ACK/Cmnt: [PATCH 0/2] [SRU] [X,B:linux] CVE-2018-13095

[Stefan Bader]

On 10.03.21 22:18, Tim Gardner wrote:
> [Impact]
> An issue was discovered in fs/xfs/libxfs/xfs_inode_buf.c in the Linux kernel
> through 4.17.3. A denial of service (memory corruption and BUG) can occur
> for a corrupted xfs image upon encountering an inode that is in extent
> format, but has more extents than fit in the inode fork.
> 
> [Test Case]
> iozone -a, misc rsync's
> 
> [Potential regression]
> There appears to be little chance of data corruption. The bigger chance is that
> some file reads/writes may be refused.
> 
> 

I wonder whether it would not have been more straight forward to make the new 
verifier return a bool and adapt the calls to if(!dinode_verify_fork()) return 
false; But ok, the current form sticks closer to upstream.

Acked-by: Stefan Bader <stefan.bader at canonical.com>

-------------- next part --------------
A non-text attachment was scrubbed...
Name: OpenPGP_signature
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL: <https://lists.ubuntu.com/archives/kernel-team/attachments/20210311/ff6d8535/attachment.sig>