| Age | Commit message (Collapse) | Author |
|---|
|
git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband
Pull InfiniBand/RDMA changes from Roland Dreier:
- Add ocrdma hardware driver for Emulex IB-over-Ethernet adapters
- Add generic and mlx4 support for "raw" QPs: allow suitably privileged
applications to send and receive arbitrary packets directly to/from
the hardware
- Add "doorbell drop" handling to the cxgb4 driver
- A fairly large batch of qib hardware driver changes
- A few fixes for lockdep-detected issues
- A few other miscellaneous fixes and cleanups
Fix up trivial conflict in drivers/net/ethernet/emulex/benet/be.h.
* tag 'rdma-for-3.5' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (53 commits)
RDMA/cxgb4: Include vmalloc.h for vmalloc and vfree
IB/mlx4: Fix mlx4_ib_add() error flow
IB/core: Fix IB_SA_COMP_MASK macro
IB/iser: Fix error flow in iser ep connection establishment
IB/mlx4: Increase the number of vectors (EQs) available for ULPs
RDMA/cxgb4: Add query_qp support
RDMA/cxgb4: Remove kfifo usage
RDMA/cxgb4: Use vmalloc() for debugfs QP dump
RDMA/cxgb4: DB Drop Recovery for RDMA and LLD queues
RDMA/cxgb4: Disable interrupts in c4iw_ev_dispatch()
RDMA/cxgb4: Add DB Overflow Avoidance
RDMA/cxgb4: Add debugfs RDMA memory stats
cxgb4: DB Drop Recovery for RDMA and LLD queues
cxgb4: Common platform specific changes for DB Drop Recovery
cxgb4: Detect DB FULL events and notify RDMA ULD
RDMA/cxgb4: Drop peer_abort when no endpoint found
RDMA/cxgb4: Always wake up waiters in c4iw_peer_abort_intr()
mlx4_core: Change bitmap allocator to work in round-robin fashion
RDMA/nes: Don't call event handler if pointer is NULL
RDMA/nes: Fix for the ORD value of the connecting peer
...
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending
Pull scsi-target changes from Nicholas Bellinger:
"There has been lots of work in existing code in a number of areas this
past cycle. The major highlights have been:
* Removal of transport_do_task_sg_chain() from core + fabrics
(Roland)
* target-core: Removal of se_task abstraction from target-core and
enforce hw_max_sectors for pSCSI backends (hch)
* Re-factoring of iscsi-target tx immediate/response queues (agrover)
* Conversion of iscsi-target back to using target core memory
allocation logic (agrover)
We've had one last minute iscsi-target patch go into for-next to
address a nasty regression bug related to the target core allocation
logic conversion from agrover that is not included in friday's
linux-next build, but has been included in this series.
On the new fabric module code front for-3.5, here is a brief status
update for the three currently in flight this round:
* usb-gadget target driver:
Sebastian Siewior's driver for supporting usb-gadget target mode
operation. This will be going out as a separate PULL request from
target-pending/usb-target-merge with subsystem maintainer ACKs. There
is one minor target-core patch in this series required to function.
* sbp ieee-1394/firewire target driver:
Chris Boot's driver for supportting the Serial Block Protocol (SBP)
across IEEE-1394 Firewire hardware. This will be going out as a
separate PULL request from target-pending/sbp-target-merge with two
additional drivers/firewire/ patches w/ subsystem maintainer ACKs.
* qla2xxx LLD target mode infrastructure changes + tcm_qla2xxx:
The Qlogic >= 24xx series HW target mode LLD infrastructure patch-set
and tcm_qla2xxx fabric driver. Support for FC target mode using
qla2xxx LLD code has been officially submitted by Qlogic to James
below, and is currently outstanding but not yet merged into
scsi.git/for-next..
[PATCH 00/22] qla2xxx: Updates for scsi "misc" branch
http://www.spinics.net/lists/linux-scsi/msg59350.html
Note there are *zero* direct dependencies upon this for-next series
for the qla2xxx LLD target + tcm_qla2xxx patches submitted above, and
over the last days the target mode team has been tracking down an
tcm_qla2xxx specific active I/O shutdown bug that appears to now be
almost squashed for 3.5-rc-fixes."
