Anthony PERARD writes ("Re: [PATCH v5 12/15] libxl: QEMU startup sync based on
QMP"):
> On Tue, Oct 16, 2018 at 03:23:26PM +0100, Ian Jackson wrote:
> > It's rather odd that neither this patch, nor anything that follows,
> > does not stop libxl from also watching the xenstore path. I think it
> >
On Tue, Oct 16, 2018 at 03:23:26PM +0100, Ian Jackson wrote:
> It's rather odd that neither this patch, nor anything that follows,
> does not stop libxl from also watching the xenstore path. I think it
> would be better to suppress that rather than leave vestigial
> behaviour. Can you add patch(e
Anthony PERARD writes ("[PATCH v5 12/15] libxl: QEMU startup sync based on
QMP"):
> This is only activated when dm_restrict=1, as explained in the previous
> patch "libxl_dm: Pre-open QMP socket for QEMU"
...
> +static void device_model_qmp_cb(libxl__egc *egc, libxl__ev_qmp *ev,
> +
This is only activated when dm_restrict=1, as explained in the previous
patch "libxl_dm: Pre-open QMP socket for QEMU"
Signed-off-by: Anthony PERARD
Reviewed-by: Roger Pau Monné
---
Notes:
v5:
removed empty success branch in device_model_qmp_cb()
call libxl__ev_qmp_init() ea
