[PATCH] D114957: [AMDGPU] Change llvm.amdgcn.image.bvh.intersect.ray to take vec3 args

Yaxun Liu via Phabricator via cfe-commits cfe-commits at lists.llvm.org
Thu Dec 2 11:59:50 PST 2021

yaxunl added a comment.

In D114957#3167703 <https://reviews.llvm.org/D114957#3167703>, @arsenm wrote:

> In D114957#3167700 <https://reviews.llvm.org/D114957#3167700>, @arsenm wrote:
>> I think this macro is purely terrible and should not be added (and at least should be all caps?). If we can't just hard break users, I would rather just leave the builtin signatures broken
> Rather than adding an ad-hoc named macro, could they just directly check the clang version?

For the previous and the next clang release, it can be determined by checking clang version. Problem is that ROCm release is not synch'ed with clang release. So for ROCm release that uses the current clang release, it could have this change or could not, which cannot be determined by checking clang version.

  rG LLVM Github Monorepo



More information about the cfe-commits mailing list