mstorsjo added a comment. In D102568#2769390 <https://reviews.llvm.org/D102568#2769390>, @MaskRay wrote:
> In D102568#2769340 <https://reviews.llvm.org/D102568#2769340>, @mstorsjo > wrote: > >> In the meantime, wouldn't it be possible to detect the presence of the other >> one and check if they match or not, to avoid passing duplicate options to >> the backend? I can give that a try. > > Should be possible! Thanks for offering the help. I think it is too much to > test contradicting values like `// NEVER_ALWAYS: "-mllvm" > "-arm-implicit-it=never" "-mllvm" "-arm-implicit-it=always"`. > We can just assume having different values is undefined behavior. (No project > should specify different values.) Yep, that sounds good to me. The old tests seem to be written under the assumption that it's ok to specify multiple values and the last one take effect, but that doesn't seem to be the case, not even when that patch was applied in February. I'll get started at making duplicate options with the same value work in one way or another. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D102568/new/ https://reviews.llvm.org/D102568 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
