[PATCH] D38090: [NVPTX] Implemented shfl.sync instruction and supporting intrinsics/builtins.

[Justin Lebar via Phabricator via llvm-commits]

jlebar added inline comments.


================
Comment at: clang/lib/Headers/__clang_cuda_intrinsics.h:161
+#endif // __CUDA_VERSION >= 9000 && (!defined(__CUDA_ARCH__) || __CUDA_ARCH__ >=
+       // 300)
+
----------------
Nit, better linebreaking in the comment?


================
Comment at: llvm/include/llvm/IR/IntrinsicsNVVM.td:3744
+  Intrinsic<[llvm_i32_ty], [llvm_i32_ty, llvm_i32_ty, llvm_i32_ty, llvm_i32_ty],
+            [IntrNoMem], "llvm.nvvm.shfl.sync.down.i32">,
+  GCCBuiltin<"__nvvm_shfl_sync_down_i32">;
----------------
IntrConvergent?


https://reviews.llvm.org/D38090