* 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending: (47 commits)
iscsi-target: Fix iov_count calculation bug in iscsit_allocate_iovecs
iscsi-target: remove dead code in iscsi_check_valuelist_for_support
target: Handle ATA_16 passthrough for pSCSI backend devices
target: Add MI_REPORT_TARGET_PGS ext. header + implict_trans_secs attribute
target: Fix MAINTENANCE_IN service action CDB checks to use lower 5 bits
target: add support for the WRITE_VERIFY command
target: make target_put_session void
target: cleanup transport_execute_tasks()
target: Remove max_sectors device attribute for modern se_task less code
target: lock => unlock typo in transport_lun_wait_for_tasks
target: Enforce hw_max_sectors for SCF_SCSI_DATA_SG_IO_CDB
target: remove the t_se_count field in struct se_cmd
target: remove the t_task_cdbs_ex_left field in struct se_cmd
target: remove the t_task_cdbs_left field in struct se_cmd
target: remove struct se_task
target: move the state and execute lists to the command
target: simplify command to task linkage
target: always allocate a single task
target: replace ->execute_task with ->execute_cmd
target: remove the task_sectors field in struct se_task
...
|
|
'ocrdma', 'qib' and 'raw-qp' into for-linus
|
|
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
We need to use a different loop index for mlx4_counter_alloc() and for
device_create_file() iterations: the mlx4_counter_alloc() loop index
is used in the error flow to free counters.
If the same loop index is used for device_create_file() and, say, the
device_create_file() loop fails on the first iteration, the allocated
counters will not be freed.
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
The current error flow code was releasing the IB connection object and
calling iscsi_destroy_endpoint() directly without going through the
reference counting mechanism introduced in commit 39ff05d ("IB/iser:
Enhance disconnection logic for multi-pathing"). This resulted in a
double free of the iscsi endpoint object, which causes a kernel NULL
pointer dereference. Fix that by plugging into the IB conn reference
counting correctly.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Enable IB ULPs to use a larger portion of the device EQs (which map to
IRQs). The mlx4_ib driver follows the mlx4_core framework of the EQs
to be divided among the device ports. In this scheme, for each IB
port, the number of allocated EQs follows the number of cores, subject
to other system constraints, such as number available MSI-X vectors.
Signed-off-by: Shlomo Pongratz <shlomop@mellanox.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
This allows querying the QP state before flushing.
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Using kfifos for ID management was limiting the number of QPs and
preventing NP384 MPI jobs. So replace it with a simple bitmap
allocator.
Remove IDs from the IDR tables before deallocating them. This bug was
causing the BUG_ON() in insert_handle() to fire because the ID was
getting reused before being removed from the IDR table.
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
This allows dumping thousands of QPs. Log active open failures of
interest.
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Add module option db_fc_threshold which is the count of active QPs
that trigger automatic db flow control mode. Automatically transition
to/from flow control mode when the active qp count crosses
db_fc_theshold.
Add more db debugfs stats
On DB DROP event from the LLD, recover all the iwarp queues.
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Use GFP_ATOMIC in _insert_handle() if ints are disabled.
Don't panic if we get an abort with no endpoint found. Just log a
warning.
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Get FULL/EMPTY/DROP events from LLD. On FULL event, disable normal
user mode DB rings.
Add modify_qp semantics to allow user processes to call into the
kernel to ring doobells without overflowing.
Add DB Full/Empty/Drop stats.
Mark queues when created indicating the doorbell state.
If we're in the middle of db overflow avoidance, then newly created
queues should start out in this mode.
Bump the C4IW_UVERBS_ABI_VERSION to 2 so the user mode library can
know if the driver supports the kernel mode db ringing.
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Log a warning and drop the abort message. Otherwise we will do a
bogus wake_up() and crash.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
This fixes a race where an ingress abort fails to wake up the thread
blocked in rdma_init() causing the app to hang.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Don't call the ibqp event_handler pointer in the case it wasn't initialized.
Signed-off-by: Tatyana Nikolova <Tatyana.E.Nikolova@intel.com>
Signed-off-by: Donald Wood <Donald.E.Wood@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Set ORD value of the connecting peer to be at least one in order to
accommodate an RDMA READ Request message.
