Skip to contentSkip to conversation

btrfs-progs: tests false alerts fixes

PATCH In review Qu Wenruo 2 participants 6 messages Aug 8, 2023 – Aug 11, 2023 Btrfs file system

lore
Share & tools
lore

Qu Wenruo

author
#0 lore raw

cover btrfs-progs: tests false alerts fixes

This is the resent and aggregate version of the fixes I sent but mostly eaten by the recent vger down time.

Those are small fixes for the false alerts I hit during my local runs.

Most of them are subtle fixes, only the last one is more like an optimization.

Qu Wenruo (4):
btrfs-progs: tests/mkfs/005: use udevadm settle to avoid false alerts
btrfs-progs: tests/misc/030: do not require v1 cache for the test case
btrfs-progs: tests/misc/046: fix false alerts on write detection
btrfs-progs: tests/misc/058: reduce the space requirement and speed up
the test

 tests/misc-tests/030-missing-device-image/test.sh     | 8 --------
 tests/misc-tests/046-seed-multi-mount/test.sh         | 4 +++-
 tests/misc-tests/058-replace-start-enqueue/test.sh    | 5 ++---
 tests/mkfs-tests/005-long-device-name-for-ssd/test.sh | 4 +++-
 4 files changed, 8 insertions(+), 13 deletions(-)

--
2.41.0

author
#1 lore raw

PATCH 1/4 btrfs-progs: tests/mkfs/005: use udevadm settle to avoid false alerts

[BUG]
During my test runs of mkfs-tests, 005-long-device-name-for-ssd failed with the following error messages:

====== RUN CHECK dmsetup remove btrfs-test-with-very-long-name-AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAQPGc
device-mapper: remove ioctl on btrfs-test-with-very-long-name-AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAQPGc failed: Device or resource busy Command failed.
failed: dmsetup remove btrfs-test-with-very-long-name-AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAQPGc
test failed for case 005-long-device-name-for-ssd

[CAUSE]
There seems to be a race between "btrfs inspect dump-super" and the dmsetup removal.

[FIX]
Add a "udevadm settle" before removing the dm devices.

Also since we're here, use the same "udevadm settle" instead of the manual sleep to wait for the new dm device to showup.

---
 tests/mkfs-tests/005-long-device-name-for-ssd/test.sh | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
tests/mkfs-tests/005-long-device-name-for-ssd/test.sh +3 -1
index 208698dce243..c32eb3a42def 100755
@@ -24,7 +24,7 @@ loopdev=`run_check_stdout $SUDO_HELPER losetup --find --show img`
 run_check $SUDO_HELPER dmsetup create "$dmname" --table "0 1048576 linear $loopdev 0"
 
 # Setting up the device may need some time to appear
-sleep 5
+run_check $SUDO_HELPER udevadm settle
 if ! [ -b "$dmdev" ]; then
 	_not_run "dm device created but not visible in /dev/mapper"
 fi
@@ -43,6 +43,8 @@ run_check_stdout $SUDO_HELPER "$TOP/mkfs.btrfs" -f "$@" "$dmdev" |
 	grep -q 'SSD detected:.*yes' || _fail 'SSD not detected'
 run_check $SUDO_HELPER "$TOP/btrfs" inspect-internal dump-super "$dmdev"
 
+run_check $SUDO_HELPER udevadm settle
+
 # cleanup
 run_check $SUDO_HELPER dmsetup remove "$dmname"
 run_mayfail $SUDO_HELPER losetup -d "$loopdev"
Signature · 1 line
2.41.0
author
#2 lore raw

PATCH 2/4 btrfs-progs: tests/misc/030: do not require v1 cache for the test case

In reply to Qu Wenruo

[PROBLEM]
Since we have migrated to default v2 cache, the test case misc/030-missing-device-image is no longer executed:

[TEST/misc] 030-missing-device-image
[NOTRUN] unable to create v1 space cache

[CAUSE]
The test case itself is trying its best to cover all paths, including the data extent read path.

Thus the test case is requiring v1 cache, as that's the only way to cover the data read path.

[FIX]
Just remove the v1 space cache requirement, it's still better to run the test even it only exercises the metadata read path.

The good news is, after commit 3ff9d352576b ("btrfs-progs: use read_data_from_disk() to replace read_extent_from_disk() and replace read_extent_data()"), all data/metadata read paths are unified. They only differ in the verification part.

Thus even if we didn't fully exercise the data read path, we didn't lose much coverage anyway.

---
 tests/misc-tests/030-missing-device-image/test.sh | 8 --------
 1 file changed, 8 deletions(-)
tests/misc-tests/030-missing-device-image/test.sh +0 -8
index be022c4bb9eb..1438f8a45229 100755
@@ -37,14 +37,6 @@ test_missing()
 	run_check $SUDO_HELPER dd if=/dev/zero of="$TEST_MNT/b" bs=4k count=1000 conv=sync
 	run_check $SUDO_HELPER umount "$TEST_MNT"
 
-	# make sure we have space cache
-	if ! run_check_stdout "$TOP/btrfs" inspect dump-tree -t root "$dev1" |
-		grep -q "EXTENT_DATA"; then
-		# Normally the above operation should create the space cache.
-		# If not, it may mean we have migrated to v2 cache by default
-		_not_run "unable to create v1 space cache"
-	fi
-
 	# now wipe the device
 	run_check wipefs -fa "$bad_dev"
 
Signature · 1 line
2.41.0
author
#3 lore raw

PATCH 3/4 btrfs-progs: tests/misc/046: fix false alerts on write detection

In reply to Qu Wenruo

[BUG]
Test case misc/046-seed-multi-mount would always fail with the following error:

    [TEST]   misc-tests.sh
    [TEST/misc]   046-seed-multi-mount

unexpected success: writable file despite read-only mount test failed for case 046-seed-multi-mount

[CAUSE]
Although mounting seed device is indeed read-only, sprouting it with a new device would always make it read-write by itself.

