[CostModel][RISCV] Model code size cost for reduction

[arcbbb]

Summary

Since code-size cost doesn't scale linearly with LMUL,
this change is to separate it from throughput.

[reames]

LGTM w/change applied.

Sorry for the delay, I missed this when first posted.

llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp

1021

This should still be LT.first - 1

1054

Same here

[arcbbb]

• Address @reames's comments

[arcbbb]

• Update tests

llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp

RISCVTTIImpl::getMinMaxReductionCost(VectorType *Ty, VectorType *CondTy,
   std::pair<InstructionCost, MVT> LT = getTypeLegalizationCost(Ty);
   if (Ty->getElementType()->isIntegerTy(1))
     // vcpop sequences, see vreduction-mask.ll.  umax, smin actually only
     // cost 2, but we don't have enough info here so we slightly over cost.
     return (LT.first - 1) + 3;
 
   // IR Reduction is composed by two vmv and one rvv reduction instruction.
   InstructionCost BaseCost = 2;
+
+  if (CostKind == TTI::TCK_CodeSize)
+    return (LT.first - 1) + BaseCost;
+
   unsigned VL = getEstimatedVLFor(Ty);
   return (LT.first - 1) + BaseCost + Log2_32_Ceil(VL);
 }
 
 InstructionCost
 RISCVTTIImpl::getArithmeticReductionCost(unsigned Opcode, VectorType *Ty,
                                          std::optional<FastMathFlags> FMF,
                                          TTI::TargetCostKind CostKind) {
RISCVTTIImpl::getArithmeticReductionCost(unsigned Opcode, VectorType *Ty,
 
   std::pair<InstructionCost, MVT> LT = getTypeLegalizationCost(Ty);
   if (Ty->getElementType()->isIntegerTy(1))
     // vcpop sequences, see vreduction-mask.ll
     return (LT.first - 1) + (ISD == ISD::AND ? 3 : 2);
 
   // IR Reduction is composed by two vmv and one rvv reduction instruction.
   InstructionCost BaseCost = 2;
+
+  if (CostKind == TTI::TCK_CodeSize)
+    return (LT.first - 1) + BaseCost;
+
   unsigned VL = getEstimatedVLFor(Ty);
   if (TTI::requiresOrderedReduction(FMF))
     return (LT.first - 1) + BaseCost + VL;
   return (LT.first - 1) + BaseCost + Log2_32_Ceil(VL);
 }
 
 InstructionCost RISCVTTIImpl::getExtendedReductionCost(
     unsigned Opcode, bool IsUnsigned, Type *ResTy, VectorType *ValTy,

llvm/test/Analysis/CostModel/RISCV/reduce-add.ll

 ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
 ; RUN: opt < %s -mtriple=riscv32 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s
 ; RUN: opt < %s -mtriple=riscv64 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s
+; RUN: opt < %s -mtriple=riscv32 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s --check-prefix=SIZE
+; RUN: opt < %s -mtriple=riscv64 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s --check-prefix=SIZE
 
 define i32 @reduce_i1(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i1'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i1 @llvm.vector.reduce.add.v1i1(<1 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i1 @llvm.vector.reduce.add.v2i1(<2 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i1 @llvm.vector.reduce.add.v4i1(<4 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i1 @llvm.vector.reduce.add.v8i1(<8 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i1 @llvm.vector.reduce.add.v16i1(<16 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i1 @llvm.vector.reduce.add.v32i1(<32 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i1 @llvm.vector.reduce.add.v64i1(<64 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i1 @llvm.vector.reduce.add.v128i1(<128 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i1'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i1 @llvm.vector.reduce.add.v1i1(<1 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i1 @llvm.vector.reduce.add.v2i1(<2 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i1 @llvm.vector.reduce.add.v4i1(<4 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i1 @llvm.vector.reduce.add.v8i1(<8 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i1 @llvm.vector.reduce.add.v16i1(<16 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i1 @llvm.vector.reduce.add.v32i1(<32 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i1 @llvm.vector.reduce.add.v64i1(<64 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i1 @llvm.vector.reduce.add.v128i1(<128 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i1 @llvm.vector.reduce.add.v1i1(<1 x i1> undef)
   %V2   = call i1 @llvm.vector.reduce.add.v2i1(<2 x i1> undef)
   %V4   = call i1 @llvm.vector.reduce.add.v4i1(<4 x i1> undef)
   %V8   = call i1 @llvm.vector.reduce.add.v8i1(<8 x i1> undef)
   %V16  = call i1 @llvm.vector.reduce.add.v16i1(<16 x i1> undef)
   %V32  = call i1 @llvm.vector.reduce.add.v32i1(<32 x i1> undef)
   %V64  = call i1 @llvm.vector.reduce.add.v64i1(<64 x i1> undef)
   %V128 = call i1 @llvm.vector.reduce.add.v128i1(<128 x i1> undef)
   ret i32 undef
 }
 
 define i32 @reduce_i8(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i8 @llvm.vector.reduce.add.v1i8(<1 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i8 @llvm.vector.reduce.add.v2i8(<2 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i8 @llvm.vector.reduce.add.v4i8(<4 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i8 @llvm.vector.reduce.add.v8i8(<8 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i8 @llvm.vector.reduce.add.v16i8(<16 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i8 @llvm.vector.reduce.add.v32i8(<32 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i8 @llvm.vector.reduce.add.v64i8(<64 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call i8 @llvm.vector.reduce.add.v128i8(<128 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i8 @llvm.vector.reduce.add.v1i8(<1 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i8 @llvm.vector.reduce.add.v2i8(<2 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i8 @llvm.vector.reduce.add.v4i8(<4 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i8 @llvm.vector.reduce.add.v8i8(<8 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i8 @llvm.vector.reduce.add.v16i8(<16 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i8 @llvm.vector.reduce.add.v32i8(<32 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i8 @llvm.vector.reduce.add.v64i8(<64 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i8 @llvm.vector.reduce.add.v128i8(<128 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i8 @llvm.vector.reduce.add.v1i8(<1 x i8> undef)
   %V2   = call i8 @llvm.vector.reduce.add.v2i8(<2 x i8> undef)
   %V4   = call i8 @llvm.vector.reduce.add.v4i8(<4 x i8> undef)
   %V8   = call i8 @llvm.vector.reduce.add.v8i8(<8 x i8> undef)
   %V16  = call i8 @llvm.vector.reduce.add.v16i8(<16 x i8> undef)
   %V32  = call i8 @llvm.vector.reduce.add.v32i8(<32 x i8> undef)
   %V64  = call i8 @llvm.vector.reduce.add.v64i8(<64 x i8> undef)
   %V128 = call i8 @llvm.vector.reduce.add.v128i8(<128 x i8> undef)
   ret i32 undef
 }
 
