Linux-NFS Archive mirror
 help / color / mirror / Atom feed
From: Amir Goldstein <amir73il@gmail.com>
To: NeilBrown <neil@brown.name>
Cc: Linus Torvalds <torvalds@linux-foundation.org>,
	Alexander Viro <viro@zeniv.linux.org.uk>,
	 Christian Brauner <brauner@kernel.org>, Jan Kara <jack@suse.cz>,
	Jeff Layton <jlayton@kernel.org>,
	 Trond Myklebust <trondmy@kernel.org>,
	Anna Schumaker <anna@kernel.org>,
	Miklos Szeredi <miklos@szeredi.hu>,  Jeremy Kerr <jk@ozlabs.org>,
	Ard Biesheuvel <ardb@kernel.org>,
	linux-efi@vger.kernel.org,  linux-fsdevel@vger.kernel.org,
	linux-nfs@vger.kernel.org,  linux-unionfs@vger.kernel.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH v3 08/19] VFS/xfs/ntfs: drop parent lock across d_alloc_parallel() in d_add_ci()
Date: Mon, 27 Apr 2026 09:49:36 +0200	[thread overview]
Message-ID: <CAOQ4uxjj6iJNnwUqV+cRtqo4qRD4v-f-ct92=ofCZWPDATXM7g@mail.gmail.com> (raw)
In-Reply-To: <20260427040517.828226-9-neilb@ownmail.net>

On Mon, Apr 27, 2026 at 6:07 AM NeilBrown <neilb@ownmail.net> wrote:
>
> From: NeilBrown <neil@brown.name>
>
> A proposed change will invert the lock ordering between
> d_alloc_parallel() and inode_lock() on the parent.
> When that happens it will not be safe to call d_alloc_parallel() while
> holding the parent lock - even shared.
>
> We don't need to keep the parent lock held when d_add_ci() is run - the
> VFS doesn't need it as dentry is exclusively held due to
> DCACHE_PAR_LOOKUP and the filesystem has finished its work.
>
> So drop and reclaim the lock (shared or exclusive as determined by
> LOOKUP_SHARED) to avoid future deadlock.
>
> Signed-off-by: NeilBrown <neil@brown.name>
> ---
>  fs/dcache.c            | 18 +++++++++++++++++-
>  fs/ntfs/namei.c        |  4 +++-
>  fs/xfs/xfs_iops.c      |  3 ++-
>  include/linux/dcache.h |  3 ++-
>  4 files changed, 24 insertions(+), 4 deletions(-)
>
> diff --git a/fs/dcache.c b/fs/dcache.c
> index 569a8ddf4c0d..a2ddfe811df3 100644
> --- a/fs/dcache.c
> +++ b/fs/dcache.c
> @@ -2294,6 +2294,7 @@ EXPORT_SYMBOL(d_obtain_root);
>   * @dentry: the negative dentry that was passed to the parent's lookup func
>   * @inode:  the inode case-insensitive lookup has found
>   * @name:   the case-exact name to be associated with the returned dentry
> + * @bool:   %true if lookup was performed with LOOKUP_SHARED

I do not understand the choice of API.
Seems better to pass the lookup flags and avoid exposing this very internal
logic in the call sites...

