IB/hfi1: Drop stale TID RDMA packets that cause TIDErr
authorKaike Wan <kaike.wan@intel.com>
Thu, 15 Aug 2019 19:20:58 +0000 (15:20 -0400)
committerDoug Ledford <dledford@redhat.com>
Tue, 20 Aug 2019 17:44:44 +0000 (13:44 -0400)
commitd9d1f5e7bb82415591e8b62b222cbb88c4797ef3
treef951837fc96bb5563cabebb442b9a75f9061fcac
parent90fdae66e72bf0381d168f12dca0259617927895
IB/hfi1: Drop stale TID RDMA packets that cause TIDErr

In a congested fabric with adaptive routing enabled, traces show that
packets could be delivered out of order. A stale TID RDMA data packet
could lead to TidErr if the TID entries have been released by duplicate
data packets generated from retries, and subsequently erroneously force
the qp into error state in the current implementation.

Since the payload has already been dropped by hardware, the packet can
be simply dropped and it is no longer necessary to put the qp into
error state.

Fixes: 9905bf06e890 ("IB/hfi1: Add functions to receive TID RDMA READ response")
Cc: <stable@vger.kernel.org>
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Kaike Wan <kaike.wan@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Link: https://lore.kernel.org/r/20190815192058.105923.72324.stgit@awfm-01.aw.intel.com
Signed-off-by: Doug Ledford <dledford@redhat.com>
drivers/infiniband/hw/hfi1/tid_rdma.c