Some firmware versions report a Local CA ACK Delay of 0. In that
case, return a more sensible default value of 12 (-> 16 msec) instead.
Signed-off-by: Joachim Fenkes <fenkes@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Certain firmware versions sometimes cause spurious PATH_MIG events to
occur during QP creation. Filter these events by making sure PATH_MIG
events are only handed down when they actually make sense (i.e. when
the QP has been armed at least once).
Signed-off-by: Joachim Fenkes <fenkes@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Enhance iser to act upon notification on network stack changes that
make its RDMA connection unaligned with the link used by the stack for
the <src,dst> IPs used to establish the connection.
When RDMA_CM_EVENT_ADDR_CHANGE arrives, just disconnect the
connection, assuming that the user space iscsid daemon will reconnect,
and the new connection will be aligned with the IP stack.
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Consumers that want to re-use their QPs in new connections need to
know when the QP has exited the timewait state. Report the timewait
event through the rdma_cm.
Signed-off-by: Amir Vadai <amirv@mellanox.co.il>
Acked-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Add an RDMA_CM_EVENT_ADDR_CHANGE event can be used by rdma-cm
consumers that wish to have their RDMA sessions always use the same
links (eg <hca/port>) as the IP stack does. In the current code, this
does not happen when bonding is used and fail-over happened but the IB
link used by an already existing session is operating fine.
Use the netevent notification for sensing that a change has happened
in the IP stack, then scan the rdma-cm ID list to see if there is an
ID that is "misaligned" with respect to the IP stack, and deliver
RDMA_CM_EVENT_ADDR_CHANGE for this ID. The consumer can act on the
event or just ignore it.
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This object really should be a struct device, or at least contain a
pointer to a struct device, as it is trying to create a separate device
tree outside of the main device tree. This patch fixes this problem.
It is needed for the class core rework that is being done in the driver
core.
Cc: Kay Sievers <kay.sievers@vrfy.org>
Cc: Roland Dreier <rolandd@cisco.com>
Cc: Sean Hefty <sean.hefty@intel.com>
Cc: Hal Rosenstock <hal.rosenstock@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This pointer really is a struct ib_device, not a struct device, so name
it properly to help prevent confusion.
This makes the followon patch in this series much smaller and easier to
understand as well.
Cc: Kay Sievers <kay.sievers@vrfy.org>
Cc: Roland Dreier <rolandd@cisco.com>
Cc: Hal Rosenstock <hal.rosenstock@gmail.com>
Acked-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
device_create() is race-prone, so use the race-free
device_create_drvdata() instead as device_create() is going away.
Cc: Roland Dreier <rolandd@cisco.com>
Cc: Sean Hefty <sean.hefty@intel.com>
Cc: Hal Rosenstock <hal.rosenstock@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
* git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-misc-2.6: (102 commits)
[SCSI] scsi_dh: fix kconfig related build errors
[SCSI] sym53c8xx: Fix bogus sym_que_entry re-implementation of container_of
[SCSI] scsi_cmnd.h: remove double inclusion of linux/blkdev.h
[SCSI] make struct scsi_{host,target}_type static
[SCSI] fix locking in host use of blk_plug_device()
[SCSI] zfcp: Cleanup external header file
[SCSI] zfcp: Cleanup code in zfcp_erp.c
[SCSI] zfcp: zfcp_fsf cleanup.
[SCSI] zfcp: consolidate sysfs things into one file.
[SCSI] zfcp: Cleanup of code in zfcp_aux.c
[SCSI] zfcp: Cleanup of code in zfcp_scsi.c
[SCSI] zfcp: Move status accessors from zfcp to SCSI include file.
[SCSI] zfcp: Small QDIO cleanups
[SCSI] zfcp: Adapter reopen for large number of unsolicited status
[SCSI] zfcp: Fix error checking for ELS ADISC requests
[SCSI] zfcp: wait until adapter is finished with ERP during auto-port
[SCSI] ibmvfc: IBM Power Virtual Fibre Channel Adapter Client Driver
[SCSI] sg: Add target reset support
[SCSI] lib: Add support for the T10 (SCSI) Data Integrity Field CRC
[SCSI] sd: Move scsi_disk() accessor function to sd.h
...
Now that we have a specific lock to protect the network
device unicast and multicast lists, remove extraneous
grabs of the TX lock in cases where the code only needs
address list protection.
Signed-off-by: David S. Miller <davem@davemloft.net>
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>
Current code uses kmalloc() and then just does a bitwise OR operation on
qp->flags in create_qp_common(), which means that qp->flags may
potentially have some unintended bits set. This patch uses kzalloc()
and avoids further explicit clearing of structure members, which also
shrinks the code:
add/remove: 0/0 grow/shrink: 0/1 up/down: 0/-65 (-65)
function old new delta
create_qp_common 2024 1959 -65
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The RDMA CM has some logic in place to make sure that callbacks on a
given CM ID are delivered to the consumer in a serialized manner.
Specifically it has code to protect against a device removal racing
with a running callback function.
This patch simplifies this logic by using a mutex per ID instead of a
wait queue and atomic variable. This means that cma_disable_remove()
now is more properly named to cma_disable_callback(), and
cma_enable_remove() can now be removed because it just would become a
trivial wrapper around mutex_unlock().
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Keep a pointer to the local (src) netdevice in struct rdma_dev_addr,
and copy it in as part of rdma_copy_addr(). Use rdma_translate_ip()
in cma_new_conn_id() to reduce some code duplication and also make
sure the src_dev member gets set.
In a high-availability configuration the netdevice pointer can be used
by the RDMA CM to align RDMA sessions to use the same links as the IP
stack does under fail-over and route change cases.
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Handling the zero STag in receive work request requires some extra
logic in the driver:
- Only set the QP_PRIV bit for kernel mode QPs.
- Add a zero STag build function for recv wrs. The uP needs a PBL
allocated and passed down in the recv WR so it can construct a HW
PBL for the zero STag S/G entries. Note: we need to place a few
restrictions on zero STag usage because of this:
1) all SGEs in a recv WR must either be zero STag or not. No mixing.
2) an individual SGE length cannot exceed 128MB for a zero-stag SGE.
This should be OK since it's not really practical to allocate
such a large chunk of pinned contiguous DMA mapped memory.
- Add an optimized non-zero-STag recv wr format for kernel users.
This is needed to optimize both zero and non-zero STag cracking in
the recv path for kernel users.
- Remove the iwch_ prefix from the static build functions.
- Bump required FW version.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
- Change the IB_DEVICE_ZERO_STAG flag to the transport-neutral name
IB_DEVICE_LOCAL_DMA_LKEY, which is used by iWARP RNICs to indicate 0
STag support and IB HCAs to indicate reserved L_Key support.
