Re: [PATCH v2] btrfs/254: test cleaning up of the stale device

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




Gentle ping? More below.

On 20/12/2021 19:06, Anand Jain wrote:


On 19/12/2021 22:02, Eryu Guan wrote:
On Sat, Dec 11, 2021 at 02:14:41AM +0800, Anand Jain wrote:
Recreating a new filesystem or adding a device to a mounted the filesystem
should remove the device entries under its previous fsid even when
confused with different device paths to the same device.

Fixed by the kernel patch (in the ml):
   btrfs: harden identification of the stale device

Signed-off-by: Anand Jain <anand.jain@xxxxxxxxxx>

I was testing with v5.16-rc2 kernel, which should not contain the kernel
fix, but test still passed for me, I was testing with three loop devices
as SCRATCH_DEV_POOL, and all default mkfs & mount options

SECTION       -- btrfs
RECREATING    -- btrfs on /dev/mapper/testvg-lv1
FSTYP         -- btrfs
PLATFORM      -- Linux/x86_64 fedoravm 5.16.0-rc2 #22 SMP PREEMPT Mon Nov 29 00:54:26 CST 2021
MKFS_OPTIONS  -- /dev/loop0
MOUNT_OPTIONS -- /dev/loop0 /mnt/scratch
btrfs/254 5s ...  5s
Ran: btrfs/254
Passed all 1 tests

Anything wrong with my setup?

And if tested with lv devices as SCRATCH_DEV_POOL

SCRATCH_DEV_POOL="/dev/mapper/testvg-lv2 /dev/mapper/testvg-lv3 /dev/mapper/testvg-lv4 /dev/mapper/testvg-lv5 /dev/mapper/testvg-lv6"

I got the following test failure

  QA output created by 254
+ERROR: cannot unregister device '/dev/mapper/254-test': No such file or directory
  Label: none  uuid: <UUID>
         Total devices <NUM> FS bytes used <SIZE>
         devid <DEVID> size <SIZE> used <SIZE> path SCRATCH_DEV

Maybe we should use _require_scratch_nolvm as well?



The test case is inconsistent because the systemd-udev block device scan interferes with the test script. There is no way to disable the systemd-udev scan (that I could find).

The same inconsistency (due to race with systemd-udev scan) would persist with/without lvm.

So when the race fails, the test case is successful to reproduce the issue. As you saw in the 2nd iteration.


There isn't any approach to stop the test case from racing with the device scan, so reproducing the issue is inconsistent. As we have had some success so IMO this test case can still integrate, it will help to verify the kernel fix on some systems.

More below.


Any suggestions?

Thanks, Anand


---
v2: Add kernel patch title in the test case
     Redirect device add output to /dev/null (avoids tirm message)
     Use the lv path for mkfs and the dm path for the device add
      so that now path used in udev scan should match with what
      we already have in the kernel memory.

-       _mkfs_dev $uuid -draid1 -mraid1 $dmdev $scratch_dev2
+       _mkfs_dev $uuid -draid1 -mraid1 $lvdev $scratch_dev2
         # Add device should free the device under $uuid in the kernel.
-       $BTRFS_UTIL_PROG device add -f $lvdev $seq_mnt > /dev/null 2>&1
+       $BTRFS_UTIL_PROG device add -f $dmdev $seq_mnt > /dev/null 2>&1


  tests/btrfs/254     | 113 ++++++++++++++++++++++++++++++++++++++++++++
  tests/btrfs/254.out |   6 +++
  2 files changed, 119 insertions(+)
  create mode 100755 tests/btrfs/254
  create mode 100644 tests/btrfs/254.out

diff --git a/tests/btrfs/254 b/tests/btrfs/254
new file mode 100755
index 000000000000..b70b9d165897
--- /dev/null
+++ b/tests/btrfs/254
@@ -0,0 +1,113 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2021 Anand Jain. All Rights Reserved.
+# Copyright (c) 2021 Oracle. All Rights Reserved.
+#
+# FS QA Test No. 254
+#
+# Test if the kernel can free the stale device entries.
+#
+# Tests bug fixed by the kernel patch:
+#    btrfs: harden identification of the stale device
+#
+. ./common/preamble
+_begin_fstest auto quick
+
+# Override the default cleanup function.
+node=$seq-test
+cleanup_dmdev()
+{
+    _dmsetup_remove $node
+}
+
+_cleanup()
+{
+    cd /
+    rm -f $tmp.*
+    rm -rf $seq_mnt > /dev/null 2>&1
+    cleanup_dmdev