>   *
>   * This is to avoid filling the dcache with case-insensitive names to the
>   * same inode, only the actual correct case is stored in the dcache for
> @@ -2306,7 +2307,7 @@ EXPORT_SYMBOL(d_obtain_root);
>   * the exact case, and return the spliced entry.
>   */
>  struct dentry *d_add_ci(struct dentry *dentry, struct inode *inode,
> -                       struct qstr *name)
> +                       struct qstr *name, bool shared)
>  {
>         struct dentry *found, *res;
>
> @@ -2319,6 +2320,17 @@ struct dentry *d_add_ci(struct dentry *dentry, struct inode *inode,
>                 iput(inode);
>                 return found;
>         }
> +       /*
> +        * We are holding parent lock and so don't want to wait for a
> +        * d_in_lookup() dentry.  We can safely drop the parent lock and
> +        * reclaim it as we have exclusive access to dentry as it is
> +        * d_in_lookup() (so ->d_parent is stable) and we are near the
> +        * end ->lookup() and will shortly drop the lock anyway.
> +        */
> +       if (shared)
> +               inode_unlock_shared(d_inode(dentry->d_parent));
> +       else
> +               inode_unlock(d_inode(dentry->d_parent));
>         if (d_in_lookup(dentry)) {
>                 found = d_alloc_parallel(dentry->d_parent, name);
>                 if (IS_ERR(found) || !d_in_lookup(found)) {
> @@ -2332,6 +2344,10 @@ struct dentry *d_add_ci(struct dentry *dentry, struct inode *inode,
>                         return ERR_PTR(-ENOMEM);
>                 }
>         }
> +       if (shared)
> +               inode_lock_shared(d_inode(dentry->d_parent));
> +       else
> +               inode_lock_nested(d_inode(dentry->d_parent), I_MUTEX_PARENT);
>         res = d_splice_alias(inode, found);
>         if (res) {
>                 d_lookup_done(found);
> diff --git a/fs/ntfs/namei.c b/fs/ntfs/namei.c
> index 10894de519c3..30dddef43300 100644
> --- a/fs/ntfs/namei.c
> +++ b/fs/ntfs/namei.c
> @@ -8,6 +8,7 @@
>
>  #include <linux/exportfs.h>
>  #include <linux/iversion.h>
> +#include <linux/namei.h> // for LOOKUP_SHARED

... this won't be needed

>
>  #include "ntfs.h"
>  #include "time.h"
> @@ -310,7 +311,8 @@ static struct dentry *ntfs_lookup(struct inode *dir_ino, struct dentry *dent,
>                 }
>                 nls_name.hash = full_name_hash(dent, nls_name.name, nls_name.len);
>
> -               dent = d_add_ci(dent, dent_inode, &nls_name);
> +               dent = d_add_ci(dent, dent_inode, &nls_name,
> +                               !!(flags & LOOKUP_SHARED));
>                 kfree(nls_name.name);
>                 return dent;
>
> diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
> index 325c2200c501..f03d832f1468 100644
> --- a/fs/xfs/xfs_iops.c
> +++ b/fs/xfs/xfs_iops.c
> @@ -35,6 +35,7 @@
>  #include <linux/security.h>
>  #include <linux/iversion.h>
>  #include <linux/fiemap.h>
> +#include <linux/namei.h> // for LOOKUP_SHARED
>
>  /*
>   * Directories have different lock order w.r.t. mmap_lock compared to regular
> @@ -369,7 +370,7 @@ xfs_vn_ci_lookup(
>         /* else case-insensitive match... */
>         dname.name = ci_name.name;
>         dname.len = ci_name.len;
> -       dentry = d_add_ci(dentry, VFS_I(ip), &dname);
> +       dentry = d_add_ci(dentry, VFS_I(ip), &dname, !!(flags & LOOKUP_SHARED));


... and this ugliness could be avoided.

Thanks,
Amir.

  reply	other threads:[~2026-04-27  7:49 UTC|newest]