- Add a u32 local_dma_lkey member to struct ib_device. Drivers fill
this in with the appropriate local DMA L_Key (if they support it).
- Fix up the drivers using this flag.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The MLX transport requires two extra gather entries for sends (one for
the header and one for the checksum at the end, as the comment says).
However the code checked that max_recv_sge was not too big, instead of
checking max_send_sge as it should have. Fix the code to check the
correct condition.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Exactly when the catastrophic error polling timer function runs is not
important, so use round_jiffies() to save unnecessary wakeups.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Since we use del_timer_sync() anyway, there's no need for an
additional flag to tell the timer not to rearm.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Increase IPoIB ring sizes to twice their original sizes (RX: 128->256,
TX: 64->128) to act as a shock absorber for high traffic peaks. With
the current settings, we have seen cases that there are many calls to
netif_stop_queue(), which causes degradation in throughput. Also,
larger receive buffer sizes help IPoIB in CM mode to avoid experiencing
RNR NAK conditions due to insufficient receive buffers at the SRQ.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Since IPoIB connected mode does not NETIF_F_SG, we only have one DMA
mapping per send, so we don't need a mapping[] array. Define a new
struct with a single u64 mapping member and use it for the CM tx_ring.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The IB spe. for SubnGet(NodeInfo) and query HCA says that the vendor
ID field should be the IEEE OUI assigned to the vendor. The ipath
driver was returning the PCI vendor ID instead. This will affect
applications which call ibv_query_device(). The old value was
0x001fc1 or 0x001077, the new value is 0x001175.
The vendor ID doesn't appear to be exported via /sys so that should
reduce possible compatibility issues. I'm only aware of Open MPI as a
major application which depends on this change, and they have made
necessary adjustments.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When the driver sets the MTU of the net device outside of its
change_mtu method, it should make use of dev_set_mtu() instead of
directly setting the mtu field of struct netdevice. Otherwise
functions registered to be called upon MTU change will not get called
(this is done through call_netdevice_notifiers() in dev_set_mtu()).
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Use of this lock is required to synchronize changes to the netdvice's
data structs. Also move the call to ipoib_flush_paths() after the
modification of the netdevice flags in set_mode().
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
ipoib_mcast_detach() does nothing except call ib_detach_mcast(), so just
use the core API in the one place that does a multicast group detach.
add/remove: 0/1 grow/shrink: 0/1 up/down: 0/-105 (-105)
function old new delta
ipoib_mcast_leave 357 319 -38
ipoib_mcast_detach 67 - -67
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The current code will set the Q_Key for any join of a non-sendonly
multicast group. The operation involves a modify QP operation, which
is fairly heavyweight, and is only really required after the join of
the broadcast group. Fix this by adding a parameter to ipoib_mcast_attach()
to control when the Q_Key is set.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
No need for a mutex around calls to ib_attach_mcast/ib_detach_mcast
since these operations are synchronized at the HW driver layer.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The IPOIB_MCAST_STARTED flag is not used at all since commit b3e2749b
("IPoIB: Don't drop multicast sends when they can be queued"), so
remove it.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Every caller of nes_post_cqp_request() passed it NES_CQP_REQUEST_RING_DOORBELL,
so just remove that parameter and always ring the doorbell.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Acked-by: Faisal Latif <flatif@neteffect.com>
cxgb3 does not currently report the page size capabilities, and
incorrectly reports them internally.
This version changes the bit-shifting to a static value (per Steve's
request).
Signed-off-by: Jon Mason <jon@opengridcomputing.com>
Acked-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The patch tries to solve the problem of device going down and paths being
flushed on an SM change event. The method is to mark the paths as candidates for
refresh (by setting the new valid flag to 0), and wait for an ARP
probe a new path record query.
The solution requires a different and less intrusive handling of SM
change event. For that, the second argument of the flush function
changes its meaning from a boolean flag to a level. In most cases, SM
failover doesn't cause LID change so traffic won't stop. In the rare
cases of LID change, the remote host (the one that hadn't changed its
LID) will lose connectivity until paths are refreshed. This is no
worse than the current state. In fact, preventing the device from
going down saves packets that otherwise would be lost.
Signed-off-by: Moni Levy <monil@voltaire.com>
Signed-off-by: Moni Shoua <monis@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This gives ehca an autogenerated modalias and therefore enables automatic loading.
Signed-off-by: Joachim Fenkes <fenkes@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Add "ipoib_use_lro" module parameter to enable LRO and an
"ipoib_lro_max_aggr" module parameter to set the max number of packets
to be aggregated. Make LRO controllable and LRO statistics accessible
through ethtool.
Signed-off-by: Vladimir Sokolovsky <vlad@mellanox.co.il>
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Set IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK for IPoIB's UD QPs if
supported by the underlying device. This creates an improvement of up
to 39% in bandwidth when sending multicast packets with IPoIB, and an
improvment of 12% in cpu usage.
Signed-off-by: Ron Livne <ronli@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Add support for handling the IB_QP_CREATE_MULTICAST_BLOCK_LOOPBACK
flag by using the per-multicast group loopback blocking feature of
mlx4 hardware.
Signed-off-by: Ron Livne <ronli@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
- Add a new rdma ctl command called RDMA_GET_MIB to the cxgb3 low
level driver to obtain the protocol mib from the rnic hardware.
- Add new iw_cxgb3 provider method to get the MIB from the low level
driver.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This patch adds a sysfs attribute group called "proto_stats" under
/sys/class/infiniband/$device/ and populates this group with protocol
statistics if they exist for a given device. Currently, only iWARP
stats are defined, but the code is designed to allow InfiniBand
protocol stats if they become available. These stats are per-device
and more importantly -not- per port.
Details:
- Add union rdma_protocol_stats in ib_verbs.h. This union allows
defining transport-specific stats. Currently only iwarp stats are
defined.
- Add struct iw_protocol_stats to define the current set of iwarp
protocol stats.
- Add new ib_device method called get_proto_stats() to return protocol
statistics.
- Add logic in core/sysfs.c to create iwarp protocol stats attributes
if the device is an RNIC and has a get_proto_stats() method.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
For devices that don't support SRQs, ipoib_cm_post_receive_nonsrq() is
called from both ipoib_cm_handle_rx_wc() and ipoib_cm_nonsrq_init_rx(),
and these two callers are not synchronized against each other.
However, ipoib_cm_post_receive_nonsrq() always reuses the same receive
work request and scatter list structures, so multiple callers can end
up stepping on each other, which leads to posting garbled work
requests.
