On 27/02/15 20:53, Ilia Mirkin wrote:
On Fri, Feb 27, 2015 at 1:04 PM, Laura Ekstrand <[email protected]> wrote:
You only need to compute tolerance[0] once, but you need to call equal()
lots of times. You should make a static variable tolerance at the top of
the file and call piglit_compute_probe_tolerance in piglit_init to
initialize it. Otherwise, this seems inefficient.
FWIW there's also an extern float piglit_tolerance[4]. All that
function does is a helper for dealing with various funky formats (like
luminance/alpha). I think you can just use piglit_tolerance[]
directly.
Thanks, I've updated the patch locally.
Can I get your R-b? I would like to push this ASAP.
_______________________________________________
Piglit mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/piglit
