New test bug/r300-readcache
This tests against a particular problem of the r300 driver, related to a readcache coherency problem. Note that while I had first found this problem via glean/texCube, the fact that I did find the problem there was pure chance, because the bug depends on the exact coordinate of the glReadPixels call (and thus on the window position). This new test loops through many coordinates to make sure the window coordinates are irrelevant.
tests/bugs/r300-readcache.c
0 → 100644
Please register or sign in to comment