From: Kaike Wan <kaike....@intel.com>
commit 35d5c8b82e2c32e8e29ca195bb4dac60ba7d97fc upstream.
When processing a TID RDMA READ RESP packet that causes KDETH EFLAGS
errors, the packet's IB PSN is checked against qp->s_last_psn and
qp->s_psn without the protection of qp->s_lock, which is not safe.
This patch fixes the issue by acquiring qp->s_lock first.
Fixes: 9905bf06e890 ("IB/hfi1: Add functions to receive TID RDMA READ response")
Cc: <sta...@vger.kernel.org>
Reviewed-by: Mike Marciniszyn <mike.marcinis...@intel.com>
Signed-off-by: Kaike Wan <kaike....@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessan...@intel.com>
Link:
https://lore.kernel.org/r/20190815192039.105923.7852.st...@awfm-01.aw.intel.com
Signed-off-by: Doug Ledford <dledf...@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gre...@linuxfoundation.org>
---
drivers/infiniband/hw/hfi1/tid_rdma.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
--- a/drivers/infiniband/hw/hfi1/tid_rdma.c
+++ b/drivers/infiniband/hw/hfi1/tid_rdma.c
@@ -2689,12 +2689,12 @@ static bool handle_read_kdeth_eflags(str
u32 fpsn;
lockdep_assert_held(&qp->r_lock);
+ spin_lock(&qp->s_lock);
/* If the psn is out of valid range, drop the packet */
if (cmp_psn(ibpsn, qp->s_last_psn) < 0 ||
cmp_psn(ibpsn, qp->s_psn) > 0)
- return ret;
+ goto s_unlock;
- spin_lock(&qp->s_lock);
/*
* Note that NAKs implicitly ACK outstanding SEND and RDMA write
* requests and implicitly NAK RDMA read and atomic requests issued