Fix this by having the caller pass in the ib_recv_wr and ib_sge
structures to use, and allocating new local structures in
ipoib_cm_nonsrq_init_rx().
Based on a patch by Pradeep Satyanarayana <pradeep@us.ibm.com> and
David Wilder <dwilder@us.ibm.com>, with debugging help from Hoang-Nam
Nguyen <hnguyen@de.ibm.com>.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The members struct iwch_rnic_attributes.vendor_id and .vendor_part_id
are write-only, so we might as well get rid of them.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Acked-by: Steve Wise <swise@opengridcomputing.com>
- set fw_ver
- set hw_ver
- set max_qp_wr to something reasonable
- set max_cqe to something reasonable
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
During corner case testing, we noticed that some versions of ehca do
not properly transition to interrupt done in special load situations.
This can be resolved by periodically triggering EOI through H_EOI, if
EQEs are pending.
Signed-off-by: Stefan Roscher <stefan.roscher@de.ibm.com>
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Commit 65adfa91 ("IB/mlx4: Fix RESET to RESET and RESET to ERROR
transitions") added some extra code to handle a QP state transition
from RESET to ERROR. However, the latest 1.2.1 version of the IB spec
has clarified that this transition is actually not allowed, so we can
remove this extra code again.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Commit b18aad71 ("IB/mthca: Fix RESET to ERROR transition") added some
extra code to handle a QP state transition from RESET to ERROR.
However, the latest 1.2.1 version of the IB spec has clarified that
this transition is actually not allowed, so we can remove this extra
code again.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
I was reviewing the QP state transition diagram in the IB 1.2.1 spec
and the code for qp_state_table[], and noticed that the code allows a
QP to be modified from IB_QPS_RESET to IB_QPS_ERR whereas the notes
for figure 124 (pg 457) specifically says that this transition isn't
allowed. This is a clarification from earlier versions of the IB
spec, which were ambiguous in this area and suggested that the RESET
to ERR transition was allowed.
Fix up the qp_state_table[] to make RESET->ERR not allowed.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
ConnectX HCAs support the IB_MGMT_CLASS_CONG_MGMT management class, so
process MADs of this class through the MAD_IFC firmware command.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
ConnectX returns the max message size it supports through the
QUERY_DEV_CAP firmware command. When modifying a QP to RTR, the max
message size for the QP must be specified. This value must not exceed
the value declared through QUERY_DEV_CAP. The current code ignores
the max allowed size and unconditionally sets the value to 2^31. This
patch sets all QPs to the max value allowed as returned from firmware.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
- set IB_DEVICE_MEM_MGT_EXTENSIONS capability bit if fw supports it.
- set max_fast_reg_page_list_len device attribute.
- add iwch_alloc_fast_reg_mr function.
- add iwch_alloc_fastreg_pbl
- add iwch_free_fastreg_pbl
- adjust the WQ depth for kernel mode work queues to account for
fastreg possibly taking 2 WR slots.
- add fastreg_mr work request support.
- add local_inv work request support.
- add send_with_inv and send_with_se_inv work request support.
- removed useless duplicate enums/defines for TPT/MW/MR stuff.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This patch adds support for the IB "base memory management extension"
(BMME) and the equivalent iWARP operations (which the iWARP verbs
mandates all devices must implement). The new operations are:
- Allocate an ib_mr for use in fast register work requests.
- Allocate/free a physical buffer lists for use in fast register work
requests. This allows device drivers to allocate this memory as
needed for use in posting send requests (eg via dma_alloc_coherent).
- New send queue work requests:
* send with remote invalidate
* fast register memory region
* local invalidate memory region
* RDMA read with invalidate local memory region (iWARP only)
Consumer interface details:
- A new device capability flag IB_DEVICE_MEM_MGT_EXTENSIONS is added
to indicate device support for these features.
- New send work request opcodes IB_WR_FAST_REG_MR, IB_WR_LOCAL_INV,
IB_WR_RDMA_READ_WITH_INV are added.
- A new consumer API function, ib_alloc_mr() is added to allocate
fast register memory regions.
- New consumer API functions, ib_alloc_fast_reg_page_list() and
ib_free_fast_reg_page_list() are added to allocate and free
device-specific memory for fast registration page lists.
- A new consumer API function, ib_update_fast_reg_key(), is added to
allow the key portion of the R_Key and L_Key of a fast registration
MR to be updated. Consumers call this if desired before posting
a IB_WR_FAST_REG_MR work request.
Consumers can use this as follows:
- MR is allocated with ib_alloc_mr().
- Page list memory is allocated with ib_alloc_fast_reg_page_list().
- MR R_Key/L_Key "key" field is updated with ib_update_fast_reg_key().
- MR made VALID and bound to a specific page list via
ib_post_send(IB_WR_FAST_REG_MR)
- MR made INVALID via ib_post_send(IB_WR_LOCAL_INV),
ib_post_send(IB_WR_RDMA_READ_WITH_INV) or an incoming send with
invalidate operation.
- MR is deallocated with ib_dereg_mr()
- page lists dealloced via ib_free_fast_reg_page_list().
Applications can allocate a fast register MR once, and then can
repeatedly bind the MR to different physical block lists (PBLs) via
posting work requests to a send queue (SQ). For each outstanding
MR-to-PBL binding in the SQ pipe, a fast_reg_page_list needs to be
allocated (the fast_reg_page_list is owned by the low-level driver
from the consumer posting a work request until the request completes).
Thus pipelining can be achieved while still allowing device-specific
page_list processing.
The 32-bit fast register memory key/STag is composed of a 24-bit index
and an 8-bit key. The application can change the key each time it
fast registers thus allowing more control over the peer's use of the
key/STag (ie it can effectively be changed each time the rkey is
rebound to a page list).
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The connected mode implementation in the IPoIB driver has a large
overhead in the way SKBs are handled in the receive flow. It usually
allocates an SKB with as big as was used in the currently received SKB
and moves unused fragments from the old SKB to the new one. This
involves a loop on all the remaining fragments and incurs overhead on
the CPU. This patch, for small SKBs, allocates an SKB just large
enough to contain the received data and copies to it the data from the
received SKB. The newly allocated SKB is passed to the stack and the
old SKB is reposted.
When running netperf, UDP small messages, without this pach I get:
UDP UNIDIRECTIONAL SEND TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to
14.4.3.178 (14.4.3.178) port 0 AF_INET
Socket Message Elapsed Messages
Size Size Time Okay Errors Throughput
bytes bytes secs # # 10^6bits/sec
114688 128 10.00 5142034 0 526.31
114688 10.00 1130489 115.71
With this patch I get both send and receive at ~315 mbps.
