https://github.com/mjklemm approved this pull request.
LGTM. One thing to note is that the AMD implementation uses a host buffer that is being transferred to perform the `memset` operation. ROCm does not have a generic kernel for that purpose. So, in the future the exact pattern of events may change. https://github.com/llvm/llvm-project/pull/194168 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