 define i32 @reduce_i16(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i16 @llvm.vector.reduce.add.v1i16(<1 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i16 @llvm.vector.reduce.add.v2i16(<2 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i16 @llvm.vector.reduce.add.v4i16(<4 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i16 @llvm.vector.reduce.add.v8i16(<8 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i16 @llvm.vector.reduce.add.v16i16(<16 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i16 @llvm.vector.reduce.add.v32i16(<32 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i16 @llvm.vector.reduce.add.v64i16(<64 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call i16 @llvm.vector.reduce.add.v128i16(<128 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i16 @llvm.vector.reduce.add.v1i16(<1 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i16 @llvm.vector.reduce.add.v2i16(<2 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i16 @llvm.vector.reduce.add.v4i16(<4 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i16 @llvm.vector.reduce.add.v8i16(<8 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i16 @llvm.vector.reduce.add.v16i16(<16 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i16 @llvm.vector.reduce.add.v32i16(<32 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i16 @llvm.vector.reduce.add.v64i16(<64 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i16 @llvm.vector.reduce.add.v128i16(<128 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i16 @llvm.vector.reduce.add.v1i16(<1 x i16> undef)
   %V2   = call i16 @llvm.vector.reduce.add.v2i16(<2 x i16> undef)
   %V4   = call i16 @llvm.vector.reduce.add.v4i16(<4 x i16> undef)
   %V8   = call i16 @llvm.vector.reduce.add.v8i16(<8 x i16> undef)
   %V16  = call i16 @llvm.vector.reduce.add.v16i16(<16 x i16> undef)
   %V32  = call i16 @llvm.vector.reduce.add.v32i16(<32 x i16> undef)
   %V64  = call i16 @llvm.vector.reduce.add.v64i16(<64 x i16> undef)
   %V128 = call i16 @llvm.vector.reduce.add.v128i16(<128 x i16> undef)
   ret i32 undef
 }
 
 define i32 @reduce_i32(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i32 @llvm.vector.reduce.add.v1i32(<1 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i32 @llvm.vector.reduce.add.v2i32(<2 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i32 @llvm.vector.reduce.add.v8i32(<8 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i32 @llvm.vector.reduce.add.v16i32(<16 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i32 @llvm.vector.reduce.add.v32i32(<32 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i32 @llvm.vector.reduce.add.v64i32(<64 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V128 = call i32 @llvm.vector.reduce.add.v128i32(<128 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i32 @llvm.vector.reduce.add.v1i32(<1 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i32 @llvm.vector.reduce.add.v2i32(<2 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i32 @llvm.vector.reduce.add.v8i32(<8 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i32 @llvm.vector.reduce.add.v16i32(<16 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i32 @llvm.vector.reduce.add.v32i32(<32 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i32 @llvm.vector.reduce.add.v64i32(<64 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i32 @llvm.vector.reduce.add.v128i32(<128 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i32 @llvm.vector.reduce.add.v1i32(<1 x i32> undef)
   %V2   = call i32 @llvm.vector.reduce.add.v2i32(<2 x i32> undef)
   %V4   = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> undef)
   %V8   = call i32 @llvm.vector.reduce.add.v8i32(<8 x i32> undef)
   %V16  = call i32 @llvm.vector.reduce.add.v16i32(<16 x i32> undef)
   %V32  = call i32 @llvm.vector.reduce.add.v32i32(<32 x i32> undef)
   %V64  = call i32 @llvm.vector.reduce.add.v64i32(<64 x i32> undef)
   %V128 = call i32 @llvm.vector.reduce.add.v128i32(<128 x i32> undef)
   ret i32 undef
 }
 
 define i32 @reduce_i64(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.vector.reduce.add.v1i64(<1 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i64 @llvm.vector.reduce.add.v2i64(<2 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i64 @llvm.vector.reduce.add.v4i64(<4 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i64 @llvm.vector.reduce.add.v8i64(<8 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i64 @llvm.vector.reduce.add.v16i64(<16 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i64 @llvm.vector.reduce.add.v32i64(<32 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V64 = call i64 @llvm.vector.reduce.add.v64i64(<64 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V128 = call i64 @llvm.vector.reduce.add.v128i64(<128 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.vector.reduce.add.v1i64(<1 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i64 @llvm.vector.reduce.add.v2i64(<2 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i64 @llvm.vector.reduce.add.v4i64(<4 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i64 @llvm.vector.reduce.add.v8i64(<8 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i64 @llvm.vector.reduce.add.v16i64(<16 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i64 @llvm.vector.reduce.add.v32i64(<32 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i64 @llvm.vector.reduce.add.v64i64(<64 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V128 = call i64 @llvm.vector.reduce.add.v128i64(<128 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i64 @llvm.vector.reduce.add.v1i64(<1 x i64> undef)
   %V2   = call i64 @llvm.vector.reduce.add.v2i64(<2 x i64> undef)
   %V4   = call i64 @llvm.vector.reduce.add.v4i64(<4 x i64> undef)
   %V8   = call i64 @llvm.vector.reduce.add.v8i64(<8 x i64> undef)
   %V16  = call i64 @llvm.vector.reduce.add.v16i64(<16 x i64> undef)
   %V32  = call i64 @llvm.vector.reduce.add.v32i64(<32 x i64> undef)
   %V64  = call i64 @llvm.vector.reduce.add.v64i64(<64 x i64> undef)
   %V128 = call i64 @llvm.vector.reduce.add.v128i64(<128 x i64> undef)

llvm/test/Analysis/CostModel/RISCV/reduce-and.ll

 ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
 ; RUN: opt < %s -mtriple=riscv32 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s
 ; RUN: opt < %s -mtriple=riscv64 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s
+; RUN: opt < %s -mtriple=riscv32 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s --check-prefix=SIZE
+; RUN: opt < %s -mtriple=riscv64 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s --check-prefix=SIZE
 
