Re: KASAN: use-after-free Write in jbd2_log_do_checkpoint

[Jan Kara]

On Mon 17-09-18 04:19:02, syzbot wrote:
> Hello,
> 
> syzbot found the following crash on:
> 
> HEAD commit:    2e2a0c961a87 Merge branch 'progarray_mapinmap_dump'
> git tree:       bpf-next
> console output: https://syzkaller.appspot.com/x/log.txt?x=17622811400000
> kernel config:  https://syzkaller.appspot.com/x/.config?x=8f59875069d721b6
> dashboard link: https://syzkaller.appspot.com/bug?extid=7f4a27091759e2fe7453
> compiler:       gcc (GCC) 8.0.1 20180413 (experimental)
> 
> Unfortunately, I don't have any reproducer for this crash yet.
> 
> IMPORTANT: if you fix the bug, please add the following tag to the commit:
> Reported-by: syzbot+7f4a27091759e2fe7453@xxxxxxxxxxxxxxxxxxxxxxxxx
> 
> ==================================================================
> BUG: KASAN: use-after-free in atomic_inc
> include/asm-generic/atomic-instrumented.h:109 [inline]
> BUG: KASAN: use-after-free in get_bh include/linux/buffer_head.h:283
> [inline]
> BUG: KASAN: use-after-free in jbd2_log_do_checkpoint+0x113d/0x13e0
> fs/jbd2/checkpoint.c:337
> Write of size 4 at addr ffff8801b6d96300 by task syz-executor1/904

This is at:

                if (buffer_locked(bh)) {
                        spin_unlock(&journal->j_list_lock);
                        get_bh(bh);
                        wait_on_buffer(bh);
                        /* the journal_head may have gone by now */
                        BUFFER_TRACE(bh, "brelse");
                        __brelse(bh);
                        spin_lock(&journal->j_list_lock);
                        goto restart2;
                }

And indeed that get_bh() should be called before we unlock j_list_lock as
otherwise we can race with jbd2_journal_try_to_free_buffers(). I'll send a
fix.

								Honza

-- 
Jan Kara <jack@xxxxxxxx>
SUSE Labs, CR