26 Apr
2023
26 Apr
'23
2:21 p.m.
On Wed Apr 26 12:07:55 2023 +0000, Giovanni Mascellani wrote:
That's great, but it doesn't seem that in those cases `filter point point point` is actually meaningful. Ideally the test should failed if that's changed to `filter linear linear linear`, so that miscompilations are detected.
It is meaningful, notice that, on the tests where there are more than one sampler, I am sampling `float(0.22, 0.22)` which only gives a rounded value for the `filter point point point` sampler, because it does a nearest neighbor sampling.