Should wipefs in cleanup as well, otherwise test fails with non-unique
UUID

-Label: none  uuid: <UUID>
-       Total devices <NUM> FS bytes used <SIZE>
-       devid <DEVID> size <SIZE> used <SIZE> path SCRATCH_DEV
-       *** Some devices missing

+ERROR: non-unique UUID: 12345678-1234-1234-1234-123456789abc


 We don't need non-unique UUID. I have fixed this in v3.

Thanks,
Anand


+btrfs-progs v5.4
+See http://btrfs.wiki.kernel.org for more information.

Thanks,
Eryu

+}
+
+# Import common functions.
+. ./common/filter
+. ./common/filter.btrfs
+
+# real QA test starts here
+_supported_fs btrfs
+_require_scratch_dev_pool 3
+_require_block_device $SCRATCH_DEV
+_require_dm_target linear
+_require_btrfs_forget_or_module_loadable
+_require_scratch_nocheck
+_require_command "$WIPEFS_PROG" wipefs
+
+_scratch_dev_pool_get 3
+
+setup_dmdev()
+{
+    # Some small size.
+    size=$((1024 * 1024 * 1024))
+    size_in_sector=$((size / 512))
+
+    table="0 $size_in_sector linear $SCRATCH_DEV 0"
+    _dmsetup_create $node --table "$table" || \
+        _fail "setup dm device failed"
+}
+
+# Use a known it is much easier to debug.
+uuid="--uuid 12345678-1234-1234-1234-123456789abc"
+lvdev=/dev/mapper/$node
+
+seq_mnt=$TEST_DIR/$seq.mnt
+mkdir -p $seq_mnt
+
+test_forget()
+{
+    setup_dmdev
+    dmdev=$(realpath $lvdev)
+
+    _mkfs_dev $uuid $dmdev
+
+    # Check if we can un-scan using the mapper device path.
+    $BTRFS_UTIL_PROG device scan --forget $lvdev
+
+    # Cleanup
+    $WIPEFS_PROG -a $lvdev > /dev/null 2>&1
+    $BTRFS_UTIL_PROG device scan --forget
+
+    cleanup_dmdev
+}
+
+test_add_device()
+{
+    setup_dmdev
+    dmdev=$(realpath $lvdev)
+    scratch_dev2=$(echo $SCRATCH_DEV_POOL | awk '{print $2}')
+    scratch_dev3=$(echo $SCRATCH_DEV_POOL | awk '{print $3}')
+
+    _mkfs_dev $scratch_dev3
+    _mount $scratch_dev3 $seq_mnt
+
+    _mkfs_dev $uuid -draid1 -mraid1 $lvdev $scratch_dev2
+
+    # Add device should free the device under $uuid in the kernel.
+    $BTRFS_UTIL_PROG device add -f $dmdev $seq_mnt > /dev/null 2>&1
+
+    _mount -o degraded $scratch_dev2 $SCRATCH_MNT
+
+    # Check if the missing device is shown.
+    $BTRFS_UTIL_PROG filesystem show -m $SCRATCH_MNT | \
+                    _filter_btrfs_filesystem_show
+
+    $UMOUNT_PROG $seq_mnt
+    _scratch_unmount
+    cleanup_dmdev
+}
+
+test_forget
+test_add_device
+
+_scratch_dev_pool_put
+
+status=0
+exit
diff --git a/tests/btrfs/254.out b/tests/btrfs/254.out
new file mode 100644
index 000000000000..20819cf5140c
--- /dev/null
+++ b/tests/btrfs/254.out
@@ -0,0 +1,6 @@
+QA output created by 254
+Label: none  uuid: <UUID>
+    Total devices <NUM> FS bytes used <SIZE>
+    devid <DEVID> size <SIZE> used <SIZE> path SCRATCH_DEV
+    *** Some devices missing
+
--
2.27.0



[Index of Archives]     [Linux Filesystems Development]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux