Linux filesystem development
 help / color / mirror / Atom feed
* [PATCH] netfs: Fix barriering when walking subrequest list
@ 2026-07-02  8:23 David Howells
  2026-07-02 12:59 ` Paulo Alcantara
  0 siblings, 1 reply; 2+ messages in thread
From: David Howells @ 2026-07-02  8:23 UTC (permalink / raw)
  To: Christian Brauner
  Cc: dhowells, Paulo Alcantara, netfs, linux-fsdevel, linux-kernel

Fix the barriering used when walking the subrequest list in retry as
there's a possibility of seeing a subreq that's just been added by the
application thread.

Fixes: ee4cdf7ba857 ("netfs: Speed up buffered reading")
Fixes: 288ace2f57c9 ("netfs: New writeback implementation")
Link: https://sashiko.dev/#/patchset/20260608145432.681865-1-dhowells%40redhat.com
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Paulo Alcantara <pc@manguebit.org>
cc: netfs@lists.linux.dev
cc: linux-fsdevel@vger.kernel.org
---
 fs/netfs/read_retry.c  |    7 ++++++-
 fs/netfs/write_retry.c |    7 ++++++-
 2 files changed, 12 insertions(+), 2 deletions(-)

diff --git a/fs/netfs/read_retry.c b/fs/netfs/read_retry.c
index f59a70f3a086..2b42758e01ec 100644
--- a/fs/netfs/read_retry.c
+++ b/fs/netfs/read_retry.c
@@ -98,7 +98,12 @@ static void netfs_retry_read_subrequests(struct netfs_io_request *rreq)
 			goto abandon;
 		}
 
-		list_for_each_continue(next, &stream->subrequests) {
+		for (;;) {
+			/* Read pointer to subreq before reading subreq state. */
+			next = smp_load_acquire(&next->next);
+			if (next == &stream->subrequests)
+				break;
+
 			subreq = list_entry(next, struct netfs_io_subrequest, rreq_link);
 			if (subreq->start + subreq->transferred != start + len ||
 			    test_bit(NETFS_SREQ_BOUNDARY, &subreq->flags) ||
diff --git a/fs/netfs/write_retry.c b/fs/netfs/write_retry.c
index 32735abfa03f..058bc7a166a5 100644
--- a/fs/netfs/write_retry.c
+++ b/fs/netfs/write_retry.c
@@ -72,7 +72,12 @@ static void netfs_retry_write_stream(struct netfs_io_request *wreq,
 		    !test_bit(NETFS_SREQ_NEED_RETRY, &from->flags))
 			return;
 
-		list_for_each_continue(next, &stream->subrequests) {
+		for (;;) {
+			/* Read pointer to subreq before reading subreq state. */
+			next = smp_load_acquire(&next->next);
+			if (next == &stream->subrequests)
+				break;
+
 			subreq = list_entry(next, struct netfs_io_subrequest, rreq_link);
 			if (subreq->start + subreq->transferred != start + len ||
 			    test_bit(NETFS_SREQ_BOUNDARY, &subreq->flags) ||


^ permalink raw reply related	[flat|nested] 2+ messages in thread

* Re: [PATCH] netfs: Fix barriering when walking subrequest list
  2026-07-02  8:23 [PATCH] netfs: Fix barriering when walking subrequest list David Howells
@ 2026-07-02 12:59 ` Paulo Alcantara
  0 siblings, 0 replies; 2+ messages in thread
From: Paulo Alcantara @ 2026-07-02 12:59 UTC (permalink / raw)
  To: David Howells, Christian Brauner
  Cc: dhowells, netfs, linux-fsdevel, linux-kernel

David Howells <dhowells@redhat.com> writes:

> Fix the barriering used when walking the subrequest list in retry as
> there's a possibility of seeing a subreq that's just been added by the
> application thread.
>
> Fixes: ee4cdf7ba857 ("netfs: Speed up buffered reading")
> Fixes: 288ace2f57c9 ("netfs: New writeback implementation")
> Link: https://sashiko.dev/#/patchset/20260608145432.681865-1-dhowells%40redhat.com
> Signed-off-by: David Howells <dhowells@redhat.com>
> cc: Paulo Alcantara <pc@manguebit.org>
> cc: netfs@lists.linux.dev
> cc: linux-fsdevel@vger.kernel.org
> ---
>  fs/netfs/read_retry.c  |    7 ++++++-
>  fs/netfs/write_retry.c |    7 ++++++-
>  2 files changed, 12 insertions(+), 2 deletions(-)

Reviewed-by: Paulo Alcantara (Red Hat) <pc@manguebit.org>

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2026-07-02 12:59 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-07-02  8:23 [PATCH] netfs: Fix barriering when walking subrequest list David Howells
2026-07-02 12:59 ` Paulo Alcantara

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox