger@osdl.org Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:41:45 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u32 vfta, index;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if ((hw->mng_cookie.status &
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
E1000_MNG_DHCP_COOKIE_STATUS_VLAN_SUPPORT) &&
(vid == adapter->mng_vlan_id))
net: make vlan ndo_vlan_rx_[add/kill]_vid return error value Let caller know the result of adding/removing vlan id to/from vlan filter. In some drivers I make those functions to just return 0. But in those where there is able to see if hw setup went correctly, return value is set appropriately. Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-12-09 00:52:37 +00:00
return 0;
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
if (!e1000_vlan_used(adapter))
e1000_vlan_filter_on_off(adapter, true);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* add VID to filter table */
index = (vid >> 5) & 0x7F;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
vfta = E1000_READ_REG_ARRAY(hw, VFTA, index);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
vfta |= (1 << (vid & 0x1F));
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_write_vfta(hw, index, vfta);
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
set_bit(vid, adapter->active_vlans);
net: make vlan ndo_vlan_rx_[add/kill]_vid return error value Let caller know the result of adding/removing vlan id to/from vlan filter. In some drivers I make those functions to just return 0. But in those where there is able to see if hw setup went correctly, return value is set appropriately. Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-12-09 00:52:37 +00:00
return 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
net: vlan: prepare for 802.1ad VLAN filtering offload Change the rx_{add,kill}_vid callbacks to take a protocol argument in preparation of 802.1ad support. The protocol argument used so far is always htons(ETH_P_8021Q). Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-04-19 02:04:28 +00:00
static int e1000_vlan_rx_kill_vid(struct net_device *netdev,
__be16 proto, u16 vid)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
[PATCH] e1000: Use netdev_priv() to get to netdev->priv Use netdev_priv() to get to netdev->priv - from shemminger@osdl.org Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:41:45 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u32 vfta, index;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: remove irq_sem irq_sem was just a hack to prevent interrupts from being enabled unexpectedly in deep call paths. Simply finding those call paths and fixing them by hand results in a driver that behaves as we expect and doesn't need the atomic at all. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-21 18:06:58 +00:00
if (!test_bit(__E1000_DOWN, &adapter->flags))
e1000_irq_disable(adapter);
if (!test_bit(__E1000_DOWN, &adapter->flags))
e1000_irq_enable(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* remove VID from filter table */
index = (vid >> 5) & 0x7F;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
vfta = E1000_READ_REG_ARRAY(hw, VFTA, index);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
vfta &= ~(1 << (vid & 0x1F));
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
e1000_write_vfta(hw, index, vfta);
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
clear_bit(vid, adapter->active_vlans);
if (!e1000_vlan_used(adapter))
e1000_vlan_filter_on_off(adapter, false);
net: make vlan ndo_vlan_rx_[add/kill]_vid return error value Let caller know the result of adding/removing vlan id to/from vlan filter. In some drivers I make those functions to just return 0. But in those where there is able to see if hw setup went correctly, return value is set appropriately. Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-12-09 00:52:37 +00:00
return 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_restore_vlan(struct e1000_adapter *adapter)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
u16 vid;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: do vlan cleanup - unify vlan and nonvlan rx path - kill adapter->vlgrp and e1000_vlan_rx_register - allow to turn on/off rx/tx vlan accel via ethtool (set_features) Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-07-21 03:26:31 +00:00
if (!e1000_vlan_used(adapter))
return;
e1000_vlan_filter_on_off(adapter, true);
for_each_set_bit(vid, adapter->active_vlans, VLAN_N_VID)
net: vlan: prepare for 802.1ad VLAN filtering offload Change the rx_{add,kill}_vid callbacks to take a protocol argument in preparation of 802.1ad support. The protocol argument used so far is always htons(ETH_P_8021Q). Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-04-19 02:04:28 +00:00
e1000_vlan_rx_add_vid(adapter->netdev, htons(ETH_P_8021Q), vid);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
net/igb/e1000/e1000e: more robust ethtool duplex/speed configuration This makes sure that one cannot request a 99Mbps full-duplex and get a 100Mbps half-duplex configuration in return due to the way the speed/duplex parameters are handled internally. Tested: e1000 works Signed-off-by: David Decotigny <decot@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-04-27 18:32:43 +00:00
int e1000_set_spd_dplx(struct e1000_adapter *adapter, u32 spd, u8 dplx)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
hw->autoneg = 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
