| Message ID | 20200911211603.5653-1-costa@purestorage.com |
|---|---|
| State | Superseded |
| Headers | show |
| Series | ext4: Fix superblock checksum calculation race | expand |
On Fri 11-09-20 15:16:03, Constantine Sapuntzakis wrote: > The race condition could cause the persisted superblock checksum > to not match the contents of the superblock, causing the > superblock to be considered corrupt. > > An example of the race follows. A first thread is interrupted in the > middle of a checksum calculation. Then, another thread changes the > superblock, calculates a new checksum, and sets it. Then, the first > thread resumes and sets the checksum based on the older superblock. > > To fix, serialize the superblock checksum calculation using the buffer > header lock. While a spinlock is sufficient, the buffer header is > already there and there is precedent for locking it (e.g. in > ext4_commit_super). > > Tested the patch by booting up a kernel with the patch, creating > a filesystem and some files (including some orphans), and then > unmounting and remounting the file system. > > Suggested-by: Jan Kara <jack@suse.cz> Thanks for the patch! Please add your Signed-off-by line to the patch to certify that you've written the patch and agree with it being included in the kernel (see "Developer's Certificate of Origin 1.1" in Documentation/process/submitting-patches.rst for more details). Without it the patch cannot be included. Otherwise it looks good so feel free to add: Reviewed-by: Jan Kara <jack@suse.cz> Thanks Honza > --- > fs/ext4/super.c | 11 +++++++++++ > 1 file changed, 11 insertions(+) > > diff --git a/fs/ext4/super.c b/fs/ext4/super.c > index ea425b49b345..3f7fdce5ab05 100644 > --- a/fs/ext4/super.c > +++ b/fs/ext4/super.c > @@ -201,7 +201,18 @@ void ext4_superblock_csum_set(struct super_block *sb) > if (!ext4_has_metadata_csum(sb)) > return; > > + /* > + * Locking the superblock prevents the scenario > + * where: > + * 1) a first thread pauses during checksum calculation. > + * 2) a second thread updates the superblock, recalculates > + * the checksum, and updates s_checksum > + * 3) the first thread resumes and finishes its checksum calculation > + * and updates s_checksum with a potentially stale or torn value. > + */ > + lock_buffer(EXT4_SB(sb)->s_sbh); > es->s_checksum = ext4_superblock_csum(sb, es); > + unlock_buffer(EXT4_SB(sb)->s_sbh); > } > > ext4_fsblk_t ext4_block_bitmap(struct super_block *sb, > -- > 2.17.1 >
diff --git a/fs/ext4/super.c b/fs/ext4/super.c index ea425b49b345..3f7fdce5ab05 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -201,7 +201,18 @@ void ext4_superblock_csum_set(struct super_block *sb) if (!ext4_has_metadata_csum(sb)) return; + /* + * Locking the superblock prevents the scenario + * where: + * 1) a first thread pauses during checksum calculation. + * 2) a second thread updates the superblock, recalculates + * the checksum, and updates s_checksum + * 3) the first thread resumes and finishes its checksum calculation + * and updates s_checksum with a potentially stale or torn value. + */ + lock_buffer(EXT4_SB(sb)->s_sbh); es->s_checksum = ext4_superblock_csum(sb, es); + unlock_buffer(EXT4_SB(sb)->s_sbh); } ext4_fsblk_t ext4_block_bitmap(struct super_block *sb,