Re: [PATCH] xfs/131: test iomap zero range via fsx and error tag

[Brian Foster <bfoster@redhat.com>]

On Tue, Aug 12, 2025 at 02:19:08AM +0800, Zorro Lang wrote:
> On Thu, Aug 07, 2025 at 10:56:26AM -0400, Brian Foster wrote:
> > iomap supports a zero range operation based on buffered writes. This
> > mechanism is used in limited spots, such as partial block zeroing,
> > etc., because usually for larger zeroing ops it is more efficient to
> > punch a hole and allocate unwritten extents.
> > 
> > This means iomap zero range has limited production test coverage
> > even though it has some particular corner cases that warrant test
> > coverage. XFS supports an error injection knob to force use of iomap
> > zero range on fallocate zero range operations. Add a test to run fsx
> > with this knob enabled to provide more zeroing test coverage.
> > 
> > Signed-off-by: Brian Foster <bfoster@redhat.com>
> > ---
> > 
> > FYI, this test is intended to go along with this patch[1] that
> > introduces the errortag.
> > 
> > [1] https://lore.kernel.org/linux-fsdevel/20250807144711.564137-8-bfoster@redhat.com/
> > 
> >  tests/xfs/131     | 26 ++++++++++++++++++++++++++
> >  tests/xfs/131.out |  2 ++
> >  2 files changed, 28 insertions(+)
> >  create mode 100755 tests/xfs/131
> >  create mode 100644 tests/xfs/131.out
> > 
> > diff --git a/tests/xfs/131 b/tests/xfs/131
> > new file mode 100755
> > index 00000000..eb8882a3
> > --- /dev/null
> > +++ b/tests/xfs/131
> > @@ -0,0 +1,26 @@
> > +#! /bin/bash
> > +# SPDX-License-Identifier: GPL-2.0
> > +# Copyright (c) 2025 Red Hat, Inc..  All Rights Reserved.
> > +#
> > +# FS QA Test 131
> > +#
> > +# Run fsx with XFS force_zero_range error injection enabled. This is a proxy
> > +# test for iomap zero range. Zero range is used in limited cases by default,
> > +# such as EOF zeroing on file extension, etc. This error tag forces use of iomap
> > +# zero range for fallocate zero range operations.
> > +#
> > +. ./common/preamble
> > +_begin_fstest auto quick zero
> > +
> > +# Import common functions.
> > +. ./common/inject
> > +
> > +# Modify as appropriate.
> > +_require_test
> > +_require_xfs_io_error_injection "force_zero_range"
> > +
> > +_test_inject_error "force_zero_range"
> > +run_fsx "-q -S 0 -N 100000"
> 
> Just check with you, if you hope to have a $TIME_FACTOR with -N at here.
> 

So do you just mean to make it 100000 * $TIME_FACTOR so the runtime can
be increased via configuration, if desired? If so, that sounds
reasonable to me (same for your tweak in the other email as well).

> This test case looks good to me with this patch:
> https://lore.kernel.org/linux-xfs/20250807144711.564137-8-bfoster@redhat.com/
> 
> Reviewed-by: Zorro Lang <zlang@redhat.com>
> 

Thanks!

Brian

> > +
> > +# success, all done
> > +_exit 0
> > diff --git a/tests/xfs/131.out b/tests/xfs/131.out
> > new file mode 100644
> > index 00000000..2ca0d932
> > --- /dev/null
> > +++ b/tests/xfs/131.out
> > @@ -0,0 +1,2 @@
> > +QA output created by 131
> > +fsx -q -S 0 -N 100000
> > -- 
> > 2.50.1
> > 
> > 
>