From 9cc3b0eda94da41efc4b4ac85b7e6f18cf72cd67 Mon Sep 17 00:00:00 2001 From: Daejun Park Date: Thu, 29 Oct 2026 11:23:06 +0900 Subject: [PATCH] xfs: make pNFS LAYOUTCOMMIT stable the way fdatasync does Every LAYOUTCOMMIT commits a synchronous transaction and forces the log. fdatasync after the same overwrite does not, because timestamps alone do not set ili_datasync_seq. Do what fdatasync does. Commit without XFS_TRANS_SYNC, force the log up to ili_datasync_seq, and flush the data device cache if that wrote no iclog. Conversions and size changes still force the log. As with fdatasync, a crash before the next log write can lose the new timestamps and the change attribute that nfsd derives from the ctime. Signed-off-by: Daejun Park Reviewed-by: Christoph Hellwig Signed-off-by: Carlos Maiolino --- fs/xfs/xfs_pnfs.c | 21 +++++++++++++++------ 1 file changed, 15 insertions(+), 6 deletions(-) diff --git a/fs/xfs/xfs_pnfs.c b/fs/xfs/xfs_pnfs.c index f4c4a02e5c9a..065875c8b7da 100644 --- a/fs/xfs/xfs_pnfs.c +++ b/fs/xfs/xfs_pnfs.c @@ -295,10 +295,12 @@ xfs_pnfs_validate_isize( * converting any unwritten extents, flushing the disk cache and updating the * time stamps. * - * Note that we rely on the caller to always send us a timestamp update so that - * we always commit a transaction here. If that stops being true we will have - * to manually flush the cache here similar to what the fsync code path does - * for datasyncs on files that have no dirty metadata. + * Like fdatasync, a commit that only updates the timestamps does not force the + * log: they are logged as XFS_ILOG_TIMESTAMP and reach the disk with a later + * log write. The log is forced up to the last change that a datasync needs, + * such as the unwritten extent conversions or a new size, and the cache of the + * device holding the file data is flushed if that log force did not write an + * iclog itself. */ static int xfs_fs_commit_blocks( @@ -312,6 +314,7 @@ xfs_fs_commit_blocks( struct xfs_trans *tp; struct timespec64 now; bool update_isize = false; + int log_flushed = 0; int error, i; loff_t size; @@ -368,7 +371,8 @@ xfs_fs_commit_blocks( xfs_ilock(ip, XFS_ILOCK_EXCL); xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); - xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); + xfs_trans_log_inode(tp, ip, + update_isize ? XFS_ILOG_CORE : XFS_ILOG_TIMESTAMP); now = inode_set_ctime_current(inode); inode_set_atime_to_ts(inode, now); @@ -379,8 +383,13 @@ xfs_fs_commit_blocks( ip->i_disk_size = new_size; } - xfs_trans_set_sync(tp); error = xfs_trans_commit(tp); + if (error) + goto out_drop_iolock; + + error = xfs_log_force_inode(ip, true, &log_flushed); + if (!error && !log_flushed) + error = blkdev_issue_flush(xfs_inode_buftarg(ip)->bt_bdev); out_drop_iolock: xfs_iunlock(ip, XFS_IOLOCK_EXCL);