On 11/14/18 8:35 AM, Kevin Wolf wrote:
Am 14.11.2018 um 00:03 hat Eric Blake geschrieben:
As the added iotests shows, we have a (corner case) data corruption
that is user triggerable, therefore, this is still appropriate for
inclusion in 3.1.
Thanks, applied to the block branch.
Patch 2 and 3 are fine; but Berto raised questions about patch 1 so I
posted a v9 of that patch, in case you want to replace it in your queue
and/or defer it out of 3.1 and instead save it for 4.0.
--
Eric Blake, Principal Software Engineer
Red Hat, Inc. +1-919-301-3266
Virtualization: qemu.org | libvirt.org