 define i32 @reduce_i1(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i1'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V1 = call i1 @llvm.vector.reduce.and.v1i1(<1 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i1 @llvm.vector.reduce.and.v2i1(<2 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i1 @llvm.vector.reduce.and.v4i1(<4 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i1 @llvm.vector.reduce.and.v8i1(<8 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i1 @llvm.vector.reduce.and.v16i1(<16 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32 = call i1 @llvm.vector.reduce.and.v32i1(<32 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i1 @llvm.vector.reduce.and.v64i1(<64 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i1 @llvm.vector.reduce.and.v128i1(<128 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V256 = call i1 @llvm.vector.reduce.and.v256i1(<256 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V512 = call i1 @llvm.vector.reduce.and.v512i1(<512 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V1024 = call i1 @llvm.vector.reduce.and.v1024i1(<1024 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i1'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V1 = call i1 @llvm.vector.reduce.and.v1i1(<1 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i1 @llvm.vector.reduce.and.v2i1(<2 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i1 @llvm.vector.reduce.and.v4i1(<4 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i1 @llvm.vector.reduce.and.v8i1(<8 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i1 @llvm.vector.reduce.and.v16i1(<16 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32 = call i1 @llvm.vector.reduce.and.v32i1(<32 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i1 @llvm.vector.reduce.and.v64i1(<64 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i1 @llvm.vector.reduce.and.v128i1(<128 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V256 = call i1 @llvm.vector.reduce.and.v256i1(<256 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V512 = call i1 @llvm.vector.reduce.and.v512i1(<512 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V1024 = call i1 @llvm.vector.reduce.and.v1024i1(<1024 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i1 @llvm.vector.reduce.and.v1i1(<1 x i1> undef)
   %V2   = call i1 @llvm.vector.reduce.and.v2i1(<2 x i1> undef)
   %V4   = call i1 @llvm.vector.reduce.and.v4i1(<4 x i1> undef)
   %V8   = call i1 @llvm.vector.reduce.and.v8i1(<8 x i1> undef)
   %V16  = call i1 @llvm.vector.reduce.and.v16i1(<16 x i1> undef)
   %V32  = call i1 @llvm.vector.reduce.and.v32i1(<32 x i1> undef)
   %V64  = call i1 @llvm.vector.reduce.and.v64i1(<64 x i1> undef)
   %V128 = call i1 @llvm.vector.reduce.and.v128i1(<128 x i1> undef)

 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i8 @llvm.vector.reduce.and.v4i8(<4 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i8 @llvm.vector.reduce.and.v8i8(<8 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i8 @llvm.vector.reduce.and.v16i8(<16 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i8 @llvm.vector.reduce.and.v32i8(<32 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i8 @llvm.vector.reduce.and.v64i8(<64 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call i8 @llvm.vector.reduce.and.v128i8(<128 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i8 @llvm.vector.reduce.and.v1i8(<1 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i8 @llvm.vector.reduce.and.v2i8(<2 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i8 @llvm.vector.reduce.and.v4i8(<4 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i8 @llvm.vector.reduce.and.v8i8(<8 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i8 @llvm.vector.reduce.and.v16i8(<16 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i8 @llvm.vector.reduce.and.v32i8(<32 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i8 @llvm.vector.reduce.and.v64i8(<64 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i8 @llvm.vector.reduce.and.v128i8(<128 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i8 @llvm.vector.reduce.and.v1i8(<1 x i8> undef)
   %V2   = call i8 @llvm.vector.reduce.and.v2i8(<2 x i8> undef)
   %V4   = call i8 @llvm.vector.reduce.and.v4i8(<4 x i8> undef)
   %V8   = call i8 @llvm.vector.reduce.and.v8i8(<8 x i8> undef)
   %V16  = call i8 @llvm.vector.reduce.and.v16i8(<16 x i8> undef)
   %V32  = call i8 @llvm.vector.reduce.and.v32i8(<32 x i8> undef)
   %V64  = call i8 @llvm.vector.reduce.and.v64i8(<64 x i8> undef)
   %V128 = call i8 @llvm.vector.reduce.and.v128i8(<128 x i8> undef)
   ret i32 undef
 }
 
 define i32 @reduce_i16(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i16 @llvm.vector.reduce.and.v1i16(<1 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i16 @llvm.vector.reduce.and.v2i16(<2 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i16 @llvm.vector.reduce.and.v4i16(<4 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i16 @llvm.vector.reduce.and.v8i16(<8 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i16 @llvm.vector.reduce.and.v16i16(<16 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i16 @llvm.vector.reduce.and.v32i16(<32 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i16 @llvm.vector.reduce.and.v64i16(<64 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call i16 @llvm.vector.reduce.and.v128i16(<128 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i16 @llvm.vector.reduce.and.v1i16(<1 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i16 @llvm.vector.reduce.and.v2i16(<2 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i16 @llvm.vector.reduce.and.v4i16(<4 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i16 @llvm.vector.reduce.and.v8i16(<8 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i16 @llvm.vector.reduce.and.v16i16(<16 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i16 @llvm.vector.reduce.and.v32i16(<32 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i16 @llvm.vector.reduce.and.v64i16(<64 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i16 @llvm.vector.reduce.and.v128i16(<128 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i16 @llvm.vector.reduce.and.v1i16(<1 x i16> undef)
   %V2   = call i16 @llvm.vector.reduce.and.v2i16(<2 x i16> undef)
   %V4   = call i16 @llvm.vector.reduce.and.v4i16(<4 x i16> undef)
   %V8   = call i16 @llvm.vector.reduce.and.v8i16(<8 x i16> undef)
   %V16  = call i16 @llvm.vector.reduce.and.v16i16(<16 x i16> undef)
   %V32  = call i16 @llvm.vector.reduce.and.v32i16(<32 x i16> undef)
   %V64  = call i16 @llvm.vector.reduce.and.v64i16(<64 x i16> undef)
   %V128 = call i16 @llvm.vector.reduce.and.v128i16(<128 x i16> undef)
   ret i32 undef
 }
 
 define i32 @reduce_i32(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i32 @llvm.vector.reduce.and.v1i32(<1 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i32 @llvm.vector.reduce.and.v2i32(<2 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i32 @llvm.vector.reduce.and.v4i32(<4 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i32 @llvm.vector.reduce.and.v8i32(<8 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i32 @llvm.vector.reduce.and.v16i32(<16 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i32 @llvm.vector.reduce.and.v32i32(<32 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i32 @llvm.vector.reduce.and.v64i32(<64 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V128 = call i32 @llvm.vector.reduce.and.v128i32(<128 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i32 @llvm.vector.reduce.and.v1i32(<1 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i32 @llvm.vector.reduce.and.v2i32(<2 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i32 @llvm.vector.reduce.and.v4i32(<4 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i32 @llvm.vector.reduce.and.v8i32(<8 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i32 @llvm.vector.reduce.and.v16i32(<16 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i32 @llvm.vector.reduce.and.v32i32(<32 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i32 @llvm.vector.reduce.and.v64i32(<64 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i32 @llvm.vector.reduce.and.v128i32(<128 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i32 @llvm.vector.reduce.and.v1i32(<1 x i32> undef)
   %V2   = call i32 @llvm.vector.reduce.and.v2i32(<2 x i32> undef)
   %V4   = call i32 @llvm.vector.reduce.and.v4i32(<4 x i32> undef)
   %V8   = call i32 @llvm.vector.reduce.and.v8i32(<8 x i32> undef)
   %V16  = call i32 @llvm.vector.reduce.and.v16i32(<16 x i32> undef)
   %V32  = call i32 @llvm.vector.reduce.and.v32i32(<32 x i32> undef)
   %V64  = call i32 @llvm.vector.reduce.and.v64i32(<64 x i32> undef)
   %V128 = call i32 @llvm.vector.reduce.and.v128i32(<128 x i32> undef)
   ret i32 undef
 }
 
