jhuber6 wrote: > > I'd like to avoid modifying this struct to avoid breaking ABI with OpenMP. > > Perhaps I should make `addr` a pointer to a struct and just GEP the two > > values. > > Are we trying to jam a square HIP peg into a round OpenMP hole, or are we > truly wanting to move to a language neutral base? If the latter, why don't we > fix the base if it isn't sufficient?
At some point I wanted to update the struct to be more generic, but it was a surprising amount of work without breaking backward compatibility w/ the OpenMP runtime. Though honestly, I forget if we ever even guaranteed that kind of backward compatibility. Long term what I wanted is that instead of using `hip_offloading_entries` and `omp_offloading_entires` we'd just have `offloading_entries` as a section and each entry has a `kind` in it when we register them. For now, it's probably easier just to put some indirection in this and pass a struct to the first argument. https://github.com/llvm/llvm-project/pull/123359 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
