On 03/29/2017 09:13 PM, Eric Blake wrote: > On 03/13/2017 05:00 PM, Max Reitz wrote: >> On 08.03.2017 03:54, Eric Blake wrote: >>> Tweak 097 and 176 to operate on an image that is not cluster-aligned, >>> to give further coverage of clearing out an entire image. >>> >>> Signed-off-by: Eric Blake >>> >>> --- >>> v6: new patch >>> --- >>> tests/qemu-iotests/097 | 7 ++++--- >>> tests/qemu-iotests/097.out | 24 ++++++++++++------------ >>> tests/qemu-iotests/176 | 7 ++++--- >>> tests/qemu-iotests/176.out | 24 ++++++++++++------------ >>> 4 files changed, 32 insertions(+), 30 deletions(-) >> >> I don't have anything against this patch, but maybe it would be even >> more useful if it wrote some data to the final partial cluster (just an >> idea). > > And not a bad one. Kevin said patch 1 and 2 are okay for freeze (if > wanted; but not essential if they miss it), so I'll post v7 soon. Urrgh. Doing this uncovered the fact that we have a latent bug: we STILL leave the unaligned tail cluster allocated, when using the slow path; or put another way, commit a3e1505 is an incomplete fix. Do I say "thanks" for your suggestion being the cause for me to find the problem, or "no thanks" for it turning into more work? ;) -- Eric Blake eblake redhat com +1-919-301-3266 Libvirt virtualization library http://libvirt.org