The reason that send performance actually slows down is as follows:
When using this patch, the overhead of the CPU for handling RX packets
is dramatically reduced. As a result, we do not experience RNR NAK
messages from the receiver which cause the connection to be closed and
reopened again; when the patch is not used, the receiver cannot handle
the packets fast enough so there is less time to post new buffers and
hence the mentioned RNR NACKs. So what happens is that the
application *thinks* it posted a certain number of packets for
transmission but these packets are flushed and do not really get
transmitted. Since the connection gets opened and closed many times,
each time netperf gets the CPU time that otherwise would have been
given to IPoIB to actually transmit the packets. This can be verified
when looking at the port counters -- the output of ifconfig and the
oputput of netperf (this is for the case without the patch):
tx packets
==========
port counter: 1,543,996
ifconfig: 1,581,426
netperf: 5,142,034
rx packets
==========
netperf 1,1304,089
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
The idea is that for QPs with fixed size work requests (eg selective
signaling QPs), before stamping the WQE, we read the value of the DS
field, which gives the effective size of the descriptor as used in the
previous post. Then we stamp only that area, since the rest of the
descriptor is already stamped.
When initializing the send queue buffer, make sure the DS field is
initialized to the max descriptor size so that the subsequent stamping
will be done on the entire descriptor area.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This patch solves a race that occurs after an event occurs that causes
the SA query module to flush its SM address handle (AH). When SM AH
becomes invalid and needs an update it is handled by the global
workqueue. On the other hand this event is also handled in the IPoIB
driver by queuing work in the ipoib_workqueue that does multicast
joins. Although queuing is in the right order, it is done to 2
different workqueues and so there is no guarantee that the first to be
queued is the first to be executed.
This causes a problem because IPoIB may end up sending an request to
the old SM, which will take a long time to time out (since the old SM
is gone); this leads to a much longer than necessary interruption in
multicast traffer.
The patch sets the SA query module's SM AH to NULL when the event
occurs, and until update_sm_ah() is done, any request that needs sm_ah
fails with -EAGAIN return status.
For consumers, the patch doesn't make things worse. Before the patch,
MADs are sent to the wrong SM so the request gets lost. Consumers can
be improved if they examine the return code and respond to EAGAIN
properly but even without an improvement the situation is not getting
worse.
Signed-off-by: Moni Levy <monil@voltaire.com>
Signed-off-by: Moni Shoua <monis@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The license text for several files references a third software license
that was inadvertently copied in. Update the license to what was
intended. This update was based on a request from HP.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Remove an explicit memset(..., 0, ...) of a 'listener' structure
allocated with kzalloc().
Signed-off-by: Christophe Jaillet <christophe.jaillet@wanadoo.fr>
Acked-by: Faisal Latif <faisal@neteffect.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The SRP initiator is currently using ib_find_cached_pkey() and
ib_get_cached_gid() in situations where the uncached ib_find_pkey()
and ib_query_gid() functions serve just as well: sleeping is allowed
and performance is not an issue. Since we want to eliminate the
cached operations in the long term, convert SRP to use the uncached
variants.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This patch fixes two bugs that are related.
1. Old tools did not set can_queue/cmds_max. This patch modifies
libiscsi so that when we add the host we catch this and set it
to the default.
2. iscsi_tcp thought that the scsi command that was passed to
the eh functions needed a iscsi_cmd_task allocated for it. It
only needed a mgmt task, and now it does not matter since it
all comes from the same pool and libiscsi handles this for the
drivers. ib_iser had copied iscsi_tcp's code and set can_queue
to its max - 1 to handle this. So this patch removes the max -1,
and just sets it to the max.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
The recv lock was defined so the iscsi layer could block
the recv path from processing IO during recovery. It
turns out iser just set a lock to that pointer which was pointless.
We now disconnect the transport connection before doing recovery
so we do not need the recv lock. For iscsi_tcp we still stop
the recv path incase older tools are being used.
This patch also has iscsi_itt_to_ctask user grab the session lock
and has the caller access the task with the lock or get a ref
to it in case the target is broken and sends a tmf success response
then sends data or a response for the command that was supposed to
be affected bty the tmf.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
This adds two new attrs used for creating initiator ports and
binding sessions to hardware.
The session level initiatorname:
Since bnx2i does a scsi_host per host device, we need to add the
iface initiator port settings on the session, so we can create
multiple initiator ports (each with different inames) per device/scsi_host.
The current iname reflects that qla4xxx can have one iname per hba, and we are
allocating a host per session for software. The iname on the host will
remain so we can export and set the hba level qla4xxx setting.
The ifacename attr:
To bind a session to a some peice of hardware in userspace we maintain
some mappings, but during boot or iscsid restart (iscsid contains the user
space part of the driver) we need to be able to figure out which of those
host mappings abstractions maps to certain sessions. This patch adds
a ifacename attr, which userspace can set to id the host side of the
endpoint across pivot_roots and iscsid restarts.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
This hooks iser into the iscsi endpoint code. Previously it handled the
lookup and allocation. This has been made generic so bnx2i and iser can
share it. It also allows us to pass iser the leading conn's ep, so we
know the ib_deivce being used and can set it as the scsi_host's parent.
And that allows scsi-ml to set the dma_mask based on those values.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Currently we duplicate the list of sessions, because we were using the
test for if a session was on the host list to indicate if the session
was bound or unbound. We can instead use the target_id and fix up
the class so that drivers like bnx2i do not have to manage the target id
space.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
This handles the iscsi_cmd_task rename and renames
the iser cmd task to iser task.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Convert ib_iser to support merged tasks.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Currently to get a ctask from the session cmd array, you have to
know to use the itt modifier. To make this easier on LLDs and
so in the future we can easilly kill the session array and use
the host shared map instead, this patch adds a nice wrapper
to strip the itt into a session->cmds index and return a ctask.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
After the stop_conn callback has returned the LLD should not
touch the scsi cmds. iscsi_tcp and libiscsi use the
conn->recv_lock and suspend_rx field to halt recv path
processing, but iser does not have any protection.
This patch modifies iser so that userspace can just
call the ep_disconnect callback, which will halt
all recv IO, before calling the stop_conn callback so
we do not have to worry about the conn->recv_lock and
suspend rx field. iser just needs to stop the send side
from accessing the ib conn.
Fixup to handle when the ep poll fails and ep disconnect
is called from Erez.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
This removes the session and conn data_size fields from the iscsi_transport.
Just pass in the value like with host allocation. This patch also makes
it so the LLD iscsi_conn data is allocated with the iscsi_cls_conn.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
This finishes the host/session unbinding, by adding some helpers
to add and remove hosts and the session they manage.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
bnx2i allocates a host per netdevice but will use libiscsi,
so this unbinds the session from the host in that code.
