From: "Darrick J. Wong" <djwong@kernel.org>
To: Christoph Hellwig <hch@lst.de>
Cc: Carlos Maiolino <cem@kernel.org>, linux-xfs@vger.kernel.org
Subject: Re: [PATCH 17/17] xfs: reduce ilock roundtrips in xfs_qm_vop_dqalloc
Date: Wed, 15 Oct 2025 14:27:07 -0700 [thread overview]
Message-ID: <20251015212707.GM2591640@frogsfrogsfrogs> (raw)
In-Reply-To: <20251013024851.4110053-18-hch@lst.de>
On Mon, Oct 13, 2025 at 11:48:18AM +0900, Christoph Hellwig wrote:
> xfs_qm_vop_dqalloc only needs the (exclusive) ilock for attaching dquots
> to the inode if not done so yet. All the other locks don't touch the inode
> and don't need the ilock - the i_rwsem / iolock protects against changes
> to the IDs while we are in a method, and the ilock would not help because
> dropping it for the dqget calls would be racy anyway.
...and I guess we no longer detach dquots from live inodes now, so we
really only need ILOCK_EXCL to prevent multiple threads from trying to
allocate and attach a new xfs_dquot object to the same inode, right?
Changing the i_dquot pointers (aka chown/chproj) is what's coordinated
under the iolock, right?
--D
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
> fs/xfs/xfs_qm.c | 32 +++-----------------------------
> 1 file changed, 3 insertions(+), 29 deletions(-)
>
> diff --git a/fs/xfs/xfs_qm.c b/fs/xfs/xfs_qm.c
> index 7fbb89fcdeb9..336de0479022 100644
> --- a/fs/xfs/xfs_qm.c
> +++ b/fs/xfs/xfs_qm.c
> @@ -1861,16 +1861,12 @@ xfs_qm_vop_dqalloc(
> struct xfs_dquot *gq = NULL;
> struct xfs_dquot *pq = NULL;
> int error;
> - uint lockflags;
>
> if (!XFS_IS_QUOTA_ON(mp))
> return 0;
>
> ASSERT(!xfs_is_metadir_inode(ip));
>
> - lockflags = XFS_ILOCK_EXCL;
> - xfs_ilock(ip, lockflags);
> -
> if ((flags & XFS_QMOPT_INHERIT) && XFS_INHERIT_GID(ip))
> gid = inode->i_gid;
>
> @@ -1879,37 +1875,22 @@ xfs_qm_vop_dqalloc(
> * if necessary. The dquot(s) will not be locked.
> */
> if (XFS_NOT_DQATTACHED(mp, ip)) {
> + xfs_ilock(ip, XFS_ILOCK_EXCL);
> error = xfs_qm_dqattach_locked(ip, true);
> - if (error) {
> - xfs_iunlock(ip, lockflags);
> + xfs_iunlock(ip, XFS_ILOCK_EXCL);
> + if (error)
> return error;
> - }
> }
>
> if ((flags & XFS_QMOPT_UQUOTA) && XFS_IS_UQUOTA_ON(mp)) {
> ASSERT(O_udqpp);
> if (!uid_eq(inode->i_uid, uid)) {
> - /*
> - * What we need is the dquot that has this uid, and
> - * if we send the inode to dqget, the uid of the inode
> - * takes priority over what's sent in the uid argument.
> - * We must unlock inode here before calling dqget if
> - * we're not sending the inode, because otherwise
> - * we'll deadlock by doing trans_reserve while
> - * holding ilock.
> - */
> - xfs_iunlock(ip, lockflags);
> error = xfs_qm_dqget(mp, from_kuid(user_ns, uid),
> XFS_DQTYPE_USER, true, &uq);
> if (error) {
> ASSERT(error != -ENOENT);
> return error;
> }
> - /*
> - * Get the ilock in the right order.
