leandrof/linux
1
0
Fork 0
forked from Minki/linux
Code Issues Pull Requests Packages Projects Releases Wiki Activity
eca3930163
linux/drivers/net/ibm_newemac/core.c

3111 lines
81 KiB
C
Raw Normal View History

Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
/*
* drivers/net/ibm_newemac/core.c
*
* Driver for PowerPC 4xx on-chip ethernet controller.
*
ibm_newemac: Update file headers copyright notices This updates the copyright notices of the new EMAC driver to avoid confusion as who is to be blamed for new bugs. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-05 00:14:33 +00:00
* Copyright 2007 Benjamin Herrenschmidt, IBM Corp.
* <benh@kernel.crashing.org>
*
* Based on the arch/ppc version of the driver:
*
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
* Copyright (c) 2004, 2005 Zultys Technologies.
* Eugene Surovegin <eugene.surovegin@zultys.com> or <ebs@ebshome.net>
*
* Based on original work by
* Matt Porter <mporter@kernel.crashing.org>
* (c) 2003 Benjamin Herrenschmidt <benh@kernel.crashing.org>
* Armin Kuster <akuster@mvista.com>
* Johnnie Peters <jpeters@mvista.com>
*
* This program is free software; you can redistribute it and/or modify it
* under the terms of the GNU General Public License as published by the
* Free Software Foundation; either version 2 of the License, or (at your
* option) any later version.
*
*/
net: Fix OF platform drivers coldplug/hotplug when compiled as modules Some OF platform drivers are missing module device tables, so they won't load automatically on boot. This patch fixes the issue by adding proper MODULE_DEVICE_TABLE() macros to the drivers. Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-14 21:54:52 +00:00
#include <linux/module.h>
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
#include <linux/sched.h>
#include <linux/string.h>
#include <linux/errno.h>
#include <linux/delay.h>
#include <linux/types.h>
#include <linux/pci.h>
#include <linux/etherdevice.h>
#include <linux/skbuff.h>
#include <linux/crc32.h>
#include <linux/ethtool.h>
#include <linux/mii.h>
#include <linux/bitops.h>
#include <linux/workqueue.h>
[POWERPC] Add of_find_matching_node() helper function Similar to of_find_compatible_node(), of_find_matching_node() and for_each_matching_node() allow you to iterate over the device tree looking for specific nodes, except that they take of_device_id tables instead of strings. This also moves of_match_node() from driver/of/device.c to driver/of/base.c to colocate it with the of_find_matching_node which depends on it. Signed-off-by: Grant Likely <grant.likely@secretlab.ca> Signed-off-by: Paul Mackerras <paulus@samba.org>
2008-01-08 19:20:40 +00:00
#include <linux/of.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 08:04:11 +00:00
#include <linux/slab.h>
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
#include <asm/processor.h>
#include <asm/io.h>
#include <asm/dma.h>
#include <asm/uaccess.h>
ibm_newemac: PowerPC 440GX EMAC PHY clock workaround The PowerPC 440GX Taishan board fails to reset EMAC3 (reset timeout error) if there's no link. Because of that it fails to find PHY chip. The older ibm_emac driver had a workaround for that: the EMAC_CLK_INTERNAL/EMAC_CLK_EXTERNAL macros, which toggle the Ethernet Clock Select bit in the SDR0_MFR register. This patch does the same for "ibm,emac-440gx" compatible chips. The workaround forces clock on -all- EMACs, so we select clock under global emac_phy_map_lock. BenH: Made that #ifdef CONFIG_PPC_DCR_NATIVE for now as dcri_* stuff doesn't exist for MMIO type DCRs like Cell. Some future rework & improvements of the DCR infrastructure will make that cleaner but for now, this makes it work. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:46 +00:00
#include <asm/dcr.h>
#include <asm/dcr-regs.h>
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
#include "core.h"
/*
* Lack of dma_unmap_???? calls is intentional.
*
* API-correct usage requires additional support state information to be
* maintained for every RX and TX buffer descriptor (BD). Unfortunately, due to
* EMAC design (e.g. TX buffer passed from network stack can be split into
* several BDs, dma_map_single/dma_map_page can be used to map particular BD),
* maintaining such information will add additional overhead.
* Current DMA API implementation for 4xx processors only ensures cache coherency
* and dma_unmap_???? routines are empty and are likely to stay this way.
* I decided to omit dma_unmap_??? calls because I don't want to add additional
* complexity just for the sake of following some abstract API, when it doesn't
* add any real benefit to the driver. I understand that this decision maybe
* controversial, but I really tried to make code API-correct and efficient
* at the same time and didn't come up with code I liked :(. --ebs
*/
#define DRV_NAME "emac"
#define DRV_VERSION "3.54"
#define DRV_DESC "PPC 4xx OCP EMAC driver"
MODULE_DESCRIPTION(DRV_DESC);
MODULE_AUTHOR
("Eugene Surovegin <eugene.surovegin@zultys.com> or <ebs@ebshome.net>");
MODULE_LICENSE("GPL");
/*
* PPC64 doesn't (yet) have a cacheable_memcpy
*/
#ifdef CONFIG_PPC64
#define cacheable_memcpy(d,s,n) memcpy((d),(s),(n))
#endif
/* minimum number of free TX descriptors required to wake up TX process */
#define EMAC_TX_WAKEUP_THRESH (NUM_TX_BUFF / 4)
/* If packet size is less than this number, we allocate small skb and copy packet
* contents into it instead of just sending original big skb up
*/
#define EMAC_RX_COPY_THRESH CONFIG_IBM_NEW_EMAC_RX_COPY_THRESHOLD
/* Since multiple EMACs share MDIO lines in various ways, we need
* to avoid re-using the same PHY ID in cases where the arch didn't
* setup precise phy_map entries
*
* XXX This is something that needs to be reworked as we can have multiple
* EMAC "sets" (multiple ASICs containing several EMACs) though we can
* probably require in that case to have explicit PHY IDs in the device-tree
*/
static u32 busy_phy_map;
static DEFINE_MUTEX(emac_phy_map_lock);
/* This is the wait queue used to wait on any event related to probe, that
* is discovery of MALs, other EMACs, ZMII/RGMIIs, etc...
*/
static DECLARE_WAIT_QUEUE_HEAD(emac_probe_wait);
/* Having stable interface names is a doomed idea. However, it would be nice
* if we didn't have completely random interface names at boot too :-) It's
* just a matter of making everybody's life easier. Since we are doing
* threaded probing, it's a bit harder though. The base idea here is that
* we make up a list of all emacs in the device-tree before we register the
* driver. Every emac will then wait for the previous one in the list to
* initialize before itself. We should also keep that list ordered by
* cell_index.
* That list is only 4 entries long, meaning that additional EMACs don't
* get ordering guarantees unless EMAC_BOOT_LIST_SIZE is increased.
*/
#define EMAC_BOOT_LIST_SIZE 4
static struct device_node *emac_boot_list[EMAC_BOOT_LIST_SIZE];
/* How long should I wait for dependent devices ? */
#define EMAC_PROBE_DEP_TIMEOUT (HZ * 5)
/* I don't want to litter system log with timeout errors
* when we have brain-damaged PHY.
*/
static inline void emac_report_timeout_error(struct emac_instance *dev,
const char *error)
{
ibm_newemac: PowerPC 440EP/440GR EMAC PHY clock workaround This patch adds ibm_newemac PHY clock workaround for 440EP/440GR EMAC attached to a PHY which doesn't generate RX clock if there is no link. The code is based on the previous ibm_emac driver stuff. The 440EP/440GR allows controlling each EMAC clock separately as opposed to global clock selection for 440GX. BenH: Made that #ifdef CONFIG_PPC_DCR_NATIVE for now as dcri_* stuff doesn't exist for MMIO type DCRs like Cell. Some future rework & improvements of the DCR infrastructure will make that cleaner but for now, this makes it work. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:48 +00:00
if (emac_has_feature(dev, EMAC_FTR_440GX_PHY_CLK_FIX |
ibm_newemac: Add support for GPCS, SGMII and M88E1112 PHY Add support for the phy types found on the Arches and other PowerPC 460 based boards. Signed-off-by: Victor Gallardo <vgallardo@amcc.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-10-02 06:37:57 +00:00
EMAC_FTR_460EX_PHY_CLK_FIX |
ibm_newemac: PowerPC 440EP/440GR EMAC PHY clock workaround This patch adds ibm_newemac PHY clock workaround for 440EP/440GR EMAC attached to a PHY which doesn't generate RX clock if there is no link. The code is based on the previous ibm_emac driver stuff. The 440EP/440GR allows controlling each EMAC clock separately as opposed to global clock selection for 440GX. BenH: Made that #ifdef CONFIG_PPC_DCR_NATIVE for now as dcri_* stuff doesn't exist for MMIO type DCRs like Cell. Some future rework & improvements of the DCR infrastructure will make that cleaner but for now, this makes it work. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:48 +00:00
EMAC_FTR_440EP_PHY_CLK_FIX))
DBG(dev, "%s" NL, error);
else if (net_ratelimit())
of: Always use 'struct device.of_node' to get device node pointer. The following structure elements duplicate the information in 'struct device.of_node' and so are being eliminated. This patch makes all readers of these elements use device.of_node instead. (struct of_device *)->node (struct dev_archdata *)->prom_node (sparc) (struct dev_archdata *)->of_node (powerpc & microblaze) Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2010-04-13 23:12:29 +00:00
printk(KERN_ERR "%s: %s\n", dev->ofdev->dev.of_node->full_name,
error);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
}
ibm_newemac: PowerPC 440EP/440GR EMAC PHY clock workaround This patch adds ibm_newemac PHY clock workaround for 440EP/440GR EMAC attached to a PHY which doesn't generate RX clock if there is no link. The code is based on the previous ibm_emac driver stuff. The 440EP/440GR allows controlling each EMAC clock separately as opposed to global clock selection for 440GX. BenH: Made that #ifdef CONFIG_PPC_DCR_NATIVE for now as dcri_* stuff doesn't exist for MMIO type DCRs like Cell. Some future rework & improvements of the DCR infrastructure will make that cleaner but for now, this makes it work. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:48 +00:00
/* EMAC PHY clock workaround:
* 440EP/440GR has more sane SDR0_MFR register implementation than 440GX,
* which allows controlling each EMAC clock
*/
static inline void emac_rx_clk_tx(struct emac_instance *dev)
{
#ifdef CONFIG_PPC_DCR_NATIVE
if (emac_has_feature(dev, EMAC_FTR_440EP_PHY_CLK_FIX))
dcri_clrset(SDR0, SDR0_MFR,
0, SDR0_MFR_ECS >> dev->cell_index);
#endif
}
static inline void emac_rx_clk_default(struct emac_instance *dev)
{
#ifdef CONFIG_PPC_DCR_NATIVE
if (emac_has_feature(dev, EMAC_FTR_440EP_PHY_CLK_FIX))
dcri_clrset(SDR0, SDR0_MFR,
SDR0_MFR_ECS >> dev->cell_index, 0);
#endif
}
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
/* PHY polling intervals */
#define PHY_POLL_LINK_ON HZ
#define PHY_POLL_LINK_OFF (HZ / 5)
/* Graceful stop timeouts in us.
* We should allow up to 1 frame time (full-duplex, ignoring collisions)
*/
#define STOP_TIMEOUT_10 1230
#define STOP_TIMEOUT_100 124
#define STOP_TIMEOUT_1000 13
#define STOP_TIMEOUT_1000_JUMBO 73
NEWEMAC: fix support for pause packets Problem Description and Fix --------------------------- When a pause packet(with destination as reserved Multicast address) is received by the EMAC hardware to control the flow of frames being transmitted by it, it is dropped by the hardware unless the reserved Multicast address is hashed in to the GAHT[1-4] registers. This code fix adds the default reserved multicast address to the GAHT[1-4] registers in the EMAC(s) present on the chip. The flow control with Pause packets will only work if the following register bits are programmed in EMAC: EMACx_MR1[APP] = 1 EMACx_RMR[BAE] = 1 EMACx_RMR[MAE] = 1 Behavior that may be observed in a running system ------------------------------------------------- A host transferring data from a PPC based system may send a Pause packet to the PPC EMAC requesting it to slow down the flow of packets. If the default reserved multicast MAC address is not programmed into the GAHT[1-4] registers this Pause packet will be dropped by PPC EMAC and no Flow Control will be done. Signed-off-by: Pravin M. Bathija <pbathija@amcc.com> Signed-off-by: Stefan Roese <sr@denx.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-14 09:52:28 +00:00
static unsigned char default_mcast_addr[] = {
0x01, 0x80, 0xC2, 0x00, 0x00, 0x01
};
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
/* Please, keep in sync with struct ibm_emac_stats/ibm_emac_error_stats */
static const char emac_stats_keys[EMAC_ETHTOOL_STATS_COUNT][ETH_GSTRING_LEN] = {
"rx_packets", "rx_bytes", "tx_packets", "tx_bytes", "rx_packets_csum",
"tx_packets_csum", "tx_undo", "rx_dropped_stack", "rx_dropped_oom",
"rx_dropped_error", "rx_dropped_resize", "rx_dropped_mtu",
"rx_stopped", "rx_bd_errors", "rx_bd_overrun", "rx_bd_bad_packet",
"rx_bd_runt_packet", "rx_bd_short_event", "rx_bd_alignment_error",
"rx_bd_bad_fcs", "rx_bd_packet_too_long", "rx_bd_out_of_range",
"rx_bd_in_range", "rx_parity", "rx_fifo_overrun", "rx_overrun",
"rx_bad_packet", "rx_runt_packet", "rx_short_event",
"rx_alignment_error", "rx_bad_fcs", "rx_packet_too_long",
"rx_out_of_range", "rx_in_range", "tx_dropped", "tx_bd_errors",
"tx_bd_bad_fcs", "tx_bd_carrier_loss", "tx_bd_excessive_deferral",
"tx_bd_excessive_collisions", "tx_bd_late_collision",
"tx_bd_multple_collisions", "tx_bd_single_collision",
"tx_bd_underrun", "tx_bd_sqe", "tx_parity", "tx_underrun", "tx_sqe",
"tx_errors"
};
static irqreturn_t emac_irq(int irq, void *dev_instance);
static void emac_clean_tx_ring(struct emac_instance *dev);
static void __emac_set_multicast_list(struct emac_instance *dev);
static inline int emac_phy_supports_gige(int phy_mode)
{
return phy_mode == PHY_MODE_GMII ||
phy_mode == PHY_MODE_RGMII ||
ibm_newemac: Add support for GPCS, SGMII and M88E1112 PHY Add support for the phy types found on the Arches and other PowerPC 460 based boards. Signed-off-by: Victor Gallardo <vgallardo@amcc.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-10-02 06:37:57 +00:00
phy_mode == PHY_MODE_SGMII ||
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
phy_mode == PHY_MODE_TBI ||
phy_mode == PHY_MODE_RTBI;
}
static inline int emac_phy_gpcs(int phy_mode)
{
ibm_newemac: Add support for GPCS, SGMII and M88E1112 PHY Add support for the phy types found on the Arches and other PowerPC 460 based boards. Signed-off-by: Victor Gallardo <vgallardo@amcc.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-10-02 06:37:57 +00:00
return phy_mode == PHY_MODE_SGMII ||
phy_mode == PHY_MODE_TBI ||
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
phy_mode == PHY_MODE_RTBI;
}
static inline void emac_tx_enable(struct emac_instance *dev)
{
struct emac_regs __iomem *p = dev->emacp;
u32 r;
DBG(dev, "tx_enable" NL);
r = in_be32(&p->mr0);
if (!(r & EMAC_MR0_TXE))
out_be32(&p->mr0, r | EMAC_MR0_TXE);
}
static void emac_tx_disable(struct emac_instance *dev)
{
struct emac_regs __iomem *p = dev->emacp;
u32 r;
DBG(dev, "tx_disable" NL);
r = in_be32(&p->mr0);
if (r & EMAC_MR0_TXE) {
int n = dev->stop_timeout;
out_be32(&p->mr0, r & ~EMAC_MR0_TXE);
while (!(in_be32(&p->mr0) & EMAC_MR0_TXI) && n) {
udelay(1);
--n;
}
if (unlikely(!n))
emac_report_timeout_error(dev, "TX disable timeout");
}
}
static void emac_rx_enable(struct emac_instance *dev)
{
struct emac_regs __iomem *p = dev->emacp;
u32 r;
if (unlikely(test_bit(MAL_COMMAC_RX_STOPPED, &dev->commac.flags)))
goto out;
DBG(dev, "rx_enable" NL);
r = in_be32(&p->mr0);
if (!(r & EMAC_MR0_RXE)) {
if (unlikely(!(r & EMAC_MR0_RXI))) {
/* Wait if previous async disable is still in progress */
int n = dev->stop_timeout;
while (!(r = in_be32(&p->mr0) & EMAC_MR0_RXI) && n) {
udelay(1);
--n;
}
if (unlikely(!n))
emac_report_timeout_error(dev,
"RX disable timeout");
}
out_be32(&p->mr0, r | EMAC_MR0_RXE);
}
out:
;
}
static void emac_rx_disable(struct emac_instance *dev)
{
struct emac_regs __iomem *p = dev->emacp;
u32 r;
DBG(dev, "rx_disable" NL);
r = in_be32(&p->mr0);
if (r & EMAC_MR0_RXE) {
int n = dev->stop_timeout;
out_be32(&p->mr0, r & ~EMAC_MR0_RXE);
while (!(in_be32(&p->mr0) & EMAC_MR0_RXI) && n) {
udelay(1);
--n;
}
if (unlikely(!n))
emac_report_timeout_error(dev, "RX disable timeout");
}
}
static inline void emac_netif_stop(struct emac_instance *dev)
{
netif_tx_lock_bh(dev->ndev);
netdev: Add netdev->addr_list_lock protection. Add netif_addr_{lock,unlock}{,_bh}() helpers. Use them to protect operations that operate on or read the network device unicast and multicast address lists. Also use them in cases where the code simply wants to block calls into the driver's ->set_rx_mode() and ->set_multicast_list() methods. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-15 07:13:44 +00:00
netif_addr_lock(dev->ndev);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
dev->no_mcast = 1;
netdev: Add netdev->addr_list_lock protection. Add netif_addr_{lock,unlock}{,_bh}() helpers. Use them to protect operations that operate on or read the network device unicast and multicast address lists. Also use them in cases where the code simply wants to block calls into the driver's ->set_rx_mode() and ->set_multicast_list() methods. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-15 07:13:44 +00:00
netif_addr_unlock(dev->ndev);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
netif_tx_unlock_bh(dev->ndev);
dev->ndev->trans_start = jiffies; /* prevent tx timeout */
mal_poll_disable(dev->mal, &dev->commac);
netif_tx_disable(dev->ndev);
}