Signed-off-by: Tatyana Nikolova <Tatyana.E.Nikolova@intel.com>
Signed-off-by: Donald Wood <Donald.E.Wood@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
This patch reorganizes the QP and devdata files to be more cache line aware.
qib_qp fields in particular are split into read-mostly, send, and receive fields.
qib_devdata fields are split into read-mostly and read/write fields
Testing has show that bidirectional tests improve by as much as 100%
with this patch.
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
If a MAD is sent directly to the local HCA rather than placed on a QP
and the MAD fails M_Key checks, there is no means to generate a
timeout for the client, which may hang. Instead we report
IB_MAD_RESULT_FAILURE, which operates the same for on-the-wire
packets, but will generate a send failure back to the client.
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Jim Foraker <foraker1@llnl.gov>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
If a port has an M_Key lease set, the M_Key protect bits set to 1, and
a SubnSet arrives with an invalid M_Key, an M_Key mismatch trap is
generated, the lease timer begins as expected, and eventually the
M_Key protect bits will be set back to 0 as per the spec. However, if
any other SMP with an invalid M_Key arrives, the lease timer is expired
and the M_Key protect bits remain in force.
This is not according to to spec. In particular, C14-17 says that
a lease timer that is underway is not affected by protection level
checks (ie, at protection level 1, a SubnGet with a bad M_Key may be
successful, but does not stop the timer), and C14-19 says that the timer
shall stop when a valid M_Key has been received. C14-19 is the only
compliance statement that specifies a stopping condition for the timer.
This behavior is magnified if the port's Master SM LID attribute
points at itself. In that case, the M_Key mismatch trap is sufficient to
expire the timer, and the mkey lease attribute is rendered useless.
Reviewed-by: Ram Vepa <ram.vepa@qlogic.com>
Signed-off-by: Jim Foraker <foraker1@llnl.gov>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@qlogic.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
The SERDES was using the incorrect Frequency Loop Bandwidth setting
causing the link to cycle through the Physical link negotiation state
machine. Fixing the Frequency Loop Bandwidth setting in the SERDES
helps the link come up faster and more reliably.
Signed-off-by: Mitko Haralanov <mitko.haralanov@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
A "fix" for a bug with the number of contexts on a single-port board
caused the calculation to be off by one, which causes problems with
the upper layers. The same problem exists for number of free
contexts, which is also fixed here.
Signed-off-by: Mitko Haralanov <mitko.haralanov@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
When a port first goes active with SMA Set(PortInfo) and reregister
bit set, the driver sends up the reregister event followed by a port
active event.
The problem is that in response to reregister event most apps try to
issue a SA query of some sort, but that fails because port is not
active.
The qib driver needs to a trivial change to correct this behavior.
This issue has been there for a while; however the recent serdes work
has probably made the delay between the reregister event and the
active event larger and hence opened the race far enough so that its
being seen more often.
The patch also changes the clientrereg local to a u8 and saves off the
rereg bit into it. The code following the nested subn_get_portinfo()
now restores that bit per o14-12.2.1 with a logical OR from that copy.
Reviewed-by: Ram Vepa <ram.vepa@qlogic.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@qlogic.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
This patch optimizes pio buffer allocation in the kernel.
For qib, kernel pio buffers are used for sending acks. The code to
allocate the buffer would always start at 0 until it found a buffer.
This means that an average of 64 comparisions were done on each
allocate, since the busy bit won't be cleared until the bits are
refreshed when buffers are exhausted.
This patch adds two new fields in the devdata struct, last_pio and
min_kernel_pio. last_pio is the last buffer that was allocated.
min_kernel_pio is the lowest potential available buffer.
min_kernel_pio is modifed as contexts are allocated and deallocted.
Reviewed-by: Ramkrishna Vepa <ramkrishna.vepa@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Add a prefetch call when a packet has been stored. The nature of the
prefetch is correctly determined by the alternatives mechanism.