> - */
> - lockflags = XFS_ILOCK_SHARED;
> - xfs_ilock(ip, lockflags);
> } else {
> /*
> * Take an extra reference, because we'll return
> @@ -1922,15 +1903,12 @@ xfs_qm_vop_dqalloc(
> if ((flags & XFS_QMOPT_GQUOTA) && XFS_IS_GQUOTA_ON(mp)) {
> ASSERT(O_gdqpp);
> if (!gid_eq(inode->i_gid, gid)) {
> - xfs_iunlock(ip, lockflags);
> error = xfs_qm_dqget(mp, from_kgid(user_ns, gid),
> XFS_DQTYPE_GROUP, true, &gq);
> if (error) {
> ASSERT(error != -ENOENT);
> goto error_rele;
> }
> - lockflags = XFS_ILOCK_SHARED;
> - xfs_ilock(ip, lockflags);
> } else {
> ASSERT(ip->i_gdquot);
> gq = xfs_qm_dqhold(ip->i_gdquot);
> @@ -1939,15 +1917,12 @@ xfs_qm_vop_dqalloc(
> if ((flags & XFS_QMOPT_PQUOTA) && XFS_IS_PQUOTA_ON(mp)) {
> ASSERT(O_pdqpp);
> if (ip->i_projid != prid) {
> - xfs_iunlock(ip, lockflags);
> error = xfs_qm_dqget(mp, prid,
> XFS_DQTYPE_PROJ, true, &pq);
> if (error) {
> ASSERT(error != -ENOENT);
> goto error_rele;
> }
> - lockflags = XFS_ILOCK_SHARED;
> - xfs_ilock(ip, lockflags);
> } else {
> ASSERT(ip->i_pdquot);
> pq = xfs_qm_dqhold(ip->i_pdquot);
> @@ -1955,7 +1930,6 @@ xfs_qm_vop_dqalloc(
> }
> trace_xfs_dquot_dqalloc(ip);
>
> - xfs_iunlock(ip, lockflags);
> if (O_udqpp)
> *O_udqpp = uq;
> else
> --
> 2.47.3
>
>
next prev parent reply other threads:[~2025-10-15 21:27 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-10-13 2:48 cleanup quota locking Christoph Hellwig
2025-10-13 2:48 ` [PATCH 01/17] xfs: make qi_dquots a 64-bit value Christoph Hellwig
2025-10-14 23:16 ` Darrick J. Wong
2025-10-15 4:48 ` Christoph Hellwig
2025-10-13 2:48 ` [PATCH 02/17] xfs: remove xfs_dqunlock and friends Christoph Hellwig
2025-10-14 23:17 ` Darrick J. Wong
2025-10-13 2:48 ` [PATCH 03/17] xfs: don't lock the dquot before return in xqcheck_commit_dquot Christoph Hellwig
2025-10-14 23:22 ` Darrick J. Wong
2025-10-15 5:00 ` Christoph Hellwig
2025-10-15 20:27 ` Darrick J. Wong
2025-10-13 2:48 ` [PATCH 04/17] xfs: don't lock the dquot before return in xrep_quota_item Christoph Hellwig
2025-10-14 23:24 ` Darrick J. Wong
2025-10-13 2:48 ` [PATCH 05/17] xfs: use a lockref for the xfs_dquot reference count Christoph Hellwig
2025-10-15 21:02 ` Darrick J. Wong
2025-10-13 2:48 ` [PATCH 06/17] xfs: remove xfs_qm_dqput and optimize dropping dquot references Christoph Hellwig
2025-10-15 21:04 ` Darrick J. Wong
2025-10-13 2:48 ` [PATCH 07/17] xfs: consolidate q_qlock locking in xfs_qm_dqget and xfs_qm_dqget_inode Christoph Hellwig
2025-10-15 21:05 ` Darrick J. Wong
2025-10-13 2:48 ` [PATCH 08/17] xfs: xfs_qm_dqattach_one is never called with a non-NULL *IO_idqpp Christoph Hellwig
2025-10-14 23:27 ` Darrick J. Wong
2025-10-13 2:48 ` [PATCH 09/17] xfs: fold xfs_qm_dqattach_one into xfs_qm_dqget_inode Christoph Hellwig
2025-10-15 21:13 ` Darrick J. Wong
2025-10-13 2:48 ` [PATCH 10/17] xfs: return the dquot unlocked from xfs_qm_dqget Christoph Hellwig
2025-10-15 21:17 ` Darrick J. Wong
2025-10-15 21:18 ` Darrick J. Wong
2025-10-16 4:21 ` Christoph Hellwig
2025-10-13 2:48 ` [PATCH 11/17] xfs: remove q_qlock locking in xfs_qm_scall_setqlim Christoph Hellwig
2025-10-15 21:17 ` Darrick J. Wong
2025-10-13 2:48 ` [PATCH 12/17] xfs: push q_qlock acquisition from xchk_dquot_iter to the callers Christoph Hellwig
2025-10-15 21:19 ` Darrick J. Wong
2025-10-13 2:48 ` [PATCH 13/17] xfs: move q_qlock locking into xchk_quota_item Christoph Hellwig
2025-10-15 21:19 ` Darrick J. Wong
2025-10-13 2:48 ` [PATCH 14/17] xfs: move q_qlock locking into xqcheck_compare_dquot Christoph Hellwig
2025-10-15 21:20 ` Darrick J. Wong
2025-10-13 2:48 ` [PATCH 15/17] xfs: move q_qlock acquisition into xqcheck_commit_dquot Christoph Hellwig
2025-10-15 21:20 ` Darrick J. Wong
2025-10-16 4:22 ` Christoph Hellwig
2025-10-13 2:48 ` [PATCH 16/17] xfs: move xfs_dquot_tree calls into xfs_qm_dqget_cache_{lookup,insert} Christoph Hellwig
2025-10-15 21:21 ` Darrick J. Wong
2025-10-13 2:48 ` [PATCH 17/17] xfs: reduce ilock roundtrips in xfs_qm_vop_dqalloc Christoph Hellwig
2025-10-15 21:27 ` Darrick J. Wong [this message]
2025-10-16 4:23 ` Christoph Hellwig
2025-10-16 15:59 ` Darrick J. Wong
2025-10-17 3:50 ` Christoph Hellwig
2025-10-17 23:09 ` Darrick J. Wong
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=20251015212707.GM2591640@frogsfrogsfrogs \
--to=djwong@kernel.org \
--cc=cem@kernel.org \
--cc=hch@lst.de \
--cc=linux-xfs@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).