arichardson wrote: Thanks for the detailed rationale, I agree with you. Omitting default base components for forward compatibility makes a lot of sense, and we shouldn't emit two single-letter base ISAs. That's what the long base format will be for.
I've updated the PR to omit i in toString() when y is present and accept rv32y/rv64y in parseNormalizedArchString(), leaving E+Y and long base names for a follow-up. A nice side effect of this is that it simplifies the overall diff, and allows merging the case statements again. https://github.com/llvm/llvm-project/pull/227865 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
