]> git.ipfire.org Git - thirdparty/kernel/stable.git/commitdiff
ext4: wait for ongoing I/O to complete before freeing blocks
authorZhang Yi <yi.zhang@huawei.com>
Tue, 16 Sep 2025 09:33:37 +0000 (17:33 +0800)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Thu, 23 Oct 2025 14:20:19 +0000 (16:20 +0200)
commit 328a782cb138029182e521c08f50eb1587db955d upstream.

When freeing metadata blocks in nojournal mode, ext4_forget() calls
bforget() to clear the dirty flag on the buffer_head and remvoe
associated mappings. This is acceptable if the metadata has not yet
begun to be written back. However, if the write-back has already started
but is not yet completed, ext4_forget() will have no effect.
Subsequently, ext4_mb_clear_bb() will immediately return the block to
the mb allocator. This block can then be reallocated immediately,
potentially causing an data corruption issue.

Fix this by clearing the buffer's dirty flag and waiting for the ongoing
I/O to complete, ensuring that no further writes to stale data will
occur.

Fixes: 16e08b14a455 ("ext4: cleanup clean_bdev_aliases() calls")
Cc: stable@kernel.org
Reported-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Closes: https://lore.kernel.org/linux-ext4/a9417096-9549-4441-9878-b1955b899b4e@huaweicloud.com/
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Message-ID: <20250916093337.3161016-3-yi.zhang@huaweicloud.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
fs/ext4/ext4_jbd2.c

index da4a8245638364614f0bbac3cd73e87f5e503d98..9e7275dd901f4989e2877232ad929ea47010f5ca 100644 (file)
@@ -276,9 +276,16 @@ int __ext4_forget(const char *where, unsigned int line, handle_t *handle,
                  bh, is_metadata, inode->i_mode,
                  test_opt(inode->i_sb, DATA_FLAGS));
 
-       /* In the no journal case, we can just do a bforget and return */
+       /*
+        * In the no journal case, we should wait for the ongoing buffer
+        * to complete and do a forget.
+        */
        if (!ext4_handle_valid(handle)) {
-               bforget(bh);
+               if (bh) {
+                       clear_buffer_dirty(bh);
+                       wait_on_buffer(bh);
+                       __bforget(bh);
+               }
                return 0;
        }