diff options
| author | Kumar Sanghvi <kumaras@chelsio.com> | 2011-10-24 19:50:21 +0400 |
|---|---|---|
| committer | Roland Dreier <roland@purestorage.com> | 2011-10-31 22:34:53 +0400 |
| commit | 581bbe2cd0694a935e0c3ccd7f011e10094f1df6 (patch) | |
| tree | 38e536efa0d05d76964b09836def2210a00b41b5 /drivers/infiniband/hw/cxgb4/ev.c | |
| parent | e14d62c05c0b8eff61c6fd46b4a78fb27c8cf38b (diff) | |
| download | linux-581bbe2cd0694a935e0c3ccd7f011e10094f1df6.tar.xz | |
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/ev.c')
| -rw-r--r-- | drivers/infiniband/hw/cxgb4/ev.c | 10 |
1 files changed, 8 insertions, 2 deletions
diff --git a/drivers/infiniband/hw/cxgb4/ev.c b/drivers/infiniband/hw/cxgb4/ev.c index c13041a0aeba..397cb36cf103 100644 --- a/drivers/infiniband/hw/cxgb4/ev.c +++ b/drivers/infiniband/hw/cxgb4/ev.c @@ -42,6 +42,7 @@ static void post_qp_event(struct c4iw_dev *dev, struct c4iw_cq *chp, { struct ib_event event; struct c4iw_qp_attributes attrs; + unsigned long flag; if ((qhp->attr.state == C4IW_QP_STATE_ERROR) || (qhp->attr.state == C4IW_QP_STATE_TERMINATE)) { @@ -72,7 +73,9 @@ static void post_qp_event(struct c4iw_dev *dev, struct c4iw_cq *chp, if (qhp->ibqp.event_handler) (*qhp->ibqp.event_handler)(&event, qhp->ibqp.qp_context); + spin_lock_irqsave(&chp->comp_handler_lock, flag); (*chp->ibcq.comp_handler)(&chp->ibcq, chp->ibcq.cq_context); + spin_unlock_irqrestore(&chp->comp_handler_lock, flag); } void c4iw_ev_dispatch(struct c4iw_dev *dev, struct t4_cqe *err_cqe) @@ -183,11 +186,14 @@ out: int c4iw_ev_handler(struct c4iw_dev *dev, u32 qid) { struct c4iw_cq *chp; + unsigned long flag; chp = get_chp(dev, qid); - if (chp) + if (chp) { + spin_lock_irqsave(&chp->comp_handler_lock, flag); (*chp->ibcq.comp_handler)(&chp->ibcq, chp->ibcq.cq_context); - else + spin_unlock_irqrestore(&chp->comp_handler_lock, flag); + } else PDBG("%s unknown cqid 0x%x\n", __func__, qid); return 0; } |