This will also be useful for the iser parent device dma settings
fixes.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
max_cmd_len and max_conn are not really used. max_cmd_len is
always 16 and can be set by the LLD. max_conn is always one
since we do not support MCS.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
iscsi offload (bnx2i and qla4xx) allocate a scsi host per hba,
so the session creation path needs a shost/host_no argument.
Software iscsi/iser will follow the same behabior as before
where it allcoates a host per session, but in the future iser
will probably look more like bnx2i where the host's parent is
the hardware (rnic for iser and for bnx2i it is the nic), because
it does not use a socket layer like how iscsi_tcp does.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Remove explicit lock_kernel() calls and document why the code is safe.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
The change to iwch_provider.c in commit f4e91eb4 ("IB: convert struct
class_device to struct device") undid the fix done in commit 7f049f2f
("RDMA/cxgb3: Hold rtnl_lock() around ethtool get_drvinfo call"). It
removed the calls to rtnl_lock() that serialized the iw_cxgb3 ethtool
ops calls into the cxgb3 driver. This locking is needed to avoid
messing up the internal state of the cxgb3 driver.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Remove explicit lock_kernel() calls and document why the code is safe.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Current memfree FW has a bug which in some cases, assumes that ICM
pages passed to it are cleared. This patch uses __GFP_ZERO to
allocate all ICM pages passed to the FW. Once firmware with a fix is
released, we can make the workaround conditional on firmware version.
This fixes the bug reported by Arthur Kepner <akepner@sgi.com> here:
http://lists.openfabrics.org/pipermail/general/2008-May/050026.html
Cc: <stable@kernel.org>
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
[ Rewritten to be a one-liner using __GFP_ZERO instead of vmap()ing
ICM memory and memset()ing it to 0. - Roland ]
Signed-off-by: Roland Dreier <rolandd@cisco.com>
All of the open() functions which don't need the BKL on their face may
still depend on its acquisition to serialize opens against driver
initialization. So make those functions acquire then release the BKL to be
on the safe side.
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
This documents the fact that somebody looked at the relevant open()
functions and concluded that, due to their trivial nature, no locking was
needed.
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Commit 1ae5c187 ("IB/uverbs: Don't store struct file * for event
files") changed the way that closed files are handled in the uverbs
code. However, after the conversion, is_closed flag is checked
incorrectly in ib_uverbs_async_handler(). As a result, no async
events are ever passed to applications.
Found by: Ronni Zimmerman <ronniz@mellanox.co.il>
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
nes_reg_user_mr() should fail if page_count becomes >= 1024 * 512
rather than just testing for strict >, because page_count is
essentially used as an index into an array with 1024 * 512 entries, so
allowing the loop to continue with page_count == 1024 * 512 means that
memory after the end of the array is corrupted. This leads to a crash
triggerable by a userspace application that requests registration of a
too-big region.
Also get rid of the call to pci_free_consistent() here to avoid
corrupting state with a double free, since the same memory will be
freed in the code jumped to at reg_user_mr_err.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
In 2.6.26, we added some support for send with invalidate work
requests, including a device capability flag to indicate whether a
device supports such requests. However, the support was incomplete:
the completion structure was not extended with a field for the key
contained in incoming send with invalidate requests.
Full support for memory management extensions (send with invalidate,
local invalidate, fast register through a send queue, etc) is planned
for 2.6.27. Since send with invalidate is not very useful by itself,
just remove the IB_DEVICE_SEND_W_INV bit before the 2.6.26 final
release; we will add an IB_DEVICE_MEM_MGT_EXTENSIONS bit in 2.6.27,
which makes things simpler for applications, since they will not have
quite as confusing an array of fine-grained bits to check.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
On a 64-bit architecture, if ib_umem_get() is called with a size value
that is so big that npages is negative when cast to int, then the
length of the page list passed to get_user_pages(), namely
min_t(int, npages, PAGE_SIZE / sizeof (struct page *))
will be negative, and get_user_pages() will immediately return 0 (at
least since 900cf086, "Be more robust about bad arguments in
get_user_pages()"). This leads to an infinite loop in ib_umem_get(),
since the code boils down to:
while (npages) {
ret = get_user_pages(...);
npages -= ret;
}
Fix this by taking the minimum as unsigned longs, so that the value of
npages is never truncated.
The impact of this bug isn't too severe, since the value of npages is
checked against RLIMIT_MEMLOCK, so a process would need to have an
astronomical limit or have CAP_IPC_LOCK to be able to trigger this,
and such a process could already cause lots of mischief. But it does
let buggy userspace code cause a kernel lock-up; for example I hit
this with code that passes a negative value into a memory registartion
function where it is promoted to a huge u64 value.
Cc: <stable@kernel.org>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
SM/SMA traps received by the ipath driver should be forwarded to the
SM if it is running on the host. The ib_ipath driver was incorrectly
replying with "bad method."
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The driver supports a few features (RNR NAK, port active event, SRQ
resize) that were not reported in the device capability flags. This
patch fixes that.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Gabriel C <nix.or.die@googlemail.com> pointed out that when the x86
bitops are updated to operate on unsigned long, the code in
sdma_abort_task() will produce warnings:
drivers/infiniband/hw/ipath/ipath_sdma.c: In function 'sdma_abort_task':
drivers/infiniband/hw/ipath/ipath_sdma.c:267: warning: passing argument 2 of 'constant_test_bit' from incompatible pointer type
and so on, because it uses test_bit() to operation on a u64 value
(returned by ipath_read_kref64() for a hardware register).
Fix up these warnings by converting the test_bit() operations to &ing
with appropriate symbolic defines of the bits within the hardware
register. This has the benign side-effect of making the code more
self-documenting as well.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
IB/mad: Fix kernel crash when .process_mad() returns SUCCESS|CONSUMED
IPoIB: Test for NULL broadcast object in ipiob_mcast_join_finish()
MAINTAINERS: Add cxgb3 and iw_cxgb3 NIC and iWARP driver entries
IB/mlx4: Fix creation of kernel QP with max number of send s/g entries
IB/mthca: Fix max_sge value returned by query_device
RDMA/cxgb3: Fix uninitialized variable warning in iwch_post_send()
IB/mlx4: Fix uninitialized-var warning in mlx4_ib_post_send()
IB/ipath: Fix UC receive completion opcode for RDMA WRITE with immediate
IB/ipath: Fix printk format for ipath_sdma_status
If a low-level driver returns IB_MAD_RESULT_SUCCESS | IB_MAD_RESULT_CONSUMED,
handle_outgoing_dr_smp() doesn't clean up properly. The fix is to
kfree the local data and break, rather than falling through. This was
observed with the ipath driver, but could happen with any driver.