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Commit bc3e53f682d9 ("mm: distinguish between mlocked and pinned
pages") introduced a separate counter for pinned pages and used it in
the IB stack. However, in ib_umem_get() the pinned counter is
incremented, but ib_umem_release() wrongly decrements the locked
counter. Fix this.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Signed-off-by: Shlomo Pongratz <shlomop@mellanox.com>
[ Replace one more printk_once() with pr_info_once(). - Roland ]
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Otherwise CM packets going over MLX QP1 get fixed scheduling priority 0.
We want CM packets to get the same scheduling priority, and therefore
map to the same SQ (Schedule Queue) and eventually TC (Traffic Class),
as the application requested for the actual QP used for the connection.
Signed-off-by: Oren Duer <oren@mellanox.co.il>
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Implement raw packet QPs for Ethernet ports using the MLX transport (as
done by the mlx4_en Ethernet netdevice driver).
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
IB_QPT_RAW_PACKET allows applications to build a complete packet,
including L2 headers, when sending; on the receive side, the HW will
not strip any headers.
This QP type is designed for userspace direct access to Ethernet; for
example by applications that do TCP/IP themselves. Only processes
with the NET_RAW capability are allowed to create raw packet QPs (the
name "raw packet QP" is supposed to suggest an analogy to AF_PACKET /
SOL_RAW sockets).
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
When IPV6 is not enabled:
ERROR: "register_inet6addr_notifier" [drivers/infiniband/hw/ocrdma/ocrdma.ko] undefined!
ERROR: "unregister_inet6addr_notifier" [drivers/infiniband/hw/ocrdma/ocrdma.ko] undefined!
Fix this by wrapping the inet6 calls in #ifdef IPV6. Also make the
ocrdma module depend on (IPV6 || IPV6=n) to forbid the case of modular
ipv6 but built-in ocrdma (which can't work, because ocrdma calls ipv6
functions).
Reported-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
We only need to disable the IRQs one time.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
[ Rename "wq_flags" to more conventional "flags." - Roland ]
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
The ocrdma_alloc_lkey() function never returns NULL pointers -- it
returns ERR_PTRs.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Events sent to ocrdma_inet6addr_event() are sent from an atomic context,
therefore we can't try to lock a mutex within the notifier callback.
We could just switch the mutex to a spinlock since all it does it
protect a list, but I've gone ahead and switched the list to use RCU
instead. I couldn't fully test it since I don't have IB hardware, so
if it doesn't fully work for some reason let me know and I'll switch
it back to using a spinlock.
Signed-off-by: Sasha Levin <levinsasha928@gmail.com>
[ Fixed locking in ocrdma_add(). - Roland ]
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
We need to set ib_evt.device, or else ib_dispatch_event() will crash
when we call it for unaffiliated events (and consumers may get
confused in their QP/CQ/SRQ event handler for affiliated events).
Also fix sparse warning:
drivers/infiniband/hw/ocrdma/ocrdma_hw.c:678:36: warning: Using plain integer as NULL pointer
There's no need to clear ib_evt, since every member is initialized.
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
First, fix
drivers/infiniband/hw/ocrdma/ocrdma_verbs.c: In function 'ocrdma_alloc_pd':
drivers/infiniband/hw/ocrdma/ocrdma_verbs.c:371:17: warning: 'dpp_page_addr' may be used uninitialized in this function [-Wuninitialized]
drivers/infiniband/hw/ocrdma/ocrdma_verbs.c:337:6: note: 'dpp_page_addr' was declared here
which seems that it may border on a bug (the call to ocrdma_del_mmap()
might conceivably do bad things if pd->dpp_enabled is not set and
dpp_page_addr ends up with just the wrong value).
Also take care of:
drivers/infiniband/hw/ocrdma/ocrdma_hw.c: In function 'ocrdma_init_hw':
drivers/infiniband/hw/ocrdma/ocrdma_hw.c:2587:5: warning: 'status' may be used uninitialized in this function [-Wuninitialized]
drivers/infiniband/hw/ocrdma/ocrdma_hw.c:2549:17: note: 'status' was declared here
which is only real if num_eq == 0, which should be impossible.
