Re: [Qemu-devel] [PATCH v4 0/4] test virtio-blk hotplug

[Amos Kong]

On Wed, Jun 18, 2014 at 10:58:52AM +0800, Amos Kong wrote:
> On Tue, Jun 17, 2014 at 03:37:08PM +0200, Andreas Färber wrote:
> > -----BEGIN PGP SIGNED MESSAGE-----
> > Hash: SHA1
> > 
> > Hi Amos,
> > 
> > Am 17.06.2014 14:54, schrieb Amos Kong:
> > > On Mon, Jun 09, 2014 at 03:22:51PM +0200, Stefan Hajnoczi wrote:
> > >> On Fri, Jun 06, 2014 at 10:33:49PM +0800, Amos Kong wrote:
> > >>> It's worth to add a hotplug test to qtest, but without 
> > >>> cooperation of guest OS, new devices can't be initialized by
> > >>> guest, and hot-unplug doesn't work.
> > >>> 
> > >>> However, the new test can cover some part of code of 
> > >>> hotplug/unplug.
> > >>> 
> > >>> I will write another subtest to test hotplug with pci support.
> > >>> 
> > >>> V2: move qmp_exec_hmp_cmd() to libqtest.c excape hmp cmd
> > >>> (stefanha) use qmp_exec_hmp_cmd() in blockdev-test V3: use
> > >>> vp_list to format string, free escaped string V4: free escaped
> > >>> string by g_free()
> > > 
> > > Hi Andreas,
> > > 
> > > Can you apply this patchset to your tree?
> > 
> > Sorry for the late reply, I was on vacation and am still fighting a
> > mail backlog...
> 
> That's OK :)
>  
> > I had a question about the new HMP helper function - I see in 4/4 that
> > drive_add was already being done via HMP, so I guess that's the
> > culprit, but still I wonder whether we can do device-add via QMP.
> 
> I didn't find device_add/del definition in qapi-schema.json, so I
> used HMP instead. Actually those two QMP commands exist and work.
> 
> { "execute": "blockdev-add", "arguments": { "options": { "driver":
>   "file", "filename": "/dev/null", "id": "id1" } } }
> 
> { "execute": "device_add", "arguments": { "driver": "virtio-blk-pci",
>   "drive": "id1", "driver": "virtio-blk-pci", "id": "id2" } }
> 
> I will convert virtio-blk-test to use QMP, send a V5.

We have QMP command to hot-unplug device:

{ "execute": "device_del", "arguments": { "id": "id2" } }

But there is not QMP command to hot-unplug drive/blockdev

drive_del / blockdev-del doesn't exist, so I will use
human-monitor-command to hot-unplug blockdev.
  
> > That said, I've reviewed the generic qtest part OK, and would assume
> > virtio-blk-test (you can drop the .c in the topic please) to go
> > through the same tree virtio-blk does, so via Stefan/Kevin.
> 
> Stefan said (offline) it belongs to you, and I found the first commit of
> this file was merged by you.
>  
> Thanks.
> 
> > Regards,
> > Andreas
> 
> -- 
>                       Amos.

-- 
                        Amos.