This fixes <https://bugs.openfabrics.org/show_bug.cgi?id=1027>.
Signed-off-by: Dave Olson <dave.olson@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Change references from for_each_cpu_mask to for_each_cpu_mask_nr
where appropriate
Reviewed-by: Paul Jackson <pj@sgi.com>
Reviewed-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
We saw a kernel oops in our regression testing when a multicast "join
finish" occurred just after the interface was -- this is
<https://bugs.openfabrics.org/show_bug.cgi?id=1040>. The test
randomly causes the HCA physical port to go down then up.
The cause of this is that ipoib_mcast_join_finish() processing happen
just after ipoib_mcast_dev_flush() was invoked (in which case the
broadcast pointer is NULL). This patch tests for and handles the case
where priv->broadcast is NULL.
Cc: <stable@kernel.org>
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When creating a kernel QP where the consumer asked for a send queue
with lots of scatter/gater entries, set_kernel_sq_size() incorrectly
returned an error if the send queue stride is larger than the
hardware's maximum send work request descriptor size. This is not a
problem; the only issue is to make sure that the actual descriptors
used do not overflow the maximum descriptor size, so check this instead.
Clamp the returned max_send_sge value to be no bigger than what
query_device returns for the max_sge to avoid confusing hapless users,
even if the hardware is capable of handling a few more s/g entries.
This bug caused NFS/RDMA mounts to fail when the server adapter used
the mlx4 driver.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
There is a race from when a device is created with device_create() and
then the drvdata is set with a call to dev_set_drvdata() in which a
sysfs file could be open, yet the drvdata will be NULL, causing all
sorts of bad things to happen.
This patch fixes the problem by using the new function,
device_create_drvdata().
Cc: Kay Sievers <kay.sievers@vrfy.org>
Reviewed-by: Roland Dreier <rolandd@cisco.com>
Cc: Sean Hefty <sean.hefty@intel.com>
Cc: Hal Rosenstock <hal.rosenstock@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Move rcu-protected lists from list.h into a new header file rculist.h.
This is done because list are a very used primitive structure all over the
kernel and it's currently impossible to include other header files in this
list.h without creating some circular dependencies.
For example, list.h implements rcu-protected list and uses rcu_dereference()
without including rcupdate.h. It actually compiles because users of
rcu_dereference() are macros. Others RCU functions could be used too but
aren't probably because of this.
Therefore this patch creates rculist.h which includes rcupdates without to
many changes/troubles.
Signed-off-by: Franck Bui-Huu <fbuihuu@gmail.com>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Josh Triplett <josh@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The mthca driver returns the maximum number of scatter/gather entries
returned by the firmware as the max_sge value when device properties
are queried. However, the firmware also reports a limit on the
maximum descriptor size allowed, and because mthca takes into account
the worst case send request overhead when checking whether to allow a
QP to be created, the largest number of scatter/gather entries that
can be used with mthca may be limited by the maximum descriptor size
rather than just by the actual s/g entry limit.
This means that applications cannot actually create QPs with
max_send_sge equal to the limit returned by ib_query_device(). Fix
this by checking if the maximum descriptor size imposes a lower limit
and if so returning that lower limit.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
drivers/infiniband/hw/cxgb3/iwch_qp.c: In function 'iwch_post_send':
drivers/infiniband/hw/cxgb3/iwch_qp.c:232: warning: 't3_wr_flit_cnt' may be used uninitialized in this function
This is what akpm describes as "the dopey
gcc-doesn't-know-that-foo(&var)-writes-to-var problem."
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Acked-by: Steve Wise <swise@opengridcomputing.com>
drivers/infiniband/hw/mlx4/qp.c: In function 'mlx4_ib_post_send':
drivers/infiniband/hw/mlx4/qp.c:1460: warning: 'seglen' may be used uninitialized in this function
This is the dopey gcc-doesn't-know-that-foo(&var)-writes-to-var problem.
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When I fixed the RC receive completion opcode in 2bfc8e9e ("IB/ipath:
Return the correct opcode for RDMA WRITE with immediate"), I forgot to
fix UC, which had the same problem for RDMA write with immediate
returning the wrong opcode.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Commit f018c7e1 ("IB/ipath: Change ipath_devdata.ipath_sdma_status to be
unsigned long") changed ipath_sdma_status to be unsigned long, but left
a few debug messages that printed it out with a %016llx format, which
generates the warnings
drivers/infiniband/hw/ipath/ipath_sdma.c:348: warning: format '%016llx' expects type 'long long unsigned int', but argument 3 has type 'long unsigned int'
drivers/infiniband/hw/ipath/ipath_sdma.c:618: warning: format '%016llx' expects type 'long long unsigned int', but argument 3 has type 'long unsigned int'
Fix this by changing the format used to print out the value to %08lx
(8 hex digits are now sufficient, because the highest bit used is 31).
Warnings reported by Randy Dunlap <randy.dunlap@oracle.com>.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
cxio_flush_sq() was failing to wrap around the software send queue
causing garbage completion entries on a flush operation.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Andrew Morton <akpm@linux-foundation.org> pointed out that bitops
should take an unsigned long * arg. However, the ipath driver was
doing bitops on struct ipath_devdata.ipath_sdma_status, which is u64.
Change this member to unsigned long to avoid tons of warnings when x86
fixes the bitops to take unsigned long * instead of void *.
Also, change the IPATH_SDMA_RUNNING and IPATH_SDMA_SHUTDOWN bit
numbers to 30 and 31 (instead of 62 and 63) so that we're not setting
another booby trap for someone who tries to make ipath work on a
32-bit architecture.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The official reason is "with the presence of pid namespaces in the
kernel using pid_t-s inside one is no longer safe."
But the reason I fix this right now is the following:
About a month ago (when 2.6.25 was not yet released) there still was a
one last caller of a to-be-deprecated-soon function find_pid() - the
kill_proc() function, which in turn was only used by nfs callback
code.
During the last merge window, this last caller was finally eliminated
by some NFS patch(es) and I was about to finally kill this kill_proc()
and find_pid(), but found, that I was late and the kill_proc is now
called from the ipath driver since commit 58411d1c ("IB/ipath: Head of
Line blocking vs forward progress of user apps").
So here's a patch that fixes this code to use struct pid * and (!)
the kill_pid routine.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
If an out of sequence RDMA read response middle or last packet is
received, we should only resend the RDMA read request on the first
out of sequence packet and drop subsequent out of sequence packets
otherwise, we get "too many retries".