Thread overview: 53+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-04-27  4:01 [PATCH v3 00/19] Prepare to lift lookup out of exclusive lock for directory ops NeilBrown
2026-04-27  4:01 ` [PATCH v3 01/19] VFS: fix various typos in documentation for start_creating start_removing etc NeilBrown
2026-04-27  4:01 ` [PATCH v3 02/19] VFS: enhance d_splice_alias() to handle in-lookup dentries NeilBrown
2026-04-27  4:01 ` [PATCH v3 03/19] VFS: allow d_alloc_name() to be used with ->d_hash NeilBrown
2026-04-28  2:10   ` Al Viro
2026-04-29  2:44     ` NeilBrown
2026-04-27  4:01 ` [PATCH v3 04/19] VFS: use wait_var_event for waiting in d_alloc_parallel() NeilBrown
2026-04-28  3:37   ` Al Viro
2026-04-28 11:18     ` NeilBrown
2026-04-28 14:22       ` Al Viro
2026-04-28 23:26         ` NeilBrown
2026-04-29  5:26           ` Al Viro
2026-04-29 17:07             ` Al Viro
2026-04-29 21:03               ` Linus Torvalds
2026-04-30 23:51               ` NeilBrown
2026-05-01  1:11                 ` Al Viro
2026-05-01  1:39                   ` NeilBrown
2026-05-01  1:45                     ` NeilBrown
2026-05-01  3:37                     ` Al Viro
2026-05-01 10:46                       ` NeilBrown
2026-05-01  1:20                 ` NeilBrown
2026-04-28 16:32     ` Linus Torvalds
2026-04-27  4:01 ` [PATCH v3 05/19] VFS: introduce d_alloc_noblock() NeilBrown
2026-04-28  2:22   ` Al Viro
2026-04-28 11:24     ` NeilBrown
2026-04-27  4:01 ` [PATCH v3 06/19] VFS: add d_duplicate() NeilBrown
2026-04-27  4:01 ` [PATCH v3 07/19] VFS: Add LOOKUP_SHARED flag NeilBrown
2026-04-27  7:43   ` Amir Goldstein
2026-04-27  8:47     ` NeilBrown
2026-04-27  9:05       ` Amir Goldstein
2026-04-27 23:51         ` NeilBrown
2026-04-27  4:01 ` [PATCH v3 08/19] VFS/xfs/ntfs: drop parent lock across d_alloc_parallel() in d_add_ci() NeilBrown
2026-04-27  7:49   ` Amir Goldstein [this message]
2026-04-27  8:48     ` NeilBrown
2026-04-27  4:01 ` [PATCH v3 09/19] ovl: stop using lookup_one() in iterate_shared() handling NeilBrown
2026-04-27 10:10   ` Amir Goldstein
2026-04-28  0:24     ` NeilBrown
2026-04-27  4:01 ` [PATCH v3 10/19] VFS/ovl: add d_alloc_noblock_return() NeilBrown
2026-04-27  9:40   ` Amir Goldstein
2026-04-28  0:34     ` NeilBrown
2026-04-28  4:35       ` Al Viro
2026-04-28 11:44         ` NeilBrown
2026-04-27  4:01 ` [PATCH v3 11/19] efivarfs: use d_alloc_name() NeilBrown
2026-04-27  4:01 ` [PATCH v3 12/19] shmem: use d_duplicate() NeilBrown
2026-04-27  4:01 ` [PATCH v3 13/19] nfs: remove d_drop()/d_alloc_parallel() from nfs_atomic_open() NeilBrown
2026-04-27  4:01 ` [PATCH v3 14/19] nfs: use d_splice_alias() in nfs_link() NeilBrown
2026-04-27  4:01 ` [PATCH v3 15/19] nfs: don't d_drop() before d_splice_alias() NeilBrown
2026-04-27  4:01 ` [PATCH v3 16/19] nfs: don't d_drop() before d_splice_alias() in atomic_create NeilBrown
2026-04-27  4:01 ` [PATCH v3 17/19] nfs: Use d_alloc_noblock() in nfs_prime_dcache() NeilBrown
2026-04-27  4:01 ` [PATCH v3 18/19] nfs: use d_alloc_noblock() in silly-rename NeilBrown
2026-04-27  4:01 ` [PATCH v3 19/19] nfs: use d_duplicate() NeilBrown
2026-04-27  8:42 ` [syzbot ci] Re: Prepare to lift lookup out of exclusive lock for directory ops syzbot ci
2026-04-28 23:16   ` NeilBrown

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='CAOQ4uxjj6iJNnwUqV+cRtqo4qRD4v-f-ct92=ofCZWPDATXM7g@mail.gmail.com' \
    --to=amir73il@gmail.com \
    --cc=anna@kernel.org \
    --cc=ardb@kernel.org \
    --cc=brauner@kernel.org \
    --cc=jack@suse.cz \
    --cc=jk@ozlabs.org \
    --cc=jlayton@kernel.org \
    --cc=linux-efi@vger.kernel.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-nfs@vger.kernel.org \
    --cc=linux-unionfs@vger.kernel.org \
    --cc=miklos@szeredi.hu \
    --cc=neil@brown.name \
    --cc=torvalds@linux-foundation.org \
    --cc=trondmy@kernel.org \
    --cc=viro@zeniv.linux.org.uk \
    /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 read-only IMAP folder(s) and NNTP newsgroup(s).