On Wed, Oct 26, 2016 at 08:21:40AM +0200, Daniel Vetter wrote:
> > +static void check_timings(int crtc_idx, const drmModeModeInfo *kmode)
> > +{
> > +#define CALIBRATE_TS_STEPS 120 /* ~2s has to be less than 128! */
>
> Because speed matters, maybe just 50?
As you can tell I picked 120 just because it was a round number close to
our maximum queue size. If we push this for BAT (and people will
complain as half our machines will then fail), 50/1s per connector per
mode is probably still a little too long :|
> > @@ -732,6 +802,7 @@ int main(int argc, char **argv)
> > enum test_flags flags;
> > const char *name;
> > } tests[] = {
> > + { TEST_TIMINGS, "basic" },
>
> If you want this in BAT, it needs to be manually added to the list. And
> maybe give it a better name, like timings or basic-timings.
Compared to everything else in this test, this is the only one that
tried to set a mode on a single CRTC (afaict)! Seems a strange oversight
for kms_setmode, so I must be missing something.
-Chris
--
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
