[PULL 21/29] iotests: cover the Format Extension against the leak check

["Denis V. Lunev" <den@openvz.org>]

From: Denis V. Lunev <den@openvz.org>

The Format Extension and the bitmap data clusters it points to are
absent from used_bmap, so the leak check has to be told about them
through s->ext_end or it reports the space they occupy as leaked and
'qemu-img check -r leaks' truncates it away while the header still
points at it.

The leak coverage in parallels-checks grows a persistent bitmap, dirtied
one granule at a time so that parallels_save_bitmap() stores a bitmap
data cluster rather than the all ones L1 entry and the extension spans
two clusters. It asserts that neither is a leak, that a genuine leak
appended after them is still reported and repaired, that the bitmap and
the guest data survive the repair, and that dropping the bitmap makes
the space reclaimable. Without s->ext_end the check reports two leaked
clusters where there should be none.

A name which is not a UUID is covered there too, as the refusal belongs
to the same command.

Cc: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Denis V. Lunev <den@openvz.org>
---
 tests/qemu-iotests/tests/parallels-checks     | 48 +++++++++++++++++++
 tests/qemu-iotests/tests/parallels-checks.out | 35 ++++++++++++++
 2 files changed, 83 insertions(+)

diff --git a/tests/qemu-iotests/tests/parallels-checks b/tests/qemu-iotests/tests/parallels-checks
index c9dcd715ac..6f60fda62b 100755
--- a/tests/qemu-iotests/tests/parallels-checks
+++ b/tests/qemu-iotests/tests/parallels-checks
@@ -362,6 +362,54 @@ echo "file size: `stat --printf="%s" "$TEST_IMG"`"
 
 echo "== the first cluster survived =="
 { $QEMU_IO -c "read -P 0x11 0 $CLUSTER_SIZE" "$TEST_IMG"; } 2>&1 | _filter_qemu_io | _filter_testdir
+_make_test_img $SIZE
+
+echo "== TEST FORMAT EXTENSION IS NOT LEAKED SPACE =="
+
+BITMAP=b2c9e1a4-5d3f-4e8b-9a7c-6f0d1e2b3a45
+
+echo "== a name which is not a UUID can not be stored, so refuse it =="
+$QEMU_IMG bitmap --add -f $IMGFMT "$TEST_IMG" bitmap0 2>&1 | _filter_testdir
+
+echo "== add a persistent dirty bitmap =="
+$QEMU_IMG bitmap --add -f $IMGFMT "$TEST_IMG" $BITMAP 2>&1 | _filter_testdir
+
+# A dirty run shorter than a cluster is stored as a bitmap data cluster
+# instead of the all-ones L1 entry, so the extension spans two clusters.
+echo "== dirty a single granule of the bitmap =="
+{ $QEMU_IO -c "write -P 0x11 0 65536" "$TEST_IMG"; } 2>&1 | _filter_qemu_io | _filter_testdir
+
+file_size=`stat --printf="%s" "$TEST_IMG"`
+echo "file size: $file_size"
+
+echo "== the extension and its bitmap data are not a leak =="
+_check_test_img
+
+echo "== extend image by 1 cluster =="
+fallocate -xl $((file_size + CLUSTER_SIZE)) "$TEST_IMG"
+
+echo "== a real leak is still reported =="
+_check_test_img
+
+echo "== repair reclaims the leak only =="
+_check_test_img -r leaks
+
+file_size=`stat --printf="%s" "$TEST_IMG"`
+echo "file size: $file_size"
+
+echo "== the bitmap survived the repair =="
+$QEMU_IMG bitmap --add -f $IMGFMT "$TEST_IMG" $BITMAP 2>&1 | _filter_testdir
+
+echo "== data survived the repair =="
+{ $QEMU_IO -r -c "read -P 0x11 0 65536" "$TEST_IMG"; } 2>&1 | _filter_qemu_io | _filter_testdir
+
+echo "== dropping the bitmap makes the space reclaimable again =="
+$QEMU_IMG bitmap --remove -f $IMGFMT "$TEST_IMG" $BITMAP 2>&1 | _filter_testdir
+
+file_size=`stat --printf="%s" "$TEST_IMG"`
+echo "file size: $file_size"
+
+_check_test_img
 
 # Clear image
 _make_test_img $SIZE
diff --git a/tests/qemu-iotests/tests/parallels-checks.out b/tests/qemu-iotests/tests/parallels-checks.out
index 6699848996..d40f865868 100644
--- a/tests/qemu-iotests/tests/parallels-checks.out
+++ b/tests/qemu-iotests/tests/parallels-checks.out
@@ -233,6 +233,41 @@ file size: 2097152
 read 1048576/1048576 bytes at offset 0
 1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=4194304
+== TEST FORMAT EXTENSION IS NOT LEAKED SPACE ==
+== a name which is not a UUID can not be stored, so refuse it ==
+qemu-img: Operation add on bitmap bitmap0 failed: Bitmap name must be a UUID to be stored in a parallels image: bitmap0
+== add a persistent dirty bitmap ==
+== dirty a single granule of the bitmap ==
+wrote 65536/65536 bytes at offset 0
+64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
+file size: 4194304
+== the extension and its bitmap data are not a leak ==
+No errors were found on the image.
+== extend image by 1 cluster ==
+== a real leak is still reported ==
+ERROR space leaked at the end of the image 1048576
+
+1 leaked clusters were found on the image.
+This means waste of disk space, but no harm to data.
+== repair reclaims the leak only ==
+Repairing space leaked at the end of the image 1048576
+The following inconsistencies were found and repaired:
+
+    1 leaked clusters
+    0 corruptions
+
+Double checking the fixed image now...
+No errors were found on the image.
+file size: 4194304
+== the bitmap survived the repair ==
+qemu-img: Operation add on bitmap b2c9e1a4-5d3f-4e8b-9a7c-6f0d1e2b3a45 failed: Bitmap already exists: b2c9e1a4-5d3f-4e8b-9a7c-6f0d1e2b3a45
+== data survived the repair ==
+read 65536/65536 bytes at offset 0
+64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
+== dropping the bitmap makes the space reclaimable again ==
+file size: 2097152
+No errors were found on the image.
+Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=4194304
 == TEST A DUPLICATE IN THE LAST ALLOCATED BAT ENTRY ==
 == write two clusters ==
 wrote 1048576/1048576 bytes at offset 0
-- 
2.53.0