[ 14/91] RDMA/cxgb4: Always wake up waiters in c4iw_peer_abort_intr()
From: Greg KH <gregkh@linuxfoundation.org>
Date: 2012-05-27 01:33:20
Also in:
lkml
From: Greg KH <gregkh@linuxfoundation.org>
Date: 2012-05-27 01:33:20
Also in:
lkml
3.4-stable review patch. If anyone has any objections, please let me know. ------------------ From: Steve Wise <redacted> commit 0f1dcfae6bc5563424346ad3a03282b8235a4c33 upstream. 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 <redacted> Signed-off-by: Roland Dreier <redacted> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> --- drivers/infiniband/hw/cxgb4/cm.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-)
--- a/drivers/infiniband/hw/cxgb4/cm.c
+++ b/drivers/infiniband/hw/cxgb4/cm.c@@ -2667,11 +2667,8 @@ static int peer_abort_intr(struct c4iw_d /* * Wake up any threads in rdma_init() or rdma_fini(). - * However, this is not needed if com state is just - * MPA_REQ_SENT */ - if (ep->com.state != MPA_REQ_SENT) - c4iw_wake_up(&ep->com.wr_wait, -ECONNRESET); + c4iw_wake_up(&ep->com.wr_wait, -ECONNRESET); sched(dev, skb); return 0; }