static inline void emac_netif_start(struct emac_instance *dev)
{
netif_tx_lock_bh(dev->ndev);
netdev: Add netdev->addr_list_lock protection. Add netif_addr_{lock,unlock}{,_bh}() helpers. Use them to protect operations that operate on or read the network device unicast and multicast address lists. Also use them in cases where the code simply wants to block calls into the driver's ->set_rx_mode() and ->set_multicast_list() methods. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-15 07:13:44 +00:00
netif_addr_lock(dev->ndev);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
dev->no_mcast = 0;
if (dev->mcast_pending && netif_running(dev->ndev))
__emac_set_multicast_list(dev);
netdev: Add netdev->addr_list_lock protection. Add netif_addr_{lock,unlock}{,_bh}() helpers. Use them to protect operations that operate on or read the network device unicast and multicast address lists. Also use them in cases where the code simply wants to block calls into the driver's ->set_rx_mode() and ->set_multicast_list() methods. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-15 07:13:44 +00:00
netif_addr_unlock(dev->ndev);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
netif_tx_unlock_bh(dev->ndev);
netif_wake_queue(dev->ndev);
/* NOTE: unconditional netif_wake_queue is only appropriate
* so long as all callers are assured to have free tx slots
* (taken from tg3... though the case where that is wrong is
* not terribly harmful)
*/
mal_poll_enable(dev->mal, &dev->commac);
}
static inline void emac_rx_disable_async(struct emac_instance *dev)
{
struct emac_regs __iomem *p = dev->emacp;
u32 r;
DBG(dev, "rx_disable_async" NL);
r = in_be32(&p->mr0);
if (r & EMAC_MR0_RXE)
out_be32(&p->mr0, r & ~EMAC_MR0_RXE);
}
static int emac_reset(struct emac_instance *dev)
{
struct emac_regs __iomem *p = dev->emacp;
int n = 20;
DBG(dev, "reset" NL);
if (!dev->reset_failed) {
/* 40x erratum suggests stopping RX channel before reset,
* we stop TX as well
*/
emac_rx_disable(dev);
emac_tx_disable(dev);
}
ibm_newemac: Add support for GPCS, SGMII and M88E1112 PHY Add support for the phy types found on the Arches and other PowerPC 460 based boards. Signed-off-by: Victor Gallardo <vgallardo@amcc.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-10-02 06:37:57 +00:00
#ifdef CONFIG_PPC_DCR_NATIVE
/* Enable internal clock source */
if (emac_has_feature(dev, EMAC_FTR_460EX_PHY_CLK_FIX))
dcri_clrset(SDR0, SDR0_ETH_CFG,
0, SDR0_ETH_CFG_ECS << dev->cell_index);
#endif
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
out_be32(&p->mr0, EMAC_MR0_SRST);
while ((in_be32(&p->mr0) & EMAC_MR0_SRST) && n)
--n;
ibm_newemac: Add support for GPCS, SGMII and M88E1112 PHY Add support for the phy types found on the Arches and other PowerPC 460 based boards. Signed-off-by: Victor Gallardo <vgallardo@amcc.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-10-02 06:37:57 +00:00
#ifdef CONFIG_PPC_DCR_NATIVE
/* Enable external clock source */
if (emac_has_feature(dev, EMAC_FTR_460EX_PHY_CLK_FIX))
dcri_clrset(SDR0, SDR0_ETH_CFG,
SDR0_ETH_CFG_ECS << dev->cell_index, 0);
#endif
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
if (n) {
dev->reset_failed = 0;
return 0;
} else {
emac_report_timeout_error(dev, "reset timeout");
dev->reset_failed = 1;
return -ETIMEDOUT;
}
}
static void emac_hash_mc(struct emac_instance *dev)
{
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
const int regs = EMAC_XAHT_REGS(dev);
u32 *gaht_base = emac_gaht_base(dev);
u32 gaht_temp[regs];
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
struct netdev_hw_addr *ha;
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
int i;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
net: use netdev_mc_count and netdev_mc_empty when appropriate This patch replaces dev->mc_count in all drivers (hopefully I didn't miss anything). Used spatch and did small tweaks and conding style changes when it was suitable. Jirka Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-02-08 04:30:35 +00:00
DBG(dev, "hash_mc %d" NL, netdev_mc_count(dev->ndev));
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
memset(gaht_temp, 0, sizeof (gaht_temp));
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, dev->ndev) {
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
int slot, reg, mask;
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
DBG2(dev, "mc %pM" NL, ha->addr);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
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
slot = EMAC_XAHT_CRC_TO_SLOT(dev,
ether_crc(ETH_ALEN, ha->addr));
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
reg = EMAC_XAHT_SLOT_TO_REG(dev, slot);
mask = EMAC_XAHT_SLOT_TO_MASK(dev, slot);
gaht_temp[reg] |= mask;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
}
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
for (i = 0; i < regs; i++)
out_be32(gaht_base + i, gaht_temp[i]);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
}
static inline u32 emac_iff2rmr(struct net_device *ndev)
{
struct emac_instance *dev = netdev_priv(ndev);
u32 r;
r = EMAC_RMR_SP | EMAC_RMR_SFCS | EMAC_RMR_IAE | EMAC_RMR_BAE;
if (emac_has_feature(dev, EMAC_FTR_EMAC4))
r |= EMAC4_RMR_BASE;
else
r |= EMAC_RMR_BASE;
if (ndev->flags & IFF_PROMISC)
r |= EMAC_RMR_PME;
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
else if (ndev->flags & IFF_ALLMULTI ||
net: use netdev_mc_count and netdev_mc_empty when appropriate This patch replaces dev->mc_count in all drivers (hopefully I didn't miss anything). Used spatch and did small tweaks and conding style changes when it was suitable. Jirka Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-02-08 04:30:35 +00:00
(netdev_mc_count(ndev) > EMAC_XAHT_SLOTS(dev)))
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
r |= EMAC_RMR_PMME;
net: use netdev_mc_count and netdev_mc_empty when appropriate This patch replaces dev->mc_count in all drivers (hopefully I didn't miss anything). Used spatch and did small tweaks and conding style changes when it was suitable. Jirka Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-02-08 04:30:35 +00:00
else if (!netdev_mc_empty(ndev))
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
r |= EMAC_RMR_MAE;
return r;
}
static u32 __emac_calc_base_mr1(struct emac_instance *dev, int tx_size, int rx_size)
{
u32 ret = EMAC_MR1_VLE | EMAC_MR1_IST | EMAC_MR1_TR0_MULT;
DBG2(dev, "__emac_calc_base_mr1" NL);
switch(tx_size) {
case 2048:
ret |= EMAC_MR1_TFS_2K;
break;
default:
ibm_newemac: Added 16K Tx FIFO size support for EMAC4 Some of the EMAC V4 implementations support 16K Tx FIFOs. This patch adds support for this functionality and fixes typos in the Tx FIFO size error messages. Signed-off-by: Dave Mitchell <dmitchell@appliedmicro.com> Acked-by: Prodyut Hazarika <phazarika@appliedmicro.com> Acked-by: Victor Gallardo <vgallardo@appliedmicro.com> Acked-by: Loc Ho <lho@appliedmicro.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-08 06:32:21 +00:00
printk(KERN_WARNING "%s: Unknown Tx FIFO size %d\n",
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
dev->ndev->name, tx_size);
}
switch(rx_size) {
case 16384:
ret |= EMAC_MR1_RFS_16K;
break;
case 4096:
ret |= EMAC_MR1_RFS_4K;
break;
default:
printk(KERN_WARNING "%s: Unknown Rx FIFO size %d\n",
dev->ndev->name, rx_size);
}
return ret;
}
static u32 __emac4_calc_base_mr1(struct emac_instance *dev, int tx_size, int rx_size)
{
u32 ret = EMAC_MR1_VLE | EMAC_MR1_IST | EMAC4_MR1_TR |
ibm_newemac: Correct opb_bus_freq value The EMAC4_MR1_OBCI(freq) macro expects freg in MHz, while opb_bus_freq is kept in Hz. Correct this. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-05 00:14:31 +00:00
EMAC4_MR1_OBCI(dev->opb_bus_freq / 1000000);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
DBG2(dev, "__emac4_calc_base_mr1" NL);
switch(tx_size) {
ibm_newemac: Added 16K Tx FIFO size support for EMAC4 Some of the EMAC V4 implementations support 16K Tx FIFOs. This patch adds support for this functionality and fixes typos in the Tx FIFO size error messages. Signed-off-by: Dave Mitchell <dmitchell@appliedmicro.com> Acked-by: Prodyut Hazarika <phazarika@appliedmicro.com> Acked-by: Victor Gallardo <vgallardo@appliedmicro.com> Acked-by: Loc Ho <lho@appliedmicro.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-08 06:32:21 +00:00
case 16384:
ret |= EMAC4_MR1_TFS_16K;
break;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
case 4096:
ret |= EMAC4_MR1_TFS_4K;
break;
case 2048:
ret |= EMAC4_MR1_TFS_2K;
break;
default:
ibm_newemac: Added 16K Tx FIFO size support for EMAC4 Some of the EMAC V4 implementations support 16K Tx FIFOs. This patch adds support for this functionality and fixes typos in the Tx FIFO size error messages. Signed-off-by: Dave Mitchell <dmitchell@appliedmicro.com> Acked-by: Prodyut Hazarika <phazarika@appliedmicro.com> Acked-by: Victor Gallardo <vgallardo@appliedmicro.com> Acked-by: Loc Ho <lho@appliedmicro.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-08 06:32:21 +00:00
printk(KERN_WARNING "%s: Unknown Tx FIFO size %d\n",
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
dev->ndev->name, tx_size);
}
switch(rx_size) {
case 16384:
ret |= EMAC4_MR1_RFS_16K;
break;
case 4096:
ret |= EMAC4_MR1_RFS_4K;
break;
case 2048:
ret |= EMAC4_MR1_RFS_2K;
break;
default:
printk(KERN_WARNING "%s: Unknown Rx FIFO size %d\n",
dev->ndev->name, rx_size);
}
return ret;
}
static u32 emac_calc_base_mr1(struct emac_instance *dev, int tx_size, int rx_size)
{
return emac_has_feature(dev, EMAC_FTR_EMAC4) ?
__emac4_calc_base_mr1(dev, tx_size, rx_size) :
__emac_calc_base_mr1(dev, tx_size, rx_size);
}
static inline u32 emac_calc_trtr(struct emac_instance *dev, unsigned int size)
{
if (emac_has_feature(dev, EMAC_FTR_EMAC4))
return ((size >> 6) - 1) << EMAC_TRTR_SHIFT_EMAC4;
else
return ((size >> 6) - 1) << EMAC_TRTR_SHIFT;
}
static inline u32 emac_calc_rwmr(struct emac_instance *dev,
unsigned int low, unsigned int high)
{
if (emac_has_feature(dev, EMAC_FTR_EMAC4))
return (low << 22) | ( (high & 0x3ff) << 6);
else
return (low << 23) | ( (high & 0x1ff) << 7);
}
static int emac_configure(struct emac_instance *dev)
{
struct emac_regs __iomem *p = dev->emacp;
struct net_device *ndev = dev->ndev;
ibm_newemac: Workaround reset timeout when no link With some PHYs, when the link goes away, the EMAC reset fails due to the loss of the RX clock I believe. The old EMAC driver worked around that using some internal chip-specific clock force bits that are different on various 44x implementations. This is an attempt at doing it differently, by avoiding the reset when there is no link, but forcing loopback mode instead. It seems to work on my Taishan 440GX based board so far. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Stefan Roese <sr@denx.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-05 00:14:27 +00:00
int tx_size, rx_size, link = netif_carrier_ok(dev->ndev);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
u32 r, mr1 = 0;
DBG(dev, "configure" NL);
ibm_newemac: Workaround reset timeout when no link With some PHYs, when the link goes away, the EMAC reset fails due to the loss of the RX clock I believe. The old EMAC driver worked around that using some internal chip-specific clock force bits that are different on various 44x implementations. This is an attempt at doing it differently, by avoiding the reset when there is no link, but forcing loopback mode instead. It seems to work on my Taishan 440GX based board so far. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Stefan Roese <sr@denx.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-05 00:14:27 +00:00
if (!link) {
out_be32(&p->mr1, in_be32(&p->mr1)
| EMAC_MR1_FDE | EMAC_MR1_ILE);
udelay(100);
} else if (emac_reset(dev) < 0)
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
return -ETIMEDOUT;
if (emac_has_feature(dev, EMAC_FTR_HAS_TAH))
tah_reset(dev->tah_dev);
ibm_newemac: Workaround reset timeout when no link With some PHYs, when the link goes away, the EMAC reset fails due to the loss of the RX clock I believe. The old EMAC driver worked around that using some internal chip-specific clock force bits that are different on various 44x implementations. This is an attempt at doing it differently, by avoiding the reset when there is no link, but forcing loopback mode instead. It seems to work on my Taishan 440GX based board so far. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Stefan Roese <sr@denx.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-05 00:14:27 +00:00
DBG(dev, " link = %d duplex = %d, pause = %d, asym_pause = %d\n",
link, dev->phy.duplex, dev->phy.pause, dev->phy.asym_pause);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
/* Default fifo sizes */
tx_size = dev->tx_fifo_size;
rx_size = dev->rx_fifo_size;
ibm_newemac: Workaround reset timeout when no link With some PHYs, when the link goes away, the EMAC reset fails due to the loss of the RX clock I believe. The old EMAC driver worked around that using some internal chip-specific clock force bits that are different on various 44x implementations. This is an attempt at doing it differently, by avoiding the reset when there is no link, but forcing loopback mode instead. It seems to work on my Taishan 440GX based board so far. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Stefan Roese <sr@denx.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-05 00:14:27 +00:00
/* No link, force loopback */
if (!link)
mr1 = EMAC_MR1_FDE | EMAC_MR1_ILE;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
/* Check for full duplex */
ibm_newemac: Workaround reset timeout when no link With some PHYs, when the link goes away, the EMAC reset fails due to the loss of the RX clock I believe. The old EMAC driver worked around that using some internal chip-specific clock force bits that are different on various 44x implementations. This is an attempt at doing it differently, by avoiding the reset when there is no link, but forcing loopback mode instead. It seems to work on my Taishan 440GX based board so far. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Stefan Roese <sr@denx.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-05 00:14:27 +00:00
else if (dev->phy.duplex == DUPLEX_FULL)
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
mr1 |= EMAC_MR1_FDE | EMAC_MR1_MWSW_001;
/* Adjust fifo sizes, mr1 and timeouts based on link speed */
dev->stop_timeout = STOP_TIMEOUT_10;
switch (dev->phy.speed) {
case SPEED_1000:
if (emac_phy_gpcs(dev->phy.mode)) {
ibm_newemac: Add support for GPCS, SGMII and M88E1112 PHY Add support for the phy types found on the Arches and other PowerPC 460 based boards. Signed-off-by: Victor Gallardo <vgallardo@amcc.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-10-02 06:37:57 +00:00
mr1 |= EMAC_MR1_MF_1000GPCS | EMAC_MR1_MF_IPPA(
(dev->phy.gpcs_address != 0xffffffff) ?
dev->phy.gpcs_address : dev->phy.address);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
/* Put some arbitrary OUI, Manuf & Rev IDs so we can
* identify this GPCS PHY later.
*/
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
out_be32(&p->u1.emac4.ipcr, 0xdeadbeef);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
} else
mr1 |= EMAC_MR1_MF_1000;
/* Extended fifo sizes */
tx_size = dev->tx_fifo_size_gige;
rx_size = dev->rx_fifo_size_gige;
if (dev->ndev->mtu > ETH_DATA_LEN) {
ibm_newemac: Fix problem with jumbo frame support and EMAC V4.patch This fixes the jumbo frame support on EMAC V4 systems. Now the correct bit is set depending on the EMAC version configured. Tested on Kilauea (405EX) and Canyonlands (460EX). Signed-off-by: Stefan Roese <sr@denx.de> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:42 +00:00
if (emac_has_feature(dev, EMAC_FTR_EMAC4))
mr1 |= EMAC4_MR1_JPSM;
else
mr1 |= EMAC_MR1_JPSM;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
dev->stop_timeout = STOP_TIMEOUT_1000_JUMBO;
} else
dev->stop_timeout = STOP_TIMEOUT_1000;
break;
case SPEED_100:
mr1 |= EMAC_MR1_MF_100;
dev->stop_timeout = STOP_TIMEOUT_100;
break;
default: /* make gcc happy */
break;
}
if (emac_has_feature(dev, EMAC_FTR_HAS_RGMII))
rgmii_set_speed(dev->rgmii_dev, dev->rgmii_port,
dev->phy.speed);
if (emac_has_feature(dev, EMAC_FTR_HAS_ZMII))
zmii_set_speed(dev->zmii_dev, dev->zmii_port, dev->phy.speed);
/* on 40x erratum forces us to NOT use integrated flow control,
* let's hope it works on 44x ;)
*/
if (!emac_has_feature(dev, EMAC_FTR_NO_FLOW_CONTROL_40x) &&
dev->phy.duplex == DUPLEX_FULL) {
if (dev->phy.pause)
mr1 |= EMAC_MR1_EIFC | EMAC_MR1_APP;
else if (dev->phy.asym_pause)
mr1 |= EMAC_MR1_APP;
}
/* Add base settings & fifo sizes & program MR1 */
mr1 |= emac_calc_base_mr1(dev, tx_size, rx_size);
out_be32(&p->mr1, mr1);
/* Set individual MAC address */
out_be32(&p->iahr, (ndev->dev_addr[0] << 8) | ndev->dev_addr[1]);
out_be32(&p->ialr, (ndev->dev_addr[2] << 24) |
(ndev->dev_addr[3] << 16) | (ndev->dev_addr[4] << 8) |
ndev->dev_addr[5]);
/* VLAN Tag Protocol ID */
out_be32(&p->vtpid, 0x8100);
/* Receive mode register */
r = emac_iff2rmr(ndev);
if (r & EMAC_RMR_MAE)
emac_hash_mc(dev);
out_be32(&p->rmr, r);
/* FIFOs thresholds */
if (emac_has_feature(dev, EMAC_FTR_EMAC4))
r = EMAC4_TMR1((dev->mal_burst_size / dev->fifo_entry_size) + 1,
tx_size / 2 / dev->fifo_entry_size);
else
r = EMAC_TMR1((dev->mal_burst_size / dev->fifo_entry_size) + 1,
tx_size / 2 / dev->fifo_entry_size);
out_be32(&p->tmr1, r);
out_be32(&p->trtr, emac_calc_trtr(dev, tx_size / 2));
/* PAUSE frame is sent when RX FIFO reaches its high-water mark,
there should be still enough space in FIFO to allow the our link
partner time to process this frame and also time to send PAUSE
frame itself.
