mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
* [PATCH v2] ext4: cancel dirty accounting for folios without buffers
@ 2026-06-26 10:07 Zhu Jia
  2026-06-27  3:47 ` Zhang Yi
                   ` (2 more replies)
  0 siblings, 3 replies; 4+ messages in thread
From: Zhu Jia @ 2026-06-26 10:07 UTC (permalink / raw)
  To: tytso, adilger.kernel
  Cc: libaokun, jack, ojaswin, ritesh.list, yi.zhang, linux-ext4,
	linux-kernel, Zhu Jia, stable, Zhang Yi

Since commit cc5095747edf ("ext4: don't BUG if someone dirty pages
without asking ext4 first"), mpage_prepare_extent_to_map() handles dirty
folios without buffer heads by warning, clearing PG_dirty, and skipping
them. ext4 cannot write these folios because there are no buffer heads to
map and submit.

That recovery leaves dirty accounting behind: folio_clear_dirty() clears
PG_dirty but does not undo the accounting charged when the folio was
dirtied. We have seen this in production as Dirty/nr_dirty staying high
while Writeback/nr_writeback and device write IO stayed near zero, with
many writer tasks blocked in balance_dirty_pages() throttling. Thus the
warning-and-skip recovery can still become a dirty-throttle DoS.

Use folio_cancel_dirty() so dropping PG_dirty also cancels the dirty
accounting. Then cycle the folio through writeback state so the generic
writeback helpers update the xarray DIRTY/TOWRITE tags.

Fixes: cc5095747edf ("ext4: don't BUG if someone dirty pages without asking ext4 first")
Cc: stable@vger.kernel.org
Suggested-by: Zhang Yi <yizhang089@gmail.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Zhu Jia <zhujia.zj@bytedance.com>
---
Changes since v1:
- After folio_cancel_dirty(), cycle the folio through writeback state so
  generic writeback helpers update PAGECACHE_TAG_DIRTY/TOWRITE, as
  suggested by Yi and Jan.

 fs/ext4/inode.c | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index c2c2d6ac7f3d1..4c25dcd47fb15 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -2715,7 +2715,15 @@ static int mpage_prepare_extent_to_map(struct mpage_da_data *mpd)
 			 */
 			if (!folio_buffers(folio)) {
 				ext4_warning_inode(mpd->inode, "page %lu does not have buffers attached", folio->index);
-				folio_clear_dirty(folio);
+				/*
+				 * folio_cancel_dirty() pairs the dropped dirty
+				 * state with dirty accounting. Cycle through
+				 * writeback state so the generic writeback
+				 * helpers update the xarray tags.
+				 */
+				folio_cancel_dirty(folio);
+				folio_start_writeback(folio);
+				folio_end_writeback(folio);
 				folio_unlock(folio);
 				continue;
 			}
-- 
2.20.1

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH v2] ext4: cancel dirty accounting for folios without buffers
  2026-06-26 10:07 [PATCH v2] ext4: cancel dirty accounting for folios without buffers Zhu Jia
@ 2026-06-27  3:47 ` Zhang Yi
  2026-08-12 10:05 ` Zhu Jia
  2026-10-09  7:23 ` Jia Zhu
  2 siblings, 0 replies; 4+ messages in thread
From: Zhang Yi @ 2026-06-27  3:47 UTC (permalink / raw)
  To: Zhu Jia, tytso, adilger.kernel
  Cc: libaokun, jack, ojaswin, ritesh.list, linux-ext4, linux-kernel,
	stable, Zhang Yi

On 6/26/2026 6:07 PM, Zhu Jia wrote:
> Since commit cc5095747edf ("ext4: don't BUG if someone dirty pages
> without asking ext4 first"), mpage_prepare_extent_to_map() handles dirty
> folios without buffer heads by warning, clearing PG_dirty, and skipping
> them. ext4 cannot write these folios because there are no buffer heads to
> map and submit.
> 
> That recovery leaves dirty accounting behind: folio_clear_dirty() clears
> PG_dirty but does not undo the accounting charged when the folio was
> dirtied. We have seen this in production as Dirty/nr_dirty staying high
> while Writeback/nr_writeback and device write IO stayed near zero, with
> many writer tasks blocked in balance_dirty_pages() throttling. Thus the
> warning-and-skip recovery can still become a dirty-throttle DoS.
> 
> Use folio_cancel_dirty() so dropping PG_dirty also cancels the dirty
> accounting. Then cycle the folio through writeback state so the generic
> writeback helpers update the xarray DIRTY/TOWRITE tags.
> 
> Fixes: cc5095747edf ("ext4: don't BUG if someone dirty pages without asking ext4 first")
> Cc: stable@vger.kernel.org
> Suggested-by: Zhang Yi <yizhang089@gmail.com>
> Reviewed-by: Jan Kara <jack@suse.cz>
> Signed-off-by: Zhu Jia <zhujia.zj@bytedance.com>

