Linux NFS development
 help / color / mirror / Atom feed
From: Mike Snitzer <snitzer@kernel.org>
To: Trond Myklebust <trondmy@kernel.org>, Anna Schumaker <anna@kernel.org>
Cc: linux-nfs@vger.kernel.org
Subject: [PATCH v2 3/3] NFSv4/pnfs: don't run the release of a LAYOUTCOMMIT in the submitter
Date: Mon, 28 Sep 2026 14:57:51 -0400	[thread overview]
Message-ID: <20260928185751.98682-4-snitzer@kernel.org> (raw)
In-Reply-To: <20260928185751.98682-1-snitzer@kernel.org>

Since commit 472e25944981 ("NFSv4.1: Pin the inode and super block in
asynchronous layoutcommit") nfs4_proc_layoutcommit() pins the inode and
the superblock for an asynchronous LAYOUTCOMMIT with
nfs_igrab_and_active(), and nfs4_layoutcommit_release() drops both with
nfs_iput_and_deactive().

The task is put with rpc_put_task(), so if it has already completed,
the release runs in the submitter. nfs4_write_inode() sends an
asynchronous LAYOUTCOMMIT from writeback, where the submitter is the
flusher. It holds sb->s_umount and has set I_SYNC on the inode, so it
can do neither the final iput() nor the final deactivate_super(). If
the task has not completed, the release runs on rpciod, because the
task has no workqueue.

Give the asynchronous task nfsiod as its workqueue and put it with
rpc_put_task_async(), as is done for WRITE and COMMIT. None of the
callers that send an asynchronous LAYOUTCOMMIT waits for its release.

A synchronous LAYOUTCOMMIT is left as it is: it pins nothing, so its
release has nothing of the kind to drop.

Fixes: 472e25944981 ("NFSv4.1: Pin the inode and super block in asynchronous layoutcommit")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-fable-5-1
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
---
 fs/nfs/nfs4proc.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 518348e87dd8..133d01ce9f08 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -10147,6 +10147,7 @@ nfs4_proc_layoutcommit(struct nfs4_layoutcommit_data *data, bool sync)
 			return -EAGAIN;
 		}
 		task_setup_data.flags = RPC_TASK_ASYNC;
+		task_setup_data.workqueue = nfsiod_workqueue;
 	}
 	nfs4_init_sequence(NFS_SERVER(data->args.inode)->nfs_client,
 			   &data->args.seq_args, &data->res.seq_res, 1, 0);
@@ -10157,7 +10158,11 @@ nfs4_proc_layoutcommit(struct nfs4_layoutcommit_data *data, bool sync)
 		status = task->tk_status;
 	trace_nfs4_layoutcommit(data->args.inode, &data->args.stateid, status);
 	dprintk("%s: status %d\n", __func__, status);
-	rpc_put_task(task);
+	/* An asynchronous caller may be writeback: don't run rpc_release() here */
+	if (sync)
+		rpc_put_task(task);
+	else
+		rpc_put_task_async(task);
 	return status;
 }
 
-- 
2.52.0


      parent reply	other threads:[~2026-09-28 18:57 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-28 18:57 [PATCH v2 0/3] NFS: don't release an open context from writeback Mike Snitzer
2026-09-28 18:57 ` [PATCH v2 1/3] NFS: don't release the open context of a failed write " Mike Snitzer
2026-09-28 18:57 ` [PATCH v2 2/3] NFS: don't run the release of a WRITE or COMMIT in the submitter Mike Snitzer
2026-09-28 18:57 ` Mike Snitzer [this message]

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=20260928185751.98682-4-snitzer@kernel.org \
    --to=snitzer@kernel.org \
    --cc=anna@kernel.org \
    --cc=linux-nfs@vger.kernel.org \
    --cc=trondmy@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