 define i32 @reduce_i64(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.vector.reduce.and.v1i64(<1 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i64 @llvm.vector.reduce.and.v2i64(<2 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i64 @llvm.vector.reduce.and.v4i64(<4 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i64 @llvm.vector.reduce.and.v8i64(<8 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i64 @llvm.vector.reduce.and.v16i64(<16 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i64 @llvm.vector.reduce.and.v32i64(<32 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V64 = call i64 @llvm.vector.reduce.and.v64i64(<64 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V128 = call i64 @llvm.vector.reduce.and.v128i64(<128 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.vector.reduce.and.v1i64(<1 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i64 @llvm.vector.reduce.and.v2i64(<2 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i64 @llvm.vector.reduce.and.v4i64(<4 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i64 @llvm.vector.reduce.and.v8i64(<8 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i64 @llvm.vector.reduce.and.v16i64(<16 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i64 @llvm.vector.reduce.and.v32i64(<32 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i64 @llvm.vector.reduce.and.v64i64(<64 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V128 = call i64 @llvm.vector.reduce.and.v128i64(<128 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i64 @llvm.vector.reduce.and.v1i64(<1 x i64> undef)
   %V2   = call i64 @llvm.vector.reduce.and.v2i64(<2 x i64> undef)
   %V4   = call i64 @llvm.vector.reduce.and.v4i64(<4 x i64> undef)
   %V8   = call i64 @llvm.vector.reduce.and.v8i64(<8 x i64> undef)
   %V16  = call i64 @llvm.vector.reduce.and.v16i64(<16 x i64> undef)
   %V32  = call i64 @llvm.vector.reduce.and.v32i64(<32 x i64> undef)
   %V64  = call i64 @llvm.vector.reduce.and.v64i64(<64 x i64> undef)
   %V128 = call i64 @llvm.vector.reduce.and.v128i64(<128 x i64> undef)

llvm/test/Analysis/CostModel/RISCV/reduce-fadd.ll

 ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
 ; RUN: opt < %s -mtriple=riscv64 -mattr=+v,+zfh,+experimental-zvfh -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s --check-prefix=FP-REDUCE
+; RUN: opt < %s -mtriple=riscv64 -mattr=+v,+zfh,+experimental-zvfh -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s  --check-prefix=SIZE
 
 define void @reduce_fadd_half() {
 ; FP-REDUCE-LABEL: 'reduce_fadd_half'
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call fast half @llvm.vector.reduce.fadd.v1f16(half 0xH0000, <1 x half> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call fast half @llvm.vector.reduce.fadd.v2f16(half 0xH0000, <2 x half> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call fast half @llvm.vector.reduce.fadd.v4f16(half 0xH0000, <4 x half> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call fast half @llvm.vector.reduce.fadd.v8f16(half 0xH0000, <8 x half> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call fast half @llvm.vector.reduce.fadd.v16f16(half 0xH0000, <16 x half> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %v32 = call fast half @llvm.vector.reduce.fadd.v32f16(half 0xH0000, <32 x half> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call fast half @llvm.vector.reduce.fadd.v64f16(half 0xH0000, <64 x half> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call fast half @llvm.vector.reduce.fadd.v128f16(half 0xH0000, <128 x half> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
 ;
+; SIZE-LABEL: 'reduce_fadd_half'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call fast half @llvm.vector.reduce.fadd.v1f16(half 0xH0000, <1 x half> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call fast half @llvm.vector.reduce.fadd.v2f16(half 0xH0000, <2 x half> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call fast half @llvm.vector.reduce.fadd.v4f16(half 0xH0000, <4 x half> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call fast half @llvm.vector.reduce.fadd.v8f16(half 0xH0000, <8 x half> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call fast half @llvm.vector.reduce.fadd.v16f16(half 0xH0000, <16 x half> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v32 = call fast half @llvm.vector.reduce.fadd.v32f16(half 0xH0000, <32 x half> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call fast half @llvm.vector.reduce.fadd.v64f16(half 0xH0000, <64 x half> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call fast half @llvm.vector.reduce.fadd.v128f16(half 0xH0000, <128 x half> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
+;
   %V1 = call fast half @llvm.vector.reduce.fadd.v1f16(half 0.0, <1 x half> undef)
   %V2 = call fast half @llvm.vector.reduce.fadd.v2f16(half 0.0, <2 x half> undef)
   %V4 = call fast half @llvm.vector.reduce.fadd.v4f16(half 0.0, <4 x half> undef)
   %V8 = call fast half @llvm.vector.reduce.fadd.v8f16(half 0.0, <8 x half> undef)
   %V16 = call fast half @llvm.vector.reduce.fadd.v16f16(half 0.0, <16 x half> undef)
   %v32 = call fast half @llvm.vector.reduce.fadd.v32f16(half 0.0, <32 x half> undef)
   %V64 = call fast half @llvm.vector.reduce.fadd.v64f16(half 0.0, <64 x half> undef)
   %V128 = call fast half @llvm.vector.reduce.fadd.v128f16(half 0.0, <128 x half> undef)
   ret void
 }
 