Here is the worst case scenario for the RX FIFO "headroom"
(from "The Switch Book") (100Mbps, without preamble, inter-frame gap):
1) One maximum-length frame on TX 1522 bytes
2) One PAUSE frame time 64 bytes
3) PAUSE frame decode time allowance 64 bytes
4) One maximum-length frame on RX 1522 bytes
5) Round-trip propagation delay of the link (100Mb) 15 bytes
----------
3187 bytes
I chose to set high-water mark to RX_FIFO_SIZE / 4 (1024 bytes)
low-water mark to RX_FIFO_SIZE / 8 (512 bytes)
*/
r = emac_calc_rwmr(dev, rx_size / 8 / dev->fifo_entry_size,
rx_size / 4 / dev->fifo_entry_size);
out_be32(&p->rwmr, r);
/* Set PAUSE timer to the maximum */
out_be32(&p->ptr, 0xffff);
/* IRQ sources */
r = EMAC_ISR_OVR | EMAC_ISR_BP | EMAC_ISR_SE |
EMAC_ISR_ALE | EMAC_ISR_BFCS | EMAC_ISR_PTLE | EMAC_ISR_ORE |
EMAC_ISR_IRE | EMAC_ISR_TE;
if (emac_has_feature(dev, EMAC_FTR_EMAC4))
r |= EMAC4_ISR_TXPE | EMAC4_ISR_RXPE /* | EMAC4_ISR_TXUE |
EMAC4_ISR_RXOE | */;
out_be32(&p->iser, r);
/* We need to take GPCS PHY out of isolate mode after EMAC reset */
ibm_newemac: Add support for GPCS, SGMII and M88E1112 PHY Add support for the phy types found on the Arches and other PowerPC 460 based boards. Signed-off-by: Victor Gallardo <vgallardo@amcc.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-10-02 06:37:57 +00:00
if (emac_phy_gpcs(dev->phy.mode)) {
if (dev->phy.gpcs_address != 0xffffffff)
emac_mii_reset_gpcs(&dev->phy);
else
emac_mii_reset_phy(&dev->phy);
}
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
return 0;
}
static void emac_reinitialize(struct emac_instance *dev)
{
DBG(dev, "reinitialize" NL);
emac_netif_stop(dev);
if (!emac_configure(dev)) {
emac_tx_enable(dev);
emac_rx_enable(dev);
}
emac_netif_start(dev);
}
static void emac_full_tx_reset(struct emac_instance *dev)
{
DBG(dev, "full_tx_reset" NL);
emac_tx_disable(dev);
mal_disable_tx_channel(dev->mal, dev->mal_tx_chan);
emac_clean_tx_ring(dev);
dev->tx_cnt = dev->tx_slot = dev->ack_slot = 0;
emac_configure(dev);
mal_enable_tx_channel(dev->mal, dev->mal_tx_chan);
emac_tx_enable(dev);
emac_rx_enable(dev);
}
static void emac_reset_work(struct work_struct *work)
{
struct emac_instance *dev = container_of(work, struct emac_instance, reset_work);
DBG(dev, "reset_work" NL);
mutex_lock(&dev->link_lock);
ibm_newemac: Fix possible lockup on close It's a bad idea to call flush_scheduled_work from within a netdev->stop because the linkwatch will occasionally take the rtnl lock from a workqueue context, and thus that can deadlock. This reworks things a bit in that area to avoid the problem. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-20 03:50:46 +00:00
if (dev->opened) {
emac_netif_stop(dev);
emac_full_tx_reset(dev);
emac_netif_start(dev);
}
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
mutex_unlock(&dev->link_lock);
}
static void emac_tx_timeout(struct net_device *ndev)
{
struct emac_instance *dev = netdev_priv(ndev);
DBG(dev, "tx_timeout" NL);
schedule_work(&dev->reset_work);
}
static inline int emac_phy_done(struct emac_instance *dev, u32 stacr)
{
int done = !!(stacr & EMAC_STACR_OC);
if (emac_has_feature(dev, EMAC_FTR_STACR_OC_INVERT))
done = !done;
return done;
};
static int __emac_mdio_read(struct emac_instance *dev, u8 id, u8 reg)
{
struct emac_regs __iomem *p = dev->emacp;
u32 r = 0;
int n, err = -ETIMEDOUT;
mutex_lock(&dev->mdio_lock);
DBG2(dev, "mdio_read(%02x,%02x)" NL, id, reg);
/* Enable proper MDIO port */
if (emac_has_feature(dev, EMAC_FTR_HAS_ZMII))
zmii_get_mdio(dev->zmii_dev, dev->zmii_port);
if (emac_has_feature(dev, EMAC_FTR_HAS_RGMII))
rgmii_get_mdio(dev->rgmii_dev, dev->rgmii_port);
/* Wait for management interface to become idle */
ibm_newemac: Increase MDIO timeouts This patch doubles the MDIO timeouts in EMAC as there are field cases where they are two short to communicate with some PHYs. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-23 05:17:14 +00:00
n = 20;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
while (!emac_phy_done(dev, in_be32(&p->stacr))) {
udelay(1);
if (!--n) {
DBG2(dev, " -> timeout wait idle\n");
goto bail;
}
}
/* Issue read command */
if (emac_has_feature(dev, EMAC_FTR_EMAC4))
r = EMAC4_STACR_BASE(dev->opb_bus_freq);
else
r = EMAC_STACR_BASE(dev->opb_bus_freq);
if (emac_has_feature(dev, EMAC_FTR_STACR_OC_INVERT))
r |= EMAC_STACR_OC;
ibm_newemac: Cleanup/fix support for STACR register variants There are a few variants of the STACR register that affect more than just the "AXON" version of EMAC. Replace the current test of various chip models with tests for generic properties in the device-tree. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Stefan Roese <sr@denx.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-05 00:14:29 +00:00
if (emac_has_feature(dev, EMAC_FTR_HAS_NEW_STACR))
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
r |= EMACX_STACR_STAC_READ;
else
r |= EMAC_STACR_STAC_READ;
r |= (reg & EMAC_STACR_PRA_MASK)
| ((id & EMAC_STACR_PCDA_MASK) << EMAC_STACR_PCDA_SHIFT);
out_be32(&p->stacr, r);
/* Wait for read to complete */
ibm_newemac: Increase MDIO timeouts This patch doubles the MDIO timeouts in EMAC as there are field cases where they are two short to communicate with some PHYs. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-23 05:17:14 +00:00
n = 200;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
while (!emac_phy_done(dev, (r = in_be32(&p->stacr)))) {
udelay(1);
if (!--n) {
DBG2(dev, " -> timeout wait complete\n");
goto bail;
}
}
if (unlikely(r & EMAC_STACR_PHYE)) {
DBG(dev, "mdio_read(%02x, %02x) failed" NL, id, reg);
err = -EREMOTEIO;
goto bail;
}
r = ((r >> EMAC_STACR_PHYD_SHIFT) & EMAC_STACR_PHYD_MASK);
DBG2(dev, "mdio_read -> %04x" NL, r);
err = 0;
bail:
if (emac_has_feature(dev, EMAC_FTR_HAS_RGMII))
rgmii_put_mdio(dev->rgmii_dev, dev->rgmii_port);
if (emac_has_feature(dev, EMAC_FTR_HAS_ZMII))
zmii_put_mdio(dev->zmii_dev, dev->zmii_port);
mutex_unlock(&dev->mdio_lock);
return err == 0 ? r : err;
}
static void __emac_mdio_write(struct emac_instance *dev, u8 id, u8 reg,
u16 val)
{
struct emac_regs __iomem *p = dev->emacp;
u32 r = 0;
int n, err = -ETIMEDOUT;
mutex_lock(&dev->mdio_lock);
DBG2(dev, "mdio_write(%02x,%02x,%04x)" NL, id, reg, val);
/* Enable proper MDIO port */
if (emac_has_feature(dev, EMAC_FTR_HAS_ZMII))
zmii_get_mdio(dev->zmii_dev, dev->zmii_port);
if (emac_has_feature(dev, EMAC_FTR_HAS_RGMII))
rgmii_get_mdio(dev->rgmii_dev, dev->rgmii_port);
/* Wait for management interface to be idle */
ibm_newemac: Increase MDIO timeouts This patch doubles the MDIO timeouts in EMAC as there are field cases where they are two short to communicate with some PHYs. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-23 05:17:14 +00:00
n = 20;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
while (!emac_phy_done(dev, in_be32(&p->stacr))) {
udelay(1);
if (!--n) {
DBG2(dev, " -> timeout wait idle\n");
goto bail;
}
}
/* Issue write command */
if (emac_has_feature(dev, EMAC_FTR_EMAC4))
r = EMAC4_STACR_BASE(dev->opb_bus_freq);
else
r = EMAC_STACR_BASE(dev->opb_bus_freq);
if (emac_has_feature(dev, EMAC_FTR_STACR_OC_INVERT))
r |= EMAC_STACR_OC;
ibm_newemac: Cleanup/fix support for STACR register variants There are a few variants of the STACR register that affect more than just the "AXON" version of EMAC. Replace the current test of various chip models with tests for generic properties in the device-tree. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Stefan Roese <sr@denx.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-05 00:14:29 +00:00
if (emac_has_feature(dev, EMAC_FTR_HAS_NEW_STACR))
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
r |= EMACX_STACR_STAC_WRITE;
else
r |= EMAC_STACR_STAC_WRITE;
r |= (reg & EMAC_STACR_PRA_MASK) |
((id & EMAC_STACR_PCDA_MASK) << EMAC_STACR_PCDA_SHIFT) |
(val << EMAC_STACR_PHYD_SHIFT);
out_be32(&p->stacr, r);
/* Wait for write to complete */
ibm_newemac: Increase MDIO timeouts This patch doubles the MDIO timeouts in EMAC as there are field cases where they are two short to communicate with some PHYs. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-23 05:17:14 +00:00
n = 200;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
while (!emac_phy_done(dev, in_be32(&p->stacr))) {
udelay(1);
if (!--n) {
DBG2(dev, " -> timeout wait complete\n");
goto bail;
}
}
err = 0;
bail:
if (emac_has_feature(dev, EMAC_FTR_HAS_RGMII))
rgmii_put_mdio(dev->rgmii_dev, dev->rgmii_port);
if (emac_has_feature(dev, EMAC_FTR_HAS_ZMII))
zmii_put_mdio(dev->zmii_dev, dev->zmii_port);
mutex_unlock(&dev->mdio_lock);
}
static int emac_mdio_read(struct net_device *ndev, int id, int reg)
{
struct emac_instance *dev = netdev_priv(ndev);
int res;
ibm_newemac: Add support for GPCS, SGMII and M88E1112 PHY Add support for the phy types found on the Arches and other PowerPC 460 based boards. Signed-off-by: Victor Gallardo <vgallardo@amcc.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-10-02 06:37:57 +00:00
res = __emac_mdio_read((dev->mdio_instance &&
dev->phy.gpcs_address != id) ?
dev->mdio_instance : dev,
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
(u8) id, (u8) reg);
return res;
}
static void emac_mdio_write(struct net_device *ndev, int id, int reg, int val)
{
struct emac_instance *dev = netdev_priv(ndev);
ibm_newemac: Add support for GPCS, SGMII and M88E1112 PHY Add support for the phy types found on the Arches and other PowerPC 460 based boards. Signed-off-by: Victor Gallardo <vgallardo@amcc.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-10-02 06:37:57 +00:00
__emac_mdio_write((dev->mdio_instance &&
dev->phy.gpcs_address != id) ?
dev->mdio_instance : dev,
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
(u8) id, (u8) reg, (u16) val);
}
/* Tx lock BH */
static void __emac_set_multicast_list(struct emac_instance *dev)
{
struct emac_regs __iomem *p = dev->emacp;
u32 rmr = emac_iff2rmr(dev->ndev);
DBG(dev, "__multicast %08x" NL, rmr);
/* I decided to relax register access rules here to avoid
* full EMAC reset.
*
* There is a real problem with EMAC4 core if we use MWSW_001 bit
* in MR1 register and do a full EMAC reset.
* One TX BD status update is delayed and, after EMAC reset, it
* never happens, resulting in TX hung (it'll be recovered by TX
* timeout handler eventually, but this is just gross).
* So we either have to do full TX reset or try to cheat here :)
*
* The only required change is to RX mode register, so I *think* all
* we need is just to stop RX channel. This seems to work on all
* tested SoCs. --ebs
*
* If we need the full reset, we might just trigger the workqueue
* and do it async... a bit nasty but should work --BenH
*/
dev->mcast_pending = 0;
emac_rx_disable(dev);
if (rmr & EMAC_RMR_MAE)
emac_hash_mc(dev);
out_be32(&p->rmr, rmr);
emac_rx_enable(dev);
}
/* Tx lock BH */
static void emac_set_multicast_list(struct net_device *ndev)
{
struct emac_instance *dev = netdev_priv(ndev);
DBG(dev, "multicast" NL);
BUG_ON(!netif_running(dev->ndev));
if (dev->no_mcast) {
dev->mcast_pending = 1;
return;
}
__emac_set_multicast_list(dev);
}
static int emac_resize_rx_ring(struct emac_instance *dev, int new_mtu)
{
int rx_sync_size = emac_rx_sync_size(new_mtu);
int rx_skb_size = emac_rx_skb_size(new_mtu);
int i, ret = 0;
mutex_lock(&dev->link_lock);
emac_netif_stop(dev);
emac_rx_disable(dev);
mal_disable_rx_channel(dev->mal, dev->mal_rx_chan);
if (dev->rx_sg_skb) {
++dev->estats.rx_dropped_resize;
dev_kfree_skb(dev->rx_sg_skb);
dev->rx_sg_skb = NULL;
}
/* Make a first pass over RX ring and mark BDs ready, dropping
* non-processed packets on the way. We need this as a separate pass
* to simplify error recovery in the case of allocation failure later.
*/
for (i = 0; i < NUM_RX_BUFF; ++i) {
if (dev->rx_desc[i].ctrl & MAL_RX_CTRL_FIRST)
++dev->estats.rx_dropped_resize;
dev->rx_desc[i].data_len = 0;
dev->rx_desc[i].ctrl = MAL_RX_CTRL_EMPTY |
(i == (NUM_RX_BUFF - 1) ? MAL_RX_CTRL_WRAP : 0);
}
/* Reallocate RX ring only if bigger skb buffers are required */
if (rx_skb_size <= dev->rx_skb_size)
goto skip;
/* Second pass, allocate new skbs */
for (i = 0; i < NUM_RX_BUFF; ++i) {
struct sk_buff *skb = alloc_skb(rx_skb_size, GFP_ATOMIC);
if (!skb) {
ret = -ENOMEM;
goto oom;
}
BUG_ON(!dev->rx_skb[i]);
dev_kfree_skb(dev->rx_skb[i]);
skb_reserve(skb, EMAC_RX_SKB_HEADROOM + 2);
dev->rx_desc[i].data_ptr =
dma_map_single(&dev->ofdev->dev, skb->data - 2, rx_sync_size,
DMA_FROM_DEVICE) + 2;
dev->rx_skb[i] = skb;
}
skip:
/* Check if we need to change "Jumbo" bit in MR1 */
if ((new_mtu > ETH_DATA_LEN) ^ (dev->ndev->mtu > ETH_DATA_LEN)) {
/* This is to prevent starting RX channel in emac_rx_enable() */
set_bit(MAL_COMMAC_RX_STOPPED, &dev->commac.flags);
dev->ndev->mtu = new_mtu;
emac_full_tx_reset(dev);
}
mal_set_rcbs(dev->mal, dev->mal_rx_chan, emac_rx_size(new_mtu));
oom:
/* Restart RX */
clear_bit(MAL_COMMAC_RX_STOPPED, &dev->commac.flags);
dev->rx_slot = 0;
mal_enable_rx_channel(dev->mal, dev->mal_rx_chan);
emac_rx_enable(dev);
emac_netif_start(dev);
mutex_unlock(&dev->link_lock);
return ret;
}
/* Process ctx, rtnl_lock semaphore */
static int emac_change_mtu(struct net_device *ndev, int new_mtu)
{
struct emac_instance *dev = netdev_priv(ndev);
int ret = 0;
if (new_mtu < EMAC_MIN_MTU || new_mtu > dev->max_mtu)
return -EINVAL;
DBG(dev, "change_mtu(%d)" NL, new_mtu);
if (netif_running(ndev)) {
/* Check if we really need to reinitalize RX ring */
if (emac_rx_skb_size(ndev->mtu) != emac_rx_skb_size(new_mtu))
ret = emac_resize_rx_ring(dev, new_mtu);
}
if (!ret) {
ndev->mtu = new_mtu;
dev->rx_skb_size = emac_rx_skb_size(new_mtu);
dev->rx_sync_size = emac_rx_sync_size(new_mtu);
}
return ret;
}
static void emac_clean_tx_ring(struct emac_instance *dev)
{
int i;
for (i = 0; i < NUM_TX_BUFF; ++i) {
if (dev->tx_skb[i]) {
dev_kfree_skb(dev->tx_skb[i]);
dev->tx_skb[i] = NULL;
if (dev->tx_desc[i].ctrl & MAL_TX_CTRL_READY)
++dev->estats.tx_dropped;
}
dev->tx_desc[i].ctrl = 0;
dev->tx_desc[i].data_ptr = 0;
}
}
static void emac_clean_rx_ring(struct emac_instance *dev)
{
int i;
for (i = 0; i < NUM_RX_BUFF; ++i)
if (dev->rx_skb[i]) {
dev->rx_desc[i].ctrl = 0;
dev_kfree_skb(dev->rx_skb[i]);
dev->rx_skb[i] = NULL;
dev->rx_desc[i].data_ptr = 0;
}
if (dev->rx_sg_skb) {
dev_kfree_skb(dev->rx_sg_skb);
dev->rx_sg_skb = NULL;
}
}
static inline int emac_alloc_rx_skb(struct emac_instance *dev, int slot,
gfp_t flags)
{
struct sk_buff *skb = alloc_skb(dev->rx_skb_size, flags);
if (unlikely(!skb))
return -ENOMEM;
dev->rx_skb[slot] = skb;
dev->rx_desc[slot].data_len = 0;
skb_reserve(skb, EMAC_RX_SKB_HEADROOM + 2);
dev->rx_desc[slot].data_ptr =
dma_map_single(&dev->ofdev->dev, skb->data - 2, dev->rx_sync_size,
DMA_FROM_DEVICE) + 2;
wmb();
dev->rx_desc[slot].ctrl = MAL_RX_CTRL_EMPTY |
(slot == (NUM_RX_BUFF - 1) ? MAL_RX_CTRL_WRAP : 0);
return 0;
}
static void emac_print_link_status(struct emac_instance *dev)
{
if (netif_carrier_ok(dev->ndev))
printk(KERN_INFO "%s: link is up, %d %s%s\n",
dev->ndev->name, dev->phy.speed,
dev->phy.duplex == DUPLEX_FULL ? "FDX" : "HDX",
dev->phy.pause ? ", pause enabled" :
dev->phy.asym_pause ? ", asymmetric pause enabled" : "");
else
printk(KERN_INFO "%s: link is down\n", dev->ndev->name);
}
/* Process ctx, rtnl_lock semaphore */
static int emac_open(struct net_device *ndev)
{
struct emac_instance *dev = netdev_priv(ndev);
int err, i;
DBG(dev, "open" NL);
/* Setup error IRQ handler */
err = request_irq(dev->emac_irq, emac_irq, 0, "EMAC", dev);
if (err) {
printk(KERN_ERR "%s: failed to request IRQ %d\n",
ndev->name, dev->emac_irq);
return err;
}
/* Allocate RX ring */
for (i = 0; i < NUM_RX_BUFF; ++i)
if (emac_alloc_rx_skb(dev, i, GFP_KERNEL)) {
printk(KERN_ERR "%s: failed to allocate RX ring\n",
ndev->name);
goto oom;
}
dev->tx_cnt = dev->tx_slot = dev->ack_slot = dev->rx_slot = 0;
clear_bit(MAL_COMMAC_RX_STOPPED, &dev->commac.flags);
dev->rx_sg_skb = NULL;
mutex_lock(&dev->link_lock);
ibm_newemac: Fix possible lockup on close It's a bad idea to call flush_scheduled_work from within a netdev->stop because the linkwatch will occasionally take the rtnl lock from a workqueue context, and thus that can deadlock. This reworks things a bit in that area to avoid the problem. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-20 03:50:46 +00:00
dev->opened = 1;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
ibm_newemac: Fix possible lockup on close It's a bad idea to call flush_scheduled_work from within a netdev->stop because the linkwatch will occasionally take the rtnl lock from a workqueue context, and thus that can deadlock. This reworks things a bit in that area to avoid the problem. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-20 03:50:46 +00:00
/* Start PHY polling now.
