diff options
| author | Ira Weiny | 2015-05-13 20:02:57 -0400 |
|---|---|---|
| committer | Doug Ledford | 2015-05-20 12:37:34 -0400 |
| commit | 26c454288a4beac774ea31c15284783fcd75721d (patch) | |
| tree | 1210551c9c4a91eac0bcbffff506c7246db392ac | |
| parent | ab8be619b81a0d83706294bd791407e0f497d646 (diff) | |
IB/user_mad: Fix buggy usage of port index
The addition of the rdma_cap_ib_mad is technically broken in ib_umad_remove_one
because the loop "i" value is not a port value.
This bug resulted in the ib_umad failing to properly remove its resources when
the core capability functions were converted to bit fields.
NOTE: e17371d73908 did not result in broken behavior on its own. It was only
an issue when the implementation of rdma_cap_ib_mad was changed.
Pass the port value to rdma_cap_ib_mad.
Fixes: e17371d73908 ("IB/Verbs: Use management helper rdma_cap_ib_mad()")
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
| -rw-r--r-- | drivers/infiniband/core/user_mad.c | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/drivers/infiniband/core/user_mad.c b/drivers/infiniband/core/user_mad.c index cc5001e850de..278cfaee9a94 100644 --- a/drivers/infiniband/core/user_mad.c +++ b/drivers/infiniband/core/user_mad.c @@ -1325,7 +1325,7 @@ static void ib_umad_remove_one(struct ib_device *device) return; for (i = 0; i <= rdma_end_port(device) - rdma_start_port(device); ++i) { - if (rdma_cap_ib_mad(device, i)) + if (rdma_cap_ib_mad(device, i + rdma_start_port(device))) ib_umad_kill_port(&umad_dev->port[i]); } |