[SLP]Do not create unused gather nodes for scalar arguments of vector intrinsics.

[ABataev]

Summary

If the vector intrinsic has scalar argument, we currently still create
a tree entry for this argument. This entry is not used, just consumes
resources and increases the cost of the tree.

[RKSimon]

LGTM

llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp

case Instruction::Call: {
           return;
         }
       }
 
       TreeEntry *TE = newTreeEntry(VL, Bundle /*vectorized*/, S, UserTreeIdx,
                                    ReuseShuffleIndicies);
       TE->setOperandsInOrder();
       for (unsigned i = 0, e = CI->arg_size(); i != e; ++i) {
+        // For scalar operands no need to to create an entry since no need to
+        // vectorize it.
+        if (hasVectorInstrinsicScalarOpd(ID, i))
+          continue;
         ValueList Operands;
         // Prepare the operand vector.
         for (Value *V : VL) {
           auto *CI2 = cast<CallInst>(V);
           Operands.push_back(CI2->getArgOperand(i));
         }
         buildTree_rec(Operands, Depth + 1, {TE, i});
       }

llvm/test/Transforms/SLPVectorizer/X86/intrinsic_with_scalar_param.ll

 ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ; RUN: opt < %s -slp-vectorizer -S -mtriple=x86_64-apple-macosx10.8.0 -slp-threshold=-8 | FileCheck %s
 
 declare float @llvm.powi.f32.i32(float, i32)
 define void @vec_powi_f32(float* %a, float* %c, i32 %P) {
 ; CHECK-LABEL: @vec_powi_f32(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[I0:%.*]] = load float, float* [[A:%.*]], align 4
-; CHECK-NEXT:    [[CALL1:%.*]] = tail call float @llvm.powi.f32.i32(float [[I0]], i32 [[P:%.*]])
-; CHECK-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds float, float* [[A]], i32 1
-; CHECK-NEXT:    [[I2:%.*]] = load float, float* [[ARRAYIDX2]], align 4
-; CHECK-NEXT:    [[CALL2:%.*]] = tail call float @llvm.powi.f32.i32(float [[I2]], i32 [[P]])
+; CHECK-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds float, float* [[A:%.*]], i32 1
 ; CHECK-NEXT:    [[ARRAYIDX4:%.*]] = getelementptr inbounds float, float* [[A]], i32 2
-; CHECK-NEXT:    [[I4:%.*]] = load float, float* [[ARRAYIDX4]], align 4
-; CHECK-NEXT:    [[CALL3:%.*]] = tail call float @llvm.powi.f32.i32(float [[I4]], i32 [[P]])
 ; CHECK-NEXT:    [[ARRAYIDX6:%.*]] = getelementptr inbounds float, float* [[A]], i32 3
-; CHECK-NEXT:    [[I6:%.*]] = load float, float* [[ARRAYIDX6]], align 4
-; CHECK-NEXT:    [[CALL4:%.*]] = tail call float @llvm.powi.f32.i32(float [[I6]], i32 [[P]])
-; CHECK-NEXT:    store float [[CALL1]], float* [[C:%.*]], align 4
-; CHECK-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, float* [[C]], i32 1
-; CHECK-NEXT:    store float [[CALL2]], float* [[ARRAYIDX8]], align 4
+; CHECK-NEXT:    [[TMP0:%.*]] = bitcast float* [[A]] to <4 x float>*
+; CHECK-NEXT:    [[TMP1:%.*]] = load <4 x float>, <4 x float>* [[TMP0]], align 4
+; CHECK-NEXT:    [[TMP2:%.*]] = call <4 x float> @llvm.powi.v4f32.i32(<4 x float> [[TMP1]], i32 [[P:%.*]])
+; CHECK-NEXT:    [[ARRAYIDX8:%.*]] = getelementptr inbounds float, float* [[C:%.*]], i32 1
 ; CHECK-NEXT:    [[ARRAYIDX9:%.*]] = getelementptr inbounds float, float* [[C]], i32 2
-; CHECK-NEXT:    store float [[CALL3]], float* [[ARRAYIDX9]], align 4
 ; CHECK-NEXT:    [[ARRAYIDX10:%.*]] = getelementptr inbounds float, float* [[C]], i32 3
-; CHECK-NEXT:    store float [[CALL4]], float* [[ARRAYIDX10]], align 4
+; CHECK-NEXT:    [[TMP3:%.*]] = bitcast float* [[C]] to <4 x float>*
+; CHECK-NEXT:    store <4 x float> [[TMP2]], <4 x float>* [[TMP3]], align 4
 ; CHECK-NEXT:    ret void
 ;
 entry:
   %i0 = load float, float* %a, align 4
   %call1 = tail call float @llvm.powi.f32.i32(float %i0,i32 %P)
 
   %arrayidx2 = getelementptr inbounds float, float* %a, i32 1
   %i2 = load float, float* %arrayidx2, align 4