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
*/
if (dev->phy.address >= 0) {
int link_poll_interval;
if (dev->phy.def->ops->poll_link(&dev->phy)) {
dev->phy.def->ops->read_link(&dev->phy);
ibm_newemac: PowerPC 440EP/440GR EMAC PHY clock workaround This patch adds ibm_newemac PHY clock workaround for 440EP/440GR EMAC attached to a PHY which doesn't generate RX clock if there is no link. The code is based on the previous ibm_emac driver stuff. The 440EP/440GR allows controlling each EMAC clock separately as opposed to global clock selection for 440GX. BenH: Made that #ifdef CONFIG_PPC_DCR_NATIVE for now as dcri_* stuff doesn't exist for MMIO type DCRs like Cell. Some future rework & improvements of the DCR infrastructure will make that cleaner but for now, this makes it work. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:48 +00:00
emac_rx_clk_default(dev);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
netif_carrier_on(dev->ndev);
link_poll_interval = PHY_POLL_LINK_ON;
} else {
ibm_newemac: PowerPC 440EP/440GR EMAC PHY clock workaround This patch adds ibm_newemac PHY clock workaround for 440EP/440GR EMAC attached to a PHY which doesn't generate RX clock if there is no link. The code is based on the previous ibm_emac driver stuff. The 440EP/440GR allows controlling each EMAC clock separately as opposed to global clock selection for 440GX. BenH: Made that #ifdef CONFIG_PPC_DCR_NATIVE for now as dcri_* stuff doesn't exist for MMIO type DCRs like Cell. Some future rework & improvements of the DCR infrastructure will make that cleaner but for now, this makes it work. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:48 +00:00
emac_rx_clk_tx(dev);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
netif_carrier_off(dev->ndev);
link_poll_interval = PHY_POLL_LINK_OFF;
}
dev->link_polling = 1;
wmb();
schedule_delayed_work(&dev->link_work, link_poll_interval);
emac_print_link_status(dev);
} else
netif_carrier_on(dev->ndev);
ibm_newemac: Don't call dev_mc_add() before device is registered We must not call dev_mc_add() from within our HW configure which happens before we initialize and register the netdev. Do it in open() instead. Thanks to Sebastian Siewior for tracking it down. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-08-17 21:29:23 +00:00
/* Required for Pause packet support in EMAC */
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
dev_mc_add_global(ndev, default_mcast_addr);
ibm_newemac: Don't call dev_mc_add() before device is registered We must not call dev_mc_add() from within our HW configure which happens before we initialize and register the netdev. Do it in open() instead. Thanks to Sebastian Siewior for tracking it down. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-08-17 21:29:23 +00:00
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
emac_configure(dev);
mal_poll_add(dev->mal, &dev->commac);
mal_enable_tx_channel(dev->mal, dev->mal_tx_chan);
mal_set_rcbs(dev->mal, dev->mal_rx_chan, emac_rx_size(ndev->mtu));
mal_enable_rx_channel(dev->mal, dev->mal_rx_chan);
emac_tx_enable(dev);
emac_rx_enable(dev);
emac_netif_start(dev);
mutex_unlock(&dev->link_lock);
return 0;
oom:
emac_clean_rx_ring(dev);
free_irq(dev->emac_irq, dev);
return -ENOMEM;
}
/* BHs disabled */
#if 0
static int emac_link_differs(struct emac_instance *dev)
{
u32 r = in_be32(&dev->emacp->mr1);
int duplex = r & EMAC_MR1_FDE ? DUPLEX_FULL : DUPLEX_HALF;
int speed, pause, asym_pause;
if (r & EMAC_MR1_MF_1000)
speed = SPEED_1000;
else if (r & EMAC_MR1_MF_100)
speed = SPEED_100;
else
speed = SPEED_10;
switch (r & (EMAC_MR1_EIFC | EMAC_MR1_APP)) {
case (EMAC_MR1_EIFC | EMAC_MR1_APP):
pause = 1;
asym_pause = 0;
break;
case EMAC_MR1_APP:
pause = 0;
asym_pause = 1;
break;
default:
pause = asym_pause = 0;
}
return speed != dev->phy.speed || duplex != dev->phy.duplex ||
pause != dev->phy.pause || asym_pause != dev->phy.asym_pause;
}
#endif
static void emac_link_timer(struct work_struct *work)
{
struct emac_instance *dev =
ibm_newemac: Fix dangerous struct assumption The ibm_newemac driver includes code which assumes that the work_struct which is included in every delayed_work is the first member of that structure. This is currently the case but might change in the future, so use to_delayed_work() instead, which doesn't make such an assumption. Signed-off-by: Jean Delvare <khali@linux-fr.org> Cc: Josh Boyer <jwboyer@linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-06 02:27:10 +00:00
container_of(to_delayed_work(work),
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
struct emac_instance, link_work);
int link_poll_interval;
mutex_lock(&dev->link_lock);
DBG2(dev, "link timer" NL);
ibm_newemac: Fix possible lockup on close It's a bad idea to call flush_scheduled_work from within a netdev->stop because the linkwatch will occasionally take the rtnl lock from a workqueue context, and thus that can deadlock. This reworks things a bit in that area to avoid the problem. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-20 03:50:46 +00:00
if (!dev->opened)
goto bail;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
if (dev->phy.def->ops->poll_link(&dev->phy)) {
if (!netif_carrier_ok(dev->ndev)) {
ibm_newemac: PowerPC 440EP/440GR EMAC PHY clock workaround This patch adds ibm_newemac PHY clock workaround for 440EP/440GR EMAC attached to a PHY which doesn't generate RX clock if there is no link. The code is based on the previous ibm_emac driver stuff. The 440EP/440GR allows controlling each EMAC clock separately as opposed to global clock selection for 440GX. BenH: Made that #ifdef CONFIG_PPC_DCR_NATIVE for now as dcri_* stuff doesn't exist for MMIO type DCRs like Cell. Some future rework & improvements of the DCR infrastructure will make that cleaner but for now, this makes it work. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:48 +00:00
emac_rx_clk_default(dev);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
/* Get new link parameters */
dev->phy.def->ops->read_link(&dev->phy);
netif_carrier_on(dev->ndev);
emac_netif_stop(dev);
emac_full_tx_reset(dev);
emac_netif_start(dev);
emac_print_link_status(dev);
}
link_poll_interval = PHY_POLL_LINK_ON;
} else {
if (netif_carrier_ok(dev->ndev)) {
ibm_newemac: PowerPC 440EP/440GR EMAC PHY clock workaround This patch adds ibm_newemac PHY clock workaround for 440EP/440GR EMAC attached to a PHY which doesn't generate RX clock if there is no link. The code is based on the previous ibm_emac driver stuff. The 440EP/440GR allows controlling each EMAC clock separately as opposed to global clock selection for 440GX. BenH: Made that #ifdef CONFIG_PPC_DCR_NATIVE for now as dcri_* stuff doesn't exist for MMIO type DCRs like Cell. Some future rework & improvements of the DCR infrastructure will make that cleaner but for now, this makes it work. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:48 +00:00
emac_rx_clk_tx(dev);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
netif_carrier_off(dev->ndev);
netif_tx_disable(dev->ndev);
ibm_newemac: Workaround reset timeout when no link With some PHYs, when the link goes away, the EMAC reset fails due to the loss of the RX clock I believe. The old EMAC driver worked around that using some internal chip-specific clock force bits that are different on various 44x implementations. This is an attempt at doing it differently, by avoiding the reset when there is no link, but forcing loopback mode instead. It seems to work on my Taishan 440GX based board so far. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Stefan Roese <sr@denx.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-05 00:14:27 +00:00
emac_reinitialize(dev);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
emac_print_link_status(dev);
}
link_poll_interval = PHY_POLL_LINK_OFF;
}
schedule_delayed_work(&dev->link_work, link_poll_interval);
ibm_newemac: Fix possible lockup on close It's a bad idea to call flush_scheduled_work from within a netdev->stop because the linkwatch will occasionally take the rtnl lock from a workqueue context, and thus that can deadlock. This reworks things a bit in that area to avoid the problem. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-20 03:50:46 +00:00
bail:
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
mutex_unlock(&dev->link_lock);
}
static void emac_force_link_update(struct emac_instance *dev)
{
netif_carrier_off(dev->ndev);
ibm_newemac: Fix possible lockup on close It's a bad idea to call flush_scheduled_work from within a netdev->stop because the linkwatch will occasionally take the rtnl lock from a workqueue context, and thus that can deadlock. This reworks things a bit in that area to avoid the problem. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-20 03:50:46 +00:00
smp_rmb();
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
if (dev->link_polling) {
cancel_rearming_delayed_work(&dev->link_work);
if (dev->link_polling)
schedule_delayed_work(&dev->link_work, PHY_POLL_LINK_OFF);
}
}
/* Process ctx, rtnl_lock semaphore */
static int emac_close(struct net_device *ndev)
{
struct emac_instance *dev = netdev_priv(ndev);
DBG(dev, "close" NL);
ibm_newemac: Fix possible lockup on close It's a bad idea to call flush_scheduled_work from within a netdev->stop because the linkwatch will occasionally take the rtnl lock from a workqueue context, and thus that can deadlock. This reworks things a bit in that area to avoid the problem. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-20 03:50:46 +00:00
if (dev->phy.address >= 0) {
dev->link_polling = 0;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
cancel_rearming_delayed_work(&dev->link_work);
ibm_newemac: Fix possible lockup on close It's a bad idea to call flush_scheduled_work from within a netdev->stop because the linkwatch will occasionally take the rtnl lock from a workqueue context, and thus that can deadlock. This reworks things a bit in that area to avoid the problem. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-20 03:50:46 +00:00
}
mutex_lock(&dev->link_lock);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
emac_netif_stop(dev);
ibm_newemac: Fix possible lockup on close It's a bad idea to call flush_scheduled_work from within a netdev->stop because the linkwatch will occasionally take the rtnl lock from a workqueue context, and thus that can deadlock. This reworks things a bit in that area to avoid the problem. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-20 03:50:46 +00:00
dev->opened = 0;
mutex_unlock(&dev->link_lock);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
emac_rx_disable(dev);
emac_tx_disable(dev);
mal_disable_rx_channel(dev->mal, dev->mal_rx_chan);
mal_disable_tx_channel(dev->mal, dev->mal_tx_chan);
mal_poll_del(dev->mal, &dev->commac);
emac_clean_tx_ring(dev);
emac_clean_rx_ring(dev);
free_irq(dev->emac_irq, dev);
ibm_newemac: emac_close() needs to call netif_carrier_off() When ibm_newemac netdev instance is shutdown with "ifconfig down", the netdev interface does not go properly down. netif_carrier_ok() keeps returning TRUE even after "ifconfig down". The problem can be seen when ibm_newemac instances are slaves of a bonding interface. The bonding interface code uses netif_carrier_ok() to determine the link status of its slaves. When ibm_newemac slave is shutdown with "ifconfig down", the bonding interface won't detect any link status change because netif_carrier_ok() keeps returning TRUE. Signed-off-by: Petri Gynther <pgynther@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-08-20 09:21:27 +00:00
netif_carrier_off(ndev);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
return 0;
}
static inline u16 emac_tx_csum(struct emac_instance *dev,
struct sk_buff *skb)
{
ibm_newemac: emac_tx_csum typo fix Move the "&& skb->ip_summed == CHECKSUM_PARTIAL" part out of emac_has_feature parameters. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-03-27 14:43:57 +00:00
if (emac_has_feature(dev, EMAC_FTR_HAS_TAH) &&
(skb->ip_summed == CHECKSUM_PARTIAL)) {
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
++dev->stats.tx_packets_csum;
return EMAC_TX_CTRL_TAH_CSUM;
}
return 0;
}
static inline int emac_xmit_finish(struct emac_instance *dev, int len)
{
struct emac_regs __iomem *p = dev->emacp;
struct net_device *ndev = dev->ndev;
/* Send the packet out. If the if makes a significant perf
* difference, then we can store the TMR0 value in "dev"
* instead
*/
if (emac_has_feature(dev, EMAC_FTR_EMAC4))
out_be32(&p->tmr0, EMAC4_TMR0_XMIT);
Fix typo in new EMAC driver. Fix an obvious typo in emac_xmit_finish. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-10-02 12:01:07 +00:00
else
out_be32(&p->tmr0, EMAC_TMR0_XMIT);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
if (unlikely(++dev->tx_cnt == NUM_TX_BUFF)) {
netif_stop_queue(ndev);
DBG2(dev, "stopped TX queue" NL);
}
ndev->trans_start = jiffies;
++dev->stats.tx_packets;
dev->stats.tx_bytes += len;
net: convert remaining non-symbolic return values in ndo_start_xmit() functions This patch converts the remaining occurences of raw return values to their symbolic counterparts in ndo_start_xmit() functions that were missed by the previous automatic conversion. Additionally code that assumed the symbolic value of NETDEV_TX_OK to be zero is changed to explicitly use NETDEV_TX_OK. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-06 02:23:38 +00:00
return NETDEV_TX_OK;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
}
/* Tx lock BH */
static int emac_start_xmit(struct sk_buff *skb, struct net_device *ndev)
{
struct emac_instance *dev = netdev_priv(ndev);
unsigned int len = skb->len;
int slot;
u16 ctrl = EMAC_TX_CTRL_GFCS | EMAC_TX_CTRL_GP | MAL_TX_CTRL_READY |
MAL_TX_CTRL_LAST | emac_tx_csum(dev, skb);
slot = dev->tx_slot++;
if (dev->tx_slot == NUM_TX_BUFF) {
dev->tx_slot = 0;
ctrl |= MAL_TX_CTRL_WRAP;
}
DBG2(dev, "xmit(%u) %d" NL, len, slot);
dev->tx_skb[slot] = skb;
dev->tx_desc[slot].data_ptr = dma_map_single(&dev->ofdev->dev,
skb->data, len,
DMA_TO_DEVICE);
dev->tx_desc[slot].data_len = (u16) len;
wmb();
dev->tx_desc[slot].ctrl = ctrl;
return emac_xmit_finish(dev, len);
}
static inline int emac_xmit_split(struct emac_instance *dev, int slot,
u32 pd, int len, int last, u16 base_ctrl)
{
while (1) {
u16 ctrl = base_ctrl;
int chunk = min(len, MAL_MAX_TX_SIZE);
len -= chunk;
slot = (slot + 1) % NUM_TX_BUFF;
if (last && !len)
ctrl |= MAL_TX_CTRL_LAST;
if (slot == NUM_TX_BUFF - 1)
ctrl |= MAL_TX_CTRL_WRAP;
dev->tx_skb[slot] = NULL;
dev->tx_desc[slot].data_ptr = pd;
dev->tx_desc[slot].data_len = (u16) chunk;
dev->tx_desc[slot].ctrl = ctrl;
++dev->tx_cnt;
if (!len)
break;
pd += chunk;
}
return slot;
}
/* Tx lock BH disabled (SG version for TAH equipped EMACs) */
static int emac_start_xmit_sg(struct sk_buff *skb, struct net_device *ndev)
{
struct emac_instance *dev = netdev_priv(ndev);
int nr_frags = skb_shinfo(skb)->nr_frags;
int len = skb->len, chunk;
int slot, i;
u16 ctrl;
u32 pd;
/* This is common "fast" path */
if (likely(!nr_frags && len <= MAL_MAX_TX_SIZE))
return emac_start_xmit(skb, ndev);
len -= skb->data_len;
/* Note, this is only an *estimation*, we can still run out of empty
* slots because of the additional fragmentation into
* MAL_MAX_TX_SIZE-sized chunks
*/
if (unlikely(dev->tx_cnt + nr_frags + mal_tx_chunks(len) > NUM_TX_BUFF))
goto stop_queue;
ctrl = EMAC_TX_CTRL_GFCS | EMAC_TX_CTRL_GP | MAL_TX_CTRL_READY |
emac_tx_csum(dev, skb);
slot = dev->tx_slot;
/* skb data */
dev->tx_skb[slot] = NULL;
chunk = min(len, MAL_MAX_TX_SIZE);
dev->tx_desc[slot].data_ptr = pd =
dma_map_single(&dev->ofdev->dev, skb->data, len, DMA_TO_DEVICE);
dev->tx_desc[slot].data_len = (u16) chunk;
len -= chunk;
if (unlikely(len))
slot = emac_xmit_split(dev, slot, pd + chunk, len, !nr_frags,
ctrl);
/* skb fragments */
for (i = 0; i < nr_frags; ++i) {
struct skb_frag_struct *frag = &skb_shinfo(skb)->frags[i];
len = frag->size;
if (unlikely(dev->tx_cnt + mal_tx_chunks(len) >= NUM_TX_BUFF))
goto undo_frame;
pd = dma_map_page(&dev->ofdev->dev, frag->page, frag->page_offset, len,
DMA_TO_DEVICE);
slot = emac_xmit_split(dev, slot, pd, len, i == nr_frags - 1,
ctrl);
}
DBG2(dev, "xmit_sg(%u) %d - %d" NL, skb->len, dev->tx_slot, slot);
/* Attach skb to the last slot so we don't release it too early */
dev->tx_skb[slot] = skb;
/* Send the packet out */
if (dev->tx_slot == NUM_TX_BUFF - 1)
ctrl |= MAL_TX_CTRL_WRAP;
wmb();
dev->tx_desc[dev->tx_slot].ctrl = ctrl;
dev->tx_slot = (slot + 1) % NUM_TX_BUFF;
return emac_xmit_finish(dev, skb->len);
undo_frame:
/* Well, too bad. Our previous estimation was overly optimistic.
* Undo everything.
