Re: [PATCH 2/2] xfs: test that a grown RT section can be filled

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

On Tue, Jun 09, 2026 at 10:00:44AM +0200, Christoph Hellwig wrote:
> This reproduces issues with zoned growfs.
> 
> Signed-off-by: Christoph Hellwig <hch@lst.de>

Looks good to me
Reviewed-by: "Darrick J. Wong" <djwong@kernel.org>

--D

> ---
>  tests/xfs/4201     | 58 ++++++++++++++++++++++++++++++++++++++++++++++
>  tests/xfs/4201.out |  3 +++
>  2 files changed, 61 insertions(+)
>  create mode 100755 tests/xfs/4201
>  create mode 100644 tests/xfs/4201.out
> 
> diff --git a/tests/xfs/4201 b/tests/xfs/4201
> new file mode 100755
> index 000000000000..1f1c8e7eb140
> --- /dev/null
> +++ b/tests/xfs/4201
> @@ -0,0 +1,58 @@
> +#! /bin/bash
> +# SPDX-License-Identifier: GPL-2.0-or-later
> +# Copyright (c) 2026 Christoph Hellwig.
> +#
> +# FS QA Test No. 4201
> +#
> +# Test that the entire device capacity can be used after resizing a realtime
> +# device.
> +#
> +. ./common/preamble
> +_begin_fstest auto quick realtime growfs zone
> +
> +. ./common/filter
> +. ./common/zoned
> +
> +_require_scratch
> +_require_external_realtime_device
> +
> +minrtdevsize=$((5 * 1024 * 1024)) #kB
> +rtdevsize=`_get_device_size $SCRATCH_RTDEV`
> +if [ $rtdevsize -lt $minrtdevsize ]; then
> +	_notrun "SCRATCH_RTDEV device too small, $rtdevsize < $minrtdevsize"
> +fi
> +
> +_scratch_mkfs -r size=1g  >> $seqres.full 2>&1
> +rblocks=`_scratch_xfs_get_sb_field rblocks`
> +
> +_scratch_mount
> +_xfs_force_bdev realtime $SCRATCH_MNT
> +free_zones_pre=$(_xfs_get_mountstat $SCRATCH_MNT "free zones:")
> +
> +echo "Growing file system"
> +$XFS_GROWFS_PROG $SCRATCH_MNT -R $((2 * rblocks)) >> $seqres.full 2>&1 || \
> +	_fail "growfs failed"
> +
> +# For zoned file systems, check that the growfs did increase the available free
> +# zones.  This avoids a hang on unfixed kernels when filling later.
> +free_zones_post=$(_xfs_get_mountstat $SCRATCH_MNT "free zones:")
> +if _has_fs_sysfs_attr $SCRATCH_DEV "zoned/max_open_zones"; then
> +	if [ "$free_zones_pre" == "$free_zones_post" ]; then
> +		_fail "growfs did not increase free zones"
> +	fi
> +fi
> +
> +echo "Filling grown file system"
> +writesize_mb=32
> +dd if=/dev/zero of=$SCRATCH_MNT/fill bs=${writesize_mb}M oflag=direct \
> +	>> $seqres.full 2>&1
> +
> +blocksize=$(_get_block_size $SCRATCH_MNT)
> +free_blocks=$(stat -f $SCRATCH_MNT -c '%f')
> +
> +# check that we really mostly filled the file system
> +_within_tolerance "free space after fill" \
> +	$free_blocks $((writesize_mb * 1024 * 1024 / blocksize)) 100%
> +
> +status=0
> +exit
> diff --git a/tests/xfs/4201.out b/tests/xfs/4201.out
> new file mode 100644
> index 000000000000..2e7ac4019f3f
> --- /dev/null
> +++ b/tests/xfs/4201.out
> @@ -0,0 +1,3 @@
> +QA output created by 4201
> +Growing file system
> +Filling grown file system
> -- 
> 2.53.0
> 
>