NetBSD-Bugs archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: kern/60621 (panic: ufs_inactive: dirty filesystem?)



> Date: Tue, 17 Nov 2026 18:37:54 +0000
> From: "J. Hannken-Illjes" <hannken%netbsd.org@localhost>
> 
> On Tue, Nov 17, 2026 at 06:11:53PM +0000, Taylor R Campbell wrote:
> > > Date: Sun, 15 Nov 2026 14:32:10 +0000
> > > From: "J. Hannken-Illjes" <hannken%netbsd.org@localhost>
> [snip]
> > > Using a persistent snapshot or no snapshot the "rm" succeeds.
> > 
> > I assume you mean `Using a *non*-persistent snapshot or no snapshot'?
> 
> I meant what I wrote: a persistent snapshot.  As long as the snapshot
> is not unlinked it will sync its data bypassing wapbl as the snapshot
> activates BEFORE wapbl on mount.

Oh, I see -- you meant a file-system-internal snapshot that has _not_
been unlinked (as in, e.g., `fssconfig fss0 / /mysnap'), as opposed to
an fs-internal snapshot that _has_ been unlinked (as in, e.g.,
`fssconfig -x fss0 / /mysnap' or simply `fssconfig fss0 / /', which
triggers the crash).  Is that right?

I was confused by the term `persistent snapshot' here because the
language in the fssconfig(8) has long used the term `persistent
snapshot' to mean any fs-internal snapshots, whether or not they have
been unlinked:

     If backup resides on the snapshotted file system a persistent snapshot
     will be created.  This snapshot is active until backup is unlinked.  Per-
     sistent snapshots save the file system as if it had been unmounted.

     https://man.netbsd.org/NetBSD-10.x-BRANCH/fssconfig.8

And in rewriting the fssconfig(8) man page (and
<https://wiki.NetBSD.org/tutorials/how_to_use_snapshots>) I made sure
not to change that terminology even though it didn't make much sense
to me -- I think `external' vs `internal' is better terminology for
the mode of operation, and `persistent' is better limited to internal
snapshots that haven't been unlinked.

So if I understand correctly, the reproducer:

- crashes with internal unlinked snapshot
- works with internal still-linked (`persistent') snapshot
- works with no snapshot
- presumably works with external snapshot, whether linked or unlinked

This suggests a (purely speculative, unconfirmed) hypothesis about the
`ufs_inactive: dirty filesystem?' panic, that the fast path of writing
to unlinked internal snapshots doesn't work with wapbl because it
allocates blocks and updates i_blocks without grouping them into a
single atomic transaction.



Home | Main Index | Thread Index | Old Index