*/
while (slot != dev->tx_slot) {
dev->tx_desc[slot].ctrl = 0;
--dev->tx_cnt;
if (--slot < 0)
slot = NUM_TX_BUFF - 1;
}
++dev->estats.tx_undo;
stop_queue:
netif_stop_queue(ndev);
DBG2(dev, "stopped TX queue" NL);
net: use symbolic values for ndo_start_xmit() return codes Convert magic values 1 and -1 to NETDEV_TX_BUSY and NETDEV_TX_LOCKED respectively. 0 (NETDEV_TX_OK) is not changed to keep the noise down, except in very few cases where its in direct proximity to one of the other values. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-06-12 06:22:29 +00:00
return NETDEV_TX_BUSY;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
}
/* Tx lock BHs */
static void emac_parse_tx_error(struct emac_instance *dev, u16 ctrl)
{
struct emac_error_stats *st = &dev->estats;
DBG(dev, "BD TX error %04x" NL, ctrl);
++st->tx_bd_errors;
if (ctrl & EMAC_TX_ST_BFCS)
++st->tx_bd_bad_fcs;
if (ctrl & EMAC_TX_ST_LCS)
++st->tx_bd_carrier_loss;
if (ctrl & EMAC_TX_ST_ED)
++st->tx_bd_excessive_deferral;
if (ctrl & EMAC_TX_ST_EC)
++st->tx_bd_excessive_collisions;
if (ctrl & EMAC_TX_ST_LC)
++st->tx_bd_late_collision;
if (ctrl & EMAC_TX_ST_MC)
++st->tx_bd_multple_collisions;
if (ctrl & EMAC_TX_ST_SC)
++st->tx_bd_single_collision;
if (ctrl & EMAC_TX_ST_UR)
++st->tx_bd_underrun;
if (ctrl & EMAC_TX_ST_SQE)
++st->tx_bd_sqe;
}
static void emac_poll_tx(void *param)
{
struct emac_instance *dev = param;
u32 bad_mask;
DBG2(dev, "poll_tx, %d %d" NL, dev->tx_cnt, dev->ack_slot);
if (emac_has_feature(dev, EMAC_FTR_HAS_TAH))
bad_mask = EMAC_IS_BAD_TX_TAH;
else
bad_mask = EMAC_IS_BAD_TX;
netif_tx_lock_bh(dev->ndev);
if (dev->tx_cnt) {
u16 ctrl;
int slot = dev->ack_slot, n = 0;
again:
ctrl = dev->tx_desc[slot].ctrl;
if (!(ctrl & MAL_TX_CTRL_READY)) {
struct sk_buff *skb = dev->tx_skb[slot];
++n;
if (skb) {
dev_kfree_skb(skb);
dev->tx_skb[slot] = NULL;
}
slot = (slot + 1) % NUM_TX_BUFF;
if (unlikely(ctrl & bad_mask))
emac_parse_tx_error(dev, ctrl);
if (--dev->tx_cnt)
goto again;
}
if (n) {
dev->ack_slot = slot;
if (netif_queue_stopped(dev->ndev) &&
dev->tx_cnt < EMAC_TX_WAKEUP_THRESH)
netif_wake_queue(dev->ndev);
DBG2(dev, "tx %d pkts" NL, n);
}
}
netif_tx_unlock_bh(dev->ndev);
}
static inline void emac_recycle_rx_skb(struct emac_instance *dev, int slot,
int len)
{
struct sk_buff *skb = dev->rx_skb[slot];
DBG2(dev, "recycle %d %d" NL, slot, len);
if (len)
dma_map_single(&dev->ofdev->dev, skb->data - 2,
EMAC_DMA_ALIGN(len + 2), DMA_FROM_DEVICE);
dev->rx_desc[slot].data_len = 0;
wmb();
dev->rx_desc[slot].ctrl = MAL_RX_CTRL_EMPTY |
(slot == (NUM_RX_BUFF - 1) ? MAL_RX_CTRL_WRAP : 0);
}
static void emac_parse_rx_error(struct emac_instance *dev, u16 ctrl)
{
struct emac_error_stats *st = &dev->estats;
DBG(dev, "BD RX error %04x" NL, ctrl);
++st->rx_bd_errors;
if (ctrl & EMAC_RX_ST_OE)
++st->rx_bd_overrun;
if (ctrl & EMAC_RX_ST_BP)
++st->rx_bd_bad_packet;
if (ctrl & EMAC_RX_ST_RP)
++st->rx_bd_runt_packet;
if (ctrl & EMAC_RX_ST_SE)
++st->rx_bd_short_event;
if (ctrl & EMAC_RX_ST_AE)
++st->rx_bd_alignment_error;
if (ctrl & EMAC_RX_ST_BFCS)
++st->rx_bd_bad_fcs;
if (ctrl & EMAC_RX_ST_PTL)
++st->rx_bd_packet_too_long;
if (ctrl & EMAC_RX_ST_ORE)
++st->rx_bd_out_of_range;
if (ctrl & EMAC_RX_ST_IRE)
++st->rx_bd_in_range;
}
static inline void emac_rx_csum(struct emac_instance *dev,
struct sk_buff *skb, u16 ctrl)
{
#ifdef CONFIG_IBM_NEW_EMAC_TAH
if (!ctrl && dev->tah_dev) {
skb->ip_summed = CHECKSUM_UNNECESSARY;
++dev->stats.rx_packets_csum;
}
#endif
}
static inline int emac_rx_sg_append(struct emac_instance *dev, int slot)
{
if (likely(dev->rx_sg_skb != NULL)) {
int len = dev->rx_desc[slot].data_len;
int tot_len = dev->rx_sg_skb->len + len;
if (unlikely(tot_len + 2 > dev->rx_skb_size)) {
++dev->estats.rx_dropped_mtu;
dev_kfree_skb(dev->rx_sg_skb);
dev->rx_sg_skb = NULL;
} else {
skb->tail in ibm_newemac should be skb_tail_pointer() ... since that sucker is not 32bit-only and on 64bit skb->tail is an offset, not a pointer. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-14 18:20:30 +00:00
cacheable_memcpy(skb_tail_pointer(dev->rx_sg_skb),
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
dev->rx_skb[slot]->data, len);
skb_put(dev->rx_sg_skb, len);
emac_recycle_rx_skb(dev, slot, len);
return 0;
}
}
emac_recycle_rx_skb(dev, slot, 0);
return -1;
}
/* NAPI poll context */
static int emac_poll_rx(void *param, int budget)
{
struct emac_instance *dev = param;
int slot = dev->rx_slot, received = 0;
DBG2(dev, "poll_rx(%d)" NL, budget);
again:
while (budget > 0) {
int len;
struct sk_buff *skb;
u16 ctrl = dev->rx_desc[slot].ctrl;
if (ctrl & MAL_RX_CTRL_EMPTY)
break;
skb = dev->rx_skb[slot];
mb();
len = dev->rx_desc[slot].data_len;
if (unlikely(!MAL_IS_SINGLE_RX(ctrl)))
goto sg;
ctrl &= EMAC_BAD_RX_MASK;
if (unlikely(ctrl && ctrl != EMAC_RX_TAH_BAD_CSUM)) {
emac_parse_rx_error(dev, ctrl);
++dev->estats.rx_dropped_error;
emac_recycle_rx_skb(dev, slot, 0);
len = 0;
goto next;
}
ibm_newemac: Fixes entry of short packets Short packets has to be discarded by the driver. So this patch addresses the issue of discarding the short packets of size lesser then ethernet header size. Signed-off-by: Sathya Narayanan <sathyan@teamf1.com> Signed-off-by: Stefan Roese <sr@denx.de> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-01 08:58:19 +00:00
if (len < ETH_HLEN) {
++dev->estats.rx_dropped_stack;
emac_recycle_rx_skb(dev, slot, len);
goto next;
}
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
if (len && len < EMAC_RX_COPY_THRESH) {
struct sk_buff *copy_skb =
alloc_skb(len + EMAC_RX_SKB_HEADROOM + 2, GFP_ATOMIC);
if (unlikely(!copy_skb))
goto oom;
skb_reserve(copy_skb, EMAC_RX_SKB_HEADROOM + 2);
cacheable_memcpy(copy_skb->data - 2, skb->data - 2,
len + 2);
emac_recycle_rx_skb(dev, slot, len);
skb = copy_skb;
} else if (unlikely(emac_alloc_rx_skb(dev, slot, GFP_ATOMIC)))
goto oom;
skb_put(skb, len);
push_packet:
skb->protocol = eth_type_trans(skb, dev->ndev);
emac_rx_csum(dev, skb, ctrl);
if (unlikely(netif_receive_skb(skb) == NET_RX_DROP))
++dev->estats.rx_dropped_stack;
next:
++dev->stats.rx_packets;
skip:
dev->stats.rx_bytes += len;
slot = (slot + 1) % NUM_RX_BUFF;
--budget;
++received;
continue;
sg:
if (ctrl & MAL_RX_CTRL_FIRST) {
BUG_ON(dev->rx_sg_skb);
if (unlikely(emac_alloc_rx_skb(dev, slot, GFP_ATOMIC))) {
DBG(dev, "rx OOM %d" NL, slot);
++dev->estats.rx_dropped_oom;
emac_recycle_rx_skb(dev, slot, 0);
} else {
dev->rx_sg_skb = skb;
skb_put(skb, len);
}
} else if (!emac_rx_sg_append(dev, slot) &&
(ctrl & MAL_RX_CTRL_LAST)) {
skb = dev->rx_sg_skb;
dev->rx_sg_skb = NULL;
ctrl &= EMAC_BAD_RX_MASK;
if (unlikely(ctrl && ctrl != EMAC_RX_TAH_BAD_CSUM)) {
emac_parse_rx_error(dev, ctrl);
++dev->estats.rx_dropped_error;
dev_kfree_skb(skb);
len = 0;
} else
goto push_packet;
}
goto skip;
oom:
DBG(dev, "rx OOM %d" NL, slot);
/* Drop the packet and recycle skb */
++dev->estats.rx_dropped_oom;
emac_recycle_rx_skb(dev, slot, 0);
goto next;
}
if (received) {
DBG2(dev, "rx %d BDs" NL, received);
dev->rx_slot = slot;
}
if (unlikely(budget && test_bit(MAL_COMMAC_RX_STOPPED, &dev->commac.flags))) {
mb();
if (!(dev->rx_desc[slot].ctrl & MAL_RX_CTRL_EMPTY)) {
DBG2(dev, "rx restart" NL);
received = 0;
goto again;
}
if (dev->rx_sg_skb) {
DBG2(dev, "dropping partial rx packet" NL);
++dev->estats.rx_dropped_error;
dev_kfree_skb(dev->rx_sg_skb);
dev->rx_sg_skb = NULL;
}
clear_bit(MAL_COMMAC_RX_STOPPED, &dev->commac.flags);
mal_enable_rx_channel(dev->mal, dev->mal_rx_chan);
emac_rx_enable(dev);
dev->rx_slot = 0;
}
return received;
}
/* NAPI poll context */
static int emac_peek_rx(void *param)
{
struct emac_instance *dev = param;
return !(dev->rx_desc[dev->rx_slot].ctrl & MAL_RX_CTRL_EMPTY);
}
/* NAPI poll context */
static int emac_peek_rx_sg(void *param)
{
struct emac_instance *dev = param;
int slot = dev->rx_slot;
while (1) {
u16 ctrl = dev->rx_desc[slot].ctrl;
if (ctrl & MAL_RX_CTRL_EMPTY)
return 0;
else if (ctrl & MAL_RX_CTRL_LAST)
return 1;
slot = (slot + 1) % NUM_RX_BUFF;
/* I'm just being paranoid here :) */
if (unlikely(slot == dev->rx_slot))
return 0;
}
}
/* Hard IRQ */
static void emac_rxde(void *param)
{
struct emac_instance *dev = param;
++dev->estats.rx_stopped;
emac_rx_disable_async(dev);
}
/* Hard IRQ */
static irqreturn_t emac_irq(int irq, void *dev_instance)
{
struct emac_instance *dev = dev_instance;
struct emac_regs __iomem *p = dev->emacp;
struct emac_error_stats *st = &dev->estats;
u32 isr;
spin_lock(&dev->lock);
isr = in_be32(&p->isr);
out_be32(&p->isr, isr);
DBG(dev, "isr = %08x" NL, isr);
if (isr & EMAC4_ISR_TXPE)
++st->tx_parity;
if (isr & EMAC4_ISR_RXPE)
++st->rx_parity;
if (isr & EMAC4_ISR_TXUE)
++st->tx_underrun;
if (isr & EMAC4_ISR_RXOE)
++st->rx_fifo_overrun;
if (isr & EMAC_ISR_OVR)
++st->rx_overrun;
if (isr & EMAC_ISR_BP)
++st->rx_bad_packet;
if (isr & EMAC_ISR_RP)
++st->rx_runt_packet;
if (isr & EMAC_ISR_SE)
++st->rx_short_event;
if (isr & EMAC_ISR_ALE)
++st->rx_alignment_error;
if (isr & EMAC_ISR_BFCS)
++st->rx_bad_fcs;
if (isr & EMAC_ISR_PTLE)
++st->rx_packet_too_long;
if (isr & EMAC_ISR_ORE)
++st->rx_out_of_range;
if (isr & EMAC_ISR_IRE)
++st->rx_in_range;
if (isr & EMAC_ISR_SQE)
++st->tx_sqe;
if (isr & EMAC_ISR_TE)
++st->tx_errors;
spin_unlock(&dev->lock);
return IRQ_HANDLED;
}
static struct net_device_stats *emac_stats(struct net_device *ndev)
{
struct emac_instance *dev = netdev_priv(ndev);
struct emac_stats *st = &dev->stats;
struct emac_error_stats *est = &dev->estats;
struct net_device_stats *nst = &dev->nstats;
unsigned long flags;
DBG2(dev, "stats" NL);
/* Compute "legacy" statistics */
spin_lock_irqsave(&dev->lock, flags);
nst->rx_packets = (unsigned long)st->rx_packets;
nst->rx_bytes = (unsigned long)st->rx_bytes;
nst->tx_packets = (unsigned long)st->tx_packets;
nst->tx_bytes = (unsigned long)st->tx_bytes;
nst->rx_dropped = (unsigned long)(est->rx_dropped_oom +
est->rx_dropped_error +
est->rx_dropped_resize +
est->rx_dropped_mtu);
nst->tx_dropped = (unsigned long)est->tx_dropped;
nst->rx_errors = (unsigned long)est->rx_bd_errors;
nst->rx_fifo_errors = (unsigned long)(est->rx_bd_overrun +
est->rx_fifo_overrun +
est->rx_overrun);
nst->rx_frame_errors = (unsigned long)(est->rx_bd_alignment_error +
est->rx_alignment_error);
nst->rx_crc_errors = (unsigned long)(est->rx_bd_bad_fcs +
est->rx_bad_fcs);
nst->rx_length_errors = (unsigned long)(est->rx_bd_runt_packet +
est->rx_bd_short_event +
est->rx_bd_packet_too_long +
est->rx_bd_out_of_range +
est->rx_bd_in_range +
est->rx_runt_packet +
est->rx_short_event +
est->rx_packet_too_long +
est->rx_out_of_range +
est->rx_in_range);
nst->tx_errors = (unsigned long)(est->tx_bd_errors + est->tx_errors);
nst->tx_fifo_errors = (unsigned long)(est->tx_bd_underrun +
est->tx_underrun);
nst->tx_carrier_errors = (unsigned long)est->tx_bd_carrier_loss;
nst->collisions = (unsigned long)(est->tx_bd_excessive_deferral +
est->tx_bd_excessive_collisions +
est->tx_bd_late_collision +
est->tx_bd_multple_collisions);
spin_unlock_irqrestore(&dev->lock, flags);
return nst;
}
static struct mal_commac_ops emac_commac_ops = {
.poll_tx = &emac_poll_tx,
.poll_rx = &emac_poll_rx,
.peek_rx = &emac_peek_rx,
.rxde = &emac_rxde,
};
static struct mal_commac_ops emac_commac_sg_ops = {
.poll_tx = &emac_poll_tx,
.poll_rx = &emac_poll_rx,
.peek_rx = &emac_peek_rx_sg,
.rxde = &emac_rxde,
};
/* Ethtool support */
static int emac_ethtool_get_settings(struct net_device *ndev,
struct ethtool_cmd *cmd)
{
struct emac_instance *dev = netdev_priv(ndev);
cmd->supported = dev->phy.features;
cmd->port = PORT_MII;
cmd->phy_address = dev->phy.address;
cmd->transceiver =
dev->phy.address >= 0 ? XCVR_EXTERNAL : XCVR_INTERNAL;
mutex_lock(&dev->link_lock);
cmd->advertising = dev->phy.advertising;
cmd->autoneg = dev->phy.autoneg;
cmd->speed = dev->phy.speed;
cmd->duplex = dev->phy.duplex;
mutex_unlock(&dev->link_lock);
return 0;
}
static int emac_ethtool_set_settings(struct net_device *ndev,
struct ethtool_cmd *cmd)
{
struct emac_instance *dev = netdev_priv(ndev);
u32 f = dev->phy.features;
DBG(dev, "set_settings(%d, %d, %d, 0x%08x)" NL,
cmd->autoneg, cmd->speed, cmd->duplex, cmd->advertising);
/* Basic sanity checks */
if (dev->phy.address < 0)
return -EOPNOTSUPP;
if (cmd->autoneg != AUTONEG_ENABLE && cmd->autoneg != AUTONEG_DISABLE)
return -EINVAL;
if (cmd->autoneg == AUTONEG_ENABLE && cmd->advertising == 0)
return -EINVAL;
if (cmd->duplex != DUPLEX_HALF && cmd->duplex != DUPLEX_FULL)
return -EINVAL;
if (cmd->autoneg == AUTONEG_DISABLE) {
switch (cmd->speed) {
case SPEED_10:
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
if (cmd->duplex == DUPLEX_HALF &&
!(f & SUPPORTED_10baseT_Half))
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
return -EINVAL;
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
if (cmd->duplex == DUPLEX_FULL &&
!(f & SUPPORTED_10baseT_Full))
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
return -EINVAL;
break;
case SPEED_100:
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
if (cmd->duplex == DUPLEX_HALF &&
!(f & SUPPORTED_100baseT_Half))
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
return -EINVAL;
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
if (cmd->duplex == DUPLEX_FULL &&
!(f & SUPPORTED_100baseT_Full))
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
return -EINVAL;
break;
case SPEED_1000:
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
if (cmd->duplex == DUPLEX_HALF &&
!(f & SUPPORTED_1000baseT_Half))
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
return -EINVAL;
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
if (cmd->duplex == DUPLEX_FULL &&
!(f & SUPPORTED_1000baseT_Full))
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
return -EINVAL;
break;
default:
return -EINVAL;
}
mutex_lock(&dev->link_lock);
dev->phy.def->ops->setup_forced(&dev->phy, cmd->speed,
cmd->duplex);
mutex_unlock(&dev->link_lock);
} else {
if (!(f & SUPPORTED_Autoneg))
return -EINVAL;
mutex_lock(&dev->link_lock);
dev->phy.def->ops->setup_aneg(&dev->phy,
(cmd->advertising & f) |
(dev->phy.advertising &
(ADVERTISED_Pause |
ADVERTISED_Asym_Pause)));
mutex_unlock(&dev->link_lock);
}
emac_force_link_update(dev);
return 0;
}
static void emac_ethtool_get_ringparam(struct net_device *ndev,
struct ethtool_ringparam *rp)
{
rp->rx_max_pending = rp->rx_pending = NUM_RX_BUFF;
rp->tx_max_pending = rp->tx_pending = NUM_TX_BUFF;
}
static void emac_ethtool_get_pauseparam(struct net_device *ndev,
struct ethtool_pauseparam *pp)
{
struct emac_instance *dev = netdev_priv(ndev);
mutex_lock(&dev->link_lock);
if ((dev->phy.features & SUPPORTED_Autoneg) &&
(dev->phy.advertising & (ADVERTISED_Pause | ADVERTISED_Asym_Pause)))
pp->autoneg = 1;
if (dev->phy.duplex == DUPLEX_FULL) {
if (dev->phy.pause)
pp->rx_pause = pp->tx_pause = 1;
else if (dev->phy.asym_pause)
pp->tx_pause = 1;
}
mutex_unlock(&dev->link_lock);
}
static u32 emac_ethtool_get_rx_csum(struct net_device *ndev)
{
struct emac_instance *dev = netdev_priv(ndev);
ibm_newemac annotations (iomem, NULL noise) Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-14 18:36:10 +00:00
return dev->tah_dev != NULL;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
}
static int emac_get_regs_len(struct emac_instance *dev)
{
if (emac_has_feature(dev, EMAC_FTR_EMAC4))
return sizeof(struct emac_ethtool_regs_subhdr) +
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
EMAC4_ETHTOOL_REGS_SIZE(dev);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
else
return sizeof(struct emac_ethtool_regs_subhdr) +
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
EMAC_ETHTOOL_REGS_SIZE(dev);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
}
static int emac_ethtool_get_regs_len(struct net_device *ndev)
{
struct emac_instance *dev = netdev_priv(ndev);
int size;
size = sizeof(struct emac_ethtool_regs_hdr) +
emac_get_regs_len(dev) + mal_get_regs_len(dev->mal);
if (emac_has_feature(dev, EMAC_FTR_HAS_ZMII))
size += zmii_get_regs_len(dev->zmii_dev);
if (emac_has_feature(dev, EMAC_FTR_HAS_RGMII))
size += rgmii_get_regs_len(dev->rgmii_dev);
if (emac_has_feature(dev, EMAC_FTR_HAS_TAH))
size += tah_get_regs_len(dev->tah_dev);
return size;
}
static void *emac_dump_regs(struct emac_instance *dev, void *buf)
{
struct emac_ethtool_regs_subhdr *hdr = buf;
