> This would be more logical above where inode content is initialized (and
> less errorprone just in case security_inode_alloc() grows dependency on
> i_state value) - like just after:
> 
>       inode->i_flags = 0;

Fixed that in-tree. Thanks!

Reply via email to