scripts: fix Windows test dm/loopback cleanup

The Windows integration test scripts tore down their device-mapper
snapshot and loop devices with 'dmsetup remove_all -f' and 'losetup
-D'. Both operate on every such device on the host, not just the ones
the script created.

On a host whose root filesystem is device-mapper backed (for example an
LVM root), 'dmsetup remove_all -f' replaces the in-use root device's
table with an error target. The host is wedged until reboot.  This is
harmless in CI, which runs in a disposable VM, but destroys a developer
machine.

It turns out windows-snapshot-base is not used at all. Drop it.

Signed-off-by: Wei Liu <liuwe@microsoft.com>
Assisted-by: Copilot:Opus-4.8
This commit is contained in:
Wei Liu
2026-07-09 00:55:56 +00:00
parent 5ce8732196
commit ae3c517368
2 changed files with 56 additions and 8 deletions

View File

@@ -31,13 +31,35 @@ if [[ ! -f ${WIN_IMAGE_FILE} || ! -f ${OVMF_FW} ]]; then
exit 1
fi
# Remove every device-mapper device this Windows test suite may have created
# (the base plus any per-test snapshot a crashed or timed-out test left
# behind). Snapshots reference the base as their origin and must be removed
# first, so retry a few times to tolerate ordering. Only 'windows-*' devices
# are touched, never any other device-mapper device on the host.
cleanup_windows_dm() {
for _ in 1 2 3; do
local devs
devs=$(dmsetup ls 2>/dev/null | awk '{print $1}' | grep '^windows-' || true)
[ -z "$devs" ] && break
for dev in $devs; do
dmsetup remove "$dev" 2>/dev/null || true
done
done
}
# Clear anything a previous crashed or killed run left behind so the dmsetup
# create below does not fail on a stale device, then detach any stale loop
# devices still backing the image.
cleanup_windows_dm
losetup -j "${WIN_IMAGE_FILE}" | cut -d : -f 1 | while read -r stale; do
losetup -d "$stale" 2>/dev/null || true
done
# Use device mapper to create a snapshot of the Windows image
img_blk_size=$(du -b -B 512 "${WIN_IMAGE_FILE}" | awk '{print $1;}')
loop_device=$(losetup --find --show --read-only "${WIN_IMAGE_FILE}")
dmsetup create windows-base --table "0 $img_blk_size linear $loop_device 0"
dmsetup mknodes
dmsetup create windows-snapshot-base --table "0 $img_blk_size snapshot-origin /dev/mapper/windows-base"
dmsetup mknodes
# Common configuration for every test run
export RUST_BACKTRACE=1
@@ -50,7 +72,9 @@ cargo build --all --release --target "$BUILD_TARGET"
time cargo nextest run -p cloud-hypervisor --retries 3 --no-tests=pass "windows::$test_filter" --target "$BUILD_TARGET" -- ${test_binary_args[*]}
RES=$?
dmsetup remove_all -f
losetup -D
# Tear down the base and any per-test snapshot devices left over from a crash,
# then detach this run's loop device.
cleanup_windows_dm
losetup -d "$loop_device" 2>/dev/null || true
exit $RES

View File

@@ -36,13 +36,35 @@ if [[ ! -f ${WIN_IMAGE_FILE} || ! -f ${OVMF_FW} ]]; then
exit 1
fi
# Remove every device-mapper device this Windows test suite may have created
# (the base plus any per-test snapshot a crashed or timed-out test left
# behind). Snapshots reference the base as their origin and must be removed
# first, so retry a few times to tolerate ordering. Only 'windows-*' devices
# are touched, never any other device-mapper device on the host.
cleanup_windows_dm() {
for _ in 1 2 3; do
local devs
devs=$(dmsetup ls 2>/dev/null | awk '{print $1}' | grep '^windows-' || true)
[ -z "$devs" ] && break
for dev in $devs; do
dmsetup remove "$dev" 2>/dev/null || true
done
done
}
# Clear anything a previous crashed or killed run left behind so the dmsetup
# create below does not fail on a stale device, then detach any stale loop
# devices still backing the image.
cleanup_windows_dm
losetup -j ${WIN_IMAGE_FILE} | cut -d : -f 1 | while read -r stale; do
losetup -d "$stale" 2>/dev/null || true
done
# Use device mapper to create a snapshot of the Windows image
img_blk_size=$(du -b -B 512 ${WIN_IMAGE_FILE} | awk '{print $1;}')
loop_device=$(losetup --find --show --read-only ${WIN_IMAGE_FILE})
dmsetup create windows-base --table "0 $img_blk_size linear $loop_device 0"
dmsetup mknodes
dmsetup create windows-snapshot-base --table "0 $img_blk_size snapshot-origin /dev/mapper/windows-base"
dmsetup mknodes
cargo build --features mshv --all --release --target "$BUILD_TARGET"
@@ -55,7 +77,9 @@ export RUSTFLAGS="$RUSTFLAGS"
time cargo nextest run -p cloud-hypervisor --retries 3 --no-tests=pass $test_features "windows::$test_filter" --target "$BUILD_TARGET" -- ${test_binary_args[*]}
RES=$?
dmsetup remove_all -f
losetup -D
# Tear down the base and any per-test snapshot devices left over from a crash,
# then detach this run's loop device.
cleanup_windows_dm
losetup -d "$loop_device" 2>/dev/null || true
exit $RES