hdr->index = dev->cell_index;
if (emac_has_feature(dev, EMAC_FTR_EMAC4)) {
hdr->version = EMAC4_ETHTOOL_REGS_VER;
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
memcpy_fromio(hdr + 1, dev->emacp, EMAC4_ETHTOOL_REGS_SIZE(dev));
return ((void *)(hdr + 1) + EMAC4_ETHTOOL_REGS_SIZE(dev));
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
} else {
hdr->version = EMAC_ETHTOOL_REGS_VER;
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
memcpy_fromio(hdr + 1, dev->emacp, EMAC_ETHTOOL_REGS_SIZE(dev));
return ((void *)(hdr + 1) + EMAC_ETHTOOL_REGS_SIZE(dev));
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
}
}
static void emac_ethtool_get_regs(struct net_device *ndev,
struct ethtool_regs *regs, void *buf)
{
struct emac_instance *dev = netdev_priv(ndev);
struct emac_ethtool_regs_hdr *hdr = buf;
hdr->components = 0;
buf = hdr + 1;
buf = mal_dump_regs(dev->mal, buf);
buf = emac_dump_regs(dev, buf);
if (emac_has_feature(dev, EMAC_FTR_HAS_ZMII)) {
hdr->components |= EMAC_ETHTOOL_REGS_ZMII;
buf = zmii_dump_regs(dev->zmii_dev, buf);
}
if (emac_has_feature(dev, EMAC_FTR_HAS_RGMII)) {
hdr->components |= EMAC_ETHTOOL_REGS_RGMII;
buf = rgmii_dump_regs(dev->rgmii_dev, buf);
}
if (emac_has_feature(dev, EMAC_FTR_HAS_TAH)) {
hdr->components |= EMAC_ETHTOOL_REGS_TAH;
buf = tah_dump_regs(dev->tah_dev, buf);
}
}
static int emac_ethtool_nway_reset(struct net_device *ndev)
{
struct emac_instance *dev = netdev_priv(ndev);
int res = 0;
DBG(dev, "nway_reset" NL);
if (dev->phy.address < 0)
return -EOPNOTSUPP;
mutex_lock(&dev->link_lock);
if (!dev->phy.autoneg) {
res = -EINVAL;
goto out;
}
dev->phy.def->ops->setup_aneg(&dev->phy, dev->phy.advertising);
out:
mutex_unlock(&dev->link_lock);
emac_force_link_update(dev);
return res;
}
net: Convert ethtool {get_stats, self_test}_count() ops to get_sset_count() These string query operations were supposed to be replaced by the generic get_sset_count() starting in 2007. Convert the remaining implementations. Also remove calls to these operations to initialise drvinfo->n_stats. The ethtool core code already does that. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Acked-by: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-01 11:58:24 +00:00
static int emac_ethtool_get_sset_count(struct net_device *ndev, int stringset)
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
{
net: Convert ethtool {get_stats, self_test}_count() ops to get_sset_count() These string query operations were supposed to be replaced by the generic get_sset_count() starting in 2007. Convert the remaining implementations. Also remove calls to these operations to initialise drvinfo->n_stats. The ethtool core code already does that. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Acked-by: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-01 11:58:24 +00:00
if (stringset == ETH_SS_STATS)
return EMAC_ETHTOOL_STATS_COUNT;
else
return -EINVAL;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
}
static void emac_ethtool_get_strings(struct net_device *ndev, u32 stringset,
u8 * buf)
{
if (stringset == ETH_SS_STATS)
memcpy(buf, &emac_stats_keys, sizeof(emac_stats_keys));
}
static void emac_ethtool_get_ethtool_stats(struct net_device *ndev,
struct ethtool_stats *estats,
u64 * tmp_stats)
{
struct emac_instance *dev = netdev_priv(ndev);
memcpy(tmp_stats, &dev->stats, sizeof(dev->stats));
tmp_stats += sizeof(dev->stats) / sizeof(u64);
memcpy(tmp_stats, &dev->estats, sizeof(dev->estats));
}
static void emac_ethtool_get_drvinfo(struct net_device *ndev,
struct ethtool_drvinfo *info)
{
struct emac_instance *dev = netdev_priv(ndev);
strcpy(info->driver, "ibm_emac");
strcpy(info->version, DRV_VERSION);
info->fw_version[0] = '\0';
sprintf(info->bus_info, "PPC 4xx EMAC-%d %s",
of: Always use 'struct device.of_node' to get device node pointer. The following structure elements duplicate the information in 'struct device.of_node' and so are being eliminated. This patch makes all readers of these elements use device.of_node instead. (struct of_device *)->node (struct dev_archdata *)->prom_node (sparc) (struct dev_archdata *)->of_node (powerpc & microblaze) Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2010-04-13 23:12:29 +00:00
dev->cell_index, dev->ofdev->dev.of_node->full_name);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
info->regdump_len = emac_ethtool_get_regs_len(ndev);
}
static const struct ethtool_ops emac_ethtool_ops = {
.get_settings = emac_ethtool_get_settings,
.set_settings = emac_ethtool_set_settings,
.get_drvinfo = emac_ethtool_get_drvinfo,
.get_regs_len = emac_ethtool_get_regs_len,
.get_regs = emac_ethtool_get_regs,
.nway_reset = emac_ethtool_nway_reset,
.get_ringparam = emac_ethtool_get_ringparam,
.get_pauseparam = emac_ethtool_get_pauseparam,
.get_rx_csum = emac_ethtool_get_rx_csum,
.get_strings = emac_ethtool_get_strings,
net: Convert ethtool {get_stats, self_test}_count() ops to get_sset_count() These string query operations were supposed to be replaced by the generic get_sset_count() starting in 2007. Convert the remaining implementations. Also remove calls to these operations to initialise drvinfo->n_stats. The ethtool core code already does that. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Acked-by: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-01 11:58:24 +00:00
.get_sset_count = emac_ethtool_get_sset_count,
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
.get_ethtool_stats = emac_ethtool_get_ethtool_stats,
.get_link = ethtool_op_get_link,
.get_tx_csum = ethtool_op_get_tx_csum,
.get_sg = ethtool_op_get_sg,
};
static int emac_ioctl(struct net_device *ndev, struct ifreq *rq, int cmd)
{
struct emac_instance *dev = netdev_priv(ndev);
netdev: Convert MDIO ioctl implementation to use struct mii_ioctl_data A few drivers still access the arguments to MDIO ioctls as an array of u16. Convert them to use struct mii_ioctl_data. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-03 10:41:17 +00:00
struct mii_ioctl_data *data = if_mii(rq);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
DBG(dev, "ioctl %08x" NL, cmd);
if (dev->phy.address < 0)
return -EOPNOTSUPP;
switch (cmd) {
case SIOCGMIIPHY:
netdev: Convert MDIO ioctl implementation to use struct mii_ioctl_data A few drivers still access the arguments to MDIO ioctls as an array of u16. Convert them to use struct mii_ioctl_data. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-03 10:41:17 +00:00
data->phy_id = dev->phy.address;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
/* Fall through */
case SIOCGMIIREG:
netdev: Convert MDIO ioctl implementation to use struct mii_ioctl_data A few drivers still access the arguments to MDIO ioctls as an array of u16. Convert them to use struct mii_ioctl_data. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-03 10:41:17 +00:00
data->val_out = emac_mdio_read(ndev, dev->phy.address,
data->reg_num);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
return 0;
case SIOCSMIIREG:
netdev: Convert MDIO ioctl implementation to use struct mii_ioctl_data A few drivers still access the arguments to MDIO ioctls as an array of u16. Convert them to use struct mii_ioctl_data. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-09-03 10:41:17 +00:00
emac_mdio_write(ndev, dev->phy.address, data->reg_num,
data->val_in);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
return 0;
default:
return -EOPNOTSUPP;
}
}
struct emac_depentry {
u32 phandle;
struct device_node *node;
struct of_device *ofdev;
void *drvdata;
};
#define EMAC_DEP_MAL_IDX 0
#define EMAC_DEP_ZMII_IDX 1
#define EMAC_DEP_RGMII_IDX 2
#define EMAC_DEP_TAH_IDX 3
#define EMAC_DEP_MDIO_IDX 4
#define EMAC_DEP_PREV_IDX 5
#define EMAC_DEP_COUNT 6
static int __devinit emac_check_deps(struct emac_instance *dev,
struct emac_depentry *deps)
{
int i, there = 0;
struct device_node *np;
for (i = 0; i < EMAC_DEP_COUNT; i++) {
/* no dependency on that item, allright */
if (deps[i].phandle == 0) {
there++;
continue;
}
/* special case for blist as the dependency might go away */
if (i == EMAC_DEP_PREV_IDX) {
np = *(dev->blist - 1);
if (np == NULL) {
deps[i].phandle = 0;
there++;
continue;
}
if (deps[i].node == NULL)
deps[i].node = of_node_get(np);
}
if (deps[i].node == NULL)
deps[i].node = of_find_node_by_phandle(deps[i].phandle);
if (deps[i].node == NULL)
continue;
if (deps[i].ofdev == NULL)
deps[i].ofdev = of_find_device_by_node(deps[i].node);
if (deps[i].ofdev == NULL)
continue;
if (deps[i].drvdata == NULL)
deps[i].drvdata = dev_get_drvdata(&deps[i].ofdev->dev);
if (deps[i].drvdata != NULL)
there++;
}
return (there == EMAC_DEP_COUNT);
}
static void emac_put_deps(struct emac_instance *dev)
{
if (dev->mal_dev)
of_dev_put(dev->mal_dev);
if (dev->zmii_dev)
of_dev_put(dev->zmii_dev);
if (dev->rgmii_dev)
of_dev_put(dev->rgmii_dev);
if (dev->mdio_dev)
of_dev_put(dev->mdio_dev);
if (dev->tah_dev)
of_dev_put(dev->tah_dev);
}
static int __devinit emac_of_bus_notify(struct notifier_block *nb,
unsigned long action, void *data)
{
/* We are only intereted in device addition */
if (action == BUS_NOTIFY_BOUND_DRIVER)
wake_up_all(&emac_probe_wait);
return 0;
}
ibm_newemac: Fix section mismatch warnings This patch fixes several section mismatch warnings in the ibm_newemac driver similar to: WARNING: vmlinux.o(.devinit.text+0x3a04): Section mismatch in reference from the function emac_probe() to the function .devexit.text:tah_detach() The function __devinit emac_probe() references a function __devexit tah_detach(). Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:43 +00:00
static struct notifier_block emac_of_bus_notifier __devinitdata = {
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
.notifier_call = emac_of_bus_notify
};
static int __devinit emac_wait_deps(struct emac_instance *dev)
{
struct emac_depentry deps[EMAC_DEP_COUNT];
int i, err;
memset(&deps, 0, sizeof(deps));
deps[EMAC_DEP_MAL_IDX].phandle = dev->mal_ph;
deps[EMAC_DEP_ZMII_IDX].phandle = dev->zmii_ph;
deps[EMAC_DEP_RGMII_IDX].phandle = dev->rgmii_ph;
if (dev->tah_ph)
deps[EMAC_DEP_TAH_IDX].phandle = dev->tah_ph;
if (dev->mdio_ph)
deps[EMAC_DEP_MDIO_IDX].phandle = dev->mdio_ph;
if (dev->blist && dev->blist > emac_boot_list)
deps[EMAC_DEP_PREV_IDX].phandle = 0xffffffffu;
bus_register_notifier(&of_platform_bus_type, &emac_of_bus_notifier);
wait_event_timeout(emac_probe_wait,
emac_check_deps(dev, deps),
EMAC_PROBE_DEP_TIMEOUT);
bus_unregister_notifier(&of_platform_bus_type, &emac_of_bus_notifier);
err = emac_check_deps(dev, deps) ? 0 : -ENODEV;
for (i = 0; i < EMAC_DEP_COUNT; i++) {
if (deps[i].node)
of_node_put(deps[i].node);
if (err && deps[i].ofdev)
of_dev_put(deps[i].ofdev);
}
if (err == 0) {
dev->mal_dev = deps[EMAC_DEP_MAL_IDX].ofdev;
dev->zmii_dev = deps[EMAC_DEP_ZMII_IDX].ofdev;
dev->rgmii_dev = deps[EMAC_DEP_RGMII_IDX].ofdev;
dev->tah_dev = deps[EMAC_DEP_TAH_IDX].ofdev;
dev->mdio_dev = deps[EMAC_DEP_MDIO_IDX].ofdev;
}
if (deps[EMAC_DEP_PREV_IDX].ofdev)
of_dev_put(deps[EMAC_DEP_PREV_IDX].ofdev);
return err;
}
static int __devinit emac_read_uint_prop(struct device_node *np, const char *name,
u32 *val, int fatal)
{
int len;
const u32 *prop = of_get_property(np, name, &len);
if (prop == NULL || len < sizeof(u32)) {
if (fatal)
printk(KERN_ERR "%s: missing %s property\n",
np->full_name, name);
return -ENODEV;
}
*val = *prop;
return 0;
}
static int __devinit emac_init_phy(struct emac_instance *dev)
{
of: Always use 'struct device.of_node' to get device node pointer. The following structure elements duplicate the information in 'struct device.of_node' and so are being eliminated. This patch makes all readers of these elements use device.of_node instead. (struct of_device *)->node (struct dev_archdata *)->prom_node (sparc) (struct dev_archdata *)->of_node (powerpc & microblaze) Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2010-04-13 23:12:29 +00:00
struct device_node *np = dev->ofdev->dev.of_node;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
struct net_device *ndev = dev->ndev;
u32 phy_map, adv;
int i;
dev->phy.dev = ndev;
dev->phy.mode = dev->phy_mode;
/* PHY-less configuration.
* XXX I probably should move these settings to the dev tree
*/
if (dev->phy_address == 0xffffffff && dev->phy_map == 0xffffffff) {
emac_reset(dev);
/* PHY-less configuration.
* XXX I probably should move these settings to the dev tree
*/
dev->phy.address = -1;
ibm_newemac: Add support for GPCS, SGMII and M88E1112 PHY Add support for the phy types found on the Arches and other PowerPC 460 based boards. Signed-off-by: Victor Gallardo <vgallardo@amcc.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-10-02 06:37:57 +00:00
dev->phy.features = SUPPORTED_MII;
if (emac_phy_supports_gige(dev->phy_mode))
dev->phy.features |= SUPPORTED_1000baseT_Full;
else
dev->phy.features |= SUPPORTED_100baseT_Full;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
dev->phy.pause = 1;
return 0;
}
mutex_lock(&emac_phy_map_lock);
phy_map = dev->phy_map | busy_phy_map;
DBG(dev, "PHY maps %08x %08x" NL, dev->phy_map, busy_phy_map);
dev->phy.mdio_read = emac_mdio_read;
dev->phy.mdio_write = emac_mdio_write;
ibm_newemac: PowerPC 440GX EMAC PHY clock workaround The PowerPC 440GX Taishan board fails to reset EMAC3 (reset timeout error) if there's no link. Because of that it fails to find PHY chip. The older ibm_emac driver had a workaround for that: the EMAC_CLK_INTERNAL/EMAC_CLK_EXTERNAL macros, which toggle the Ethernet Clock Select bit in the SDR0_MFR register. This patch does the same for "ibm,emac-440gx" compatible chips. The workaround forces clock on -all- EMACs, so we select clock under global emac_phy_map_lock. BenH: Made that #ifdef CONFIG_PPC_DCR_NATIVE for now as dcri_* stuff doesn't exist for MMIO type DCRs like Cell. Some future rework & improvements of the DCR infrastructure will make that cleaner but for now, this makes it work. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:46 +00:00
/* Enable internal clock source */
#ifdef CONFIG_PPC_DCR_NATIVE
if (emac_has_feature(dev, EMAC_FTR_440GX_PHY_CLK_FIX))
dcri_clrset(SDR0, SDR0_MFR, 0, SDR0_MFR_ECS);
ibm_newemac: PowerPC 440EP/440GR EMAC PHY clock workaround This patch adds ibm_newemac PHY clock workaround for 440EP/440GR EMAC attached to a PHY which doesn't generate RX clock if there is no link. The code is based on the previous ibm_emac driver stuff. The 440EP/440GR allows controlling each EMAC clock separately as opposed to global clock selection for 440GX. BenH: Made that #ifdef CONFIG_PPC_DCR_NATIVE for now as dcri_* stuff doesn't exist for MMIO type DCRs like Cell. Some future rework & improvements of the DCR infrastructure will make that cleaner but for now, this makes it work. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:48 +00:00
#endif
/* PHY clock workaround */
emac_rx_clk_tx(dev);
/* Enable internal clock source on 440GX*/
#ifdef CONFIG_PPC_DCR_NATIVE
if (emac_has_feature(dev, EMAC_FTR_440GX_PHY_CLK_FIX))
dcri_clrset(SDR0, SDR0_MFR, 0, SDR0_MFR_ECS);
ibm_newemac: PowerPC 440GX EMAC PHY clock workaround The PowerPC 440GX Taishan board fails to reset EMAC3 (reset timeout error) if there's no link. Because of that it fails to find PHY chip. The older ibm_emac driver had a workaround for that: the EMAC_CLK_INTERNAL/EMAC_CLK_EXTERNAL macros, which toggle the Ethernet Clock Select bit in the SDR0_MFR register. This patch does the same for "ibm,emac-440gx" compatible chips. The workaround forces clock on -all- EMACs, so we select clock under global emac_phy_map_lock. BenH: Made that #ifdef CONFIG_PPC_DCR_NATIVE for now as dcri_* stuff doesn't exist for MMIO type DCRs like Cell. Some future rework & improvements of the DCR infrastructure will make that cleaner but for now, this makes it work. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:46 +00:00
#endif
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
/* Configure EMAC with defaults so we can at least use MDIO
* This is needed mostly for 440GX
*/
if (emac_phy_gpcs(dev->phy.mode)) {
/* XXX
* Make GPCS PHY address equal to EMAC index.
* We probably should take into account busy_phy_map
* and/or phy_map here.
*
* Note that the busy_phy_map is currently global
* while it should probably be per-ASIC...
