bob80905 wrote: > > Do we want to require the same decoration be added within the context of a > > library shader? > > There wouldn't be decorations to the library shaders. > > > In this case, the target is library, yet we could be processing a pixel > > shader function with the pixel shader attribute. Separately, it might be > > worth adding a library case to document / test this behavior > > While i'm not explicitly testing that the attribute is not applied, i did > notice during development of this pr that before I limited the change via > `ShaderAttr->getType() == llvm::Triple::EnvironmentType::Pixel` I was getting > test failures. So I don't think we need an explicit these attributes are not > here test, but can add one if you feel strongly about it.
Are they failing because the validator is failing due to the missing decoration? Or another reason? If the former, then I do think we should add the test and update the PR logic too. https://github.com/llvm/llvm-project/pull/210116 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
