Message ID | 1308163478-21136-1-git-send-email-alevy@redhat.com |
---|---|
State | New |
Headers | show |
On 06/15/11 20:44, Alon Levy wrote: > This fixes a problem where on windows 7 startup phase, before the qxl driver > is loaded, the drawables are sufficiently large and video like to trigger a > stream, but the lack of a filled mm time field triggers a warning in spice-gtk. Looks good to me. Why this is tagged RFC? cheers, Gerd
On Thu, Jun 16, 2011 at 03:34:52PM +0200, Gerd Hoffmann wrote: > On 06/15/11 20:44, Alon Levy wrote: > >This fixes a problem where on windows 7 startup phase, before the qxl driver > >is loaded, the drawables are sufficiently large and video like to trigger a > >stream, but the lack of a filled mm time field triggers a warning in spice-gtk. > > Looks good to me. Why this is tagged RFC? I wasn't sure if the right thing is to set the mm_time or just not warn if it is missing. This can be committed as is if you think it's good. > > cheers, > Gerd >
diff --git a/ui/spice-display.c b/ui/spice-display.c index 15f0704..feeee73 100644 --- a/ui/spice-display.c +++ b/ui/spice-display.c @@ -70,6 +70,7 @@ static SimpleSpiceUpdate *qemu_spice_create_update(SimpleSpiceDisplay *ssd) QXLCommand *cmd; uint8_t *src, *dst; int by, bw, bh; + struct timespec time_space; if (qemu_spice_rect_is_empty(&ssd->dirty)) { return NULL; @@ -96,6 +97,10 @@ static SimpleSpiceUpdate *qemu_spice_create_update(SimpleSpiceDisplay *ssd) drawable->surfaces_dest[0] = -1; drawable->surfaces_dest[1] = -1; drawable->surfaces_dest[2] = -1; + clock_gettime(CLOCK_MONOTONIC, &time_space); + /* time in milliseconds from epoch. */ + drawable->mm_time = time_space.tv_sec * 1000 + + time_space.tv_nsec / 1000 / 1000; drawable->u.copy.rop_descriptor = SPICE_ROPD_OP_PUT; drawable->u.copy.src_bitmap = (intptr_t)image;