From: Christian Brauner <brauner@kernel.org>
To: linux-fsdevel@vger.kernel.org, Jan Kara <jack@suse.cz>
Cc: Josef Bacik <josef@toxicpanda.com>,
Chris Mason <mason@kernel.org>,
linux-ext4@vger.kernel.org,
OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>,
Andreas Dilger <adilger.kernel@dilger.ca>,
Baokun Li <libaokun@linux.alibaba.com>,
Ojaswin Mujoo <ojaswin@linux.ibm.com>,
"Ritesh Harjani (IBM)" <ritesh.list@gmail.com>,
Namjae Jeon <namjae.jeon@samsung.com>,
"Christian Brauner (Amutable)" <brauner@kernel.org>,
stable@vger.kernel.org
Subject: [PATCH RFC 4/9] ext4: don't read the inode block in ext4_sync_inode_metadata()
Date: Thu, 29 Oct 2026 14:28:19 +0100 [thread overview]
Message-ID: <20261029-work-inode-metadata_writeback-v1-4-7d6c678a585d@kernel.org> (raw)
In-Reply-To: <20261029-work-inode-metadata_writeback-v1-0-7d6c678a585d@kernel.org>
In nojournal mode ext4_sync_inode_metadata() looks up the inode table
block with __ext4_get_inode_loc_noinmem() which reads it from disk, with
readahead of the inode table on top, if it isn't cached. But a block
that isn't in the buffer cache is neither dirty nor under writeback. So
there's nothing to write or to wait for and reading it in is pointless.
It's also not something we can afford once inode eviction calls
->sync_inode_metadata() from reclaim. So look the block up with
sb_find_get_block_nonatomic() instead.
Note that ext4_get_inode_loc_noio() isn't usable here. It uses the
atomic lookup and that one gives up on a folio that is being migrated.
That's fine for a best-effort fast commit snapshot but not for a data
integrity sync.
Cc: stable@vger.kernel.org # v7.3
Signed-off-by: Christian Brauner (Amutable) <brauner@kernel.org>
---
fs/ext4/inode.c | 10 +++++++---
1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index b34b99d15db6..cab26ec01aba 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -6824,28 +6824,32 @@ int ext4_sync_inode_metadata(struct inode *inode, struct writeback_control *wbc)
{
struct ext4_iloc iloc;
struct mapping_metadata_bhs *mmb;
+ ext4_fsblk_t block;
int err;
/* We should only get here in nojournal mode */
if (WARN_ON_ONCE(EXT4_SB(inode->i_sb)->s_journal))
return -EFSCORRUPTED;
- err = __ext4_get_inode_loc_noinmem(inode, &iloc);
+ err = ext4_inode_loc_block(inode, &iloc, &block);
if (err)
return err;
mmb = READ_ONCE(EXT4_I(inode)->i_metadata_bhs);
if (mmb) {
err = mmb_sync(mmb);
if (err)
- goto out;
+ return err;
}
+ /* A block that isn't cached is neither dirty nor under writeback. */
+ iloc.bh = sb_find_get_block_nonatomic(inode->i_sb, block);
+ if (!iloc.bh)
+ return 0;
sync_dirty_buffer(iloc.bh);
if (buffer_write_io_error(iloc.bh)) {
ext4_error_inode_block(inode, iloc.bh->b_blocknr, EIO,
"IO error syncing inode");
err = -EIO;
}
-out:
brelse(iloc.bh);
return err;
}
--
2.53.0
next prev parent reply other threads:[~2026-10-29 13:28 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-10-29 13:28 [PATCH RFC 0/9] fs: reclaim inodes with I_METADATA_WRITEBACK Christian Brauner
2026-10-29 13:28 ` [PATCH RFC 1/9] ext2: add ext2_inode_block() helper Christian Brauner
2026-10-29 13:31 ` sashiko-bot
2026-10-29 13:28 ` [PATCH RFC 2/9] ext2: don't read the inode block in ext2_sync_inode_metadata() Christian Brauner
2026-10-29 13:33 ` sashiko-bot
2026-10-29 13:28 ` [PATCH RFC 3/9] ext4: add ext4_inode_loc_block() helper Christian Brauner
2026-10-29 13:32 ` sashiko-bot
2026-10-29 13:28 ` Christian Brauner [this message]
2026-10-29 13:34 ` [PATCH RFC 4/9] ext4: don't read the inode block in ext4_sync_inode_metadata() sashiko-bot
2026-10-29 13:28 ` [PATCH RFC 5/9] minix: add minix_inode_block() helper Christian Brauner
2026-10-29 13:34 ` sashiko-bot
2026-10-29 13:28 ` [PATCH RFC 6/9] minix: don't read the inode block in minix_sync_inode_metadata() Christian Brauner
2026-10-29 13:32 ` sashiko-bot
2026-10-29 13:28 ` [PATCH RFC 7/9] udf: don't allocate the inode block in udf_sync_inode_metadata() Christian Brauner
2026-10-29 13:35 ` sashiko-bot
2026-10-29 13:28 ` [PATCH RFC 8/9] fat: sync the inode in fat_free_eofblocks() Christian Brauner
2026-10-29 13:37 ` sashiko-bot
2026-10-29 13:28 ` [PATCH RFC 9/9] fs: don't strand I_METADATA_WRITEBACK inodes outside the LRU Christian Brauner
2026-10-29 13:46 ` sashiko-bot
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=20261029-work-inode-metadata_writeback-v1-4-7d6c678a585d@kernel.org \
--to=brauner@kernel.org \
--cc=adilger.kernel@dilger.ca \
--cc=hirofumi@mail.parknet.co.jp \
--cc=jack@suse.cz \
--cc=josef@toxicpanda.com \
--cc=libaokun@linux.alibaba.com \
--cc=linux-ext4@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=mason@kernel.org \
--cc=namjae.jeon@samsung.com \
--cc=ojaswin@linux.ibm.com \
--cc=ritesh.list@gmail.com \
--cc=stable@vger.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.