Re: [PATCH 03/11] xfs_scrub: handle media scans of internal rt devices correctly

["Darrick J. Wong" <djwong@kernel.org>]

On Thu, Jun 25, 2026 at 02:01:19PM +0200, Christoph Hellwig wrote:
> On Wed, Jun 24, 2026 at 11:15:17AM -0700, Darrick J. Wong wrote:
> > From: Darrick J. Wong <djwong@kernel.org>
> > 
> > Codex noticed media scans of internal rt devices don't work at all
> > correctly.  First, we fail to allocate a ctx->verify_disks[XFS_DEV_RT]
> > for the internal rt section, and even if we did, phase 6 doesn't
> > allocate media_verify_state.rvp[XFS_DEV_RT] if there's a media error on
> > an internal rt volume, so we'll crash there too.
> > 
> > Fix both issues to make it work properly.
> 
> Do you have a test case for this?

Sort of -- xfs/556 is the media verification test, but it _notruns on
zoned filesystems because of _require_xfs_scratch_non_zoned.  The
_notrun logic isn't quite right because while dm-error doesn't work for
zoned disks, it works fine with -rzoned=1 on conventional storage.

Well, not entirely -- with zoned=1 and conventional external rt storage
it works fine, but with an internal rt volume, it does:

--- /run/fstests/bin/tests/xfs/556.out  2025-07-15 14:41:40.342952932 -0700
+++ /run/fstests/logs/xfs/556.out.bad   2026-06-25 15:35:59.593600834 -0700
@@ -1,12 +1,7 @@
 QA output created by 556
+blockdev: cannot open : No such file or directory
 Scrub for injected media error (single threaded)
-Unfixable Error: SCRATCH_MNT/a: media error at data offset 2FSB length 1FSB.
-SCRATCH_MNT: unfixable errors found: 1
 Scrub for injected media error (multi threaded)
-Unfixable Error: SCRATCH_MNT/a: media error at data offset 2FSB length 1FSB.
-SCRATCH_MNT: unfixable errors found: 1
-od: SCRATCH_MNT/a: Input/output error
 Scrub for injected media error (after remount)
-Unfixable Error: SCRATCH_MNT/a: media error at data offset 2FSB length 1FSB.
-SCRATCH_MNT: unfixable errors found: 1
+blockdev: cannot open : No such file or directory
 Scrub after removing injected media error

So I have to patch fstests further to avoid setting errordev=RT when
there's no external volume.

--D

> The changes look good, but I'd love to have coverage for it...
> 
> Reviewed-by: Christoph Hellwig <hch@lst.de>
> 
>