On Tue, Aug 30, 2022 at 11:05 PM Darrick J. Wong <djwong@xxxxxxxxxx> wrote:
>
> On Tue, Aug 30, 2022 at 12:44:32PM +0800, Murphy Zhou wrote:
> > Since this xfsprogs commit:
> > 6e0ed3d19c54 mkfs: stop allowing tiny filesystems
> > Single-AG xfs is not allowed.
> >
> > Remove agcount=1 from mkfs options and xfs/202 entirely.
>
> It's not supported for /new/ filesystems, but the rest of the tools must
> continue the same levels of support for existing filesystems, even if
> they cannot be created today.
All these changes, all of them, only fix mkfs complaints, not others.
So it is ONLY about creating new filesystems, not the existing ones.
>
> Second, there exist fstests that need to create a specific layout to
> test some part of the code. Single-AG filesystems sometimes make this
> much easier.
But it's not allowed to be made by mkfs.xfs. Most of fstests create fs
from scratch, it's hard to cover in this situation.
Thanks,
Murphy
>
> Both of these reasons are why fstests (and LTP) get a special pass on
> all the new checks in mkfs 5.19.
>
> IOWs, we still need to check that xfs_repair works ok for existing
> single AG filesystems. We can perhaps drop these tests in a decade or
> so, but now is premature.
>
> --D
>
> > Signed-off-by: Murphy Zhou <jencce.kernel@xxxxxxxxx>
> > ---
> > tests/xfs/179 | 2 +-
> > tests/xfs/202 | 40 ----------------------------------------
> > tests/xfs/202.out | 29 -----------------------------
> > tests/xfs/520 | 2 +-
> > 4 files changed, 2 insertions(+), 71 deletions(-)
> > delete mode 100755 tests/xfs/202
> > delete mode 100644 tests/xfs/202.out
> >
> > diff --git a/tests/xfs/179 b/tests/xfs/179
> > index ec0cb7e5..f0169717 100755
> > --- a/tests/xfs/179
> > +++ b/tests/xfs/179
> > @@ -22,7 +22,7 @@ _require_cp_reflink
> > _require_test_program "punch-alternating"
> >
> > echo "Format and mount"
> > -_scratch_mkfs -d agcount=1 > $seqres.full 2>&1
> > +_scratch_mkfs > $seqres.full 2>&1
> > _scratch_mount >> $seqres.full 2>&1
> >
> > testdir=$SCRATCH_MNT/test-$seq
> > diff --git a/tests/xfs/202 b/tests/xfs/202
> > deleted file mode 100755
> > index 5075d3a1..00000000
> > --- a/tests/xfs/202
> > +++ /dev/null
> > @@ -1,40 +0,0 @@
> > -#! /bin/bash
> > -# SPDX-License-Identifier: GPL-2.0
> > -# Copyright (c) 2009 Christoph Hellwig.
> > -#
> > -# FS QA Test No. 202
> > -#
> > -# Test out the xfs_repair -o force_geometry option on single-AG filesystems.
> > -#
> > -. ./common/preamble
> > -_begin_fstest repair auto quick
> > -
> > -# Import common functions.
> > -. ./common/filter
> > -. ./common/repair
> > -
> > -# real QA test starts here
> > -_supported_fs xfs
> > -
> > -# single AG will cause default xfs_repair to fail. This test is actually
> > -# testing the special corner case option needed to repair a single AG fs.
> > -_require_scratch_nocheck
> > -
> > -#
> > -# The AG size is limited to 1TB (or even less with historic xfsprogs),
> > -# so chose a small enough filesystem to make sure we can actually create
> > -# a single AG filesystem.
> > -#
> > -echo "== Creating single-AG filesystem =="
> > -_scratch_mkfs_xfs -d agcount=1 -d size=$((1024*1024*1024)) >/dev/null 2>&1 \
> > - || _fail "!!! failed to make filesystem with single AG"
> > -
> > -echo "== Trying to repair it (should fail) =="
> > -_scratch_xfs_repair
> > -
> > -echo "== Trying to repair it with -o force_geometry =="
> > -_scratch_xfs_repair -o force_geometry 2>&1 | _filter_repair
> > -
> > -# success, all done
> > -echo "*** done"
> > -status=0
> > diff --git a/tests/xfs/202.out b/tests/xfs/202.out
> > deleted file mode 100644
> > index c2c5c881..00000000
> > --- a/tests/xfs/202.out
> > +++ /dev/null
> > @@ -1,29 +0,0 @@
> > -QA output created by 202
> > -== Creating single-AG filesystem ==
> > -== Trying to repair it (should fail) ==
> > -Phase 1 - find and verify superblock...
> > -Only one AG detected - cannot validate filesystem geometry.
> > -Use the -o force_geometry option to proceed.
> > -== Trying to repair it with -o force_geometry ==
> > -Phase 1 - find and verify superblock...
> > -Phase 2 - using <TYPEOF> log
> > - - zero log...
> > - - scan filesystem freespace and inode maps...
> > - - found root inode chunk
> > -Phase 3 - for each AG...
> > - - scan and clear agi unlinked lists...
> > - - process known inodes and perform inode discovery...
> > - - process newly discovered inodes...
> > -Phase 4 - check for duplicate blocks...
> > - - setting up duplicate extent list...
> > - - check for inodes claiming duplicate blocks...
> > -Phase 5 - rebuild AG headers and trees...
> > - - reset superblock...
> > -Phase 6 - check inode connectivity...
> > - - resetting contents of realtime bitmap and summary inodes
> > - - traversing filesystem ...
> > - - traversal finished ...
> > - - moving disconnected inodes to lost+found ...
> > -Phase 7 - verify and correct link counts...
> > -done
> > -*** done
> > diff --git a/tests/xfs/520 b/tests/xfs/520
> > index d9e252bd..de70db60 100755
> > --- a/tests/xfs/520
> > +++ b/tests/xfs/520
> > @@ -60,7 +60,7 @@ force_crafted_metadata() {
> > }
> >
> > bigval=100000000
> > -fsdsopt="-d agcount=1,size=512m"
> > +fsdsopt="-d size=512m"
> >
> > force_crafted_metadata freeblks 0 "agf 0"
> > force_crafted_metadata longest $bigval "agf 0"
> > --
> > 2.31.1
> >
