From: Jan Kara <jack@suse.cz>
To: <linux-fsdevel@vger.kernel.org>
Cc: Christian Brauner <brauner@kernel.org>,
aivazian.tigran@gmail.com,
OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>,
Ted Tso <tytso@mit.edu>, <linux-ext4@vger.kernel.org>,
Jan Kara <jack@suse.cz>
Subject: [PATCH 9/9] ext4: Use mmb infrastructure for inode buffer writeout
Date: Mon, 11 May 2026 14:13:59 +0200 [thread overview]
Message-ID: <20260511121356.241821-18-jack@suse.cz> (raw)
In-Reply-To: <20260511115725.28441-1-jack@suse.cz>
Use mmb inode buffer writeout infrastructure to reliably write out
inode's inode table block on fsync(2) in nojournal mode (from
ext4_sync_parent() and ext4_fsync_nojournal()). This significantly
simplifies the code as we don't have to explicitely handle inode buffer
writeback in ext4_write_inode() and thus we can also remove
sync_inode_metadata() calls from ext4_sync_parent() and
ext4_write_inode() call from ext4_fsync_nojournal().
Signed-off-by: Jan Kara <jack@suse.cz>
---
fs/ext4/ext4_jbd2.c | 2 +-
fs/ext4/ext4_jbd2.h | 2 ++
fs/ext4/fsync.c | 12 ------------
fs/ext4/inode.c | 24 +++++-------------------
4 files changed, 8 insertions(+), 32 deletions(-)
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index 74f05bd0cdde..6bbaf72108fd 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -350,7 +350,7 @@ int __ext4_journal_get_create_access(const char *where, unsigned int line,
return 0;
}
-static void ext4_inode_attach_mmb(struct inode *inode)
+void ext4_inode_attach_mmb(struct inode *inode)
{
struct mapping_metadata_bhs *mmb;
diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
index 63d17c5201b5..2a01b8279c88 100644
--- a/fs/ext4/ext4_jbd2.h
+++ b/fs/ext4/ext4_jbd2.h
@@ -122,6 +122,8 @@
#define EXT4_HT_EXT_CONVERT 11
#define EXT4_HT_MAX 12
+void ext4_inode_attach_mmb(struct inode *inode);
+
int
ext4_mark_iloc_dirty(handle_t *handle,
struct inode *inode,
diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c
index e25d365e1179..af84489e57c6 100644
--- a/fs/ext4/fsync.c
+++ b/fs/ext4/fsync.c
@@ -75,9 +75,6 @@ static int ext4_sync_parent(struct inode *inode)
if (ret)
break;
}
- ret = sync_inode_metadata(inode, 1);
- if (ret)
- break;
}
dput(dentry);
return ret;
@@ -87,10 +84,6 @@ static int ext4_fsync_nojournal(struct file *file, loff_t start, loff_t end,
int datasync, bool *needs_barrier)
{
struct inode *inode = file->f_inode;
- struct writeback_control wbc = {
- .sync_mode = WB_SYNC_ALL,
- .nr_to_write = 0,
- };
int ret;
ret = mmb_fsync_noflush(file, EXT4_I(inode)->i_metadata_bhs,
@@ -98,11 +91,6 @@ static int ext4_fsync_nojournal(struct file *file, loff_t start, loff_t end,
if (ret)
return ret;
- /* Force writeout of inode table buffer to disk */
- ret = ext4_write_inode(inode, &wbc);
- if (ret)
- return ret;
-
ret = ext4_sync_parent(inode);
if (test_opt(inode->i_sb, BARRIER))
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 3e66e9510909..09506b4de1b2 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -5786,24 +5786,6 @@ int ext4_write_inode(struct inode *inode, struct writeback_control *wbc)
err = ext4_fc_commit(EXT4_SB(inode->i_sb)->s_journal,
EXT4_I(inode)->i_sync_tid);
- } else {
- struct ext4_iloc iloc;
-
- err = __ext4_get_inode_loc_noinmem(inode, &iloc);
- if (err)
- return err;
- /*
- * sync(2) will flush the whole buffer cache. No need to do
- * it here separately for each inode.
- */
- if (wbc->sync_mode == WB_SYNC_ALL && !wbc->for_sync)
- sync_dirty_buffer(iloc.bh);
- if (buffer_req(iloc.bh) && !buffer_uptodate(iloc.bh)) {
- ext4_error_inode_block(inode, iloc.bh->b_blocknr, EIO,
- "IO error syncing inode");
- err = -EIO;
- }
- brelse(iloc.bh);
}
return err;
}
@@ -6348,7 +6330,11 @@ int ext4_mark_iloc_dirty(handle_t *handle,
/* the do_update_inode consumes one bh->b_count */
get_bh(iloc->bh);
-
+ if (!ext4_handle_valid(handle)) {
+ if (!EXT4_I(inode)->i_metadata_bhs)
+ ext4_inode_attach_mmb(inode);
+ EXT4_I(inode)->i_metadata_bhs->inode_blk = iloc->bh->b_blocknr;
+ }
/* ext4_do_update_inode() does jbd2_journal_dirty_metadata */
err = ext4_do_update_inode(handle, inode, iloc);
put_bh(iloc->bh);
--
2.51.0
next prev parent reply other threads:[~2026-05-11 12:14 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-05-11 12:13 [PATCH 0/9] fs: Fix missed inode write during fsync Jan Kara
2026-05-11 12:13 ` [PATCH 1/9] affs: Drop support for metadata bh tracking Jan Kara
2026-05-11 12:13 ` [PATCH 2/9] ext4: Allocate mapping_metadata_bhs struct on demand Jan Kara
2026-05-11 12:13 ` [PATCH 3/9] fs: Writeout inode buffer from mmb_sync() Jan Kara
2026-05-11 13:27 ` Christian Brauner
2026-05-11 12:13 ` [PATCH 4/9] ext2: Fix possibly missing inode write on fsync(2) Jan Kara
2026-05-11 12:13 ` [PATCH 5/9] udf: " Jan Kara
2026-05-11 12:13 ` [PATCH 6/9] fat: " Jan Kara
2026-05-11 14:32 ` OGAWA Hirofumi
2026-05-11 17:03 ` Jan Kara
2026-05-11 18:02 ` OGAWA Hirofumi
2026-05-12 7:29 ` Jan Kara
2026-05-12 14:17 ` OGAWA Hirofumi
2026-05-13 9:41 ` Jan Kara
2026-05-11 12:13 ` [PATCH 7/9] minix: " Jan Kara
2026-05-11 12:13 ` [PATCH 8/9] bfs: " Jan Kara
2026-05-11 12:13 ` Jan Kara [this message]
2026-05-11 13:30 ` [PATCH 9/9] ext4: Use mmb infrastructure for inode buffer writeout Christian Brauner
2026-05-11 20:49 ` [syzbot ci] Re: fs: Fix missed inode write during fsync syzbot ci
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=20260511121356.241821-18-jack@suse.cz \
--to=jack@suse.cz \
--cc=aivazian.tigran@gmail.com \
--cc=brauner@kernel.org \
--cc=hirofumi@mail.parknet.co.jp \
--cc=linux-ext4@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=tytso@mit.edu \
/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