 define void @reduce_fadd_float() {
 ; FP-REDUCE-LABEL: 'reduce_fadd_float'
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call fast float @llvm.vector.reduce.fadd.v1f32(float 0.000000e+00, <1 x float> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call fast float @llvm.vector.reduce.fadd.v2f32(float 0.000000e+00, <2 x float> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call fast float @llvm.vector.reduce.fadd.v4f32(float 0.000000e+00, <4 x float> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call fast float @llvm.vector.reduce.fadd.v8f32(float 0.000000e+00, <8 x float> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call fast float @llvm.vector.reduce.fadd.v16f32(float 0.000000e+00, <16 x float> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %v32 = call fast float @llvm.vector.reduce.fadd.v32f32(float 0.000000e+00, <32 x float> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call fast float @llvm.vector.reduce.fadd.v64f32(float 0.000000e+00, <64 x float> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V128 = call fast float @llvm.vector.reduce.fadd.v128f32(float 0.000000e+00, <128 x float> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
 ;
+; SIZE-LABEL: 'reduce_fadd_float'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call fast float @llvm.vector.reduce.fadd.v1f32(float 0.000000e+00, <1 x float> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call fast float @llvm.vector.reduce.fadd.v2f32(float 0.000000e+00, <2 x float> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call fast float @llvm.vector.reduce.fadd.v4f32(float 0.000000e+00, <4 x float> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call fast float @llvm.vector.reduce.fadd.v8f32(float 0.000000e+00, <8 x float> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call fast float @llvm.vector.reduce.fadd.v16f32(float 0.000000e+00, <16 x float> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v32 = call fast float @llvm.vector.reduce.fadd.v32f32(float 0.000000e+00, <32 x float> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call fast float @llvm.vector.reduce.fadd.v64f32(float 0.000000e+00, <64 x float> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call fast float @llvm.vector.reduce.fadd.v128f32(float 0.000000e+00, <128 x float> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
+;
   %V1 = call fast float @llvm.vector.reduce.fadd.v1f32(float 0.0, <1 x float> undef)
   %V2 = call fast float @llvm.vector.reduce.fadd.v2f32(float 0.0, <2 x float> undef)
   %V4 = call fast float @llvm.vector.reduce.fadd.v4f32(float 0.0, <4 x float> undef)
   %V8 = call fast float @llvm.vector.reduce.fadd.v8f32(float 0.0, <8 x float> undef)
   %V16 = call fast float @llvm.vector.reduce.fadd.v16f32(float 0.0, <16 x float> undef)
   %v32 = call fast float @llvm.vector.reduce.fadd.v32f32(float 0.0, <32 x float> undef)
   %V64 = call fast float @llvm.vector.reduce.fadd.v64f32(float 0.0, <64 x float> undef)
   %V128 = call fast float @llvm.vector.reduce.fadd.v128f32(float 0.0, <128 x float> undef)
   ret void
 }
 
 define void @reduce_fadd_double() {
 ; FP-REDUCE-LABEL: 'reduce_fadd_double'
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call fast double @llvm.vector.reduce.fadd.v1f64(double 0.000000e+00, <1 x double> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call fast double @llvm.vector.reduce.fadd.v2f64(double 0.000000e+00, <2 x double> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call fast double @llvm.vector.reduce.fadd.v4f64(double 0.000000e+00, <4 x double> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call fast double @llvm.vector.reduce.fadd.v8f64(double 0.000000e+00, <8 x double> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call fast double @llvm.vector.reduce.fadd.v16f64(double 0.000000e+00, <16 x double> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %v32 = call fast double @llvm.vector.reduce.fadd.v32f64(double 0.000000e+00, <32 x double> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V64 = call fast double @llvm.vector.reduce.fadd.v64f64(double 0.000000e+00, <64 x double> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V128 = call fast double @llvm.vector.reduce.fadd.v128f64(double 0.000000e+00, <128 x double> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
 ;
+; SIZE-LABEL: 'reduce_fadd_double'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call fast double @llvm.vector.reduce.fadd.v1f64(double 0.000000e+00, <1 x double> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call fast double @llvm.vector.reduce.fadd.v2f64(double 0.000000e+00, <2 x double> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call fast double @llvm.vector.reduce.fadd.v4f64(double 0.000000e+00, <4 x double> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call fast double @llvm.vector.reduce.fadd.v8f64(double 0.000000e+00, <8 x double> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call fast double @llvm.vector.reduce.fadd.v16f64(double 0.000000e+00, <16 x double> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v32 = call fast double @llvm.vector.reduce.fadd.v32f64(double 0.000000e+00, <32 x double> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call fast double @llvm.vector.reduce.fadd.v64f64(double 0.000000e+00, <64 x double> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V128 = call fast double @llvm.vector.reduce.fadd.v128f64(double 0.000000e+00, <128 x double> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
+;
   %V1 = call fast double @llvm.vector.reduce.fadd.v1f64(double 0.0, <1 x double> undef)
   %V2 = call fast double @llvm.vector.reduce.fadd.v2f64(double 0.0, <2 x double> undef)
   %V4 = call fast double @llvm.vector.reduce.fadd.v4f64(double 0.0, <4 x double> undef)
   %V8 = call fast double @llvm.vector.reduce.fadd.v8f64(double 0.0, <8 x double> undef)
   %V16 = call fast double @llvm.vector.reduce.fadd.v16f64(double 0.0, <16 x double> undef)
   %v32 = call fast double @llvm.vector.reduce.fadd.v32f64(double 0.0, <32 x double> undef)
   %V64 = call fast double @llvm.vector.reduce.fadd.v64f64(double 0.0, <64 x double> undef)
   %V128 = call fast double @llvm.vector.reduce.fadd.v128f64(double 0.0, <128 x double> undef)
   ret void
 }
 
 define void @reduce_oredered_fadd_half() {
 ; FP-REDUCE-LABEL: 'reduce_oredered_fadd_half'
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V1 = call half @llvm.vector.reduce.fadd.v1f16(half 0xH0000, <1 x half> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2 = call half @llvm.vector.reduce.fadd.v2f16(half 0xH0000, <2 x half> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4 = call half @llvm.vector.reduce.fadd.v4f16(half 0xH0000, <4 x half> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8 = call half @llvm.vector.reduce.fadd.v8f16(half 0xH0000, <8 x half> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V16 = call half @llvm.vector.reduce.fadd.v16f16(half 0xH0000, <16 x half> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %v32 = call half @llvm.vector.reduce.fadd.v32f16(half 0xH0000, <32 x half> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 66 for instruction: %V64 = call half @llvm.vector.reduce.fadd.v64f16(half 0xH0000, <64 x half> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 130 for instruction: %V128 = call half @llvm.vector.reduce.fadd.v128f16(half 0xH0000, <128 x half> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
 ;
+; SIZE-LABEL: 'reduce_oredered_fadd_half'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call half @llvm.vector.reduce.fadd.v1f16(half 0xH0000, <1 x half> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call half @llvm.vector.reduce.fadd.v2f16(half 0xH0000, <2 x half> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call half @llvm.vector.reduce.fadd.v4f16(half 0xH0000, <4 x half> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call half @llvm.vector.reduce.fadd.v8f16(half 0xH0000, <8 x half> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call half @llvm.vector.reduce.fadd.v16f16(half 0xH0000, <16 x half> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v32 = call half @llvm.vector.reduce.fadd.v32f16(half 0xH0000, <32 x half> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call half @llvm.vector.reduce.fadd.v64f16(half 0xH0000, <64 x half> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call half @llvm.vector.reduce.fadd.v128f16(half 0xH0000, <128 x half> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
+;
   %V1 = call half @llvm.vector.reduce.fadd.v1f16(half 0.0, <1 x half> undef)
   %V2 = call half @llvm.vector.reduce.fadd.v2f16(half 0.0, <2 x half> undef)
   %V4 = call half @llvm.vector.reduce.fadd.v4f16(half 0.0, <4 x half> undef)
   %V8 = call half @llvm.vector.reduce.fadd.v8f16(half 0.0, <8 x half> undef)
   %V16 = call half @llvm.vector.reduce.fadd.v16f16(half 0.0, <16 x half> undef)
   %v32 = call half @llvm.vector.reduce.fadd.v32f16(half 0.0, <32 x half> undef)
   %V64 = call half @llvm.vector.reduce.fadd.v64f16(half 0.0, <64 x half> undef)
   %V128 = call half @llvm.vector.reduce.fadd.v128f16(half 0.0, <128 x half> undef)
   ret void
 }
 