*/
ibm_newemac: Add support for GPCS, SGMII and M88E1112 PHY Add support for the phy types found on the Arches and other PowerPC 460 based boards. Signed-off-by: Victor Gallardo <vgallardo@amcc.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-10-02 06:37:57 +00:00
dev->phy.gpcs_address = dev->gpcs_address;
if (dev->phy.gpcs_address == 0xffffffff)
dev->phy.address = dev->cell_index;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
}
emac_configure(dev);
if (dev->phy_address != 0xffffffff)
phy_map = ~(1 << dev->phy_address);
for (i = 0; i < 0x20; phy_map >>= 1, ++i)
if (!(phy_map & 1)) {
int r;
busy_phy_map |= 1 << i;
/* Quick check if there is a PHY at the address */
r = emac_mdio_read(dev->ndev, i, MII_BMCR);
if (r == 0xffff || r < 0)
continue;
if (!emac_mii_phy_probe(&dev->phy, i))
break;
}
ibm_newemac: PowerPC 440GX EMAC PHY clock workaround The PowerPC 440GX Taishan board fails to reset EMAC3 (reset timeout error) if there's no link. Because of that it fails to find PHY chip. The older ibm_emac driver had a workaround for that: the EMAC_CLK_INTERNAL/EMAC_CLK_EXTERNAL macros, which toggle the Ethernet Clock Select bit in the SDR0_MFR register. This patch does the same for "ibm,emac-440gx" compatible chips. The workaround forces clock on -all- EMACs, so we select clock under global emac_phy_map_lock. BenH: Made that #ifdef CONFIG_PPC_DCR_NATIVE for now as dcri_* stuff doesn't exist for MMIO type DCRs like Cell. Some future rework & improvements of the DCR infrastructure will make that cleaner but for now, this makes it work. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:46 +00:00
/* Enable external clock source */
#ifdef CONFIG_PPC_DCR_NATIVE
if (emac_has_feature(dev, EMAC_FTR_440GX_PHY_CLK_FIX))
dcri_clrset(SDR0, SDR0_MFR, SDR0_MFR_ECS, 0);
#endif
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
mutex_unlock(&emac_phy_map_lock);
if (i == 0x20) {
printk(KERN_WARNING "%s: can't find PHY!\n", np->full_name);
return -ENXIO;
}
/* Init PHY */
if (dev->phy.def->ops->init)
dev->phy.def->ops->init(&dev->phy);
/* Disable any PHY features not supported by the platform */
dev->phy.def->features &= ~dev->phy_feat_exc;
/* Setup initial link parameters */
if (dev->phy.features & SUPPORTED_Autoneg) {
adv = dev->phy.features;
if (!emac_has_feature(dev, EMAC_FTR_NO_FLOW_CONTROL_40x))
adv |= ADVERTISED_Pause | ADVERTISED_Asym_Pause;
/* Restart autonegotiation */
dev->phy.def->ops->setup_aneg(&dev->phy, adv);
} else {
u32 f = dev->phy.def->features;
int speed = SPEED_10, fd = DUPLEX_HALF;
/* Select highest supported speed/duplex */
if (f & SUPPORTED_1000baseT_Full) {
speed = SPEED_1000;
fd = DUPLEX_FULL;
} else if (f & SUPPORTED_1000baseT_Half)
speed = SPEED_1000;
else if (f & SUPPORTED_100baseT_Full) {
speed = SPEED_100;
fd = DUPLEX_FULL;
} else if (f & SUPPORTED_100baseT_Half)
speed = SPEED_100;
else if (f & SUPPORTED_10baseT_Full)
fd = DUPLEX_FULL;
/* Force link parameters */
dev->phy.def->ops->setup_forced(&dev->phy, speed, fd);
}
return 0;
}
static int __devinit emac_init_config(struct emac_instance *dev)
{
of: Always use 'struct device.of_node' to get device node pointer. The following structure elements duplicate the information in 'struct device.of_node' and so are being eliminated. This patch makes all readers of these elements use device.of_node instead. (struct of_device *)->node (struct dev_archdata *)->prom_node (sparc) (struct dev_archdata *)->of_node (powerpc & microblaze) Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2010-04-13 23:12:29 +00:00
struct device_node *np = dev->ofdev->dev.of_node;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
const void *p;
unsigned int plen;
const char *pm, *phy_modes[] = {
[PHY_MODE_NA] = "",
[PHY_MODE_MII] = "mii",
[PHY_MODE_RMII] = "rmii",
[PHY_MODE_SMII] = "smii",
[PHY_MODE_RGMII] = "rgmii",
[PHY_MODE_TBI] = "tbi",
[PHY_MODE_GMII] = "gmii",
[PHY_MODE_RTBI] = "rtbi",
[PHY_MODE_SGMII] = "sgmii",
};
/* Read config from device-tree */
if (emac_read_uint_prop(np, "mal-device", &dev->mal_ph, 1))
return -ENXIO;
if (emac_read_uint_prop(np, "mal-tx-channel", &dev->mal_tx_chan, 1))
return -ENXIO;
if (emac_read_uint_prop(np, "mal-rx-channel", &dev->mal_rx_chan, 1))
return -ENXIO;
if (emac_read_uint_prop(np, "cell-index", &dev->cell_index, 1))
return -ENXIO;
if (emac_read_uint_prop(np, "max-frame-size", &dev->max_mtu, 0))
dev->max_mtu = 1500;
if (emac_read_uint_prop(np, "rx-fifo-size", &dev->rx_fifo_size, 0))
dev->rx_fifo_size = 2048;
if (emac_read_uint_prop(np, "tx-fifo-size", &dev->tx_fifo_size, 0))
dev->tx_fifo_size = 2048;
if (emac_read_uint_prop(np, "rx-fifo-size-gige", &dev->rx_fifo_size_gige, 0))
dev->rx_fifo_size_gige = dev->rx_fifo_size;
if (emac_read_uint_prop(np, "tx-fifo-size-gige", &dev->tx_fifo_size_gige, 0))
dev->tx_fifo_size_gige = dev->tx_fifo_size;
if (emac_read_uint_prop(np, "phy-address", &dev->phy_address, 0))
dev->phy_address = 0xffffffff;
if (emac_read_uint_prop(np, "phy-map", &dev->phy_map, 0))
dev->phy_map = 0xffffffff;
ibm_newemac: Add support for GPCS, SGMII and M88E1112 PHY Add support for the phy types found on the Arches and other PowerPC 460 based boards. Signed-off-by: Victor Gallardo <vgallardo@amcc.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-10-02 06:37:57 +00:00
if (emac_read_uint_prop(np, "gpcs-address", &dev->gpcs_address, 0))
dev->gpcs_address = 0xffffffff;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
if (emac_read_uint_prop(np->parent, "clock-frequency", &dev->opb_bus_freq, 1))
return -ENXIO;
if (emac_read_uint_prop(np, "tah-device", &dev->tah_ph, 0))
dev->tah_ph = 0;
if (emac_read_uint_prop(np, "tah-channel", &dev->tah_port, 0))
ibm_newemac: Fix typo reading TAH channel info This patch fixes a typo in ibm_newemac/core.c (tah_port should be used instead of tah_ph) Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-05 00:14:31 +00:00
dev->tah_port = 0;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
if (emac_read_uint_prop(np, "mdio-device", &dev->mdio_ph, 0))
dev->mdio_ph = 0;
if (emac_read_uint_prop(np, "zmii-device", &dev->zmii_ph, 0))
trivial: remove unnecessary semicolons Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2009-08-18 18:18:35 +00:00
dev->zmii_ph = 0;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
if (emac_read_uint_prop(np, "zmii-channel", &dev->zmii_port, 0))
trivial: remove unnecessary semicolons Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2009-08-18 18:18:35 +00:00
dev->zmii_port = 0xffffffff;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
if (emac_read_uint_prop(np, "rgmii-device", &dev->rgmii_ph, 0))
trivial: remove unnecessary semicolons Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2009-08-18 18:18:35 +00:00
dev->rgmii_ph = 0;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
if (emac_read_uint_prop(np, "rgmii-channel", &dev->rgmii_port, 0))
trivial: remove unnecessary semicolons Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2009-08-18 18:18:35 +00:00
dev->rgmii_port = 0xffffffff;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
if (emac_read_uint_prop(np, "fifo-entry-size", &dev->fifo_entry_size, 0))
dev->fifo_entry_size = 16;
if (emac_read_uint_prop(np, "mal-burst-size", &dev->mal_burst_size, 0))
dev->mal_burst_size = 256;
/* PHY mode needs some decoding */
dev->phy_mode = PHY_MODE_NA;
pm = of_get_property(np, "phy-mode", &plen);
if (pm != NULL) {
int i;
for (i = 0; i < ARRAY_SIZE(phy_modes); i++)
if (!strcasecmp(pm, phy_modes[i])) {
dev->phy_mode = i;
break;
}
}
/* Backward compat with non-final DT */
if (dev->phy_mode == PHY_MODE_NA && pm != NULL && plen == 4) {
u32 nmode = *(const u32 *)pm;
if (nmode > PHY_MODE_NA && nmode <= PHY_MODE_SGMII)
dev->phy_mode = nmode;
}
/* Check EMAC version */
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
if (of_device_is_compatible(np, "ibm,emac4sync")) {
dev->features |= (EMAC_FTR_EMAC4 | EMAC_FTR_EMAC4SYNC);
ibm_newemac: Add support for GPCS, SGMII and M88E1112 PHY Add support for the phy types found on the Arches and other PowerPC 460 based boards. Signed-off-by: Victor Gallardo <vgallardo@amcc.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-10-02 06:37:57 +00:00
if (of_device_is_compatible(np, "ibm,emac-460ex") ||
of_device_is_compatible(np, "ibm,emac-460gt"))
dev->features |= EMAC_FTR_460EX_PHY_CLK_FIX;
emac: Fix clock control for 405EX and 405EXr chips The EMAC variant in the 405EX and 405EXr chips needs the "440EP" type clock control workaround to avoid lockups of the Rx side during reset. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Tested-by: Felix Radensky <felix@embedded-sol.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-13 20:48:46 +00:00
if (of_device_is_compatible(np, "ibm,emac-405ex") ||
of_device_is_compatible(np, "ibm,emac-405exr"))
dev->features |= EMAC_FTR_440EP_PHY_CLK_FIX;
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
} else if (of_device_is_compatible(np, "ibm,emac4")) {
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
dev->features |= EMAC_FTR_EMAC4;
ibm_newemac: PowerPC 440GX EMAC PHY clock workaround The PowerPC 440GX Taishan board fails to reset EMAC3 (reset timeout error) if there's no link. Because of that it fails to find PHY chip. The older ibm_emac driver had a workaround for that: the EMAC_CLK_INTERNAL/EMAC_CLK_EXTERNAL macros, which toggle the Ethernet Clock Select bit in the SDR0_MFR register. This patch does the same for "ibm,emac-440gx" compatible chips. The workaround forces clock on -all- EMACs, so we select clock under global emac_phy_map_lock. BenH: Made that #ifdef CONFIG_PPC_DCR_NATIVE for now as dcri_* stuff doesn't exist for MMIO type DCRs like Cell. Some future rework & improvements of the DCR infrastructure will make that cleaner but for now, this makes it work. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:46 +00:00
if (of_device_is_compatible(np, "ibm,emac-440gx"))
dev->features |= EMAC_FTR_440GX_PHY_CLK_FIX;
ibm_newemac: PowerPC 440EP/440GR EMAC PHY clock workaround This patch adds ibm_newemac PHY clock workaround for 440EP/440GR EMAC attached to a PHY which doesn't generate RX clock if there is no link. The code is based on the previous ibm_emac driver stuff. The 440EP/440GR allows controlling each EMAC clock separately as opposed to global clock selection for 440GX. BenH: Made that #ifdef CONFIG_PPC_DCR_NATIVE for now as dcri_* stuff doesn't exist for MMIO type DCRs like Cell. Some future rework & improvements of the DCR infrastructure will make that cleaner but for now, this makes it work. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:48 +00:00
} else {
if (of_device_is_compatible(np, "ibm,emac-440ep") ||
of_device_is_compatible(np, "ibm,emac-440gr"))
dev->features |= EMAC_FTR_440EP_PHY_CLK_FIX;
ibm_newemac: Fix new MAL feature handling Support for new features needed by the PPC 405EZ boards introduced some errors in the MAL and EMAC feature handling. This broke 'allmodconfig' builds as CONFIG_PPC_DCR_NATIVE is not set for those. This patch fixes these errors by wrapping the code in the appropriate #ifdefs. Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-10-16 04:38:40 +00:00
if (of_device_is_compatible(np, "ibm,emac-405ez")) {
ibm_newemac: Fix typo in flow control config option The recent build fix for ibm_newemac has a typo in the config option #ifdef used for disabling flow control. This corrects it to the proper Kconfig option name. Reported-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-10-24 13:53:14 +00:00
#ifdef CONFIG_IBM_NEW_EMAC_NO_FLOW_CTRL
ibm_newemac: Allow the "no flow control" EMAC feature to work Some PowerPC 40x chips have errata that force us not to use the integrated flow control. We have the feature defined, but it currently can't be used because it is never added to EMAC_FTRS_POSSIBLE. This adds a Kconfig option for affected platforms to select and puts the feature in the EMAC_FTRS_POSSIBLE list. This is set for PowerPC 405EZ platforms as well. Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-09-04 02:57:57 +00:00
dev->features |= EMAC_FTR_NO_FLOW_CONTROL_40x;
ibm_newemac: Fix new MAL feature handling Support for new features needed by the PPC 405EZ boards introduced some errors in the MAL and EMAC feature handling. This broke 'allmodconfig' builds as CONFIG_PPC_DCR_NATIVE is not set for those. This patch fixes these errors by wrapping the code in the appropriate #ifdefs. Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-10-16 04:38:40 +00:00
#else
printk(KERN_ERR "%s: Flow control not disabled!\n",
np->full_name);
return -ENXIO;
#endif
}
ibm_newemac: PowerPC 440GX EMAC PHY clock workaround The PowerPC 440GX Taishan board fails to reset EMAC3 (reset timeout error) if there's no link. Because of that it fails to find PHY chip. The older ibm_emac driver had a workaround for that: the EMAC_CLK_INTERNAL/EMAC_CLK_EXTERNAL macros, which toggle the Ethernet Clock Select bit in the SDR0_MFR register. This patch does the same for "ibm,emac-440gx" compatible chips. The workaround forces clock on -all- EMACs, so we select clock under global emac_phy_map_lock. BenH: Made that #ifdef CONFIG_PPC_DCR_NATIVE for now as dcri_* stuff doesn't exist for MMIO type DCRs like Cell. Some future rework & improvements of the DCR infrastructure will make that cleaner but for now, this makes it work. Signed-off-by: Valentine Barshak <vbarshak@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:46 +00:00
}
ibm_newemac: Cleanup/fix support for STACR register variants There are a few variants of the STACR register that affect more than just the "AXON" version of EMAC. Replace the current test of various chip models with tests for generic properties in the device-tree. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Stefan Roese <sr@denx.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-05 00:14:29 +00:00
/* Fixup some feature bits based on the device tree */
if (of_get_property(np, "has-inverted-stacr-oc", NULL))
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
dev->features |= EMAC_FTR_STACR_OC_INVERT;
ibm_newemac: Cleanup/fix support for STACR register variants There are a few variants of the STACR register that affect more than just the "AXON" version of EMAC. Replace the current test of various chip models with tests for generic properties in the device-tree. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Stefan Roese <sr@denx.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-05 00:14:29 +00:00
if (of_get_property(np, "has-new-stacr-staopc", NULL))
dev->features |= EMAC_FTR_HAS_NEW_STACR;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
ibm_newemac: Cleanup/fix support for STACR register variants There are a few variants of the STACR register that affect more than just the "AXON" version of EMAC. Replace the current test of various chip models with tests for generic properties in the device-tree. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Stefan Roese <sr@denx.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-05 00:14:29 +00:00
/* CAB lacks the appropriate properties */
if (of_device_is_compatible(np, "ibm,emac-axon"))
dev->features |= EMAC_FTR_HAS_NEW_STACR |
EMAC_FTR_STACR_OC_INVERT;
/* Enable TAH/ZMII/RGMII features as found */
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
if (dev->tah_ph != 0) {
#ifdef CONFIG_IBM_NEW_EMAC_TAH
dev->features |= EMAC_FTR_HAS_TAH;
#else
printk(KERN_ERR "%s: TAH support not enabled !\n",
np->full_name);
return -ENXIO;
#endif
}
if (dev->zmii_ph != 0) {
#ifdef CONFIG_IBM_NEW_EMAC_ZMII
dev->features |= EMAC_FTR_HAS_ZMII;
#else
printk(KERN_ERR "%s: ZMII support not enabled !\n",
np->full_name);
return -ENXIO;
#endif
}
if (dev->rgmii_ph != 0) {
#ifdef CONFIG_IBM_NEW_EMAC_RGMII
dev->features |= EMAC_FTR_HAS_RGMII;
#else
printk(KERN_ERR "%s: RGMII support not enabled !\n",
np->full_name);
return -ENXIO;
#endif
}
/* Read MAC-address */
p = of_get_property(np, "local-mac-address", NULL);
if (p == NULL) {
printk(KERN_ERR "%s: Can't find local-mac-address property\n",
np->full_name);
return -ENXIO;
}
memcpy(dev->ndev->dev_addr, p, 6);
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
/* IAHT and GAHT filter parameterization */
if (emac_has_feature(dev, EMAC_FTR_EMAC4SYNC)) {
dev->xaht_slots_shift = EMAC4SYNC_XAHT_SLOTS_SHIFT;
dev->xaht_width_shift = EMAC4SYNC_XAHT_WIDTH_SHIFT;
} else {
dev->xaht_slots_shift = EMAC4_XAHT_SLOTS_SHIFT;
dev->xaht_width_shift = EMAC4_XAHT_WIDTH_SHIFT;
}
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
DBG(dev, "features : 0x%08x / 0x%08x\n", dev->features, EMAC_FTRS_POSSIBLE);
DBG(dev, "tx_fifo_size : %d (%d gige)\n", dev->tx_fifo_size, dev->tx_fifo_size_gige);
DBG(dev, "rx_fifo_size : %d (%d gige)\n", dev->rx_fifo_size, dev->rx_fifo_size_gige);
DBG(dev, "max_mtu : %d\n", dev->max_mtu);
DBG(dev, "OPB freq : %d\n", dev->opb_bus_freq);
return 0;
}
ibm_newemac: convert to netdev_ops Reported-by: Subrata Modak <subrata@linux.vnet.ibm.com> Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-09 15:56:14 +00:00
static const struct net_device_ops emac_netdev_ops = {
.ndo_open = emac_open,
.ndo_stop = emac_close,
.ndo_get_stats = emac_stats,
.ndo_set_multicast_list = emac_set_multicast_list,
.ndo_do_ioctl = emac_ioctl,
.ndo_tx_timeout = emac_tx_timeout,
.ndo_validate_addr = eth_validate_addr,
.ndo_set_mac_address = eth_mac_addr,
.ndo_start_xmit = emac_start_xmit,
.ndo_change_mtu = eth_change_mtu,
};
static const struct net_device_ops emac_gige_netdev_ops = {
.ndo_open = emac_open,
.ndo_stop = emac_close,
.ndo_get_stats = emac_stats,
.ndo_set_multicast_list = emac_set_multicast_list,
.ndo_do_ioctl = emac_ioctl,
.ndo_tx_timeout = emac_tx_timeout,
.ndo_validate_addr = eth_validate_addr,
.ndo_set_mac_address = eth_mac_addr,
.ndo_start_xmit = emac_start_xmit_sg,
.ndo_change_mtu = emac_change_mtu,
};
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
static int __devinit emac_probe(struct of_device *ofdev,
const struct of_device_id *match)
{
struct net_device *ndev;
struct emac_instance *dev;
of: Always use 'struct device.of_node' to get device node pointer. The following structure elements duplicate the information in 'struct device.of_node' and so are being eliminated. This patch makes all readers of these elements use device.of_node instead. (struct of_device *)->node (struct dev_archdata *)->prom_node (sparc) (struct dev_archdata *)->of_node (powerpc & microblaze) Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2010-04-13 23:12:29 +00:00
struct device_node *np = ofdev->dev.of_node;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
struct device_node **blist = NULL;
int err, i;
ibm_newemac Use status property for unused/unwired EMACs Convert ibm_newemac to use the of_device_is_available function when checking for unused/unwired EMACs. We leave the current check for an "unused" property to maintain backwards compatibility for older device trees. Newer device trees should simply use the standard "status" property in the EMAC node. Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-22 00:46:44 +00:00
/* Skip unused/unwired EMACS. We leave the check for an unused
* property here for now, but new flat device trees should set a
* status property to "disabled" instead.
