On Fri, 12 Apr 2024 at 21:09, Stefan Berger <stefanb@xxxxxxxxxxxxx> wrote:
> I was hoping that this would be sufficiently generic to work with
> potential future stacked filesystems as well that would need to also
> provide support for D_REAL_FILEDATA.
I also have very bad feelings from IMA digging in the internals of overlayfs.
We should strive to get rid of remaining d_real() instances, not add more.
On a related note, D_REAL_METADATA was apparently added for IMA
(commit 11b3f8ae7081 ("fs: remove the inode argument to ->d_real()
method")), but there's no current user. What's up with this?
Thanks,
Miklos
,