 define void @reduce_oredered_fadd_float() {
 ; FP-REDUCE-LABEL: 'reduce_oredered_fadd_float'
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V1 = call float @llvm.vector.reduce.fadd.v1f32(float 0.000000e+00, <1 x float> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2 = call float @llvm.vector.reduce.fadd.v2f32(float 0.000000e+00, <2 x float> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4 = call float @llvm.vector.reduce.fadd.v4f32(float 0.000000e+00, <4 x float> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8 = call float @llvm.vector.reduce.fadd.v8f32(float 0.000000e+00, <8 x float> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V16 = call float @llvm.vector.reduce.fadd.v16f32(float 0.000000e+00, <16 x float> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %v32 = call float @llvm.vector.reduce.fadd.v32f32(float 0.000000e+00, <32 x float> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 66 for instruction: %V64 = call float @llvm.vector.reduce.fadd.v64f32(float 0.000000e+00, <64 x float> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 131 for instruction: %V128 = call float @llvm.vector.reduce.fadd.v128f32(float 0.000000e+00, <128 x float> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
 ;
+; SIZE-LABEL: 'reduce_oredered_fadd_float'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call float @llvm.vector.reduce.fadd.v1f32(float 0.000000e+00, <1 x float> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call float @llvm.vector.reduce.fadd.v2f32(float 0.000000e+00, <2 x float> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call float @llvm.vector.reduce.fadd.v4f32(float 0.000000e+00, <4 x float> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call float @llvm.vector.reduce.fadd.v8f32(float 0.000000e+00, <8 x float> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call float @llvm.vector.reduce.fadd.v16f32(float 0.000000e+00, <16 x float> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v32 = call float @llvm.vector.reduce.fadd.v32f32(float 0.000000e+00, <32 x float> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call float @llvm.vector.reduce.fadd.v64f32(float 0.000000e+00, <64 x float> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call float @llvm.vector.reduce.fadd.v128f32(float 0.000000e+00, <128 x float> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
+;
   %V1 = call float @llvm.vector.reduce.fadd.v1f32(float 0.0, <1 x float> undef)
   %V2 = call float @llvm.vector.reduce.fadd.v2f32(float 0.0, <2 x float> undef)
   %V4 = call float @llvm.vector.reduce.fadd.v4f32(float 0.0, <4 x float> undef)
   %V8 = call float @llvm.vector.reduce.fadd.v8f32(float 0.0, <8 x float> undef)
   %V16 = call float @llvm.vector.reduce.fadd.v16f32(float 0.0, <16 x float> undef)
   %v32 = call float @llvm.vector.reduce.fadd.v32f32(float 0.0, <32 x float> undef)
   %V64 = call float @llvm.vector.reduce.fadd.v64f32(float 0.0, <64 x float> undef)
   %V128 = call float @llvm.vector.reduce.fadd.v128f32(float 0.0, <128 x float> undef)
   ret void
 }
 
 define void @reduce_oredered_fadd_double() {
 ; FP-REDUCE-LABEL: 'reduce_oredered_fadd_double'
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V1 = call double @llvm.vector.reduce.fadd.v1f64(double 0.000000e+00, <1 x double> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2 = call double @llvm.vector.reduce.fadd.v2f64(double 0.000000e+00, <2 x double> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4 = call double @llvm.vector.reduce.fadd.v4f64(double 0.000000e+00, <4 x double> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8 = call double @llvm.vector.reduce.fadd.v8f64(double 0.000000e+00, <8 x double> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V16 = call double @llvm.vector.reduce.fadd.v16f64(double 0.000000e+00, <16 x double> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %v32 = call double @llvm.vector.reduce.fadd.v32f64(double 0.000000e+00, <32 x double> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 67 for instruction: %V64 = call double @llvm.vector.reduce.fadd.v64f64(double 0.000000e+00, <64 x double> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 133 for instruction: %V128 = call double @llvm.vector.reduce.fadd.v128f64(double 0.000000e+00, <128 x double> undef)
 ; FP-REDUCE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
 ;
+; SIZE-LABEL: 'reduce_oredered_fadd_double'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call double @llvm.vector.reduce.fadd.v1f64(double 0.000000e+00, <1 x double> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call double @llvm.vector.reduce.fadd.v2f64(double 0.000000e+00, <2 x double> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call double @llvm.vector.reduce.fadd.v4f64(double 0.000000e+00, <4 x double> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call double @llvm.vector.reduce.fadd.v8f64(double 0.000000e+00, <8 x double> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call double @llvm.vector.reduce.fadd.v16f64(double 0.000000e+00, <16 x double> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %v32 = call double @llvm.vector.reduce.fadd.v32f64(double 0.000000e+00, <32 x double> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call double @llvm.vector.reduce.fadd.v64f64(double 0.000000e+00, <64 x double> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V128 = call double @llvm.vector.reduce.fadd.v128f64(double 0.000000e+00, <128 x double> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret void
+;
   %V1 = call double @llvm.vector.reduce.fadd.v1f64(double 0.0, <1 x double> undef)
   %V2 = call double @llvm.vector.reduce.fadd.v2f64(double 0.0, <2 x double> undef)
   %V4 = call double @llvm.vector.reduce.fadd.v4f64(double 0.0, <4 x double> undef)
   %V8 = call double @llvm.vector.reduce.fadd.v8f64(double 0.0, <8 x double> undef)
   %V16 = call double @llvm.vector.reduce.fadd.v16f64(double 0.0, <16 x double> undef)
   %v32 = call double @llvm.vector.reduce.fadd.v32f64(double 0.0, <32 x double> undef)
   %V64 = call double @llvm.vector.reduce.fadd.v64f64(double 0.0, <64 x double> undef)
   %V128 = call double @llvm.vector.reduce.fadd.v128f64(double 0.0, <128 x double> undef)

llvm/test/Analysis/CostModel/RISCV/reduce-max.ll

 ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
 ; RUN: opt < %s -mtriple=riscv32 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s
 ; RUN: opt < %s -mtriple=riscv64 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s
+; RUN: opt < %s -mtriple=riscv32 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s --check-prefix=SIZE
+; RUN: opt < %s -mtriple=riscv64 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s --check-prefix=SIZE
 