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The send DMA hardware queue voided a number of prior assumptions about
when a send is complete which led to completions being generated out of
order. There were also a number of locking issues when switching the QP
to the error or reset states, and we implement the IB_QPS_SQD state.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When errors are detected in RC, the QP should transition to the
IB_QPS_ERR state, not the IB_QPS_SQE state. Also, when the error is on
the responder side, the receive work completion error was incorrect
(remote vs. local).
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Fix some bugs with the max_aggr module parameter added with LRO support:
- The module parameter value ignored and not actually used to set
lro_mgr.max_aggr.
- MODULE_PARM_DESC had a typo "_mro_" instead of "_lro_" so it didn't
end up describing the actual module parameter.
- The nes_lro_max_aggr variable was declared as unsigned, but the
module_param line said "int" instead of "uint" for the type.
- The default value for the parameter was stuck in the permissions
field of module_param, which led to nonsensical permissions for the
file under /sys/module/iw_nes/param.
- The parameter was used in only one file but defined in another, which
led to the variable being global for no good reason. Move everything
related to the parameter to the file nes_hw.c where it is actually
used.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This is necessary because, in a multicore environment, a race between
uverbs async handler and destroy QP could occur.
Signed-off-by: Stefan Roscher <stefan.roscher at de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
What's fixed:
in ipath_cancel_sends()
We need to unconditionally set ABORTING. So, swap the tests
so the set_bit() isn't shadowed by the &&.
If we've disarmed the piobufs, then we need to unconditionally
set DISARMED. So, move it out from the overly protective if
at the bottom.
in sdma_abort_task()
Abort_task was written knowing that the SDMA engine would always
be reset (and restarted) on error. A recent change broke that
fundamental assumption by taking the restart portion and making
it conditional on a link status change. But, SDMA can go boom
without a link status change in some conditions.
Signed-off-by: John Gregor <john.gregor@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Now that we always use PIO for vl15 on 7220, we could get stuck forever
if we happened to run out of PIO buffers from the verbs code, because
the setup code wouldn't run; the interrupt was also ignored if SDMA was
supported. We also have to reduce the pio update threshold if we have
fewer kernel buffers than the existing threshold.
Clean up the initialization a bit to get ordering safer and more
sensible, and use the existing ipath_chg_kernavail call to do init,
rather than doing it separately.
Drop unnecessary clearing of pio buffer on pio parity error.
Drop incorrect updating of pioavailshadow when exitting freeze mode
(software state may not match chip state if buffer has been allocated
and not yet written).
If we couldn't get a kernel buffer for a while, make sure we are
in sync with hardware, mainly to handle the exitting freeze case.
Signed-off-by: Dave Olson <dave.olson@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The loop in ipath_kreceive() that processes packets increments the
loop-index 'i' once too often, because the exit condition does not
depend on it, and is checked after the increment. By adding a check for
!last to the iterator in the for loop, we correct that in a way that is
not so likely to be re-broken by changes in the loop body.
Signed-off-by: Michael Albaugh <micheal.albaugh@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This patch fixes a bug in the RC responder which generates a completion
entry with the wrong opcode when an RDMA WRITE with immediate is received.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The semantics of cancel_sends changed, but the code using it was missed.
Don't leave sends and pioavail updates disabled, and add a comment as to
why the force update is needed.
Signed-off-by: Dave Olson <dave.olson@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
If a send work request has immediate errors and is not put on the
send queue, we shouldn't update any of the QP state.
The increment of the SSN wasn't obeying this.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
We warn about prototype chips, but the function that checks for
support is also called as a result of a get_portinfo request, which
can clutter the logs.
Restrict warning to only appear during initialization.
Signed-off-by: Michael Albaugh <michael.albaugh@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Currently, iw_cxgb3 is severely limited on the amount of userspace
memory that can be registered in in a single memory region, which
causes big problems for applications that expect to be able to
register 100s of MB.
The problem is that the driver uses a single kmalloc()ed buffer to
hold the physical buffer list (PBL) for the entire memory region
during registration, which means that 8 bytes of contiguous memory are
required for each page of memory being registered. For example, a 64
MB registration will require 128 KB of contiguous memory with 4 KB
pages, and it unlikely that such an allocation will succeed on a busy
system.
This is purely a driver problem: the temporary page list buffer is not
needed by the hardware, so we can fix this by writing the PBL to the
hardware in page-sized chunks rather than all at once. We do this by
splitting the memory registration operation up into several steps:
- Allocate PBL space in adapter memory for the full registration
- Copy PBL to adapter memory in chunks
- Allocate STag and enable memory region
This also allows several other cleanups to the __cxio_tpt_op()
interface and related parts of the driver.
This change leaves the reregister memory region and memory window
operations broken, but they already didn't work due to other
longstanding bugs, so fixing them will be left to a later patch.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Current iw_cxgb3 code adds PBL memory to the driver's gen_pool in 2 MB
chunks. This limits the largest single allocation that can be done to
the same size, which means that with 4 KB pages, each of which takes 8
bytes of PBL memory, the largest memory region that can be allocated
is 1 GB (256K PBL entries * 4 KB/entry).
Remove this limit by adding all the PBL memory in a single gen_pool
chunk, if possible. Add code that falls back to smaller chunks if
gen_pool_add() fails, which can happen if there is not sufficient
contiguous lowmem for the internal gen_pool bitmap.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Also remove duplicate assignment of local_ca_ack_delay and change
min_t check for local_ca_ack_delay to u8 instead of int.
Signed-off-by: Stefan Roscher <stefan.roscher at de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Testing on large clusters shows its way too short at 10 secs.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Remove bad BUG_ON() that can trigger in correct operation from
close_con_rpl(). It is possible to get a close_rpl message on a dead
connection. The sequence is:
- host refs ep for close exchange
- host posts close_req
- hw posts PEER_ABORT from incoming RST
- host marks ep DEAD
- host posts ABORT_RPL and releases ep resources
- hw posts CLOSE_RPL
- host derefs ep and ep freed.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
- Flush the QP only after the HW disables the connection. Currently
we flush the QP when transitioning to CLOSING. This exposes a race
condition where the HW can complete a RECV WR, for instance, -and-
the SW can flush that same WR.
- Only call CQ event handlers on flush IFF we actually flushed something.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Commit f56bcd80 ("IPoIB: Use separate CQ for UD send completions")
introduced a bug where the transmit queue could get stopped and never
woken up. The problem is that send completions are only polled at the
end of the xmit function, so if the send queue fills up and the xmit
path stops the queue, then there is no way for send completions to
ever get polled, and so the transmit queue stays stopped forever.
