[LoopVectorize] Loop vectorization for minimum and maximum

Summary: Depends on D52766.

Reviewers: aheejin, dschuff

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D52767

llvm-svn: 344816

diff --git a/llvm/lib/Analysis/VectorUtils.cpp b/llvm/lib/Analysis/VectorUtils.cpp
index e14449b..5fd6fe0 100644 (file)
--- a/llvm/lib/Analysis/VectorUtils.cpp
+++ b/llvm/lib/Analysis/VectorUtils.cpp
@@ -54,6 +54,8 @@ bool llvm::isTriviallyVectorizable(Intrinsic::ID ID) {
   case Intrinsic::fabs:
   case Intrinsic::minnum:
   case Intrinsic::maxnum:
+  case Intrinsic::minimum:
+  case Intrinsic::maximum:
   case Intrinsic::copysign:
   case Intrinsic::floor:
   case Intrinsic::ceil:

diff --git a/llvm/test/Transforms/LoopVectorize/intrinsic.ll b/llvm/test/Transforms/LoopVectorize/intrinsic.ll
index 178d602..203c443 100644 (file)
--- a/llvm/test/Transforms/LoopVectorize/intrinsic.ll
+++ b/llvm/test/Transforms/LoopVectorize/intrinsic.ll
@@ -1247,3 +1247,59 @@ for.body:                                         ; preds = %entry, %for.body
 for.end:                                          ; preds = %for.body, %entry
   ret void
 }
+
+declare float @llvm.minimum.f32(float, float) nounwind readnone
+
+;CHECK-LABEL: @minimum_f32(
+;CHECK: llvm.minimum.v4f32
+;CHECK: ret void
+define void @minimum_f32(i32 %n, float* noalias %y, float* noalias %x, float* noalias %z) nounwind uwtable {
+entry:
+  %cmp9 = icmp sgt i32 %n, 0
+  br i1 %cmp9, label %for.body, label %for.end
+
+for.body:                                         ; preds = %entry, %for.body
+  %indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ 0, %entry ]
+  %arrayidx = getelementptr inbounds float, float* %y, i64 %indvars.iv
+  %0 = load float, float* %arrayidx, align 4
+  %arrayidx2 = getelementptr inbounds float, float* %z, i64 %indvars.iv
+  %1 = load float, float* %arrayidx2, align 4
+  %call = tail call float @llvm.minimum.f32(float %0, float %1) nounwind readnone
+  %arrayidx4 = getelementptr inbounds float, float* %x, i64 %indvars.iv
+  store float %call, float* %arrayidx4, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, %n
+  br i1 %exitcond, label %for.end, label %for.body
+
+for.end:                                          ; preds = %for.body, %entry
+  ret void
+}
+
+declare float @llvm.maximum.f32(float, float) nounwind readnone
+
+;CHECK-LABEL: @maximum_f32(
+;CHECK: llvm.maximum.v4f32
+;CHECK: ret void
+define void @maximum_f32(i32 %n, float* noalias %y, float* noalias %x, float* noalias %z) nounwind uwtable {
+entry:
+  %cmp9 = icmp sgt i32 %n, 0
+  br i1 %cmp9, label %for.body, label %for.end
+
+for.body:                                         ; preds = %entry, %for.body
+  %indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ 0, %entry ]
+  %arrayidx = getelementptr inbounds float, float* %y, i64 %indvars.iv
+  %0 = load float, float* %arrayidx, align 4
+  %arrayidx2 = getelementptr inbounds float, float* %z, i64 %indvars.iv
+  %1 = load float, float* %arrayidx2, align 4
+  %call = tail call float @llvm.maximum.f32(float %0, float %1) nounwind readnone
+  %arrayidx4 = getelementptr inbounds float, float* %x, i64 %indvars.iv
+  store float %call, float* %arrayidx4, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, %n
+  br i1 %exitcond, label %for.end, label %for.body
+
+for.end:                                          ; preds = %for.body, %entry
+  ret void
+}

diff --git a/llvm/test/Transforms/Scalarizer/intrinsics.ll b/llvm/test/Transforms/Scalarizer/intrinsics.ll
index 6c85ac3..7cebdff 100644 (file)
--- a/llvm/test/Transforms/Scalarizer/intrinsics.ll
+++ b/llvm/test/Transforms/Scalarizer/intrinsics.ll
@@ -5,6 +5,8 @@ declare <2 x float> @llvm.sqrt.v2f32(<2 x float>)
 
 ; Binary fp
 declare <2 x float> @llvm.minnum.v2f32(<2 x float>, <2 x float>)
+declare <2 x float> @llvm.minimum.v2f32(<2 x float>, <2 x float>)
+declare <2 x float> @llvm.maximum.v2f32(<2 x float>, <2 x float>)
 
 ; Ternary fp
 declare <2 x float> @llvm.fma.v2f32(<2 x float>, <2 x float>, <2 x float>)
@@ -40,6 +42,28 @@ define <2 x float> @scalarize_minnum_v2f32(<2 x float> %x, <2 x float> %y) #0 {
   ret <2 x float> %minnum
 }
 
+; CHECK-LABEL: @scalarize_minimum_v2f32(
+; CHECK: %minimum.i0 = call float @llvm.minimum.f32(float %x.i0, float %y.i0)
+; CHECK: %minimum.i1 = call float @llvm.minimum.f32(float %x.i1, float %y.i1)
+; CHECK: %minimum.upto0 = insertelement <2 x float> undef, float %minimum.i0, i32 0
+; CHECK: %minimum = insertelement <2 x float> %minimum.upto0, float %minimum.i1, i32 1
+; CHECK: ret <2 x float> %minimum
+define <2 x float> @scalarize_minimum_v2f32(<2 x float> %x, <2 x float> %y) #0 {
+  %minimum = call <2 x float> @llvm.minimum.v2f32(<2 x float> %x, <2 x float> %y)
+  ret <2 x float> %minimum
+}
+
+; CHECK-LABEL: @scalarize_maximum_v2f32(
+; CHECK: %maximum.i0 = call float @llvm.maximum.f32(float %x.i0, float %y.i0)
+; CHECK: %maximum.i1 = call float @llvm.maximum.f32(float %x.i1, float %y.i1)
+; CHECK: %maximum.upto0 = insertelement <2 x float> undef, float %maximum.i0, i32 0
+; CHECK: %maximum = insertelement <2 x float> %maximum.upto0, float %maximum.i1, i32 1
+; CHECK: ret <2 x float> %maximum
+define <2 x float> @scalarize_maximum_v2f32(<2 x float> %x, <2 x float> %y) #0 {
+  %maximum = call <2 x float> @llvm.maximum.v2f32(<2 x float> %x, <2 x float> %y)
+  ret <2 x float> %maximum
+}
+
 ; CHECK-LABEL: @scalarize_fma_v2f32(
 ; CHECK: %fma.i0 = call float @llvm.fma.f32(float %x.i0, float %y.i0, float %z.i0)
 ; CHECK: %fma.i1 = call float @llvm.fma.f32(float %x.i1, float %y.i1, float %z.i1)