Commit d967df65 authored Sep 05, 2022 by Zhang Qilong Committed by Greg Kroah-Hartman Oct 26, 2022

f2fs: fix race condition on setting FI_NO_EXTENT flag



[ Upstream commit 07725adc55c0a414c10acb5c8c86cea34b95ddef ]

The following scenarios exist.
process A:               process B:
->f2fs_drop_extent_tree  ->f2fs_update_extent_cache_range
                          ->f2fs_update_extent_tree_range
                           ->write_lock
 ->set_inode_flag
                           ->is_inode_flag_set
                           ->__free_extent_tree // Shouldn't
                                                // have been
                                                // cleaned up
                                                // here
  ->write_lock

In this case, the "FI_NO_EXTENT" flag is set between
f2fs_update_extent_tree_range and is_inode_flag_set
by other process. it leads to clearing the whole exten
tree which should not have happened. And we fix it by
move the setting it to the range of write_lock.

Fixes:5f281fab ("f2fs: disable extent_cache for fcollapse/finsert inodes")
Signed-off-by: Zhang Qilong <zhangqilong3@huawei.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

parent 90e483e7

fs/f2fs/extent_cache.c

+1 −2

Original line number	Diff line number	Diff line
		@@ -717,9 +717,8 @@ void f2fs_drop_extent_tree(struct inode *inode)
		if (!f2fs_may_extent_tree(inode))
		return;

		set_inode_flag(inode, FI_NO_EXTENT);

		write_lock(&et->lock);
		set_inode_flag(inode, FI_NO_EXTENT);
		__free_extent_tree(sbi, et);
		if (et->largest.len) {
		et->largest.len = 0;