On Sun, Jan 19, 2020 at 01:45:15PM +0800, Murphy Zhou wrote: > Hi Eric, > > On Tue, Oct 15, 2019 at 11:16:40AM -0700, Eric Biggers wrote: > > From: Eric Biggers <ebiggers@xxxxxxxxxx> > > > > Test non-root use of the fscrypt filesystem-level encryption keyring and > > v2 encryption policies. > > This testcase now fails on latest Linus tree with latest e2fsprogs > on ext4: > > FSTYP -- ext4 > PLATFORM -- Linux/x86_64 dell-pesc430-01 5.4.0+ #1 SMP Mon Nov 25 16:40:55 EST 2019 > MKFS_OPTIONS -- /dev/sda3 > MOUNT_OPTIONS -- -o acl,user_xattr -o context=system_u:object_r:nfs_t:s0 /dev/sda3 /mnt/xfstests/mnt2 > generic/581 - output mismatch (see /var/lib/xfstests/results//generic/581.out.bad) > --- tests/generic/581.out 2019-11-25 20:30:04.536051638 -0500 > +++ /var/lib/xfstests/results//generic/581.out.bad 2019-11-26 01:04:17.318332220 -0500 > @@ -33,7 +33,7 @@ > Added encryption key > Added encryption key > Added encryption key > -Error adding encryption key: Disk quota exceeded > +Added encryption key > > # Adding key as root > ... > > A rough looking back shows that this probably started since your fscrypt > update for 5.5, added support for IV_INO_LBLK_64 encryption policies. > > I guess you are aware of this :-) > Nope, this has been passing for me. I don't see how this can have anything to do with the fscrypt update for 5.5. I'm guessing this is a race condition in the test caused by the kernel's keyrings subsystem reclaiming the keys quota asynchronously. I'll see if I can find a way to reproduce it and make the test more reliable. - Eric
