mesa 505a8fbdf8f2b6d2aaab5a04244cd3329f9dbe97 fixed 24 dEQP tests and regressed 2 piglit tests. One of those tests, piglit.spec.arb_gpu_shader5.execution.built-in-functions.fs-ldexp, was fixed by a subsequent change to piglit. The vertex shader ldexp test continues to fail on all platform IVB and later. This failure should be designated as a regression or a piglit bug before the Mesa 12.0 release.
Matt, Ken suggests you might be able to take a quick look at this and determine if there is something wrong with the piglit test.
Whoops. Looks like I neglected to update the VS tests at the same time I fixed the FS ones. I've pushed the following two commits commit 86891da91b1dc6f8852775b1d0ef504a89a1cf81 Author: Matt Turner <mattst88@gmail.com> Date: Mon Jun 6 14:35:56 2016 -0700 arb_gpu_shader5: Remove out of spec ldexp() test case. Forgot to update the vs test in commit c8dfb932f. commit 41c535c95581a7edc7da48a4e3aeacaf37f30f92 Author: Matt Turner <mattst88@gmail.com> Date: Mon Jun 6 14:35:29 2016 -0700 arb_gpu_shader5: Test ldexp(0.0f, non-zero-exp). Forgot to update the vs test in commit 5fae1a7ca4d. which bring the vs test in line with the fs test.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.