Skip to content

Commit f65e57e

Browse files
mikijoysmb49
authored andcommitted
RDMA/mlx5: Fix cache entry update on dereg error
BugLink: https://bugs.launchpad.net/bugs/2122072 [ Upstream commit 24d693cf6c89d216a68634d44fa93e4400775d94 ] Fix double decrement of 'in_use' counter on push_mkey_locked() failure while deregistering an MR. If we fail to return an mkey to the cache in cache_ent_find_and_store() it'll update the 'in_use' counter. Its caller, revoke_mr(), also updates it, thus having double decrement. Wrong value of 'in_use' counter will be exposed through debugfs and can also cause wrong resizing of the cache when users try to set cache entry size using the 'size' debugfs. To address this issue, the 'in_use' counter is now decremented within mlx5_revoke_mr() also after a successful call to cache_ent_find_and_store() and not within cache_ent_find_and_store(). Other success or failure flows remains unchanged where it was also decremented. Fixes: 8c1185f ("RDMA/mlx5: Change check for cacheable mkeys") Signed-off-by: Michael Guralnik <michaelgur@nvidia.com> Reviewed-by: Yishai Hadas <yishaih@nvidia.com> Link: https://patch.msgid.link/97e979dff636f232ff4c83ce709c17c727da1fdb.1741875692.git.leon@kernel.org Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Noah Wager <noah.wager@canonical.com> Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
1 parent d741c27 commit f65e57e

File tree

1 file changed

+3
-1
lines changed
  • drivers/infiniband/hw/mlx5

1 file changed

+3
-1
lines changed

drivers/infiniband/hw/mlx5/mr.c

Lines changed: 3 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1968,7 +1968,6 @@ static int cache_ent_find_and_store(struct mlx5_ib_dev *dev,
19681968

19691969
if (mr->mmkey.cache_ent) {
19701970
spin_lock_irq(&mr->mmkey.cache_ent->mkeys_queue.lock);
1971-
mr->mmkey.cache_ent->in_use--;
19721971
goto end;
19731972
}
19741973

@@ -2036,6 +2035,7 @@ static int mlx5_revoke_mr(struct mlx5_ib_mr *mr)
20362035
bool is_odp = is_odp_mr(mr);
20372036
bool is_odp_dma_buf = is_dmabuf_mr(mr) &&
20382037
!to_ib_umem_dmabuf(mr->umem)->pinned;
2038+
bool from_cache = !!ent;
20392039
int ret = 0;
20402040

20412041
if (is_odp)
@@ -2048,6 +2048,8 @@ static int mlx5_revoke_mr(struct mlx5_ib_mr *mr)
20482048
ent = mr->mmkey.cache_ent;
20492049
/* upon storing to a clean temp entry - schedule its cleanup */
20502050
spin_lock_irq(&ent->mkeys_queue.lock);
2051+
if (from_cache)
2052+
ent->in_use--;
20512053
if (ent->is_tmp && !ent->tmp_cleanup_scheduled) {
20522054
mod_delayed_work(ent->dev->cache.wq, &ent->dwork,
20532055
msecs_to_jiffies(30 * 1000));

0 commit comments

Comments
 (0)