net/igb/e1000/e1000e: more robust ethtool duplex/speed configuration This makes sure that one cannot request a 99Mbps full-duplex and get a 100Mbps half-duplex configuration in return due to the way the speed/duplex parameters are handled internally. Tested: e1000 works Signed-off-by: David Decotigny <decot@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-04-27 18:32:43 +00:00
/* Make sure dplx is at most 1 bit and lsb of speed is not set
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
* for the switch() below to work
*/
net/igb/e1000/e1000e: more robust ethtool duplex/speed configuration This makes sure that one cannot request a 99Mbps full-duplex and get a 100Mbps half-duplex configuration in return due to the way the speed/duplex parameters are handled internally. Tested: e1000 works Signed-off-by: David Decotigny <decot@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-04-27 18:32:43 +00:00
if ((spd & 1) || (dplx & ~1))
goto err_inval;
[PATCH] e1000: Ethtool set speed/duplex validates parameters for consistency Ethtool set speed/duplex validates parameters for consistency Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:44:20 +00:00
/* Fiber NICs only allow 1000 gbps Full duplex */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if ((hw->media_type == e1000_media_type_fiber) &&
net/igb/e1000/e1000e: more robust ethtool duplex/speed configuration This makes sure that one cannot request a 99Mbps full-duplex and get a 100Mbps half-duplex configuration in return due to the way the speed/duplex parameters are handled internally. Tested: e1000 works Signed-off-by: David Decotigny <decot@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-04-27 18:32:43 +00:00
spd != SPEED_1000 &&
dplx != DUPLEX_FULL)
goto err_inval;
[PATCH] e1000: Ethtool set speed/duplex validates parameters for consistency Ethtool set speed/duplex validates parameters for consistency Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:44:20 +00:00
net/igb/e1000/e1000e: more robust ethtool duplex/speed configuration This makes sure that one cannot request a 99Mbps full-duplex and get a 100Mbps half-duplex configuration in return due to the way the speed/duplex parameters are handled internally. Tested: e1000 works Signed-off-by: David Decotigny <decot@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-04-27 18:32:43 +00:00
switch (spd + dplx) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
case SPEED_10 + DUPLEX_HALF:
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
hw->forced_speed_duplex = e1000_10_half;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
case SPEED_10 + DUPLEX_FULL:
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
hw->forced_speed_duplex = e1000_10_full;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
case SPEED_100 + DUPLEX_HALF:
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
hw->forced_speed_duplex = e1000_100_half;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
case SPEED_100 + DUPLEX_FULL:
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
hw->forced_speed_duplex = e1000_100_full;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
case SPEED_1000 + DUPLEX_FULL:
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
hw->autoneg = 1;
hw->autoneg_advertised = ADVERTISE_1000_FULL;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
case SPEED_1000 + DUPLEX_HALF: /* not supported */
default:
net/igb/e1000/e1000e: more robust ethtool duplex/speed configuration This makes sure that one cannot request a 99Mbps full-duplex and get a 100Mbps half-duplex configuration in return due to the way the speed/duplex parameters are handled internally. Tested: e1000 works Signed-off-by: David Decotigny <decot@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-04-27 18:32:43 +00:00
goto err_inval;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: configure and read MDI settings This is the implementation in e1000 to allow ethtool to force MDI state, allowing users to work around some improperly behaving switches. Forcing in this driver is for now only allowed when auto-neg is enabled. To use must have the matching version of ethtool app that supports this functionality. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> CC: Tushar Dave <tushar.n.dave@intel.com> Tested-by: Aaron Brown aaron.f.brown@intel.com Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-07-26 02:31:09 +00:00
/* clear MDI, MDI(-X) override is only allowed when autoneg enabled */
hw->mdix = AUTO_ALL_MODES;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return 0;
net/igb/e1000/e1000e: more robust ethtool duplex/speed configuration This makes sure that one cannot request a 99Mbps full-duplex and get a 100Mbps half-duplex configuration in return due to the way the speed/duplex parameters are handled internally. Tested: e1000 works Signed-off-by: David Decotigny <decot@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-04-27 18:32:43 +00:00
err_inval:
e_err(probe, "Unsupported Speed/Duplex configuration\n");
return -EINVAL;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
NET/e1000: Fix powering off during shutdown Prevent e1000 from putting the adapter into D3 during shutdown except when we're going to power off the system, since doing that may generally cause problems with kexec to happen (such problems were observed for igb and forcedeth). For this purpose seperate e1000_shutdown() from e1000_suspend() and use the appropriate PCI PM callbacks in both of them. Signed-off-by: "Rafael J. Wysocki" <rjw@sisk.pl> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-15 17:43:24 +00:00