Fix this by arming the send CQ just before posting the last send
request that fills the send queue. Then, when the completion event
handler is called, drain the send CQ. Since it is possible that not
enough send completions are in the CQ, verify that the the net queue
has been woken up after draining the send CQ, and if not arm a timer
and drain again at the timer function.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When I merged bbf8eed1 ("IB/mlx4: Add support for resizing CQs") I
changed things around so that mlx4_ib_alloc_cq_buf() and
mlx4_ib_free_cq_buf() were used everywhere they could be. However, I
screwed up the number of entries passed into mlx4_ib_alloc_cq_buf()
in a couple places -- the function bumps the number of entries
internally, so the caller shouldn't add 1 as well.
Passing a too-big value for the number of entries to mlx4_ib_free_cq_buf()
can cause the cleanup to go off the end of an array and corrupt
allocator state in interesting ways.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Various cleanups:
- Change // to /* .. */
- Place whitespace around binary operators.
- Trim down a few long lines.
- Some minor alignment formatting for better readability.
- Remove some silly tabs.
Signed-off-by: Glenn Streiff <gstreiff@neteffect.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This patch enables the iw_nes module for NetEffect RNICs to support
additional PHYs including SFP+ (referred to as ARGUS in the code).
Signed-off-by: Eric Schneider <eric.schneider@neteffect.com>
Signed-off-by: Glenn Streiff <gstreiff@neteffect.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When creating a child interface, copy the MTU information from the
parent. Otherwise when the child's multicast join completes, the MTU
will not be updated since the code does
dev->mtu = min(priv->mcast_mtu, priv->admin_mtu);
and priv->admin_mtu will be set to 0.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Commit cb9fbc5c ("IB: expand ib_umem_get() prototype") changed the
mthca userspace ABI to provide a way for userspace to indicate which
memory regions need the DMA write barrier attribute. However, it is
possible to handle this without breaking existing userspace, by having
the mthca kernel driver recognize whether it is talking to old or new
userspace, depending on the size of the register MR structure passed in.
The only potential drawback of this is that is allows old userspace
(which has a bug with DMA ordering on large SGI Altix systems) to
continue to run on new kernels, but the advantage of allowing old
userspace to continue to work on unaffected systems seems to outweigh
this, and we can print a warning to push people to upgrade their
userspace.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When a FMR is unmapped, mthca resets the map count to 0, and clears
the upper part of the R_Key which is used as the sequence counter.
This poses a problem for RDS, which uses ib_fmr_unmap as a fence
operation. RDS assumes that after issuing an unmap, the old R_Keys
will be invalid for a "reasonable" period of time. For instance,
Oracle processes uses shared memory buffers allocated from a pool of
buffers. When a process dies, we want to reclaim these buffers -- but
we must make sure there are no pending RDMA operations to/from those
buffers. The only way to achieve that is by using unmap and sync the
TPT.
However, when the sequence count is reset on unmap, there is a high
likelihood that a new mapping will be given the same R_Key that was
issued a few milliseconds ago.
To prevent this, don't reset the sequence count when unmapping a FMR.
Signed-off-by: Olaf Kirch <olaf.kirch@oracle.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
If a lot of QPs fall into Error state at once and the EQ of the
respective HCA is too small, it might overrun, causing the eHCA driver
to stop processing completion events and calling the application's
completion handlers, effectively causing traffic to stop.
Fix this by limiting available QPs and CQs to a customizable max
count, and determining EQ size based on these counts and a worst-case
assumption.
Signed-off-by: Stefan Roscher <stefan.roscher@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Use a dedicated CQ for UD send completions. Also, do not arm the UD
send CQ, which reduces the number of interrupts generated. This patch
farther reduces overhead by not calling poll CQ for every posted send
WR -- it does polls only when there 16 or more outstanding work requests.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Count FMR alignment violations per session as part of the iscsi
statistics.
Signed-off-by: Eli Dorfman <elid@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
ehca_create_eq() was assigning a signed return value to an unsiged
local variable and then checking if the variable was < 0, which meant
that errors were always ignored. Fix this by using one variable for
signed integer return values and another for u64 hcall return values.
Bug originally found by Roel Kluin <12o3l@tiscali.nl>.
Signed-off-by: Hoang-Nam Nguyen <hnguyen@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Open MPI, Intel MPI and other applications don't respect the iWARP
requirement that the client (active) side of the connection send the
first RDMA message. This class of application connection setup is
called peer-to-peer. Typically once the connection is setup, _both_
sides want to send data.
This patch enables supporting peer-to-peer over the chelsio RNIC by
enforcing this iWARP requirement in the driver itself as part of RDMA
connection setup.
Connection setup is extended, when the peer2peer module option is 1,
such that the MPA initiator will send a 0B Read (the RTR) just after
connection setup. The MPA responder will suspend SQ processing until
the RTR message is received and reply-to.
In the longer term, this will be handled in a standardized way by
enhancing the MPA negotiation so peers can indicate whether they
want/need the RTR and what type of RTR (0B read, 0B write, or 0B send)
should be sent. This will be done by standardizing a few bits of the
private data in order to negotiate all this. However this patch
enables peer-to-peer applications now and allows most of the required
firmware and driver changes to be done and tested now.
Design:
- Add a module option, peer2peer, to enable this mode.
- New firmware support for peer-to-peer mode:
- a new bit in the rdma_init WR to tell it to do peer-2-peer
and what form of RTR message to send or expect.
- process _all_ preposted recvs before moving the connection
into rdma mode.
- passive side: defer completing the rdma_init WR until all
pre-posted recvs are processed. Suspend SQ processing until
the RTR is received.
- active side: expect and process the 0B read WR on offload TX
queue. Defer completing the rdma_init WR until all
pre-posted recvs are processed. Suspend SQ processing until
the 0B read WR is processed from the offload TX queue.
- If peer2peer is set, driver posts 0B read request on offload TX
queue just after posting the rdma_init WR to the offload TX queue.
- Add CQ poll logic to ignore unsolicitied read responses.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
cxgb3 only supports 4GB memory regions. The lustre RDMA code uses
this attribute and currently has to code around our bad setting.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Open MPI and other stress testing exposed a few bad bugs in handling
aborts in the middle of a normal close. Fix these by:
- serializing abort reply and peer abort processing with disconnect
processing
- warning (and ignoring) if ep timer is stopped when it wasn't running
- cleaning up disconnect path to correctly deal with aborting and
dead endpoints
- in iwch_modify_qp(), taking a ref on the ep before releasing the qp
lock if iwch_ep_disconnect() will be called. The ref is dropped
after calling disconnect.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
