linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Christoph Hellwig <hch@lst.de>
To: Christian Brauner <brauner@kernel.org>
Cc: "Darrick J. Wong" <djwong@kernel.org>,
	Chandan Babu R <chandan.babu@oracle.com>,
	Zhang Yi <yi.zhang@huaweicloud.com>,
	Ritesh Harjani <ritesh.list@gmail.com>,
	Jens Axboe <axboe@kernel.dk>,
	Andreas Gruenbacher <agruenba@redhat.com>,
	Damien Le Moal <dlemoal@kernel.org>,
	Naohiro Aota <naohiro.aota@wdc.com>,
	Johannes Thumshirn <jth@kernel.org>,
	linux-xfs@vger.kernel.org, linux-fsdevel@vger.kernel.org,
	linux-block@vger.kernel.org, gfs2@lists.linux.dev
Subject: [PATCH 01/14] iomap: clear the per-folio dirty bits on all writeback failures
Date: Thu,  7 Dec 2023 08:26:57 +0100	[thread overview]
Message-ID: <20231207072710.176093-2-hch@lst.de> (raw)
In-Reply-To: <20231207072710.176093-1-hch@lst.de>

write_cache_pages always clear the page dirty bit before calling into the
file systems, and leaves folios with a writeback failure without the
dirty bit after return.  We also clear the per-block writeback bits for
writeback failures unless no I/O has submitted, which will leave the
folio in an inconsistent state where it doesn't have the folio dirty,
but one or more per-block dirty bits.  This seems to be due the place
where the iomap_clear_range_dirty call was inserted into the existing
not very clearly structured code when adding per-block dirty bit support
and not actually intentional.  Switch to always clearing the dirty on
writeback failure.

Fixes: 4ce02c679722 ("iomap: Add per-block dirty state tracking to improve performance")
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 fs/iomap/buffered-io.c | 18 +++++++++++-------
 1 file changed, 11 insertions(+), 7 deletions(-)

diff --git a/fs/iomap/buffered-io.c b/fs/iomap/buffered-io.c
index f72df2babe561a..fc5c64712318aa 100644
--- a/fs/iomap/buffered-io.c
+++ b/fs/iomap/buffered-io.c
@@ -1843,16 +1843,10 @@ iomap_writepage_map(struct iomap_writepage_ctx *wpc,
 	if (unlikely(error)) {
 		/*
 		 * Let the filesystem know what portion of the current page
-		 * failed to map. If the page hasn't been added to ioend, it
-		 * won't be affected by I/O completion and we must unlock it
-		 * now.
+		 * failed to map.
 		 */
 		if (wpc->ops->discard_folio)
 			wpc->ops->discard_folio(folio, pos);
-		if (!count) {
-			folio_unlock(folio);
-			goto done;
-		}
 	}
 
 	/*
@@ -1861,6 +1855,16 @@ iomap_writepage_map(struct iomap_writepage_ctx *wpc,
 	 * all the dirty bits in the folio here.
 	 */
 	iomap_clear_range_dirty(folio, 0, folio_size(folio));
+
+	/*
+	 * If the page hasn't been added to the ioend, it won't be affected by
+	 * I/O completion and we must unlock it now.
+	 */
+	if (error && !count) {
+		folio_unlock(folio);
+		goto done;
+	}
+
 	folio_start_writeback(folio);
 	folio_unlock(folio);
 
-- 
2.39.2


  reply	other threads:[~2023-12-07  7:27 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-12-07  7:26 map multiple blocks per ->map_blocks in iomap writeback Christoph Hellwig
2023-12-07  7:26 ` Christoph Hellwig [this message]
2023-12-07  7:26 ` [PATCH 02/14] iomap: treat inline data in iomap_writepage_map as an I/O error Christoph Hellwig
2023-12-07  7:26 ` [PATCH 03/14] iomap: move the io_folios field out of struct iomap_ioend Christoph Hellwig
2023-12-07  7:27 ` [PATCH 04/14] iomap: move the PF_MEMALLOC check to iomap_writepages Christoph Hellwig
2023-12-07  7:27 ` [PATCH 05/14] iomap: factor out a iomap_writepage_handle_eof helper Christoph Hellwig
2023-12-07  7:27 ` [PATCH 06/14] iomap: move all remaining per-folio logic into iomap_writepage_map Christoph Hellwig
2023-12-07  7:27 ` [PATCH 07/14] iomap: clean up the iomap_alloc_ioend calling convention Christoph Hellwig
2023-12-07  7:27 ` [PATCH 08/14] iomap: move the iomap_sector sector calculation out of iomap_add_to_ioend Christoph Hellwig
2023-12-07  7:27 ` [PATCH 09/14] iomap: don't chain bios Christoph Hellwig
2023-12-07  7:27 ` [PATCH 10/14] iomap: only call mapping_set_error once for each failed bio Christoph Hellwig
2023-12-07  7:27 ` [PATCH 11/14] iomap: factor out a iomap_writepage_map_block helper Christoph Hellwig
2023-12-07  7:27 ` [PATCH 12/14] iomap: submit ioends immediately Christoph Hellwig
2023-12-07  7:27 ` [PATCH 13/14] iomap: map multiple blocks at a time Christoph Hellwig
2023-12-07 13:39   ` Zhang Yi
2023-12-07 15:03     ` Christoph Hellwig
2023-12-08  7:33       ` Zhang Yi
2023-12-07  7:27 ` [PATCH 14/14] iomap: pass the length of the dirty region to ->map_blocks Christoph Hellwig
2023-12-11 10:45 ` map multiple blocks per ->map_blocks in iomap writeback Christian Brauner
2024-02-01  6:07   ` Christoph Hellwig
2024-02-01  6:18     ` Darrick J. Wong
2024-02-01 13:23       ` Christian Brauner

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=20231207072710.176093-2-hch@lst.de \
    --to=hch@lst.de \
    --cc=agruenba@redhat.com \
    --cc=axboe@kernel.dk \
    --cc=brauner@kernel.org \
    --cc=chandan.babu@oracle.com \
    --cc=djwong@kernel.org \
    --cc=dlemoal@kernel.org \
    --cc=gfs2@lists.linux.dev \
    --cc=jth@kernel.org \
    --cc=linux-block@vger.kernel.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-xfs@vger.kernel.org \
    --cc=naohiro.aota@wdc.com \
    --cc=ritesh.list@gmail.com \
    --cc=yi.zhang@huaweicloud.com \
    /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).