Re: [PATCH 2/2] ext4/065 encryption + casefold + dirdata feature combination
"Theodore Tso" <[email protected]>
| Newsgroups | org.kernel.vger.fstests |
|---|---|
| Message-ID | <[email protected]> |
On Sun, Apr 19, 2026 at 02:52:09PM -0400, Artem Blagodarenko wrote: > From: Artem Blagodarenko <[email protected]> > > Test ext4 encryption + casefold + dirdata feature combination. > This test verifies that files created in directories with encryption, > case-insensitive (casefold), and dirdata attributes work correctly. > See ext4/064 for the same test WITHOUT dirdata feature. > > Signed-off-by: Artem Blagodarenko <[email protected]> Note that the dirdata feature patches have not landed, and I've expressed concerns of this patch set[1]. [1] https://lore.kernel.org/all/[email protected]/ From manual code inspection, the dirdata patch was never changing how the data was stored. So before we land the xfstests changes, could you please add support in debugfs to set and get the dirdata stored for a directory entry, and then change the test to actually _validate_ that when enabling encryption + casefold + dirdata, that the encrypted hash is stored as a dirdata attribute? Thanks, - Ted