From: Jan Kara <jack@suse.cz>
To: <linux-fsdevel@vger.kernel.org>
Cc: Wang Shilong <wangshilong1991@gmail.com>, Jan Kara <jack@suse.cz>
Subject: [PATCH 16/27] quota: Fix possible corruption of dqi_flags
Date: Wed, 16 Aug 2017 17:41:16 +0200 [thread overview]
Message-ID: <20170816154127.7048-17-jack@suse.cz> (raw)
In-Reply-To: <20170816154127.7048-1-jack@suse.cz>
dqi_flags modifications are protected by dq_data_lock. However the
modifications in vfs_load_quota_inode() and in mark_info_dirty() were
not which could lead to corruption of dqi_flags. Since modifications to
dqi_flags are rare, this is hard to observe in practice but in theory it
could happen. Fix the problem by always using dq_data_lock for
protection.
Signed-off-by: Jan Kara <jack@suse.cz>
---
fs/quota/dquot.c | 9 +++++++--
1 file changed, 7 insertions(+), 2 deletions(-)
diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c
index 5e77c4da69a6..e1a155e8db15 100644
--- a/fs/quota/dquot.c
+++ b/fs/quota/dquot.c
@@ -389,7 +389,9 @@ static inline int clear_dquot_dirty(struct dquot *dquot)
void mark_info_dirty(struct super_block *sb, int type)
{
- set_bit(DQF_INFO_DIRTY_B, &sb_dqopt(sb)->info[type].dqi_flags);
+ spin_lock(&dq_data_lock);
+ sb_dqopt(sb)->info[type].dqi_flags |= DQF_INFO_DIRTY;
+ spin_unlock(&dq_data_lock);
}
EXPORT_SYMBOL(mark_info_dirty);
@@ -2316,8 +2318,11 @@ static int vfs_load_quota_inode(struct inode *inode, int type, int format_id,
error = dqopt->ops[type]->read_file_info(sb, type);
if (error < 0)
goto out_file_init;
- if (dqopt->flags & DQUOT_QUOTA_SYS_FILE)
+ if (dqopt->flags & DQUOT_QUOTA_SYS_FILE) {
+ spin_lock(&dq_data_lock);
dqopt->info[type].dqi_flags |= DQF_SYS_FILE;
+ spin_unlock(&dq_data_lock);
+ }
spin_lock(&dq_state_lock);
dqopt->flags |= dquot_state_flag(flags, type);
spin_unlock(&dq_state_lock);
--
2.12.3
next prev parent reply other threads:[~2017-08-16 15:41 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-08-16 15:41 [PATCH 0/27 v1] Quota scalability patches Jan Kara
2017-08-16 15:41 ` [PATCH 01/27] quota: Convert dqio_mutex to rwsem Jan Kara
2017-08-16 16:23 ` Andreas Dilger
2017-08-17 16:47 ` Jan Kara
2017-08-16 15:41 ` [PATCH 02/27] quota: Do more fine-grained locking in dquot_acquire() Jan Kara
2017-08-16 16:35 ` Andreas Dilger
2017-08-17 16:58 ` Jan Kara
2017-08-16 15:41 ` [PATCH 03/27] quota: Acquire dqio_sem for reading in dquot_get_next_id() Jan Kara
2017-08-16 16:37 ` Andreas Dilger
2017-08-16 15:41 ` [PATCH 04/27] quota: Acquire dqio_sem for reading in vfs_load_quota_inode() Jan Kara
2017-08-16 16:38 ` Andreas Dilger
2017-08-16 15:41 ` [PATCH 05/27] quota: Protect dquot writeout with dq_lock Jan Kara
2017-08-16 16:44 ` Andreas Dilger
2017-08-16 15:41 ` [PATCH 06/27] quota: Push dqio_sem down to ->read_dqblk() Jan Kara
2017-08-16 16:46 ` Andreas Dilger
2017-08-16 15:41 ` [PATCH 07/27] quota: Remove locking for reading from the old quota format Jan Kara
2017-08-16 16:47 ` Andreas Dilger
2017-08-16 15:41 ` [PATCH 08/27] quota: Push dqio_sem down to ->write_dqblk() Jan Kara
2017-08-16 16:48 ` Andreas Dilger
2017-08-16 15:41 ` [PATCH 09/27] quota: Do not acquire dqio_sem for dquot overwrites in v2 format Jan Kara
2017-08-16 16:49 ` Andreas Dilger
2017-08-16 15:41 ` [PATCH 10/27] quota: Remove locking for writing to the old quota format Jan Kara
2017-08-16 16:50 ` Andreas Dilger
2017-08-16 15:41 ` [PATCH 11/27] quota: Push dqio_sem down to ->release_dqblk() Jan Kara
2017-08-16 16:56 ` Andreas Dilger
2017-08-16 15:41 ` [PATCH 12/27] quota: Push dqio_sem down to ->get_next_id() Jan Kara
2017-08-16 17:08 ` Andreas Dilger
2017-08-17 17:09 ` Jan Kara
2017-08-16 15:41 ` [PATCH 13/27] quota: Push dqio_sem down to ->write_file_info() Jan Kara
2017-08-16 17:33 ` Andreas Dilger
2017-08-17 17:13 ` Jan Kara
2017-08-16 15:41 ` [PATCH 14/27] quota: Push dqio_sem down to ->read_file_info() Jan Kara
2017-08-16 17:57 ` Andreas Dilger
2017-08-16 15:41 ` [PATCH 15/27] quota: Fix error codes in v2_read_file_info() Jan Kara
2017-08-16 18:00 ` Andreas Dilger
2017-08-16 15:41 ` Jan Kara [this message]
2017-08-16 18:14 ` [PATCH 16/27] quota: Fix possible corruption of dqi_flags Andreas Dilger
2017-08-16 15:41 ` [PATCH 17/27] quota: Drop return value of mark_all_dquot_dirty() Jan Kara
2017-08-16 15:41 ` [PATCH 18/27] quota: Do not dirty bad dquots Jan Kara
2017-08-16 15:41 ` [PATCH 19/27] quota: Move locking into clear_dquot_dirty() Jan Kara
2017-08-16 18:29 ` Andreas Dilger
2017-08-16 15:41 ` [PATCH 20/27] quota: Remove dq_wait_unused from dquot Jan Kara
2017-08-16 15:41 ` [PATCH 21/27] quota: Allow disabling tracking of dirty dquots in a list Jan Kara
2017-08-16 15:41 ` [PATCH 22/27] ext4: Disable dirty list tracking of dquots when journalling quotas Jan Kara
2017-08-16 15:41 ` [PATCH 23/27] quota: Inline functions into their callsites Jan Kara
2017-08-16 15:41 ` [PATCH 24/27] quota: Inline inode_{incr,decr}_space() into callsites Jan Kara
2017-08-16 15:41 ` [PATCH 25/27] quota: Inline dquot_[re]claim_reserved_space() into callsite Jan Kara
2017-08-16 19:52 ` Andreas Dilger
2017-08-16 15:41 ` [PATCH 26/27] fs: Provide __inode_get_bytes() Jan Kara
2017-08-16 16:12 ` Andreas Dilger
2017-08-17 20:04 ` Jan Kara
2017-08-16 15:41 ` [PATCH 27/27] quota: Reduce contention on dq_data_lock Jan Kara
2017-08-16 20:17 ` Andreas Dilger
2017-08-17 20:08 ` Jan Kara
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20170816154127.7048-17-jack@suse.cz \
--to=jack@suse.cz \
--cc=linux-fsdevel@vger.kernel.org \
--cc=wangshilong1991@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).