[PATCH] D139701: [Clang] Don't emit "min-legal-vector-width"="0" for AMDGPU

[Phoebe Wang via Phabricator via cfe-commits]

pengfei created this revision.
pengfei added reviewers: craig.topper, arsenm, RKSimon.
Herald added subscribers: kosarev, StephenFan, jdoerfert, kerbowa, tpr, dstuttard, yaxunl, jvesely, kzhuravl.
Herald added a project: All.
pengfei requested review of this revision.
Herald added subscribers: cfe-commits, wdng.
Herald added a project: clang.

This is an alternative way of D139627 <https://reviews.llvm.org/D139627> suggested by Craig. I'm not sure
if there are other targets care about this attribute. Let's just start
from AMDGPU.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D139701

Files:
  clang/include/clang/Basic/TargetInfo.h
  clang/lib/Basic/Targets/AMDGPU.h
  clang/lib/CodeGen/CodeGenFunction.cpp
  clang/test/OpenMP/amdgcn-attributes.cpp


Index: clang/test/OpenMP/amdgcn-attributes.cpp
===================================================================
--- clang/test/OpenMP/amdgcn-attributes.cpp
+++ clang/test/OpenMP/amdgcn-attributes.cpp
@@ -32,12 +32,12 @@
   return x + 1;
 }
 
-// DEFAULT: attributes #0 = { convergent noinline norecurse nounwind optnone "frame-pointer"="none" "kernel" "min-legal-vector-width"="0" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "uniform-work-group-size"="true" }
-// CPU: attributes #0 = { convergent noinline norecurse nounwind optnone "frame-pointer"="none" "kernel" "min-legal-vector-width"="0" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-cpu"="gfx900" "target-features"="+16-bit-insts,+ci-insts,+dpp,+flat-address-space,+gfx8-insts,+gfx9-insts,+s-memrealtime,+s-memtime-inst" "uniform-work-group-size"="true" }
-// NOIEEE: attributes #0 = { convergent noinline norecurse nounwind optnone "amdgpu-ieee"="false" "frame-pointer"="none" "kernel" "min-legal-vector-width"="0" "no-nans-fp-math"="true" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "uniform-work-group-size"="true" }
-// UNSAFEATOMIC: attributes #0 = { convergent noinline norecurse nounwind optnone "amdgpu-unsafe-fp-atomics"="true" "frame-pointer"="none" "kernel" "min-legal-vector-width"="0" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "uniform-work-group-size"="true" }
-
-// DEFAULT: attributes #1 = { convergent mustprogress noinline nounwind optnone "frame-pointer"="none" "min-legal-vector-width"="0" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
-// CPU: attributes #1 = { convergent mustprogress noinline nounwind optnone "frame-pointer"="none" "min-legal-vector-width"="0" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-cpu"="gfx900" "target-features"="+16-bit-insts,+ci-insts,+dpp,+flat-address-space,+gfx8-insts,+gfx9-insts,+s-memrealtime,+s-memtime-inst" }
-// NOIEEE: attributes #1 = { convergent mustprogress noinline nounwind optnone "amdgpu-ieee"="false" "frame-pointer"="none" "min-legal-vector-width"="0" "no-nans-fp-math"="true" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
-// UNSAFEATOMIC: attributes #1 = { convergent mustprogress noinline nounwind optnone "amdgpu-unsafe-fp-atomics"="true" "frame-pointer"="none" "min-legal-vector-width"="0" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
+// DEFAULT: attributes #0 = { convergent noinline norecurse nounwind optnone "frame-pointer"="none" "kernel" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "uniform-work-group-size"="true" }
+// CPU: attributes #0 = { convergent noinline norecurse nounwind optnone "frame-pointer"="none" "kernel" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-cpu"="gfx900" "target-features"="+16-bit-insts,+ci-insts,+dpp,+flat-address-space,+gfx8-insts,+gfx9-insts,+s-memrealtime,+s-memtime-inst" "uniform-work-group-size"="true" }
+// NOIEEE: attributes #0 = { convergent noinline norecurse nounwind optnone "amdgpu-ieee"="false" "frame-pointer"="none" "kernel" "no-nans-fp-math"="true" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "uniform-work-group-size"="true" }
+// UNSAFEATOMIC: attributes #0 = { convergent noinline norecurse nounwind optnone "amdgpu-unsafe-fp-atomics"="true" "frame-pointer"="none" "kernel" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "uniform-work-group-size"="true" }
+
+// DEFAULT: attributes #1 = { convergent mustprogress noinline nounwind optnone "frame-pointer"="none" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
+// CPU: attributes #1 = { convergent mustprogress noinline nounwind optnone "frame-pointer"="none" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-cpu"="gfx900" "target-features"="+16-bit-insts,+ci-insts,+dpp,+flat-address-space,+gfx8-insts,+gfx9-insts,+s-memrealtime,+s-memtime-inst" }
+// NOIEEE: attributes #1 = { convergent mustprogress noinline nounwind optnone "amdgpu-ieee"="false" "frame-pointer"="none" "no-nans-fp-math"="true" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
+// UNSAFEATOMIC: attributes #1 = { convergent mustprogress noinline nounwind optnone "amdgpu-unsafe-fp-atomics"="true" "frame-pointer"="none" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
Index: clang/lib/CodeGen/CodeGenFunction.cpp
===================================================================
--- clang/lib/CodeGen/CodeGenFunction.cpp
+++ clang/lib/CodeGen/CodeGenFunction.cpp
@@ -497,7 +497,9 @@
   // 4. Width of vector arguments and return types for this function.
   // 5. Width of vector aguments and return types for functions called by this
   //    function.
-  CurFn->addFnAttr("min-legal-vector-width", llvm::utostr(LargestVectorWidth));
+  if (getContext().getTargetInfo().supportMinLegalVectorWidth())
+    CurFn->addFnAttr("min-legal-vector-width",
+                     llvm::utostr(LargestVectorWidth));
 
   // Add vscale_range attribute if appropriate.
   Optional<std::pair<unsigned, unsigned>> VScaleRange =
Index: clang/lib/Basic/Targets/AMDGPU.h
===================================================================
--- clang/lib/Basic/Targets/AMDGPU.h
+++ clang/lib/Basic/Targets/AMDGPU.h
@@ -461,6 +461,8 @@
     return getCanonicalTargetID(getArchNameAMDGCN(GPUKind),
                                 OffloadArchFeatures);
   }
+
+  bool supportMinLegalVectorWidth() const override { return false; }
 };
 
 } // namespace targets
Index: clang/include/clang/Basic/TargetInfo.h
===================================================================
--- clang/include/clang/Basic/TargetInfo.h
+++ clang/include/clang/Basic/TargetInfo.h
@@ -769,6 +769,8 @@
 
   virtual bool supportSourceEvalMethod() const { return true; }
 
+  virtual bool supportMinLegalVectorWidth() const { return true; }
+
   // getLargeArrayMinWidth/Align - Return the minimum array size that is
   // 'large' and its alignment.
   unsigned getLargeArrayMinWidth() const { return LargeArrayMinWidth; }


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D139701.481585.patch
Type: text/x-patch
Size: 6068 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20221209/8008a3d5/attachment-0001.bin>