static int __e1000_shutdown(struct pci_dev *pdev, bool *enable_wake)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct net_device *netdev = pci_get_drvdata(pdev);
[PATCH] e1000: Use netdev_priv() to get to netdev->priv Use netdev_priv() to get to netdev->priv - from shemminger@osdl.org Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:41:45 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u32 ctrl, ctrl_ext, rctl, status;
u32 wufc = adapter->wol;
e1000: fix CONFIG_PM blocks e1000_suspend is called even when !CONFIG_PM. The non-PM code inside of it is properly #ifdef'd. This fixes the compiler warnings when !CONFIG_PM. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:36 +00:00
#ifdef CONFIG_PM
[PATCH] e1000: Added variable to handle return values for pci_enable_* functions This was to fix compilation warnings. Also added log messages when pci_enable_* functions return with an error. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:28 +00:00
int retval = 0;
e1000: fix CONFIG_PM blocks e1000_suspend is called even when !CONFIG_PM. The non-PM code inside of it is properly #ifdef'd. This fixes the compiler warnings when !CONFIG_PM. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:36 +00:00
#endif
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
netif_device_detach(netdev);
e1000: rework driver hardware reset locking After studying the driver mac reset code it was found that there were multiple race conditions possible to reset the unit twice or bring it e1000_up() double. This fixes all occurences where the driver needs to reset the mac. We also remove irq requesting/releasing into _open and _close so that while the device is _up we will never touch the irq's. This fixes the double free irq bug that people saw. To make sure that the watchdog task doesn't cause another race we let it run as a non-scheduled task. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:28 +00:00
if (netif_running(netdev)) {
e1000: prevent oops when adapter is being closed and reset simultaneously This change is based on a similar change made to e1000e support in commit bb9e44d0d0f4 ("e1000e: prevent oops when adapter is being closed and reset simultaneously"). The same issue has also been observed on the older e1000 cards. Here, we have increased the RESET_COUNT value to 50 because there are too many accesses to e1000 nic on stress tests to e1000 nic, it is not enough to set RESET_COUT 25. Experimentation has shown that it is enough to set RESET_COUNT 50. Signed-off-by: yzhu1 <yanjun.zhu@windriver.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2013-11-23 07:07:40 +00:00
int count = E1000_CHECK_RESET_COUNT;
while (test_bit(__E1000_RESETTING, &adapter->flags) && count--)
usleep_range(10000, 20000);
e1000: rework driver hardware reset locking After studying the driver mac reset code it was found that there were multiple race conditions possible to reset the unit twice or bring it e1000_up() double. This fixes all occurences where the driver needs to reset the mac. We also remove irq requesting/releasing into _open and _close so that while the device is _up we will never touch the irq's. This fixes the double free irq bug that people saw. To make sure that the watchdog task doesn't cause another race we let it run as a non-scheduled task. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:28 +00:00
WARN_ON(test_bit(__E1000_RESETTING, &adapter->flags));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_down(adapter);
e1000: rework driver hardware reset locking After studying the driver mac reset code it was found that there were multiple race conditions possible to reset the unit twice or bring it e1000_up() double. This fixes all occurences where the driver needs to reset the mac. We also remove irq requesting/releasing into _open and _close so that while the device is _up we will never touch the irq's. This fixes the double free irq bug that people saw. To make sure that the watchdog task doesn't cause another race we let it run as a non-scheduled task. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-27 16:06:28 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[PATCH] e1000: Added functions to save and restore config These functions help restore the driver to active configuration when coming out of resume for power management. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:34 +00:00
#ifdef CONFIG_PM
e1000: remove obsolete custom pci_save_state code Now that 2.6.19 provides a proper implementation that saves MSI, PCI-E config space, we can have the e1000 driver use those instead of it's custom implementation. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-02-16 22:39:28 +00:00
retval = pci_save_state(pdev);
e1000: fix lockdep splat in shutdown handler As reported by Steven Rostedt, e1000 has a lockdep splat added during the recent merge window. The issue is that cancel_delayed_work is called while holding our private mutex. There is no reason that I can see to hold the mutex during pci shutdown, it was more just paranoia that I put the mutex_lock around the call to e1000_down. In a quick survey lots of drivers handle locking differently when being called by the pci layer. The assumption here is that we don't need the mutexes' protection in this function because the driver could not be unloaded while in the shutdown handler which is only called at reboot or poweroff. Reported-by: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: Steven Rostedt <rostedt@goodmis.org> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-01-04 20:23:33 +00:00
