"Works on an app" is not and never has been an acceptable testing plan for landing a new feature in Mesa. There need to be tests in piglit, dEQP, or one of the conformance suites.
--Jason On Tue, Aug 28, 2018 at 11:15 AM Rogovin, Kevin <kevin.rogo...@intel.com> wrote: > Hi, > > On the questions of tests: If people want, I can adapt the test in piglit > for ARB_fragment_shader_interlock to this INTEL one. In general, I have an > app/library that uses the extension and testing of that does definitely > work on i965 (which should be utterly unsurprising since the produced > assembly is identical). Indeed the current implementation in Mesa of > ARB_fragment_shader_interlock is as flexible as the INTEL one since the > compiler accepts code with the begin/end interlock anywhere since the only > backend that supports it, i965, can easily accept it. I view the > INTEL_fragment_shader_ordering implementation in a similar light as the > NV_fragment_shader_interlock where it actually does not really do anything > new, but signals to an application that the Mesa/i965 implementation allows > more (and does it correctly) than the ARB/NV extensions restrict to. > > -Kevin >
_______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev
