diff options
author | Jan Kara <jack@suse.cz> | 2006-03-23 03:00:17 -0800 |
---|---|---|
committer | Linus Torvalds <torvalds@g5.osdl.org> | 2006-03-23 07:38:10 -0800 |
commit | 6362e4d4eda61efb04ac1cdae32e48ac6d90b701 (patch) | |
tree | be13a8b673d117ec3d186f2911f4ebae3da857dd /kernel/cpuset.c | |
parent | 804f1594cc3deb161e531a43d90c501f0db2635a (diff) | |
download | linux-6362e4d4eda61efb04ac1cdae32e48ac6d90b701.tar.gz linux-6362e4d4eda61efb04ac1cdae32e48ac6d90b701.tar.bz2 linux-6362e4d4eda61efb04ac1cdae32e48ac6d90b701.zip |
[PATCH] Fix oops in invalidate_dquots()
When quota is being turned off we assumed that all the references to dquots
were already dropped. That need not be true as inodes being deleted are
not on superblock's inodes list and hence we need not reach it when
removing quota references from inodes. So invalidate_dquots() has to wait
for all the users of dquots (as quota is already marked as turned off, no
new references can be acquired and so this is bound to happen rather
early). When we do this, we can also remove the iprune_sem locking as it
was protecting us against exactly the same problem when freeing inodes
icache memory.
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Diffstat (limited to 'kernel/cpuset.c')
0 files changed, 0 insertions, 0 deletions