if (retval)
[PATCH] e1000: Added functions to save and restore config These functions help restore the driver to active configuration when coming out of resume for power management. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:34 +00:00
return retval;
#endif
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
status = er32(STATUS);
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (status & E1000_STATUS_LU)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
wufc &= ~E1000_WUFC_LNKC;
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (wufc) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_setup_rctl(adapter);
[E1000]: Secondary unicast address support Add support for configuring secondary unicast addresses. Unicast addresses take precendece over multicast addresses when filling the exact address filters to avoid going to promiscous mode. When more unicast addresses are present than filter slots, unicast filtering is disabled and all slots can be used for multicast addresses. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-11-14 04:54:59 +00:00
e1000_set_rx_mode(netdev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: add dropped DMA receive enable back in for WoL Commit d5bc77a223b0e9b9dfb002048d2b34a79e7d0b48 broke Wake-on-LAN by inadvertently dropping the enabling of DMA receives. Restore the enabling of DMA receives for WoL. This is applicable to 3.1+ stable trees. CC: stable@vger.stable.org Reported-by: Tobias Klausmann <klausman@schwarzvogel.de> Signed-off-by: Dean Nelson <dnelson@redhat.com> Tested-by: Tobias Klausmann <klausman@schwarzvogel.de> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-19 17:47:24 +00:00
rctl = er32(RCTL);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* turn on all-multi mode if wake on multicast is enabled */
e1000: add dropped DMA receive enable back in for WoL Commit d5bc77a223b0e9b9dfb002048d2b34a79e7d0b48 broke Wake-on-LAN by inadvertently dropping the enabling of DMA receives. Restore the enabling of DMA receives for WoL. This is applicable to 3.1+ stable trees. CC: stable@vger.stable.org Reported-by: Tobias Klausmann <klausman@schwarzvogel.de> Signed-off-by: Dean Nelson <dnelson@redhat.com> Tested-by: Tobias Klausmann <klausman@schwarzvogel.de> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-19 17:47:24 +00:00
if (wufc & E1000_WUFC_MC)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
rctl |= E1000_RCTL_MPE;
e1000: add dropped DMA receive enable back in for WoL Commit d5bc77a223b0e9b9dfb002048d2b34a79e7d0b48 broke Wake-on-LAN by inadvertently dropping the enabling of DMA receives. Restore the enabling of DMA receives for WoL. This is applicable to 3.1+ stable trees. CC: stable@vger.stable.org Reported-by: Tobias Klausmann <klausman@schwarzvogel.de> Signed-off-by: Dean Nelson <dnelson@redhat.com> Tested-by: Tobias Klausmann <klausman@schwarzvogel.de> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2012-01-19 17:47:24 +00:00
/* enable receives in the hardware */
ew32(RCTL, rctl | E1000_RCTL_EN);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (hw->mac_type >= e1000_82540) {
ctrl = er32(CTRL);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* advertise wake from D3Cold */
#define E1000_CTRL_ADVD3WUC 0x00100000
/* phy power management enable */
#define E1000_CTRL_EN_PHY_PWR_MGMT 0x00200000
ctrl |= E1000_CTRL_ADVD3WUC |
E1000_CTRL_EN_PHY_PWR_MGMT;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(CTRL, ctrl);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
if (hw->media_type == e1000_media_type_fiber ||
e1000: drop dead pcie code from e1000 this patch is the first in a series of clean up patches for e1000 to drop unused code, and update the driver to kernel spec, and then, to update the driver to have all available bug fixes. Call it the e1000 weight loss plan. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 12:16:14 +00:00
hw->media_type == e1000_media_type_internal_serdes) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* keep the laser running in D3 */
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ctrl_ext = er32(CTRL_EXT);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
ctrl_ext |= E1000_CTRL_EXT_SDP7_DATA;
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(CTRL_EXT, ctrl_ext);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(WUC, E1000_WUC_PME_EN);
ew32(WUFC, wufc);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
} else {
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(WUC, 0);
ew32(WUFC, 0);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
e1000: consolidate managability enabling/disabling Several bugs existed in how we handle manageability issues all over the driver. This patch consolidates all the managability release and init code in two single functions and call them from appropriate locations. This fixes several BMC packet redirect issues and powerup/down hiccups. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:56:10 +00:00
e1000_release_manageability(adapter);
NET/e1000: Fix powering off during shutdown Prevent e1000 from putting the adapter into D3 during shutdown except when we're going to power off the system, since doing that may generally cause problems with kexec to happen (such problems were observed for igb and forcedeth). For this purpose seperate e1000_shutdown() from e1000_suspend() and use the appropriate PCI PM callbacks in both of them. Signed-off-by: "Rafael J. Wysocki" <rjw@sisk.pl> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-15 17:43:24 +00:00
