[gvn] PRE needs to skip convergent intrinsics/calls.

- As convergent intrinsics/calls could only be moved to
  control-equivalent blocks, or more precisely the same divergent
  branch, PRE needs to skip them.

Differential Revision: https://reviews.llvm.org/D90391

diff --git a/llvm/lib/Transforms/Scalar/GVN.cpp b/llvm/lib/Transforms/Scalar/GVN.cpp
index 3615316..d5fc020 100644 (file)
--- a/llvm/lib/Transforms/Scalar/GVN.cpp
+++ b/llvm/lib/Transforms/Scalar/GVN.cpp
@@ -2460,10 +2460,14 @@ bool GVN::performScalarPRE(Instruction *CurInst) {
   if (isa<GetElementPtrInst>(CurInst))
     return false;
 
-  // We don't currently value number ANY inline asm calls.
-  if (auto *CallB = dyn_cast<CallBase>(CurInst))
+  if (auto *CallB = dyn_cast<CallBase>(CurInst)) {
+    // We don't currently value number ANY inline asm calls.
     if (CallB->isInlineAsm())
       return false;
+    // Don't do PRE on convergent calls.
+    if (CallB->isConvergent())
+      return false;
+  }
 
   uint32_t ValNo = VN.lookup(CurInst);
 

diff --git a/llvm/test/Transforms/GVN/pre-skip-convergent.ll b/llvm/test/Transforms/GVN/pre-skip-convergent.ll
new file mode 100644 (file)
index 0000000..fc0d095
--- /dev/null
+++ b/llvm/test/Transforms/GVN/pre-skip-convergent.ll
@@ -0,0 +1,36 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt -S -gvn -o - %s | FileCheck %s
+; RUN: opt -S -passes=gvn -o - %s | FileCheck %s
+
+define i32 @foo(i1 %cond, i32* %q, i32* %p) {
+; CHECK-LABEL: @foo(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[V0:%.*]] = call i32 @llvm.convergent(i32 0)
+; CHECK-NEXT:    store i32 [[V0]], i32* [[Q:%.*]], align 4
+; CHECK-NEXT:    br i1 [[COND:%.*]], label [[PRE:%.*]], label [[MERGE:%.*]]
+; CHECK:       pre:
+; CHECK-NEXT:    [[T0:%.*]] = load i32, i32* [[P:%.*]], align 4
+; CHECK-NEXT:    br label [[MERGE]]
+; CHECK:       merge:
+; CHECK-NEXT:    [[M0:%.*]] = phi i32 [ [[T0]], [[PRE]] ], [ 0, [[ENTRY:%.*]] ]
+; CHECK-NEXT:    [[R0:%.*]] = call i32 @llvm.convergent(i32 [[M0]])
+; CHECK-NEXT:    ret i32 [[R0]]
+;
+entry:
+  %v0 = call i32 @llvm.convergent(i32 0)
+  store i32 %v0, i32* %q
+  br i1 %cond, label %pre, label %merge
+
+pre:
+  %t0 = load i32, i32* %p
+  br label %merge
+
+merge:
+  %m0 = phi i32 [ %t0, %pre ], [ 0, %entry ]
+  %r0 = call i32 @llvm.convergent(i32 %m0)
+  ret i32 %r0
+}
+
+declare i32 @llvm.convergent(i32) #0
+
+attributes #0 = { convergent nounwind readnone }