From: Jan Kara <jack@suse.cz>
To: Andrew Morton <akpm@linux-foundation.org>
Cc: linux-fsdevel@vger.kernel.org, Jan Kara <jack@suse.cz>
Subject: [PATCH 10/14] quota: Introduce DQUOT_QUOTA_SYS_FILE flag
Date: Mon, 27 Oct 2008 13:11:48 +0100 [thread overview]
Message-ID: <12251095131431-git-send-email-jack@suse.cz> (raw)
In-Reply-To: <12251095133281-git-send-email-jack@suse.cz>
If filesystem can handle quota files as system files hidden from users, we can
skip a lot of cache invalidation, syncing, inode flags setting etc. when
turning quotas on, off and quota_sync. Allow filesystem to indicate that it is
hiding quota files from users by DQUOT_QUOTA_SYS_FILE flag.
Signed-off-by: Jan Kara <jack@suse.cz>
---
fs/dquot.c | 45 ++++++++++++++++++++++++++++++---------------
fs/quota.c | 3 +++
include/linux/quota.h | 7 +++++++
3 files changed, 40 insertions(+), 15 deletions(-)
diff --git a/fs/dquot.c b/fs/dquot.c
index 4bedea9..45ac11c 100644
--- a/fs/dquot.c
+++ b/fs/dquot.c
@@ -1625,6 +1625,11 @@ int vfs_quota_disable(struct super_block *sb, int type, unsigned int flags)
dqopt->ops[cnt] = NULL;
}
mutex_unlock(&dqopt->dqonoff_mutex);
+
+ /* Skip syncing and setting flags if quota files are hidden */
+ if (dqopt->flags & DQUOT_QUOTA_SYS_FILE)
+ goto put_inodes;
+
/* Sync the superblock so that buffers with quota data are written to
* disk (and so userspace sees correct data afterwards). */
if (sb->s_op->sync_fs)
@@ -1649,6 +1654,12 @@ int vfs_quota_disable(struct super_block *sb, int type, unsigned int flags)
mark_inode_dirty(toputinode[cnt]);
}
mutex_unlock(&dqopt->dqonoff_mutex);
+ }
+ if (sb->s_bdev)
+ invalidate_bdev(sb->s_bdev);
+put_inodes:
+ for (cnt = 0; cnt < MAXQUOTAS; cnt++)
+ if (toputinode[cnt]) {
/* On remount RO, we keep the inode pointer so that we
* can reenable quota on the subsequent remount RW. We
* have to check 'flags' variable and not use sb_has_
@@ -1661,8 +1672,6 @@ int vfs_quota_disable(struct super_block *sb, int type, unsigned int flags)
else if (!toputinode[cnt]->i_nlink)
ret = -EBUSY;
}
- if (sb->s_bdev)
- invalidate_bdev(sb->s_bdev);
return ret;
}
@@ -1709,25 +1718,31 @@ static int vfs_load_quota_inode(struct inode *inode, int type, int format_id,
goto out_fmt;
}
- /* As we bypass the pagecache we must now flush the inode so that
- * we see all the changes from userspace... */
- write_inode_now(inode, 1);
- /* And now flush the block cache so that kernel sees the changes */
- invalidate_bdev(sb->s_bdev);
+ if (!(dqopt->flags & DQUOT_QUOTA_SYS_FILE)) {
+ /* As we bypass the pagecache we must now flush the inode so
+ * that we see all the changes from userspace... */
+ write_inode_now(inode, 1);
+ /* And now flush the block cache so that kernel sees the
+ * changes */
+ invalidate_bdev(sb->s_bdev);
+ }
mutex_lock(&inode->i_mutex);
mutex_lock(&dqopt->dqonoff_mutex);
if (sb_has_quota_loaded(sb, type)) {
error = -EBUSY;
goto out_lock;
}
- /* We don't want quota and atime on quota files (deadlocks possible)
- * Also nobody should write to the file - we use special IO operations
- * which ignore the immutable bit. */
- down_write(&dqopt->dqptr_sem);
- oldflags = inode->i_flags & (S_NOATIME | S_IMMUTABLE | S_NOQUOTA);
- inode->i_flags |= S_NOQUOTA | S_NOATIME | S_IMMUTABLE;
- up_write(&dqopt->dqptr_sem);
- sb->dq_op->drop(inode);
+
+ if (!(dqopt->flags & DQUOT_QUOTA_SYS_FILE)) {
+ /* We don't want quota and atime on quota files (deadlocks
+ * possible) Also nobody should write to the file - we use
+ * special IO operations which ignore the immutable bit. */
+ down_write(&dqopt->dqptr_sem);
+ oldflags = inode->i_flags & (S_NOATIME | S_IMMUTABLE | S_NOQUOTA);
+ inode->i_flags |= S_NOQUOTA | S_NOATIME | S_IMMUTABLE;
+ up_write(&dqopt->dqptr_sem);
+ sb->dq_op->drop(inode);
+ }
error = -EIO;