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Signed-off-by: Parav Pandit <parav.pandit@emulex.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
The following lockdep problem was reported by Or Gerlitz <ogerlitz@mellanox.com>:
[ INFO: possible recursive locking detected ]
3.3.0-32035-g1b2649e-dirty #4 Not tainted
---------------------------------------------
kworker/5:1/418 is trying to acquire lock:
(&id_priv->handler_mutex){+.+.+.}, at: [<ffffffffa0138a41>] rdma_destroy_i d+0x33/0x1f0 [rdma_cm]
but task is already holding lock:
(&id_priv->handler_mutex){+.+.+.}, at: [<ffffffffa0135130>] cma_disable_ca llback+0x24/0x45 [rdma_cm]
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0
----
lock(&id_priv->handler_mutex);
lock(&id_priv->handler_mutex);
*** DEADLOCK ***
May be due to missing lock nesting notation
3 locks held by kworker/5:1/418:
#0: (ib_cm){.+.+.+}, at: [<ffffffff81042ac1>] process_one_work+0x210/0x4a 6
#1: ((&(&work->work)->work)){+.+.+.}, at: [<ffffffff81042ac1>] process_on e_work+0x210/0x4a6
#2: (&id_priv->handler_mutex){+.+.+.}, at: [<ffffffffa0135130>] cma_disab le_callback+0x24/0x45 [rdma_cm]
stack backtrace:
Pid: 418, comm: kworker/5:1 Not tainted 3.3.0-32035-g1b2649e-dirty #4
Call Trace:
[<ffffffff8102b0fb>] ? console_unlock+0x1f4/0x204
[<ffffffff81068771>] __lock_acquire+0x16b5/0x174e
[<ffffffff8106461f>] ? save_trace+0x3f/0xb3
[<ffffffff810688fa>] lock_acquire+0xf0/0x116
[<ffffffffa0138a41>] ? rdma_destroy_id+0x33/0x1f0 [rdma_cm]
[<ffffffff81364351>] mutex_lock_nested+0x64/0x2ce
[<ffffffffa0138a41>] ? rdma_destroy_id+0x33/0x1f0 [rdma_cm]
[<ffffffff81065a78>] ? trace_hardirqs_on_caller+0x11e/0x155
[<ffffffff81065abc>] ? trace_hardirqs_on+0xd/0xf
[<ffffffffa0138a41>] rdma_destroy_id+0x33/0x1f0 [rdma_cm]
[<ffffffffa0139c02>] cma_req_handler+0x418/0x644 [rdma_cm]
[<ffffffffa012ee88>] cm_process_work+0x32/0x119 [ib_cm]
[<ffffffffa0130299>] cm_req_handler+0x928/0x982 [ib_cm]
[<ffffffffa01302f3>] ? cm_req_handler+0x982/0x982 [ib_cm]
[<ffffffffa0130326>] cm_work_handler+0x33/0xfe5 [ib_cm]
[<ffffffff81065a78>] ? trace_hardirqs_on_caller+0x11e/0x155
[<ffffffffa01302f3>] ? cm_req_handler+0x982/0x982 [ib_cm]
[<ffffffff81042b6e>] process_one_work+0x2bd/0x4a6
[<ffffffff81042ac1>] ? process_one_work+0x210/0x4a6
[<ffffffff813669f3>] ? _raw_spin_unlock_irq+0x2b/0x40
[<ffffffff8104316e>] worker_thread+0x1d6/0x350
[<ffffffff81042f98>] ? rescuer_thread+0x241/0x241
[<ffffffff81046a32>] kthread+0x84/0x8c
[<ffffffff8136e854>] kernel_thread_helper+0x4/0x10
[<ffffffff81366d59>] ? retint_restore_args+0xe/0xe
[<ffffffff810469ae>] ? __init_kthread_worker+0x56/0x56
[<ffffffff8136e850>] ? gs_change+0xb/0xb
The actual locking is fine, since we're dealing with different locks,
but from the same lock class. cma_disable_callback() acquires the
listening id mutex, whereas rdma_destroy_id() acquires the mutex for
the new connection id. To fix this, delay the call to
rdma_destroy_id() until we've released the listening id mutex.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Since XRC support was added, the uverbs code has locked SRQ, CQ and PD
objects needed during QP and SRQ creation in different orders
depending on the the code path. This leads to the (at least
theoretical) possibility of deadlock, and triggers the lockdep splat
below.
