On Sat, Nov 16, 2024 at 09:39:50AM +0800, Lizhi Xu wrote: > On Fri, 15 Nov 2024 13:24:55 +0000, Al Viro wrote: > > On Fri, Nov 15, 2024 at 01:06:15PM +0000, Al Viro wrote: > > > On Fri, Nov 15, 2024 at 05:49:08PM +0800, Lizhi Xu wrote: > > > > syzbot reported a null-ptr-deref in pick_link. [1] > > > > When symlink's inode is corrupted, the value of the i_link is 2 in this case, > > > > it will trigger null pointer deref when accessing *res in pick_link(). > > > > > > > > To avoid this issue, add a check for inode mode, return -EINVAL when it's > > > > not symlink. > > > > > > NAK. Don't paper over filesystem bugs at pathwalk time - it's the wrong > > > place for that. Fix it at in-core inode creation time. > > > > BTW, seeing that ntfs doesn't even touch ->i_link, you are dealing > Yes, ntfs3 does not handle the relevant code of i_link. > > with aftermath of memory corruption, so it's definitely papering over > > the actual bug here. > I see that finding out how the value of i_link becomes 2 is the key. How about 'how the memory currently pointed to by inode had come to be available for use by something that stored 2 at that particular offset'?
