zygoloid wrote: > Richard, I'm sorry to contradict you, but Aaron is correct to guess that this > is ABI-affecting: ABIs should and sometimes do specify whether an indirect > return address is allowed to be aliased.
My claim was narrowly that the choice to perform or not perform NRVO for an indirect return is not an ABI difference, which I'm not yet sure this contradicts. But you make a good point -- we shouldn't be turning off NRVO if it's only causing issues because we're also not correctly implementing an ABI rule on the caller side, and should instead fix Clang to obey the ABI rule. Do we support ABIs that don't have a rule like x86_64's? https://github.com/llvm/llvm-project/pull/101038 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
