diff options
| author | Kumar Sanghvi <kumaras@chelsio.com> | 2011-10-24 21:20:21 +0530 |
|---|---|---|
| committer | Roland Dreier <roland@purestorage.com> | 2011-10-31 11:34:53 -0700 |
| commit | 581bbe2cd0694a935e0c3ccd7f011e10094f1df6 (patch) | |
| tree | 38e536efa0d05d76964b09836def2210a00b41b5 /drivers/infiniband/hw/cxgb4/iw_cxgb4.h | |
| parent | e14d62c05c0b8eff61c6fd46b4a78fb27c8cf38b (diff) | |
| download | lwn-581bbe2cd0694a935e0c3ccd7f011e10094f1df6.tar.gz lwn-581bbe2cd0694a935e0c3ccd7f011e10094f1df6.zip | |
RDMA/cxgb4: Serialize calls to CQ's comp_handler
Commit 01e7da6ba53c ("RDMA/cxgb4: Make sure flush CQ entries are
collected on connection close") introduced a potential problem where a
CQ's comp_handler can get called simultaneously from different places
in the iw_cxgb4 driver. This does not comply with
Documentation/infiniband/core_locking.txt, which states that at a
given point of time, there should be only one callback per CQ should
be active.
This problem was reported by Parav Pandit <Parav.Pandit@Emulex.Com>.
Based on discussion between Parav Pandit and Steve Wise, this patch
fixes the above problem by serializing the calls to a CQ's
comp_handler using a spin_lock.
Reported-by: Parav Pandit <Parav.Pandit@Emulex.Com>
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
Acked-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Diffstat (limited to 'drivers/infiniband/hw/cxgb4/iw_cxgb4.h')
| -rw-r--r-- | drivers/infiniband/hw/cxgb4/iw_cxgb4.h | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h index 4f045375c8e2..02f015fc3ed2 100644 --- a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h +++ b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h @@ -309,6 +309,7 @@ struct c4iw_cq { struct c4iw_dev *rhp; struct t4_cq cq; spinlock_t lock; + spinlock_t comp_handler_lock; atomic_t refcnt; wait_queue_head_t wait; }; |