[PATCH] D138651: [CUDA][HIP] Don't diagnose use for __bf16

[Pierre van Houtryve via Phabricator via cfe-commits]

Pierre-vh created this revision.
Pierre-vh added reviewers: arsenm, rjmccall, tra.
Herald added subscribers: kosarev, mattd, kerbowa, pengfei, tpr, yaxunl, jvesely.
Herald added a project: All.
Pierre-vh requested review of this revision.
Herald added subscribers: cfe-commits, sstefan1, wdng.
Herald added a reviewer: jdoerfert.
Herald added a project: clang.

e0fb01e97b6b7d2fe66b17b36eeb98aa78c6e3bb <https://reviews.llvm.org/rGe0fb01e97b6b7d2fe66b17b36eeb98aa78c6e3bb> caused issues in some of our HIP projects. Builds were failing because "__bf16" wasn't allowed on the target. This is because in those cases, the main target is AMDGPU (which doesn't have bf16), and the aux target is X86 (which has bf16).

This implements a fix similar to D57369 <https://reviews.llvm.org/D57369> but for bf16 which prevents Clang from diagnosing uses of bf16 when compiling heterogenous applications.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D138651

Files:
  clang/lib/AST/ASTContext.cpp
  clang/lib/Sema/SemaType.cpp
  clang/test/SemaCUDA/amdgpu-bf16.cu


Index: clang/test/SemaCUDA/amdgpu-bf16.cu
===================================================================
--- /dev/null
+++ clang/test/SemaCUDA/amdgpu-bf16.cu
@@ -0,0 +1,14 @@
+// RUN: %clang_cc1 -fsyntax-only -triple amdgcn-amd-amdhsa -aux-triple x86_64-unknown-linux-gnu -verify %s
+// expected-no-diagnostics
+
+// If AMDGPU is the main target and X86 the aux target, ensure we
+// don't complain about unsupported BF16 types in x86 code.
+
+#include "Inputs/cuda.h"
+
+__device__ void devicefn() {
+}
+
+__bf16 hostfn(__bf16 a) {
+  return a;
+}
\ No newline at end of file
Index: clang/lib/Sema/SemaType.cpp
===================================================================
--- clang/lib/Sema/SemaType.cpp
+++ clang/lib/Sema/SemaType.cpp
@@ -1518,7 +1518,9 @@
     break;
   case DeclSpec::TST_half:    Result = Context.HalfTy; break;
   case DeclSpec::TST_BFloat16:
-    if (!S.Context.getTargetInfo().hasBFloat16Type())
+    // Likewise, CUDA host and device may have different __bf16 support.
+    if (!S.Context.getTargetInfo().hasBFloat16Type() && !S.getLangOpts().CUDA &&
+        !(S.getLangOpts().OpenMP && S.getLangOpts().OpenMPIsDevice))
       S.Diag(DS.getTypeSpecTypeLoc(), diag::err_type_unsupported)
         << "__bf16";
     Result = Context.BFloat16Ty;
Index: clang/lib/AST/ASTContext.cpp
===================================================================
--- clang/lib/AST/ASTContext.cpp
+++ clang/lib/AST/ASTContext.cpp
@@ -2171,9 +2171,15 @@
       Align = Target->getLongFractAlign();
       break;
     case BuiltinType::BFloat16:
-      if (Target->hasBFloat16Type()) {
+      if (Target->hasBFloat16Type() || !getLangOpts().OpenMP ||
+          !getLangOpts().OpenMPIsDevice) {
         Width = Target->getBFloat16Width();
         Align = Target->getBFloat16Align();
+      } else {
+        assert(getLangOpts().OpenMP && getLangOpts().OpenMPIsDevice &&
+               "Expected OpenMP device compilation.");
+        Width = AuxTarget->getBFloat16Width();
+        Align = AuxTarget->getBFloat16Align();
       }
       break;
     case BuiltinType::Float16:


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D138651.477733.patch
Type: text/x-patch
Size: 2100 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20221124/afef3920/attachment.bin>