Fix this by making sure we always lock the SRQ first, then CQs and
finally the PD.
======================================================
[ INFO: possible circular locking dependency detected ]
3.4.0-rc5+ #34 Not tainted
-------------------------------------------------------
ibv_srq_pingpon/2484 is trying to acquire lock:
(SRQ-uobj){+++++.}, at: [<ffffffffa00af51b>] idr_read_uobj+0x2f/0x4d [ib_uverbs]
but task is already holding lock:
(CQ-uobj){+++++.}, at: [<ffffffffa00af51b>] idr_read_uobj+0x2f/0x4d [ib_uverbs]
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #2 (CQ-uobj){+++++.}:
[<ffffffff81070fd0>] lock_acquire+0xbf/0xfe
[<ffffffff81384f28>] down_read+0x34/0x43
[<ffffffffa00af51b>] idr_read_uobj+0x2f/0x4d [ib_uverbs]
[<ffffffffa00af542>] idr_read_obj+0x9/0x19 [ib_uverbs]
[<ffffffffa00b16c3>] ib_uverbs_create_qp+0x180/0x684 [ib_uverbs]
[<ffffffffa00ae3dd>] ib_uverbs_write+0xb7/0xc2 [ib_uverbs]
[<ffffffff810fe47f>] vfs_write+0xa7/0xee
[<ffffffff810fe65f>] sys_write+0x45/0x69
[<ffffffff8138cdf9>] system_call_fastpath+0x16/0x1b
-> #1 (PD-uobj){++++++}:
[<ffffffff81070fd0>] lock_acquire+0xbf/0xfe
[<ffffffff81384f28>] down_read+0x34/0x43
[<ffffffffa00af51b>] idr_read_uobj+0x2f/0x4d [ib_uverbs]
[<ffffffffa00af542>] idr_read_obj+0x9/0x19 [ib_uverbs]
[<ffffffffa00af8ad>] __uverbs_create_xsrq+0x96/0x386 [ib_uverbs]
[<ffffffffa00b31b9>] ib_uverbs_detach_mcast+0x1cd/0x1e6 [ib_uverbs]
[<ffffffffa00ae3dd>] ib_uverbs_write+0xb7/0xc2 [ib_uverbs]
[<ffffffff810fe47f>] vfs_write+0xa7/0xee
[<ffffffff810fe65f>] sys_write+0x45/0x69
[<ffffffff8138cdf9>] system_call_fastpath+0x16/0x1b
-> #0 (SRQ-uobj){+++++.}:
[<ffffffff81070898>] __lock_acquire+0xa29/0xd06
[<ffffffff81070fd0>] lock_acquire+0xbf/0xfe
[<ffffffff81384f28>] down_read+0x34/0x43
[<ffffffffa00af51b>] idr_read_uobj+0x2f/0x4d [ib_uverbs]
[<ffffffffa00af542>] idr_read_obj+0x9/0x19 [ib_uverbs]
[<ffffffffa00b1728>] ib_uverbs_create_qp+0x1e5/0x684 [ib_uverbs]
[<ffffffffa00ae3dd>] ib_uverbs_write+0xb7/0xc2 [ib_uverbs]
[<ffffffff810fe47f>] vfs_write+0xa7/0xee
[<ffffffff810fe65f>] sys_write+0x45/0x69
[<ffffffff8138cdf9>] system_call_fastpath+0x16/0x1b
other info that might help us debug this:
Chain exists of:
SRQ-uobj --> PD-uobj --> CQ-uobj
Possible unsafe locking scenario:
CPU0 CPU1
---- ----
lock(CQ-uobj);
lock(PD-uobj);
lock(CQ-uobj);
lock(SRQ-uobj);
*** DEADLOCK ***
3 locks held by ibv_srq_pingpon/2484:
#0: (QP-uobj){+.+...}, at: [<ffffffffa00b162c>] ib_uverbs_create_qp+0xe9/0x684 [ib_uverbs]
#1: (PD-uobj){++++++}, at: [<ffffffffa00af51b>] idr_read_uobj+0x2f/0x4d [ib_uverbs]
#2: (CQ-uobj){+++++.}, at: [<ffffffffa00af51b>] idr_read_uobj+0x2f/0x4d [ib_uverbs]
stack backtrace:
Pid: 2484, comm: ibv_srq_pingpon Not tainted 3.4.0-rc5+ #34
Call Trace:
[<ffffffff8137eff0>] print_circular_bug+0x1f8/0x209
[<ffffffff81070898>] __lock_acquire+0xa29/0xd06
[<ffffffffa00af37c>] ? __idr_get_uobj+0x20/0x5e [ib_uverbs]
[<ffffffffa00af51b>] ? idr_read_uobj+0x2f/0x4d [ib_uverbs]
[<ffffffff81070fd0>] lock_acquire+0xbf/0xfe