Looks good to me.

Reviewed-by: Zhang Yi <yi.zhang@huawei.com>

> ---
> Changes since v1:
> - After folio_cancel_dirty(), cycle the folio through writeback state so
>   generic writeback helpers update PAGECACHE_TAG_DIRTY/TOWRITE, as
>   suggested by Yi and Jan.
> 
>  fs/ext4/inode.c | 10 +++++++++-
>  1 file changed, 9 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
> index c2c2d6ac7f3d1..4c25dcd47fb15 100644
> --- a/fs/ext4/inode.c
> +++ b/fs/ext4/inode.c
> @@ -2715,7 +2715,15 @@ static int mpage_prepare_extent_to_map(struct mpage_da_data *mpd)
>  			 */
>  			if (!folio_buffers(folio)) {
>  				ext4_warning_inode(mpd->inode, "page %lu does not have buffers attached", folio->index);
> -				folio_clear_dirty(folio);
> +				/*
> +				 * folio_cancel_dirty() pairs the dropped dirty
> +				 * state with dirty accounting. Cycle through
> +				 * writeback state so the generic writeback
> +				 * helpers update the xarray tags.
> +				 */
> +				folio_cancel_dirty(folio);
> +				folio_start_writeback(folio);
> +				folio_end_writeback(folio);
>  				folio_unlock(folio);
>  				continue;
>  			}


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH v2] ext4: cancel dirty accounting for folios without buffers
  2026-06-26 10:07 [PATCH v2] ext4: cancel dirty accounting for folios without buffers Zhu Jia
  2026-06-27  3:47 ` Zhang Yi
@ 2026-08-12 10:05 ` Zhu Jia
  2026-10-09  7:23 ` Jia Zhu
  2 siblings, 0 replies; 4+ messages in thread
From: Zhu Jia @ 2026-08-12 10:05 UTC (permalink / raw)
  To: Theodore Ts'o
  Cc: Zhu Jia, Andreas Dilger, Jan Kara, Zhang Yi, linux-ext4,
	linux-kernel, stable

Hi Ted,

Gentle ping on this ext4 fix:

  https://lore.kernel.org/all/20260626100740.52455-1-zhujia.zj@bytedance.com/

Could you please consider picking it up for the ext4 tree?

Thanks,
Jia

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH v2] ext4: cancel dirty accounting for folios without buffers
  2026-06-26 10:07 [PATCH v2] ext4: cancel dirty accounting for folios without buffers Zhu Jia
  2026-06-27  3:47 ` Zhang Yi
  2026-08-12 10:05 ` Zhu Jia
@ 2026-10-09  7:23 ` Jia Zhu
  2 siblings, 0 replies; 4+ messages in thread
From: Jia Zhu @ 2026-10-09  7:23 UTC (permalink / raw)
  To: Theodore Ts'o
  Cc: Jia Zhu, Andreas Dilger, Jan Kara, Zhang Yi, linux-ext4,
	linux-kernel, stable

Hi Ted,

Gentle ping again on this ext4 fix:

  https://lore.kernel.org/all/20260626100740.52455-1-zhujia.zj@bytedance.com/

It has Reviewed-by tags from Jan Kara and Zhang Yi. Could you please
consider picking it up for the ext4 tree, or let me know if anything
else is needed?

Thanks,
Jia

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2026-10-09  7:23 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2026-06-26 10:07 [PATCH v2] ext4: cancel dirty accounting for folios without buffers Zhu Jia
2026-06-27  3:47 ` Zhang Yi
2026-08-12 10:05 ` Zhu Jia
2026-10-09  7:23 ` Jia Zhu

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

all inboxes | Powered by JetHome®