The behavior is already there for a long time, thus expecting a new behavior (not changing the read-only flag) is a little weird.

[FIX]
Instead of doing the write check after the sprout, do it before the sprout.

This looks more correct, and would not rely on the kernel behavior change (if we determine to go that path).

---
 tests/misc-tests/046-seed-multi-mount/test.sh | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
tests/misc-tests/046-seed-multi-mount/test.sh +3 -1
index 83f7a91effdd..95654fd904ff 100755
@@ -42,9 +42,11 @@ nextdevice() {
 	run_check mkdir -p "$mnt"
 	TEST_MNT="$mnt"
 	run_check_mount_test_dev
-	run_check $SUDO_HELPER "$TOP/btrfs" device add ${loopdevs[$nextdev]} "$TEST_MNT"
 	run_mustfail "writable file despite read-only mount" \
 		$SUDO_HELPER dd if=/dev/zero of="$TEST_MNT/file$nextdevice" bs=1M count=1 status=none
+	run_check $SUDO_HELPER "$TOP/btrfs" device add ${loopdevs[$nextdev]} "$TEST_MNT"
+	# Although seed sprout would make the fs RW, explicitly remount it RW
+	# just in case of future behavior change.
 	run_check $SUDO_HELPER mount -o remount,rw "$TEST_MNT"
 	# Rewrite the file
 	md5sum=$(run_check_stdout md5sum "$TEST_MNT/file$nextdev" | awk '{print $1}')
Signature · 1 line
2.41.0
author
#4 lore raw

PATCH 4/4 btrfs-progs: tests/misc/058: reduce the space requirement and speed up the test

In reply to Qu Wenruo

[BUG]
When I was testing misc/058, the fs still has around 7GiB free space, but during that test case, btrfs kernel module reports write failures and even git commands failed inside that fs.

And obviously the test case failed.

[CAUSE]
It turns out that, the test case itself would require 6GiB (4 data disks) + 1.5GiB x 2 (the two replace target), thus it requires 9 GiB free space.

And obviously my partition is not that large and failed.

[FIX]
In fact, we really don't need that much space at all.

Our objective is to test "btrfs device replace --enqueue" functionality, there is not much need to wait for 1 second, we can just do the enqueue immediately.

So this patch would reduce the file size to a more sane (and rounded) 2GiB, and do the enqueue immediately.

---
 tests/misc-tests/058-replace-start-enqueue/test.sh | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)
tests/misc-tests/058-replace-start-enqueue/test.sh +2 -3
index 1a24d5ec7ecb..bdbc87b4090d 100755
@@ -21,16 +21,15 @@ run_check_mount_test_dev
 run_check $SUDO_HELPER "$TOP/btrfs" device remove "$REPLACE1" "$TEST_MNT"
 run_check $SUDO_HELPER "$TOP/btrfs" device remove "$REPLACE2" "$TEST_MNT"
 
-for i in `seq 48`; do
+for i in `seq 16`; do
 	run_check $SUDO_HELPER dd if=/dev/zero of="$TEST_MNT/file$i" bs=1M count=128 status=noxfer
 done
 # Sync so replace start does not block in unwritten IO
 run_check "$TOP/btrfs" filesystem sync "$TEST_MNT"
 run_check "$TOP/btrfs" filesystem usage -T "$TEST_MNT"
 
-# Go background, should not be that fast, estimated 10 seconds
+# Go background, should not be that fast.
 run_check $SUDO_HELPER "$TOP/btrfs" replace start 2 "$REPLACE1" "$TEST_MNT"
-run_check sleep 1
 # No background, should wait
 run_check $SUDO_HELPER "$TOP/btrfs" replace start --enqueue 3 "$REPLACE2" "$TEST_MNT"
 run_check $SUDO_HELPER "$TOP/btrfs" replace status "$TEST_MNT"
Signature · 1 line
2.41.0
#5 lore raw
In reply to Qu Wenruo
Quoted 5 lines · Qu Wenruo

This is the resent and aggregate version of the fixes I sent but mostly eaten by the recent vger down time.

Those are small fixes for the false alerts I hit during my local runs.

Quoted 7 lines

Most of them are subtle fixes, only the last one is more like an optimization.

Qu Wenruo (4):
btrfs-progs: tests/mkfs/005: use udevadm settle to avoid false alerts
btrfs-progs: tests/misc/030: do not require v1 cache for the test case
btrfs-progs: tests/misc/046: fix false alerts on write detection

This one was not in the first batch so I've added it now, thanks.

Quoted 2 lines

btrfs-progs: tests/misc/058: reduce the space requirement and speed up
the test