[<ffffffffa00af51b>] ? idr_read_uobj+0x2f/0x4d [ib_uverbs]
[<ffffffff81070eee>] ? lock_release+0x166/0x189
[<ffffffff81384f28>] down_read+0x34/0x43
[<ffffffffa00af51b>] ? idr_read_uobj+0x2f/0x4d [ib_uverbs]
[<ffffffffa00af51b>] idr_read_uobj+0x2f/0x4d [ib_uverbs]
[<ffffffffa00af542>] idr_read_obj+0x9/0x19 [ib_uverbs]
[<ffffffffa00b1728>] ib_uverbs_create_qp+0x1e5/0x684 [ib_uverbs]
[<ffffffff81070fec>] ? lock_acquire+0xdb/0xfe
[<ffffffff81070c09>] ? lock_release_non_nested+0x94/0x213
[<ffffffff810d470f>] ? might_fault+0x40/0x90
[<ffffffff810d470f>] ? might_fault+0x40/0x90
[<ffffffffa00ae3dd>] ib_uverbs_write+0xb7/0xc2 [ib_uverbs]
[<ffffffff810fe47f>] vfs_write+0xa7/0xee
[<ffffffff810ff736>] ? fget_light+0x3b/0x99
[<ffffffff810fe65f>] sys_write+0x45/0x69
[<ffffffff8138cdf9>] system_call_fastpath+0x16/0x1b
Reported-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Add names for our lockdep classes, so instead of having to decipher
lockdep output with mysterious names:
Chain exists of:
key#14 --> key#11 --> key#13
lockdep will give us something nicer:
Chain exists of:
SRQ-uobj --> PD-uobj --> CQ-uobj
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Change sizeof(array)/sizeof(array[0]) to ARRAY_SIZE.
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@qlogic.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Acked-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Function import_ep() is incorrectly using ep->dst instead of the dst
ptr passed in. This causes a crash when accepting new rdma connections
becase ep->dst is not initialized yet.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Just as we don't allow PDs, CQs, etc. to be destroyed if there are QPs
that are attached to them, don't let a QP be destroyed if there are
multicast group(s) attached to it. Use the existing usecnt field of
struct ib_qp which was added by commit 0e0ec7e ("RDMA/core: Export
ib_open_qp() to share XRC TGT QPs") to track this.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
|
|
Conflicts:
drivers/net/ethernet/intel/e1000e/param.c
drivers/net/wireless/iwlwifi/iwl-agn-rx.c
drivers/net/wireless/iwlwifi/iwl-trans-pcie-rx.c
drivers/net/wireless/iwlwifi/iwl-trans.h
Resolved the iwlwifi conflict with mainline using 3-way diff posted
by John Linville and Stephen Rothwell. In 'net' we added a bug
fix to make iwlwifi report a more accurate skb->truesize but this
conflicted with RX path changes that happened meanwhile in net-next.
In e1000e a conflict arose in the validation code for settings of
adapter->itr. 'net-next' had more sophisticated logic so that
logic was used.
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband
Pull infiniband fixes from Roland Dreier:
"A few fixes for regressions introduced in 3.4-rc1:
- fix memory leak in mlx4
- fix two problems with new MAD response generation code"
* tag 'ib-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
IB/mlx4: Fix memory leaks in ib_link_query_port()
IB/mad: Don't send response for failed MADs
IB/mad: Set 'D' bit in response for unhandled MADs
|
|
|
