[Qemu-block] [PATCH v2 7/7] qcow2: Make qemu-img check detect corrupted L1 tables in snapshots

[Alberto Garcia]

'qemu-img check' cannot detect if a snapshot's L1 table is corrupted.
This patch checks the table's offset and size and reports corruption
if the values are not valid.

This patch doesn't add code to fix that corruption yet, only to detect
and report it.

Signed-off-by: Alberto Garcia <address@hidden>
Reviewed-by: Eric Blake <address@hidden>
---
 block/qcow2-refcount.c     | 14 ++++++++++++++
 tests/qemu-iotests/080     |  2 ++
 tests/qemu-iotests/080.out | 20 ++++++++++++++++++++
 3 files changed, 36 insertions(+)

diff --git a/block/qcow2-refcount.c b/block/qcow2-refcount.c
index 443ec0a6e5..9d4b4c29d6 100644
--- a/block/qcow2-refcount.c
+++ b/block/qcow2-refcount.c
@@ -2019,6 +2019,20 @@ static int calculate_refcounts(BlockDriverState *bs, 
BdrvCheckResult *res,
     /* snapshots */
     for (i = 0; i < s->nb_snapshots; i++) {
         sn = s->snapshots + i;
+        if (offset_into_cluster(s, sn->l1_table_offset)) {
+            fprintf(stderr, "ERROR snapshot %s (%s) l1_offset=%#" PRIx64 ": "
+                    "L1 table is not cluster aligned; snapshot table entry "
+                    "corrupted\n", sn->id_str, sn->name, sn->l1_table_offset);
+            res->corruptions++;
+            continue;
+        }
+        if (sn->l1_size > QCOW_MAX_L1_SIZE / sizeof(uint64_t)) {
+            fprintf(stderr, "ERROR snapshot %s (%s) l1_size=%#" PRIx32 ": "
+                    "L1 table is too large; snapshot table entry corrupted\n",
+                    sn->id_str, sn->name, sn->l1_size);
+            res->corruptions++;
+            continue;
+        }
         ret = check_refcounts_l1(bs, res, refcount_table, nb_clusters,
                                  sn->l1_table_offset, sn->l1_size, 0, fix);
         if (ret < 0) {
diff --git a/tests/qemu-iotests/080 b/tests/qemu-iotests/080
index f8e7d6f4df..4dbe68e950 100755
--- a/tests/qemu-iotests/080
+++ b/tests/qemu-iotests/080
@@ -182,6 +182,7 @@ poke_file "$TEST_IMG" "$offset_snap1_l1_offset" 
"\x00\x00\x00\x00\x00\x40\x02\x0
            -c 'write 0 4k'; } 2>&1 | _filter_qemu_io | _filter_testdir
 { $QEMU_IMG snapshot -a test $TEST_IMG; } 2>&1 | _filter_testdir
 { $QEMU_IMG snapshot -d test $TEST_IMG; } 2>&1 | _filter_testdir
+_check_test_img
 
 echo
 echo "== Invalid snapshot L1 table size =="
@@ -195,6 +196,7 @@ poke_file "$TEST_IMG" "$offset_snap1_l1_size" 
"\x10\x00\x00\x00"
            -c 'write 0 4k'; } 2>&1 | _filter_qemu_io | _filter_testdir
 { $QEMU_IMG snapshot -a test $TEST_IMG; } 2>&1 | _filter_testdir
 { $QEMU_IMG snapshot -d test $TEST_IMG; } 2>&1 | _filter_testdir
+_check_test_img
 
 # success, all done
 echo "*** done"
diff --git a/tests/qemu-iotests/080.out b/tests/qemu-iotests/080.out
index 89bcd27172..4e0f7f7b92 100644
--- a/tests/qemu-iotests/080.out
+++ b/tests/qemu-iotests/080.out
@@ -71,6 +71,16 @@ write failed: Invalid argument
 qemu-img: Snapshot L1 table offset invalid
 qemu-img: Could not apply snapshot 'test': Failed to load snapshot: Invalid 
argument
 qemu-img: Could not delete snapshot 'test': Snapshot L1 table offset invalid
+ERROR snapshot 1 (test) l1_offset=0x400200: L1 table is not cluster aligned; 
snapshot table entry corrupted
+Leaked cluster 4 refcount=2 reference=1
+Leaked cluster 5 refcount=2 reference=1
+Leaked cluster 6 refcount=1 reference=0
+
+1 errors were found on the image.
+Data may be corrupted, or further writes to the image may corrupt it.
+
+3 leaked clusters were found on the image.
+This means waste of disk space, but no harm to data.
 
 == Invalid snapshot L1 table size ==
 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
@@ -84,4 +94,14 @@ write failed: File too large
 qemu-img: Snapshot L1 table too large
 qemu-img: Could not apply snapshot 'test': Failed to load snapshot: File too 
large
 qemu-img: Could not delete snapshot 'test': Snapshot L1 table too large
+ERROR snapshot 1 (test) l1_size=0x10000000: L1 table is too large; snapshot 
table entry corrupted
+Leaked cluster 4 refcount=2 reference=1
+Leaked cluster 5 refcount=2 reference=1
+Leaked cluster 6 refcount=1 reference=0
+
+1 errors were found on the image.
+Data may be corrupted, or further writes to the image may corrupt it.
+
+3 leaked clusters were found on the image.
+This means waste of disk space, but no harm to data.
 *** done
-- 
2.11.0