dqopt->files[type] = igrab(inode);
diff --git a/fs/quota.c b/fs/quota.c
index a8026f1..2c6ea78 100644
--- a/fs/quota.c
+++ b/fs/quota.c
@@ -160,6 +160,9 @@ static void quota_sync_sb(struct super_block *sb, int type)
int cnt;
sb->s_qcop->quota_sync(sb, type);
+
+ if (sb_dqopt(sb)->flags & DQUOT_QUOTA_SYS_FILE)
+ return;
/* This is not very clever (and fast) but currently I don't know about
* any other simple way of getting quota data to disk and we must get
* them there for userspace to be visible... */
diff --git a/include/linux/quota.h b/include/linux/quota.h
index 6d4f8b0..c63f29b 100644
--- a/include/linux/quota.h
+++ b/include/linux/quota.h
@@ -332,6 +332,13 @@ enum {
#define DQUOT_SUSPENDED (1 << _DQUOT_SUSPENDED)
#define DQUOT_STATE_FLAGS (DQUOT_USAGE_ENABLED | DQUOT_LIMITS_ENABLED | \
DQUOT_SUSPENDED)
+/* Other quota flags */
+#define DQUOT_QUOTA_SYS_FILE (1 << 6) /* Quota file is a special
+ * system file and user cannot
+ * touch it. Filesystem is
+ * responsible for setting
+ * S_NOQUOTA, S_NOATIME flags
+ */
static inline unsigned int dquot_state_flag(unsigned int flags, int type)
{
--
1.5.6
next prev parent reply other threads:[~2008-10-27 12:11 UTC|newest]
Thread overview: 33+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-10-27 12:11 [PATCH 00/14] 64-bit quotas and preparations for OCFS2 quotas Jan Kara
2008-10-27 12:11 ` [PATCH 01/14] quota: Add callbacks for allocating and destroying dquot structures Jan Kara
2008-10-27 12:11 ` [PATCH 02/14] quota: Increase size of variables for limits and inode usage Jan Kara
2008-10-27 12:11 ` [PATCH 03/14] quota: Remove bogus 'optimization' in check_idq() and check_bdq() Jan Kara
2008-10-27 12:11 ` [PATCH 04/14] quota: Make _SUSPENDED just a flag Jan Kara
2008-10-27 12:11 ` [PATCH 05/14] quota: Allow to separately enable quota accounting and enforcing limits Jan Kara
2008-10-27 12:11 ` [PATCH 06/14] ext3: Use sb_any_quota_loaded() instead of sb_any_quota_enabled() Jan Kara
2008-10-27 12:11 ` [PATCH 07/14] ext4: " Jan Kara
2008-10-27 12:11 ` [PATCH 08/14] reiserfs: " Jan Kara
2008-10-27 12:11 ` [PATCH 09/14] quota: Remove compatibility function sb_any_quota_enabled() Jan Kara
2008-10-27 12:11 ` Jan Kara [this message]
2008-10-27 12:11 ` [PATCH 11/14] quota: Move quotaio_v[12].h from include/linux/ to fs/ Jan Kara
2008-10-27 12:11 ` [PATCH 12/14] quota: Split off quota tree handling into a separate file Jan Kara
2008-10-27 12:11 ` [PATCH 13/14] quota: Convert union in mem_dqinfo to a pointer Jan Kara
2008-10-27 12:11 ` [PATCH 14/14] quota: Support 64-bit quota format Jan Kara
2008-10-28 5:29 ` Andrew Morton
2008-10-29 2:39 ` Jan Kara
2008-10-29 0:41 ` Mingming Cao
2008-10-29 1:02 ` Andrew Morton
2008-10-29 21:31 ` Mingming Cao
2008-10-28 5:55 ` [PATCH 13/14] quota: Convert union in mem_dqinfo to a pointer Andrew Morton
2008-10-29 2:40 ` Jan Kara
2008-10-28 5:52 ` [PATCH 12/14] quota: Split off quota tree handling into a separate file Andrew Morton
2008-10-29 2:43 ` Jan Kara
2008-10-28 22:32 ` [PATCH 11/14] quota: Move quotaio_v[12].h from include/linux/ to fs/ Andrew Morton
2008-10-29 3:21 ` Jan Kara
2008-10-28 5:44 ` [PATCH 10/14] quota: Introduce DQUOT_QUOTA_SYS_FILE flag Andrew Morton
2008-10-28 17:51 ` Joel Becker
2008-10-28 5:40 ` [PATCH 02/14] quota: Increase size of variables for limits and inode usage Andrew Morton
2008-10-28 9:11 ` Boaz Harrosh
2008-10-29 3:01 ` Jan Kara
2008-10-28 5:33 ` [PATCH 01/14] quota: Add callbacks for allocating and destroying dquot structures Andrew Morton
2008-10-29 3:05 ` 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=12251095131431-git-send-email-jack@suse.cz \
--to=jack@suse.cz \
--cc=akpm@linux-foundation.org \
--cc=linux-fsdevel@vger.kernel.org \
/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).