rjodinchr wrote: > > Note that non-vulkan spirv was not using that clspv implementation before > > (this implementation makes sure not to add a dependency on int-64 by using > > vec2 of int-32). > > Otherwise LGTM > > restored clspv status by keeping only float fma.
This PR is fine as far as clspv is concerned. But for spirv (non-vulkan), it is not the same implementation as before https://github.com/llvm/llvm-project/commit/39f3d72b72fdb6831886148f1172d2f5b0ab2d49. I don't know if that's a problem, I'll let the people concerned by it have their say. https://github.com/llvm/llvm-project/pull/199626 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