*enable_wake = !!wufc;
e1000: consolidate managability enabling/disabling Several bugs existed in how we handle manageability issues all over the driver. This patch consolidates all the managability release and init code in two single functions and call them from appropriate locations. This fixes several BMC packet redirect issues and powerup/down hiccups. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:56:10 +00:00
/* make sure adapter isn't asleep if manageability is enabled */
NET/e1000: Fix powering off during shutdown Prevent e1000 from putting the adapter into D3 during shutdown except when we're going to power off the system, since doing that may generally cause problems with kexec to happen (such problems were observed for igb and forcedeth). For this purpose seperate e1000_shutdown() from e1000_suspend() and use the appropriate PCI PM callbacks in both of them. Signed-off-by: "Rafael J. Wysocki" <rjw@sisk.pl> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-15 17:43:24 +00:00
if (adapter->en_mng_pt)
*enable_wake = true;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[PATCH] e1000: Fix regression: garbled stats and irq allocation during swsusp e1000: Fix suspend/resume powerup and irq allocation From: Auke Kok <auke-jan.h.kok@intel.com> After 7.0.33/2.6.16, e1000 suspend/resume left the user with an enabled device showing garbled statistics and undetermined irq allocation state, where `ifconfig eth0 down` would display `trying to free already freed irq`. Explicitly free and allocate irq as well as powerup the PHY during resume fixes when needed. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-11-06 16:57:12 +00:00
if (netif_running(netdev))
e1000_free_irq(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
pci_disable_device(pdev);
[PATCH] e1000: Added variable to handle return values for pci_enable_* functions This was to fix compilation warnings. Also added log messages when pci_enable_* functions return with an error. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-13 00:51:28 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return 0;
}
[PATCH] e1000: Added functions to save and restore config These functions help restore the driver to active configuration when coming out of resume for power management. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:34 +00:00
#ifdef CONFIG_PM
NET/e1000: Fix powering off during shutdown Prevent e1000 from putting the adapter into D3 during shutdown except when we're going to power off the system, since doing that may generally cause problems with kexec to happen (such problems were observed for igb and forcedeth). For this purpose seperate e1000_shutdown() from e1000_suspend() and use the appropriate PCI PM callbacks in both of them. Signed-off-by: "Rafael J. Wysocki" <rjw@sisk.pl> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-15 17:43:24 +00:00
static int e1000_suspend(struct pci_dev *pdev, pm_message_t state)
{
int retval;
bool wake;
retval = __e1000_shutdown(pdev, &wake);
if (retval)
return retval;
if (wake) {
pci_prepare_to_sleep(pdev);
} else {
pci_wake_from_d3(pdev, false);
pci_set_power_state(pdev, PCI_D3hot);
}
return 0;
}
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static int e1000_resume(struct pci_dev *pdev)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct net_device *netdev = pci_get_drvdata(pdev);
[PATCH] e1000: Use netdev_priv() to get to netdev->priv Use netdev_priv() to get to netdev->priv - from shemminger@osdl.org Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:41:45 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: convert uint16_t style integers to u16 Conglomerate from 4 separate patches from Joe. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-03 17:06:32 +00:00
u32 err;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Remove PM warning DPRINTKs breaking 2.4.x kernels remove DPRINTKs that were printing warnings about power management on 2.4 kernels. Since we really don't react differently these printk statements are not needed. This code was originally added to fix some compile time warnings that got fixed by newer kernels. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-04-15 02:04:40 +00:00
pci_set_power_state(pdev, PCI_D0);
e1000: remove obsolete custom pci_save_state code Now that 2.6.19 provides a proper implementation that saves MSI, PCI-E config space, we can have the e1000 driver use those instead of it's custom implementation. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-02-16 22:39:28 +00:00
pci_restore_state(pdev);
e1000: call pci_save_state after pci_restore_state This patch adds a call to pci_save_state() immediately after the call to pci_restore_state(). Due to a change in the behavior of pci_restore_state() it is necessary to call pci_save_state() to keep the state_saved flag. This patch is based on a similar patch for ixgbe. Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-02-03 14:49:48 +00:00
pci_save_state(pdev);
e1000: make ioport free This patch makes e1000 driver ioport-free. This corrects behavior in probe function so as not to request ioport resources as long as they are not really needed. This is based on the ioport-free patch of e1000 driver from Auke Kok and Tomohiro Kusumi. Signed-off-by: Tomohiro Kusumi <kusumi.tomohiro@jp.fujitsu.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Taku Izumi <izumi.taku@jp.fujitsu.com> Signed-off-by: Jeff Kirsher<jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:44 +00:00
