[Bug 256712] UFS: kernel panic: ffs_blk free_cg: freeing free frag – in response to pkg-delete(8) soon after login

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=256712

--- Comment #7 from andrew@tao11.riddles.org.uk ---
I've been trying to help Graham reproduce this one (per discussions on the
libera #freebsd irc channel) but so far what I seem to have managed is to
produce a different error:

** SU+J Recovering /dev/vtbd1
** Reading 4194304 byte journal from inode 4.
** Building recovery table.
** Resolving unreferenced inode list.
** Processing journal entries.
/dev/vtbd1: ino_remref: ino 2767 mode 0100000 about to go negative
/dev/vtbd1: UNEXPECTED SU+J INCONSISTENCY
/dev/vtbd1: INTERNAL ERROR: GOT TO reply()
/dev/vtbd1: UNEXPECTED SOFT UPDATE INCONSISTENCY; RUN fsck MANUALLY.

I've managed to reproduce this one fairly reliably (3 times out of 4 tries) and
I have a copy of a filesystem image taken before the fsck. The steps I used to
reproduce were (this is in a 13-stable install in a bhyve vm, where vtbd0 is
the system image and vtbd1 is a 256MB md device on the host):

newfs -O2 -U -j /dev/vtbd1
mount /dev/vtbd1 /mnt
tar -xf python38-3.8.10.tgz -C /mnt
find /mnt/usr -type f -print0 | xargs -0 rm
tar -xf python38-3.8.10.tgz -C /mnt
(when the prompt returns at this point, immediately destroy the vm)

On booting the vm again and then doing  fsck -p /dev/vtbd1, the above error
appears. (vtbd1 is not mentioned in fstab)

Should I split this off into a separate bug? (I will continue to try to
reproduce the original issue)

-- 
You are receiving this mail because:
You are the assignee for the bug.