diff options
| author | Jason Gunthorpe <jgg@mellanox.com> | 2019-08-19 14:17:05 +0300 |
|---|---|---|
| committer | Jason Gunthorpe <jgg@mellanox.com> | 2019-08-21 14:08:42 -0300 |
| commit | 0446cad9ca385c3b8d6e5a1184e59650fa7a7a6d (patch) | |
| tree | 24c7fd844838087c179c73c4bb6114f524f7d7f8 /drivers/infiniband/hw/mlx5/mr.c | |
| parent | 261dc53f8ee037bf2fbf68f90c319b04062a126c (diff) | |
| download | linux-0446cad9ca385c3b8d6e5a1184e59650fa7a7a6d.tar.bz2 | |
RDMA/odp: Provide ib_umem_odp_release() to undo the allocs
Now that there are allocator APIs that return the ib_umem_odp directly
it should be freed through a umem_odp free'er as well.
Link: https://lore.kernel.org/r/20190819111710.18440-8-leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Diffstat (limited to 'drivers/infiniband/hw/mlx5/mr.c')
| -rw-r--r-- | drivers/infiniband/hw/mlx5/mr.c | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c index fc1106ddc3f6..b7da619614e4 100644 --- a/drivers/infiniband/hw/mlx5/mr.c +++ b/drivers/infiniband/hw/mlx5/mr.c @@ -1629,7 +1629,7 @@ static void dereg_mr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr) * so that there will not be any invalidations in * flight, looking at the *mr struct. */ - ib_umem_release(umem); + ib_umem_odp_release(umem_odp); atomic_sub(npages, &dev->mdev->priv.reg_pages); /* Avoid double-freeing the umem. */ |