if (adapter->need_ioport)
err = pci_enable_device(pdev);
else
err = pci_enable_device_mem(pdev);
e1000: Move assignments in tests before test Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:28 +00:00
if (err) {
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
pr_err("Cannot enable PCI device from suspend\n");
e1000: error out if we cannot enable PCI device on resume Do not ignore errors returned by pci_enable_device, instead error out. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-08-28 21:56:27 +00:00
return err;
}
[PATCH] e1000: e1000 stops working after resume e1000 stops working after resume, call pci_enable_device after pci_restore_state - Modified Andrew Morton's patch Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> diff -up net-drivers-2.6/drivers/net/e1000/e1000_main.c net-drivers-2.6/drivers/net/e1000.new/e1000_main.c
2005-04-29 02:41:28 +00:00
pci_set_master(pdev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Remove PM warning DPRINTKs breaking 2.4.x kernels remove DPRINTKs that were printing warnings about power management on 2.4 kernels. Since we really don't react differently these printk statements are not needed. This code was originally added to fix some compile time warnings that got fixed by newer kernels. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2006-04-15 02:04:40 +00:00
pci_enable_wake(pdev, PCI_D3hot, 0);
pci_enable_wake(pdev, PCI_D3cold, 0);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: Move assignments in tests before test Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:28 +00:00
if (netif_running(netdev)) {
err = e1000_request_irq(adapter);
if (err)
return err;
}
[PATCH] e1000: Fix regression: garbled stats and irq allocation during swsusp e1000: Fix suspend/resume powerup and irq allocation From: Auke Kok <auke-jan.h.kok@intel.com> After 7.0.33/2.6.16, e1000 suspend/resume left the user with an enabled device showing garbled statistics and undetermined irq allocation state, where `ifconfig eth0 down` would display `trying to free already freed irq`. Explicitly free and allocate irq as well as powerup the PHY during resume fixes when needed. Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-11-06 16:57:12 +00:00
e1000_power_up_phy(adapter);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_reset(adapter);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(WUS, ~0);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000: consolidate managability enabling/disabling Several bugs existed in how we handle manageability issues all over the driver. This patch consolidates all the managability release and init code in two single functions and call them from appropriate locations. This fixes several BMC packet redirect issues and powerup/down hiccups. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:56:10 +00:00
e1000_init_manageability(adapter);
[PATCH] e1000: Fix whitespace Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com> Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2006-01-18 21:01:39 +00:00
if (netif_running(netdev))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
e1000_up(adapter);
netif_device_attach(netdev);
return 0;
}
#endif
e1000: add shutdown handler back to fix WOL Someone was waaay too aggressive and removed e1000's reboot notifier instead of porting it to the new way of the shutdown handler. This change broke wake on lan. Add the shutdown handler back in using the same method as e100 uses. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-05-23 20:35:57 +00:00
static void e1000_shutdown(struct pci_dev *pdev)
{
NET/e1000: Fix powering off during shutdown Prevent e1000 from putting the adapter into D3 during shutdown except when we're going to power off the system, since doing that may generally cause problems with kexec to happen (such problems were observed for igb and forcedeth). For this purpose seperate e1000_shutdown() from e1000_suspend() and use the appropriate PCI PM callbacks in both of them. Signed-off-by: "Rafael J. Wysocki" <rjw@sisk.pl> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-15 17:43:24 +00:00
bool wake;
__e1000_shutdown(pdev, &wake);
if (system_state == SYSTEM_POWER_OFF) {
pci_wake_from_d3(pdev, wake);
pci_set_power_state(pdev, PCI_D3hot);
}
e1000: add shutdown handler back to fix WOL Someone was waaay too aggressive and removed e1000's reboot notifier instead of porting it to the new way of the shutdown handler. This change broke wake on lan. Add the shutdown handler back in using the same method as e100 uses. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-05-23 20:35:57 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#ifdef CONFIG_NET_POLL_CONTROLLER
e1000: fix whitespace issues and multi-line comments Fixes whitespace issues, such as lines exceeding 80 chars, needless blank lines and the use of spaces where tabs are needed. In addition, fix multi-line comments to align with the networking standard. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com>
2013-02-09 12:49:21 +00:00
/* Polling 'interrupt' - used by things like netconsole to send skbs
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* without having to re-enable interrupts. It's not called while
* the interrupt routine is executing.
