On 03/07/2014 03:58 AM, Michal Privoznik wrote:
On 07.03.2014 00:47, Eric Blake wrote:
> While investigating
https://bugzilla.redhat.com/show_bug.cgi?id=1061827
> I noticed that we pass user input unscathed for block-pull, but
> always pass a canonical absolute name through for block-commit.
>
> * src/qemu/qemu_driver.c (qemuDomainBlockCommit): Preserve user's
> spelling, since absolute vs. relative matters to qemu.
> * src/qemu/qemu_monitor.h (qemuMonitorBlockCommit): Base is never
> null.
> * src/qemu/qemu_monitor.c (qemuMonitorBlockCommit): Likewise.
> * src/qemu/qemu_monitor_json.h (qemuMonitorJSONBlockCommit):
> Likewise.
> * src/qemu/qemu_monitor_json.c (qemuMonitorJSONBlockCommit):
> Likewise.
>
> Signed-off-by: Eric Blake <eblake(a)redhat.com>
> ---
>
> I was _hoping_ that this would solve the mentioned bugzilla. But
> even with this applied, qemu still ended up writing an absolute
> backing file name into the active file in the backing chain, so that
> bug has been reassigned back to qemu - it probably has to do with
> the fact that libvirt always spawns qemu with -drive pointing to
> an absolute name, which is unrelated to what this patch fixes.
>
> Therefore, I'm a little bit hesitant to apply this patch, but
> wanted to post it for review anyway.
ACK
After sleeping on this, I went ahead and pushed it. Even if it doesn't
fix the behavior that we currently see with qemu, it definitely makes it
easier to test how qemu behaves with relative names.
--
Eric Blake eblake redhat com +1-919-301-3266
Libvirt virtualization library
http://libvirt.org