RDMA/ucma: Add missing locking around rdma_leave_multicast()
authorJason Gunthorpe <jgg@nvidia.com>
Tue, 18 Aug 2020 12:05:22 +0000 (15:05 +0300)
committerJason Gunthorpe <jgg@nvidia.com>
Thu, 27 Aug 2020 11:38:15 +0000 (08:38 -0300)
commit38e03d092699891c3237b5aee9e8029d4ede0956
tree5e02d9e4fe7103b50cc68f3956ac45d208a8b7c4
parent98837c6c3d7285f6eca86480b6f7fac6880e27a8
RDMA/ucma: Add missing locking around rdma_leave_multicast()

All entry points to the rdma_cm from a ULP must be single threaded,
even this error unwinds. Add the missing locking.

Fixes: 7c11910783a1 ("RDMA/ucma: Put a lock around every call to the rdma_cm layer")
Link: https://lore.kernel.org/r/20200818120526.702120-11-leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
drivers/infiniband/core/ucma.c