 define i32 @reduce_umin_i1(i32 %arg) {
 ; CHECK-LABEL: 'reduce_umin_i1'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V1 = call i1 @llvm.vector.reduce.umax.v1i1(<1 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i1 @llvm.vector.reduce.umax.v2i1(<2 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i1 @llvm.vector.reduce.umax.v4i1(<4 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i1 @llvm.vector.reduce.umax.v8i1(<8 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i1 @llvm.vector.reduce.umax.v16i1(<16 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32 = call i1 @llvm.vector.reduce.umax.v32i1(<32 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i1 @llvm.vector.reduce.umax.v64i1(<64 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i1 @llvm.vector.reduce.umax.v128i1(<128 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_umin_i1'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V1 = call i1 @llvm.vector.reduce.umax.v1i1(<1 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i1 @llvm.vector.reduce.umax.v2i1(<2 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i1 @llvm.vector.reduce.umax.v4i1(<4 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i1 @llvm.vector.reduce.umax.v8i1(<8 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i1 @llvm.vector.reduce.umax.v16i1(<16 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32 = call i1 @llvm.vector.reduce.umax.v32i1(<32 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i1 @llvm.vector.reduce.umax.v64i1(<64 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i1 @llvm.vector.reduce.umax.v128i1(<128 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i1 @llvm.vector.reduce.umax.v1i1(<1 x i1> undef)
   %V2   = call i1 @llvm.vector.reduce.umax.v2i1(<2 x i1> undef)
   %V4   = call i1 @llvm.vector.reduce.umax.v4i1(<4 x i1> undef)
   %V8   = call i1 @llvm.vector.reduce.umax.v8i1(<8 x i1> undef)
   %V16  = call i1 @llvm.vector.reduce.umax.v16i1(<16 x i1> undef)
   %V32  = call i1 @llvm.vector.reduce.umax.v32i1(<32 x i1> undef)
   %V64  = call i1 @llvm.vector.reduce.umax.v64i1(<64 x i1> undef)
   %V128 = call i1 @llvm.vector.reduce.umax.v128i1(<128 x i1> undef)
   ret i32 undef
 }
 
 define i32 @reduce_umax_i8(i32 %arg) {
 ; CHECK-LABEL: 'reduce_umax_i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i8 @llvm.vector.reduce.umax.v1i8(<1 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i8 @llvm.vector.reduce.umax.v2i8(<2 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i8 @llvm.vector.reduce.umax.v4i8(<4 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i8 @llvm.vector.reduce.umax.v8i8(<8 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i8 @llvm.vector.reduce.umax.v16i8(<16 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i8 @llvm.vector.reduce.umax.v32i8(<32 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i8 @llvm.vector.reduce.umax.v64i8(<64 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call i8 @llvm.vector.reduce.umax.v128i8(<128 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_umax_i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i8 @llvm.vector.reduce.umax.v1i8(<1 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i8 @llvm.vector.reduce.umax.v2i8(<2 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i8 @llvm.vector.reduce.umax.v4i8(<4 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i8 @llvm.vector.reduce.umax.v8i8(<8 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i8 @llvm.vector.reduce.umax.v16i8(<16 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i8 @llvm.vector.reduce.umax.v32i8(<32 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i8 @llvm.vector.reduce.umax.v64i8(<64 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i8 @llvm.vector.reduce.umax.v128i8(<128 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i8 @llvm.vector.reduce.umax.v1i8(<1 x i8> undef)
   %V2   = call i8 @llvm.vector.reduce.umax.v2i8(<2 x i8> undef)
   %V4   = call i8 @llvm.vector.reduce.umax.v4i8(<4 x i8> undef)
   %V8   = call i8 @llvm.vector.reduce.umax.v8i8(<8 x i8> undef)
   %V16  = call i8 @llvm.vector.reduce.umax.v16i8(<16 x i8> undef)
   %V32  = call i8 @llvm.vector.reduce.umax.v32i8(<32 x i8> undef)
   %V64  = call i8 @llvm.vector.reduce.umax.v64i8(<64 x i8> undef)
   %V128 = call i8 @llvm.vector.reduce.umax.v128i8(<128 x i8> undef)
   ret i32 undef
 }
 
 define i32 @reduce_umax_i16(i32 %arg) {
 ; CHECK-LABEL: 'reduce_umax_i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i16 @llvm.vector.reduce.umax.v1i16(<1 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i16 @llvm.vector.reduce.umax.v2i16(<2 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i16 @llvm.vector.reduce.umax.v4i16(<4 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i16 @llvm.vector.reduce.umax.v8i16(<8 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i16 @llvm.vector.reduce.umax.v16i16(<16 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i16 @llvm.vector.reduce.umax.v32i16(<32 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i16 @llvm.vector.reduce.umax.v64i16(<64 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call i16 @llvm.vector.reduce.umax.v128i16(<128 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_umax_i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i16 @llvm.vector.reduce.umax.v1i16(<1 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i16 @llvm.vector.reduce.umax.v2i16(<2 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i16 @llvm.vector.reduce.umax.v4i16(<4 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i16 @llvm.vector.reduce.umax.v8i16(<8 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i16 @llvm.vector.reduce.umax.v16i16(<16 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i16 @llvm.vector.reduce.umax.v32i16(<32 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i16 @llvm.vector.reduce.umax.v64i16(<64 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i16 @llvm.vector.reduce.umax.v128i16(<128 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i16 @llvm.vector.reduce.umax.v1i16(<1 x i16> undef)
   %V2   = call i16 @llvm.vector.reduce.umax.v2i16(<2 x i16> undef)
   %V4   = call i16 @llvm.vector.reduce.umax.v4i16(<4 x i16> undef)
   %V8   = call i16 @llvm.vector.reduce.umax.v8i16(<8 x i16> undef)
   %V16  = call i16 @llvm.vector.reduce.umax.v16i16(<16 x i16> undef)
   %V32  = call i16 @llvm.vector.reduce.umax.v32i16(<32 x i16> undef)
   %V64  = call i16 @llvm.vector.reduce.umax.v64i16(<64 x i16> undef)
   %V128 = call i16 @llvm.vector.reduce.umax.v128i16(<128 x i16> undef)
   ret i32 undef
 }
 
