ChuanqiXu added a comment. In D115222#3175577 <https://reviews.llvm.org/D115222#3175577>, @rjmccall wrote:
> I imagine Gor hoped for this optimization to be implemented someday, assuming > it's still allowed by the language specification. Maybe you would be > interested in pursuing that? It sounds like it's really just (1) emitting > the intrinsic in the frontend and then (2) checking if the copied parameter > variable is actually used after reaching a suspend point, other than in ways > that wouldn't happen if the intrinsic returned false. To my knowledge of language specification, the optimization is not allowed now. Both the construction and destruction is required now. But I am indeed interested to implement this. Since I know that many implementation of C++ features comes before they become standard. So I am wondering if this could be a proposal demo to the language specification : ) BTW, I am still prefer to remove the intrinsic to avoid misunderstanding. Or is It sufficient to add a TODO/FIXME marker in the Documentation? How do you think about it? Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D115222/new/ https://reviews.llvm.org/D115222 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
