qcow2: check_errors are fatal

When trying to repair a dirty image, qcow2_check() may apparently
succeed (no really fatal error occurred that would prevent the check
from continuing), but if check_errors in the result object is non-zero,
we cannot trust the image to be usable.

Reported-by: R. Nageswara Sastry <nasastry@in.ibm.com>
Buglink: https://bugs.launchpad.net/qemu/+bug/1728639
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-id: 20171110203111.7666-2-mreitz@redhat.com
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>

3 files changed, 47 insertions, 1 deletions

diff --git a/block/qcow2.c b/block/qcow2.c
index 811b913..1914a94 100644
--- a/block/qcow2.c
+++ b/block/qcow2.c
@@ -1477,7 +1477,10 @@ static int qcow2_do_open(BlockDriverState *bs, QDict *options, int flags,
         BdrvCheckResult result = {0};
 
         ret = qcow2_check(bs, &result, BDRV_FIX_ERRORS | BDRV_FIX_LEAKS);
-        if (ret < 0) {
+        if (ret < 0 || result.check_errors) {
+            if (ret >= 0) {
+                ret = -EIO;
+            }
             error_setg_errno(errp, -ret, "Could not repair dirty image");
             goto fail;
         }
diff --git a/tests/qemu-iotests/060 b/tests/qemu-iotests/060
index fae08b03..56bdf1e 100755
--- a/tests/qemu-iotests/060
+++ b/tests/qemu-iotests/060
@@ -301,6 +301,26 @@ _make_test_img 64M
 poke_file "$TEST_IMG" "48"                "\x00\x00\x00\x00\x00\x00\x00\x00"
 $QEMU_IO -c "write 0 64k" "$TEST_IMG" | _filter_qemu_io
 
+echo
+echo "=== Testing dirty corrupt image ==="
+echo
+
+_make_test_img 64M
+
+# Let the refblock appear unaligned
+poke_file "$TEST_IMG" "$rt_offset"        "\x00\x00\x00\x00\xff\xff\x2a\x00"
+# Mark the image dirty, thus forcing an automatic check when opening it
+poke_file "$TEST_IMG" 72 "\x00\x00\x00\x00\x00\x00\x00\x01"
+# Open the image (qemu should refuse to do so)
+$QEMU_IO -c close "$TEST_IMG" 2>&1 | _filter_testdir | _filter_imgfmt
+
+echo '--- Repairing ---'
+
+# The actual repair should have happened (because of the dirty bit),
+# but some cleanup may have failed (like freeing the old reftable)
+# because the image was already marked corrupt by that point
+_check_test_img -r all
+
 # success, all done
 echo "*** done"
 rm -f $seq.full
diff --git a/tests/qemu-iotests/060.out b/tests/qemu-iotests/060.out
index 62c2270..f013fe7 100644
--- a/tests/qemu-iotests/060.out
+++ b/tests/qemu-iotests/060.out
@@ -284,4 +284,27 @@ No errors were found on the image.
 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
 qcow2: Marking image as corrupt: Preventing invalid allocation of L2 table at offset 0; further corruption events will be suppressed
 write failed: Input/output error
+
+=== Testing dirty corrupt image ===
+
+Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
+ERROR refcount block 0 is not cluster aligned; refcount table entry corrupted
+IMGFMT: Marking image as corrupt: Refblock offset 0xffff2a00 unaligned (reftable index: 0); further corruption events will be suppressed
+Can't get refcount for cluster 0: Input/output error
+Can't get refcount for cluster 1: Input/output error
+Can't get refcount for cluster 2: Input/output error
+Can't get refcount for cluster 3: Input/output error
+Rebuilding refcount structure
+Repairing cluster 1 refcount=1 reference=0
+can't open device TEST_DIR/t.IMGFMT: Could not repair dirty image: Input/output error
+--- Repairing ---
+Leaked cluster 1 refcount=1 reference=0
+Repairing cluster 1 refcount=1 reference=0
+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.
 *** done