diff options
| author | Steve Wise <swise@opengridcomputing.com> | 2017-11-30 18:41:56 +0100 |
|---|---|---|
| committer | Jason Gunthorpe <jgg@mellanox.com> | 2017-12-07 22:09:59 +0100 |
| commit | 335ebf6fa35ca1c59b73f76fad19b249d3550e86 (patch) | |
| tree | 1f25f1ebd3cc579081cc66861f19b0ffee31d3ee | |
| parent | RDMA/netlink: Fix general protection fault (diff) | |
| download | linux-335ebf6fa35ca1c59b73f76fad19b249d3550e86.tar.xz linux-335ebf6fa35ca1c59b73f76fad19b249d3550e86.zip | |
iw_cxgb4: only clear the ARMED bit if a notification is needed
In __flush_qp(), the CQ ARMED bit was being cleared regardless of
whether any notification is actually needed. This resulted in the iser
termination logic getting stuck in ib_drain_sq() because the CQ was not
marked ARMED and thus the drain CQE notification wasn't triggered.
This new bug was exposed when this commit was merged:
commit cbb40fadd31c ("iw_cxgb4: only call the cq comp_handler when the
cq is armed")
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
| -rw-r--r-- | drivers/infiniband/hw/cxgb4/qp.c | 8 |
1 files changed, 4 insertions, 4 deletions
diff --git a/drivers/infiniband/hw/cxgb4/qp.c b/drivers/infiniband/hw/cxgb4/qp.c index 5ee7fe433136..355e288ec969 100644 --- a/drivers/infiniband/hw/cxgb4/qp.c +++ b/drivers/infiniband/hw/cxgb4/qp.c @@ -1285,21 +1285,21 @@ static void __flush_qp(struct c4iw_qp *qhp, struct c4iw_cq *rchp, spin_unlock_irqrestore(&rchp->lock, flag); if (schp == rchp) { - if (t4_clear_cq_armed(&rchp->cq) && - (rq_flushed || sq_flushed)) { + if ((rq_flushed || sq_flushed) && + t4_clear_cq_armed(&rchp->cq)) { spin_lock_irqsave(&rchp->comp_handler_lock, flag); (*rchp->ibcq.comp_handler)(&rchp->ibcq, rchp->ibcq.cq_context); spin_unlock_irqrestore(&rchp->comp_handler_lock, flag); } } else { - if (t4_clear_cq_armed(&rchp->cq) && rq_flushed) { + if (rq_flushed && t4_clear_cq_armed(&rchp->cq)) { spin_lock_irqsave(&rchp->comp_handler_lock, flag); (*rchp->ibcq.comp_handler)(&rchp->ibcq, rchp->ibcq.cq_context); spin_unlock_irqrestore(&rchp->comp_handler_lock, flag); } - if (t4_clear_cq_armed(&schp->cq) && sq_flushed) { + if (sq_flushed && t4_clear_cq_armed(&schp->cq)) { spin_lock_irqsave(&schp->comp_handler_lock, flag); (*schp->ibcq.comp_handler)(&schp->ibcq, schp->ibcq.cq_context); |