Oh, you're right, and I think there is a patch of mine related to the fail:
"[PATCH 02/12] VMDK: bugfix, align offset to cluster in get_whole_cluster" Can you test if this fixes it? On Wed, Jun 8, 2011 at 9:52 PM, Kevin Wolf <kw...@redhat.com> wrote: > Am 08.06.2011 15:29, schrieb Fam Zheng: >> Yes I noticed the failure, but it's trivial mismatch between `qemu-img >> create` output. The old test case has "cluster_size=65536" but I find >> the latest qemu-img omits this info, as the diff: >> >> -Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=6442450944 >> cluster_size=65536 >> +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=6442450944 >> >> Is this your case too? > > No, I get real errors, pattern mismatches: > > > [ Before this, there are some successful writes ] > > Testing conversion with -B TEST_DIR/t.IMGFMT.base > > Checking if backing clusters are allocated when they shouldn't > > qemu-io> 0/128 sectors allocated at offset 1 MiB > qemu-io> qemu-io> 0/128 sectors allocated at offset 4.001 GiB > qemu-io> Reading > > === IO: pattern 42 > qemu-io> Pattern verification failed at offset 0, 512 bytes > read 512/512 bytes at offset 0 > 512.000000 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) > qemu-io> Pattern verification failed at offset 1024, 512 bytes > read 512/512 bytes at offset 1024 > 512.000000 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) > qemu-io> Pattern verification failed at offset 2048, 512 bytes > > [ More pattern verification errors follow ] > -- Best regards! Fam Zheng
