On 30.03.2017 20:39, Eric Blake wrote:
> 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 <eblake@redhat.com>
>>>>
>>>> ---
>>>> 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? ;)

Just pretend you didn't see any problem. Whoops, you've already posted a
mail to the list, this will make it a bit awkward. ;-)

Max