*/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static void e1000_netpoll(struct net_device *netdev)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
[PATCH] e1000: Use netdev_priv() to get to netdev->priv Use netdev_priv() to get to netdev->priv - from shemminger@osdl.org Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com> Signed-off-by: Ganesh Venkatesan <ganesh.venkatesan@intel.com> Signed-off-by: John Ronciak <john.ronciak@intel.com>
2005-06-18 00:41:45 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
e1000: Redo netpoll fix to address community concerns The original suggested fix for netpoll was found to be racy on SMP kernels. While it is highly unlikely that this race would ever be seen in the real world due to current netpoll usage models, we implemented this updated fix to address concerns. Signed-off-by: Mitch Williams <mitch.a.williams@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-07-14 23:14:23 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
disable_irq(adapter->pdev->irq);
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 13:55:46 +00:00
e1000_intr(adapter->pdev->irq, netdev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
enable_irq(adapter->pdev->irq);
}
#endif
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
/**
* e1000_io_error_detected - called when PCI error is detected
* @pdev: Pointer to PCI device
e1000: updated whitespace and comments A large whitespace change to e1000_hw.[ch] in order to update it to kernel coding style (by running lindent). Updated function header comments into kdoc style. Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-25 22:19:46 +00:00
* @state: The current pci connection state
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
*
* This function is called after a PCI bus error affecting
* this device has been detected.
*/
e1000: neaten function declarations Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:02 +00:00
static pci_ers_result_t e1000_io_error_detected(struct pci_dev *pdev,
pci_channel_state_t state)
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
{
struct net_device *netdev = pci_get_drvdata(pdev);
netdevice: safe convert to netdev_priv() #part-2 We have some reasons to kill netdev->priv: 1. netdev->priv is equal to netdev_priv(). 2. netdev_priv() wraps the calculation of netdev->priv's offset, obviously netdev_priv() is more flexible than netdev->priv. But we cann't kill netdev->priv, because so many drivers reference to it directly. This patch is a safe convert for netdev->priv to netdev_priv(netdev). Since all of the netdev->priv is only for read. But it is too big to be sent in one mail. I split it to 4 parts and make every part smaller than 100,000 bytes, which is max size allowed by vger. Signed-off-by: Wang Chen <wangchen@cn.fujitsu.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-11-13 07:38:14 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
netif_device_detach(netdev);
e1000: return PCI_ERS_RESULT_DISCONNECT on permanent error PCI drivers that implement the io_error_detected callback should return PCI_ERS_RESULT_DISCONNECT if the state passed in is pci_channel_io_perm_failure. This state is not checked in many of the network drivers. The patch fixes the omission in the e1000 driver. Based on Mike Mason's similar patch for e1000e. Signed-off-by: Andre Detsch <adetsch@br.ibm.com> CC: Mike Mason <mmlnx@us.ibm.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-06-30 12:46:13 +00:00
if (state == pci_channel_io_perm_failure)
return PCI_ERS_RESULT_DISCONNECT;
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
if (netif_running(netdev))
e1000_down(adapter);
[PATCH] e1000 disable device on PCI error A recent patch in -mm3 titled "gregkh-pci-pci-don-t-enable-device-if-already-enabled.patch" causes pci_enable_device() to be a no-op if the kernel thinks that the device is already enabled. This change breaks the PCI error recovery mechanism in the e1000 device driver, since, after PCI slot reset, the card is no longer enabled. This is a trivial fix for this problem. Tested. Signed-off-by: Linas Vepstas <linas@austin.ibm.com> Cc: John Ronciak <john.ronciak@intel.com> Cc: Jesse Brandeburg <jesse.brandeburg@intel.com> Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Acked-by: Auke Kok <auke-jan.h.kok@intel.com> Cc: Jeff Garzik <jeff@garzik.org> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-09-19 03:58:06 +00:00
pci_disable_device(pdev);
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
/* Request a slot slot reset. */
return PCI_ERS_RESULT_NEED_RESET;
}
/**
* e1000_io_slot_reset - called after the pci bus has been reset.
* @pdev: Pointer to PCI device
*
* Restart the card from scratch, as if from a cold-boot. Implementation
* resembles the first-half of the e1000_resume routine.
