schittir wrote:
> Thanks, @schittir, this looks good to me so long as we make the additional > changes to diagnose invalid host/device combinations in the near future. > Those `HostTarget->getPointerWidth()` checks are difficult to explain. > > When committing the changes, I suggest avoiding mention of the calling > convention concerns since this PR doesn't address them. Thank you for reviewing this as thoroughly as you did, @tahonermann ! Yes, I plan to get rid of those checks ASAP in a follow-up PR. https://github.com/llvm/llvm-project/pull/208196 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