*/
if (of_get_property(np, "unused", NULL) || !of_device_is_available(np))
ibm_newemac: Skip EMACs that are marked unused by the firmware Depending on how the 44x processors are wired, some EMAC cells might not be useable (and not connected to a PHY). However, some device-trees may choose to still expose them (since their registers are present in the MMIO space) but with an "unused" property in them. Signed-off-by: Hugh Blemings <hugh@blemings.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-05 00:14:30 +00:00
return -ENODEV;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
/* Find ourselves in the bootlist if we are there */
for (i = 0; i < EMAC_BOOT_LIST_SIZE; i++)
if (emac_boot_list[i] == np)
blist = &emac_boot_list[i];
/* Allocate our net_device structure */
err = -ENOMEM;
ndev = alloc_etherdev(sizeof(struct emac_instance));
if (!ndev) {
printk(KERN_ERR "%s: could not allocate ethernet device!\n",
np->full_name);
goto err_gone;
}
dev = netdev_priv(ndev);
dev->ndev = ndev;
dev->ofdev = ofdev;
dev->blist = blist;
SET_NETDEV_DEV(ndev, &ofdev->dev);
/* Initialize some embedded data structures */
mutex_init(&dev->mdio_lock);
mutex_init(&dev->link_lock);
spin_lock_init(&dev->lock);
INIT_WORK(&dev->reset_work, emac_reset_work);
/* Init various config data based on device-tree */
err = emac_init_config(dev);
if (err != 0)
goto err_free;
/* Get interrupts. EMAC irq is mandatory, WOL irq is optional */
dev->emac_irq = irq_of_parse_and_map(np, 0);
dev->wol_irq = irq_of_parse_and_map(np, 1);
if (dev->emac_irq == NO_IRQ) {
printk(KERN_ERR "%s: Can't map main interrupt\n", np->full_name);
goto err_free;
}
ndev->irq = dev->emac_irq;
/* Map EMAC regs */
if (of_address_to_resource(np, 0, &dev->rsrc_regs)) {
printk(KERN_ERR "%s: Can't get registers address\n",
np->full_name);
goto err_irq_unmap;
}
// TODO : request_mem_region
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
dev->emacp = ioremap(dev->rsrc_regs.start,
dev->rsrc_regs.end - dev->rsrc_regs.start + 1);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
if (dev->emacp == NULL) {
printk(KERN_ERR "%s: Can't map device registers!\n",
np->full_name);
err = -ENOMEM;
goto err_irq_unmap;
}
/* Wait for dependent devices */
err = emac_wait_deps(dev);
if (err) {
printk(KERN_ERR
"%s: Timeout waiting for dependent devices\n",
np->full_name);
/* display more info about what's missing ? */
goto err_reg_unmap;
}
dev->mal = dev_get_drvdata(&dev->mal_dev->dev);
if (dev->mdio_dev != NULL)
dev->mdio_instance = dev_get_drvdata(&dev->mdio_dev->dev);
/* Register with MAL */
dev->commac.ops = &emac_commac_ops;
dev->commac.dev = dev;
dev->commac.tx_chan_mask = MAL_CHAN_MASK(dev->mal_tx_chan);
dev->commac.rx_chan_mask = MAL_CHAN_MASK(dev->mal_rx_chan);
err = mal_register_commac(dev->mal, &dev->commac);
if (err) {
printk(KERN_ERR "%s: failed to register with mal %s!\n",
of: Always use 'struct device.of_node' to get device node pointer. The following structure elements duplicate the information in 'struct device.of_node' and so are being eliminated. This patch makes all readers of these elements use device.of_node instead. (struct of_device *)->node (struct dev_archdata *)->prom_node (sparc) (struct dev_archdata *)->of_node (powerpc & microblaze) Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2010-04-13 23:12:29 +00:00
np->full_name, dev->mal_dev->dev.of_node->full_name);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
goto err_rel_deps;
}
dev->rx_skb_size = emac_rx_skb_size(ndev->mtu);
dev->rx_sync_size = emac_rx_sync_size(ndev->mtu);
/* Get pointers to BD rings */
dev->tx_desc =
dev->mal->bd_virt + mal_tx_bd_offset(dev->mal, dev->mal_tx_chan);
dev->rx_desc =
dev->mal->bd_virt + mal_rx_bd_offset(dev->mal, dev->mal_rx_chan);
DBG(dev, "tx_desc %p" NL, dev->tx_desc);
DBG(dev, "rx_desc %p" NL, dev->rx_desc);
/* Clean rings */
memset(dev->tx_desc, 0, NUM_TX_BUFF * sizeof(struct mal_descriptor));
memset(dev->rx_desc, 0, NUM_RX_BUFF * sizeof(struct mal_descriptor));
ibm_newemac: Fixes kernel crashes when speed of cable connected changes The descriptor pointers were not initialized to NIL values, so it was poiniting to some random addresses which was completely invalid. This fix takes care of initializing the descriptor to NIL values and clearing the valid descriptors on clean ring operation. Signed-off-by: Sathya Narayanan <sathyan@teamf1.com> Signed-off-by: Stefan Roese <sr@denx.de> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-07-01 08:58:05 +00:00
memset(dev->tx_skb, 0, NUM_TX_BUFF * sizeof(struct sk_buff *));
memset(dev->rx_skb, 0, NUM_RX_BUFF * sizeof(struct sk_buff *));
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
/* Attach to ZMII, if needed */
if (emac_has_feature(dev, EMAC_FTR_HAS_ZMII) &&
(err = zmii_attach(dev->zmii_dev, dev->zmii_port, &dev->phy_mode)) != 0)
goto err_unreg_commac;
/* Attach to RGMII, if needed */
if (emac_has_feature(dev, EMAC_FTR_HAS_RGMII) &&
(err = rgmii_attach(dev->rgmii_dev, dev->rgmii_port, dev->phy_mode)) != 0)
goto err_detach_zmii;
/* Attach to TAH, if needed */
if (emac_has_feature(dev, EMAC_FTR_HAS_TAH) &&
(err = tah_attach(dev->tah_dev, dev->tah_port)) != 0)
goto err_detach_rgmii;
/* Set some link defaults before we can find out real parameters */
dev->phy.speed = SPEED_100;
dev->phy.duplex = DUPLEX_FULL;
dev->phy.autoneg = AUTONEG_DISABLE;
dev->phy.pause = dev->phy.asym_pause = 0;
dev->stop_timeout = STOP_TIMEOUT_100;
INIT_DELAYED_WORK(&dev->link_work, emac_link_timer);
/* Find PHY if any */
err = emac_init_phy(dev);
if (err != 0)
goto err_detach_tah;
net: NEWEMAC: Fix problem with mtu > 4080 on non TAH equipped 4xx PPC's Currently, all non TAH equipped 4xx PPC's call emac_start_xmit() upon xmit. This routine doesn't check if the frame length exceeds the max. MAL buffer size. This patch now changes the driver to call emac_start_xmit_sg() on all GigE platforms and not only the TAH equipped ones (440GX). This enables an MTU of 9000 instead 4080. Signed-off-by: Stefan Roese <sr@denx.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-01-16 07:11:15 +00:00
if (dev->tah_dev)
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
ndev->features |= NETIF_F_IP_CSUM | NETIF_F_SG;
ndev->watchdog_timeo = 5 * HZ;
if (emac_phy_supports_gige(dev->phy_mode)) {
ibm_newemac: convert to netdev_ops Reported-by: Subrata Modak <subrata@linux.vnet.ibm.com> Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-09 15:56:14 +00:00
ndev->netdev_ops = &emac_gige_netdev_ops;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
dev->commac.ops = &emac_commac_sg_ops;
ibm_newemac: convert to netdev_ops Reported-by: Subrata Modak <subrata@linux.vnet.ibm.com> Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-09 15:56:14 +00:00
} else
ndev->netdev_ops = &emac_netdev_ops;
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
SET_ETHTOOL_OPS(ndev, &emac_ethtool_ops);
netif_carrier_off(ndev);
netif_stop_queue(ndev);
err = register_netdev(ndev);
if (err) {
printk(KERN_ERR "%s: failed to register net device (%d)!\n",
np->full_name, err);
goto err_detach_tah;
}
/* Set our drvdata last as we don't want them visible until we are
* fully initialized
*/
wmb();
dev_set_drvdata(&ofdev->dev, dev);
/* There's a new kid in town ! Let's tell everybody */
wake_up_all(&emac_probe_wait);
net: convert more to %pM A number of places still use %02x:...:%02x because it's in debug statements or for no real reason. Make a few of them use %pM. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-28 00:47:26 +00:00
printk(KERN_INFO "%s: EMAC-%d %s, MAC %pM\n",
ndev->name, dev->cell_index, np->full_name, ndev->dev_addr);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
ibm_newemac: Add support for GPCS, SGMII and M88E1112 PHY Add support for the phy types found on the Arches and other PowerPC 460 based boards. Signed-off-by: Victor Gallardo <vgallardo@amcc.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
2008-10-02 06:37:57 +00:00
if (dev->phy_mode == PHY_MODE_SGMII)
printk(KERN_NOTICE "%s: in SGMII mode\n", ndev->name);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
if (dev->phy.address >= 0)
printk("%s: found %s PHY (0x%02x)\n", ndev->name,
dev->phy.def->name, dev->phy.address);
emac_dbg_register(dev);
/* Life is good */
return 0;
/* I have a bad feeling about this ... */
err_detach_tah:
if (emac_has_feature(dev, EMAC_FTR_HAS_TAH))
tah_detach(dev->tah_dev, dev->tah_port);
err_detach_rgmii:
if (emac_has_feature(dev, EMAC_FTR_HAS_RGMII))
rgmii_detach(dev->rgmii_dev, dev->rgmii_port);
err_detach_zmii:
if (emac_has_feature(dev, EMAC_FTR_HAS_ZMII))
zmii_detach(dev->zmii_dev, dev->zmii_port);
err_unreg_commac:
mal_unregister_commac(dev->mal, &dev->commac);
err_rel_deps:
emac_put_deps(dev);
err_reg_unmap:
iounmap(dev->emacp);
err_irq_unmap:
if (dev->wol_irq != NO_IRQ)
irq_dispose_mapping(dev->wol_irq);
if (dev->emac_irq != NO_IRQ)
irq_dispose_mapping(dev->emac_irq);
err_free:
kfree(ndev);
err_gone:
/* if we were on the bootlist, remove us as we won't show up and
* wake up all waiters to notify them in case they were waiting
* on us
*/
if (blist) {
*blist = NULL;
wake_up_all(&emac_probe_wait);
}
return err;
}
static int __devexit emac_remove(struct of_device *ofdev)
{
struct emac_instance *dev = dev_get_drvdata(&ofdev->dev);
DBG(dev, "remove" NL);
dev_set_drvdata(&ofdev->dev, NULL);
unregister_netdev(dev->ndev);
ibm_newemac: Fix possible lockup on close It's a bad idea to call flush_scheduled_work from within a netdev->stop because the linkwatch will occasionally take the rtnl lock from a workqueue context, and thus that can deadlock. This reworks things a bit in that area to avoid the problem. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-20 03:50:46 +00:00
flush_scheduled_work();
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
if (emac_has_feature(dev, EMAC_FTR_HAS_TAH))
tah_detach(dev->tah_dev, dev->tah_port);
if (emac_has_feature(dev, EMAC_FTR_HAS_RGMII))
rgmii_detach(dev->rgmii_dev, dev->rgmii_port);
if (emac_has_feature(dev, EMAC_FTR_HAS_ZMII))
zmii_detach(dev->zmii_dev, dev->zmii_port);
mal_unregister_commac(dev->mal, &dev->commac);
emac_put_deps(dev);
emac_dbg_unregister(dev);
iounmap(dev->emacp);
if (dev->wol_irq != NO_IRQ)
irq_dispose_mapping(dev->wol_irq);
if (dev->emac_irq != NO_IRQ)
irq_dispose_mapping(dev->emac_irq);
kfree(dev->ndev);
return 0;
}
/* XXX Features in here should be replaced by properties... */
static struct of_device_id emac_match[] =
{
{
.type = "network",
.compatible = "ibm,emac",
},
{
.type = "network",
.compatible = "ibm,emac4",
},
ibm_newemac: Parameterize EMAC Multicast Match Handling Various instances of the EMAC core have varying: 1) number of address match slots, 2) width of the registers for handling address match slots, 3) number of registers for handling address match slots and 4) base offset for those registers. As the driver stands today, it assumes that all EMACs have 4 IAHT and GAHT 32-bit registers, starting at offset 0x30 from the register base, with only 16-bits of each used for a total of 64 match slots. The 405EX(r) and 460EX now use the EMAC4SYNC core rather than the EMAC4 core. This core has 8 IAHT and GAHT registers, starting at offset 0x80 from the register base, with ALL 32-bits of each used for a total of 256 match slots. This adds a new compatible device tree entry "emac4sync" and a new, related feature flag "EMAC_FTR_EMAC4SYNC" along with a series of macros and inlines which supply the appropriate parameterized value based on the presence or absence of the EMAC4SYNC feature. The code has further been reworked where appropriate to use those macros and inlines. In addition, the register size passed to ioremap is now taken from the device tree: c4 for EMAC4SYNC cores 74 for EMAC4 cores 70 for EMAC cores rather than sizeof (emac_regs). Finally, the device trees have been updated with the appropriate compatible entries and resource sizes. This has been tested on an AMCC Haleakala board such that: 1) inbound ICMP requests to 'haleakala.local' via MDNS from both Mac OS X 10.4.11 and Ubuntu 8.04 systems as well as 2) outbound ICMP requests from 'haleakala.local' to those same systems in the '.local' domain via MDNS now work. Signed-off-by: Grant Erickson <gerickson@nuovations.com> Acked-by: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2008-07-07 22:03:11 +00:00
{
.type = "network",
.compatible = "ibm,emac4sync",
},
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
{},
};
net: Fix OF platform drivers coldplug/hotplug when compiled as modules Some OF platform drivers are missing module device tables, so they won't load automatically on boot. This patch fixes the issue by adding proper MODULE_DEVICE_TABLE() macros to the drivers. Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-14 21:54:52 +00:00
MODULE_DEVICE_TABLE(of, emac_match);
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
static struct of_platform_driver emac_driver = {
of: Remove duplicate fields from of_platform_driver .name, .match_table and .owner are duplicated in both of_platform_driver and device_driver. This patch is a removes the extra copies from struct of_platform_driver and converts all users to the device_driver members. This patch is a pretty mechanical change. The usage model doesn't change and if any drivers have been missed, or if anything has been fixed up incorrectly, then it will fail with a compile time error, and the fixup will be trivial. This patch looks big and scary because it touches so many files, but it should be pretty safe. Signed-off-by: Grant Likely <grant.likely@secretlab.ca> Acked-by: Sean MacLennan <smaclennan@pikatech.com>
2010-04-13 23:13:02 +00:00
.driver = {
.name = "emac",
.owner = THIS_MODULE,
.of_match_table = emac_match,
},
Device tree aware EMAC driver Based on BenH's earlier work, this is a new version of the EMAC driver for the built-in ethernet found on PowerPC 4xx embedded CPUs. The same ASIC is also found in the Axon bridge chip. This new version is designed to work in the arch/powerpc tree, using the device tree to probe the device, rather than the old and ugly arch/ppc OCP layer. This driver is designed to sit alongside the old driver (that lies in drivers/net/ibm_emac and this one in drivers/net/ibm_newemac). The old driver is left in place to support arch/ppc until arch/ppc itself reaches its final demise (not too long now, with luck). This driver still has a number of things that could do with cleaning up, but I think they can be fixed up after merging. Specifically: - Should be adjusted to properly use the dma mapping API. Axon needs this. - Probe logic needs reworking, in conjuction with the general probing code for of_platform devices. The dependencies here between EMAC, MAL, ZMII etc. make this complicated. At present, it usually works, because we initialize and register the sub-drivers before the EMAC driver itself, and (being in driver code) runs after the devices themselves have been instantiated from the device tree. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-23 03:56:01 +00:00
.probe = emac_probe,
.remove = emac_remove,
};
static void __init emac_make_bootlist(void)
{
struct device_node *np = NULL;
int j, max, i = 0, k;
int cell_indices[EMAC_BOOT_LIST_SIZE];
/* Collect EMACs */
while((np = of_find_all_nodes(np)) != NULL) {
const u32 *idx;
if (of_match_node(emac_match, np) == NULL)
continue;
if (of_get_property(np, "unused", NULL))
continue;
idx = of_get_property(np, "cell-index", NULL);
if (idx == NULL)
continue;
cell_indices[i] = *idx;
emac_boot_list[i++] = of_node_get(np);
if (i >= EMAC_BOOT_LIST_SIZE) {
of_node_put(np);
break;
}
}
max = i;
/* Bubble sort them (doh, what a creative algorithm :-) */
for (i = 0; max > 1 && (i < (max - 1)); i++)
for (j = i; j < max; j++) {
if (cell_indices[i] > cell_indices[j]) {
np = emac_boot_list[i];
emac_boot_list[i] = emac_boot_list[j];
emac_boot_list[j] = np;
k = cell_indices[i];
cell_indices[i] = cell_indices[j];
cell_indices[j] = k;
}
}
}
static int __init emac_init(void)
{
int rc;
printk(KERN_INFO DRV_DESC ", version " DRV_VERSION "\n");
/* Init debug stuff */
emac_init_debug();
/* Build EMAC boot list */
emac_make_bootlist();
/* Init submodules */
rc = mal_init();
if (rc)
goto err;
rc = zmii_init();
if (rc)
goto err_mal;
rc = rgmii_init();
if (rc)
goto err_zmii;
rc = tah_init();
if (rc)
goto err_rgmii;
rc = of_register_platform_driver(&emac_driver);
if (rc)
goto err_tah;
return 0;
err_tah:
tah_exit();
err_rgmii:
rgmii_exit();
err_zmii:
zmii_exit();
err_mal:
mal_exit();
err:
return rc;
}
static void __exit emac_exit(void)
{
int i;
of_unregister_platform_driver(&emac_driver);
tah_exit();
rgmii_exit();
zmii_exit();
mal_exit();
emac_fini_debug();
/* Destroy EMAC boot list */
for (i = 0; i < EMAC_BOOT_LIST_SIZE; i++)
if (emac_boot_list[i])
of_node_put(emac_boot_list[i]);
}
module_init(emac_init);
module_exit(emac_exit);
Reference in New Issue Copy Permalink