*/
static pci_ers_result_t e1000_io_slot_reset(struct pci_dev *pdev)
{
struct net_device *netdev = pci_get_drvdata(pdev);
netdevice: safe convert to netdev_priv() #part-2 We have some reasons to kill netdev->priv: 1. netdev->priv is equal to netdev_priv(). 2. netdev_priv() wraps the calculation of netdev->priv's offset, obviously netdev_priv() is more flexible than netdev->priv. But we cann't kill netdev->priv, because so many drivers reference to it directly. This patch is a safe convert for netdev->priv to netdev_priv(netdev). Since all of the netdev->priv is only for read. But it is too big to be sent in one mail. I split it to 4 parts and make every part smaller than 100,000 bytes, which is max size allowed by vger. Signed-off-by: Wang Chen <wangchen@cn.fujitsu.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-11-13 07:38:14 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
struct e1000_hw *hw = &adapter->hw;
e1000: make ioport free This patch makes e1000 driver ioport-free. This corrects behavior in probe function so as not to request ioport resources as long as they are not really needed. This is based on the ioport-free patch of e1000 driver from Auke Kok and Tomohiro Kusumi. Signed-off-by: Tomohiro Kusumi <kusumi.tomohiro@jp.fujitsu.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Taku Izumi <izumi.taku@jp.fujitsu.com> Signed-off-by: Jeff Kirsher<jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:44 +00:00
int err;
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
e1000: make ioport free This patch makes e1000 driver ioport-free. This corrects behavior in probe function so as not to request ioport resources as long as they are not really needed. This is based on the ioport-free patch of e1000 driver from Auke Kok and Tomohiro Kusumi. Signed-off-by: Tomohiro Kusumi <kusumi.tomohiro@jp.fujitsu.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Taku Izumi <izumi.taku@jp.fujitsu.com> Signed-off-by: Jeff Kirsher<jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:44 +00:00
if (adapter->need_ioport)
err = pci_enable_device(pdev);
else
err = pci_enable_device_mem(pdev);
if (err) {
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
pr_err("Cannot re-enable PCI device after reset.\n");
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
return PCI_ERS_RESULT_DISCONNECT;
}
pci_set_master(pdev);
e1000: Janitor: Use #defined values for literals Minor janitorial patch: use #defines for literal values. Signed-off-by: Linas Vepstas <linas@austin.ibm.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-09-27 19:54:11 +00:00
pci_enable_wake(pdev, PCI_D3hot, 0);
pci_enable_wake(pdev, PCI_D3cold, 0);
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
e1000_reset(adapter);
e1000: Use hw, er32, and ew32 Use struct e1000_hw *hw = adapter->hw; where necessary Change macros E1000_READ_REG and E1000_WRITE_REG to er32 and ew32 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-11 22:17:08 +00:00
ew32(WUS, ~0);
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
return PCI_ERS_RESULT_RECOVERED;
}
/**
* e1000_io_resume - called when traffic can start flowing again.
* @pdev: Pointer to PCI device
*
* This callback is called when the error recovery driver tells us that
* its OK to resume normal operation. Implementation resembles the
* second-half of the e1000_resume routine.
*/
static void e1000_io_resume(struct pci_dev *pdev)
{
struct net_device *netdev = pci_get_drvdata(pdev);
netdevice: safe convert to netdev_priv() #part-2 We have some reasons to kill netdev->priv: 1. netdev->priv is equal to netdev_priv(). 2. netdev_priv() wraps the calculation of netdev->priv's offset, obviously netdev_priv() is more flexible than netdev->priv. But we cann't kill netdev->priv, because so many drivers reference to it directly. This patch is a safe convert for netdev->priv to netdev_priv(netdev). Since all of the netdev->priv is only for read. But it is too big to be sent in one mail. I split it to 4 parts and make every part smaller than 100,000 bytes, which is max size allowed by vger. Signed-off-by: Wang Chen <wangchen@cn.fujitsu.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-11-13 07:38:14 +00:00
struct e1000_adapter *adapter = netdev_priv(netdev);
e1000: consolidate managability enabling/disabling Several bugs existed in how we handle manageability issues all over the driver. This patch consolidates all the managability release and init code in two single functions and call them from appropriate locations. This fixes several BMC packet redirect issues and powerup/down hiccups. Originally from Jesse Brandeburg <jesse.brandeburg@intel.com>, rewritten to use feature flags by me. Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-12-15 15:56:10 +00:00
e1000_init_manageability(adapter);
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
if (netif_running(netdev)) {
if (e1000_up(adapter)) {
e1000: Use netdev_<level>, pr_<level> and dev_<level> This patch is an alternative to similar patch provided by Joe Perches. Substitute DPRINTK macro for e_<level> that uses netdev_<level> and dev_<level> similar to e1000e. - Convert printk to pr_<level> where applicable. - Use common #define pr_fmt for the driver. - Use dev_<level> for displaying text in parts of the driver where the interface name is not assigned (like e1000_param.c). - Better align test with the new macros. CC: Joe Perches <joe@perches.com> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-27 14:02:58 +00:00
pr_info("can't bring device back up after reset\n");
e1000: add PCI Error Recovery Various PCI bus errors can be signaled by newer PCI controllers. This patch adds the PCI error recovery callbacks to the intel gigabit ethernet e1000 device driver. The patch has been tested, and appears to work well. Signed-off-by: Linas Vepstas <linas@linas.org> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com>
2006-06-08 16:30:24 +00:00
return;
}
}
netif_device_attach(netdev);
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* e1000_main.c */
Reference in New Issue Copy Permalink