]> git.baikalelectronics.ru Git - kernel.git/commit
gfs2: Don't call dlm after protocol is unmounted
authorBob Peterson <rpeterso@redhat.com>
Fri, 30 Jul 2021 17:41:49 +0000 (12:41 -0500)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 22 Sep 2021 10:26:33 +0000 (12:26 +0200)
commit39d8f2bc977f425ff803fa1fffb4cf4579bed9b8
tree4d191f19e0fbf0aed6e4368749f7681830cf9573
parent423cd36992b77ce496eb6642a05f1efdab196cfd
gfs2: Don't call dlm after protocol is unmounted

[ Upstream commit d1340f80f0b8066321b499a376780da00560e857 ]

In the gfs2 withdraw sequence, the dlm protocol is unmounted with a call
to lm_unmount. After a withdraw, users are allowed to unmount the
withdrawn file system. But at that point we may still have glocks left
over that we need to free via unmount's call to gfs2_gl_hash_clear.
These glocks may have never been completed because of whatever problem
caused the withdraw (IO errors or whatever).

Before this patch, function gdlm_put_lock would still try to call into
dlm to unlock these leftover glocks, which resulted in dlm returning
-EINVAL because the lock space was abandoned. These glocks were never
freed because there was no mechanism after that to free them.

This patch adds a check to gdlm_put_lock to see if the locking protocol
was inactive (DFL_UNMOUNT flag) and if so, free the glock and not
make the invalid call into dlm.

I could have combined this "if" with the one that follows, related to
leftover glock LVBs, but I felt the code was more readable with its own
if clause.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
fs/gfs2/lock_dlm.c