 define i32 @reduce_umax_i32(i32 %arg) {
 ; CHECK-LABEL: 'reduce_umax_i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i32 @llvm.vector.reduce.umax.v1i32(<1 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i32 @llvm.vector.reduce.umax.v2i32(<2 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i32 @llvm.vector.reduce.umax.v4i32(<4 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i32 @llvm.vector.reduce.umax.v8i32(<8 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i32 @llvm.vector.reduce.umax.v16i32(<16 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i32 @llvm.vector.reduce.umax.v32i32(<32 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i32 @llvm.vector.reduce.umax.v64i32(<64 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V128 = call i32 @llvm.vector.reduce.umax.v128i32(<128 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_umax_i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i32 @llvm.vector.reduce.umax.v1i32(<1 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i32 @llvm.vector.reduce.umax.v2i32(<2 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i32 @llvm.vector.reduce.umax.v4i32(<4 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i32 @llvm.vector.reduce.umax.v8i32(<8 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i32 @llvm.vector.reduce.umax.v16i32(<16 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i32 @llvm.vector.reduce.umax.v32i32(<32 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i32 @llvm.vector.reduce.umax.v64i32(<64 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i32 @llvm.vector.reduce.umax.v128i32(<128 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i32 @llvm.vector.reduce.umax.v1i32(<1 x i32> undef)
   %V2   = call i32 @llvm.vector.reduce.umax.v2i32(<2 x i32> undef)
   %V4   = call i32 @llvm.vector.reduce.umax.v4i32(<4 x i32> undef)
   %V8   = call i32 @llvm.vector.reduce.umax.v8i32(<8 x i32> undef)
   %V16  = call i32 @llvm.vector.reduce.umax.v16i32(<16 x i32> undef)
   %V32  = call i32 @llvm.vector.reduce.umax.v32i32(<32 x i32> undef)
   %V64  = call i32 @llvm.vector.reduce.umax.v64i32(<64 x i32> undef)
   %V128 = call i32 @llvm.vector.reduce.umax.v128i32(<128 x i32> undef)
   ret i32 undef
 }
 
 define i32 @reduce_umax_i64(i32 %arg) {
 ; CHECK-LABEL: 'reduce_umax_i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.vector.reduce.umax.v1i64(<1 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i64 @llvm.vector.reduce.umax.v2i64(<2 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i64 @llvm.vector.reduce.umax.v4i64(<4 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i64 @llvm.vector.reduce.umax.v8i64(<8 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i64 @llvm.vector.reduce.umax.v16i64(<16 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i64 @llvm.vector.reduce.umax.v32i64(<32 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V64 = call i64 @llvm.vector.reduce.umax.v64i64(<64 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V128 = call i64 @llvm.vector.reduce.umax.v128i64(<128 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_umax_i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.vector.reduce.umax.v1i64(<1 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i64 @llvm.vector.reduce.umax.v2i64(<2 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i64 @llvm.vector.reduce.umax.v4i64(<4 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i64 @llvm.vector.reduce.umax.v8i64(<8 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i64 @llvm.vector.reduce.umax.v16i64(<16 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i64 @llvm.vector.reduce.umax.v32i64(<32 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i64 @llvm.vector.reduce.umax.v64i64(<64 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V128 = call i64 @llvm.vector.reduce.umax.v128i64(<128 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i64 @llvm.vector.reduce.umax.v1i64(<1 x i64> undef)
   %V2   = call i64 @llvm.vector.reduce.umax.v2i64(<2 x i64> undef)
   %V4   = call i64 @llvm.vector.reduce.umax.v4i64(<4 x i64> undef)
   %V8   = call i64 @llvm.vector.reduce.umax.v8i64(<8 x i64> undef)
   %V16  = call i64 @llvm.vector.reduce.umax.v16i64(<16 x i64> undef)
   %V32  = call i64 @llvm.vector.reduce.umax.v32i64(<32 x i64> undef)
   %V64  = call i64 @llvm.vector.reduce.umax.v64i64(<64 x i64> undef)
   %V128 = call i64 @llvm.vector.reduce.umax.v128i64(<128 x i64> undef)
   ret i32 undef
 }
 
 define i32 @reduce_smin_i1(i32 %arg) {
 ; CHECK-LABEL: 'reduce_smin_i1'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V1 = call i1 @llvm.vector.reduce.smax.v1i1(<1 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i1 @llvm.vector.reduce.smax.v2i1(<2 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i1 @llvm.vector.reduce.smax.v4i1(<4 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i1 @llvm.vector.reduce.smax.v8i1(<8 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i1 @llvm.vector.reduce.smax.v16i1(<16 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32 = call i1 @llvm.vector.reduce.smax.v32i1(<32 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i1 @llvm.vector.reduce.smax.v64i1(<64 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i1 @llvm.vector.reduce.smax.v128i1(<128 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_smin_i1'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V1 = call i1 @llvm.vector.reduce.smax.v1i1(<1 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i1 @llvm.vector.reduce.smax.v2i1(<2 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i1 @llvm.vector.reduce.smax.v4i1(<4 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i1 @llvm.vector.reduce.smax.v8i1(<8 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i1 @llvm.vector.reduce.smax.v16i1(<16 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32 = call i1 @llvm.vector.reduce.smax.v32i1(<32 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i1 @llvm.vector.reduce.smax.v64i1(<64 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i1 @llvm.vector.reduce.smax.v128i1(<128 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i1 @llvm.vector.reduce.smax.v1i1(<1 x i1> undef)
   %V2   = call i1 @llvm.vector.reduce.smax.v2i1(<2 x i1> undef)
   %V4   = call i1 @llvm.vector.reduce.smax.v4i1(<4 x i1> undef)
   %V8   = call i1 @llvm.vector.reduce.smax.v8i1(<8 x i1> undef)
   %V16  = call i1 @llvm.vector.reduce.smax.v16i1(<16 x i1> undef)
   %V32  = call i1 @llvm.vector.reduce.smax.v32i1(<32 x i1> undef)
   %V64  = call i1 @llvm.vector.reduce.smax.v64i1(<64 x i1> undef)
   %V128 = call i1 @llvm.vector.reduce.smax.v128i1(<128 x i1> undef)
   ret i32 undef
 }
 
 define i32 @reduce_smax_i8(i32 %arg) {
 ; CHECK-LABEL: 'reduce_smax_i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i8 @llvm.vector.reduce.smax.v1i8(<1 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i8 @llvm.vector.reduce.smax.v2i8(<2 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i8 @llvm.vector.reduce.smax.v4i8(<4 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i8 @llvm.vector.reduce.smax.v8i8(<8 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i8 @llvm.vector.reduce.smax.v16i8(<16 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i8 @llvm.vector.reduce.smax.v32i8(<32 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i8 @llvm.vector.reduce.smax.v64i8(<64 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call i8 @llvm.vector.reduce.smax.v128i8(<128 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_smax_i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i8 @llvm.vector.reduce.smax.v1i8(<1 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i8 @llvm.vector.reduce.smax.v2i8(<2 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i8 @llvm.vector.reduce.smax.v4i8(<4 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i8 @llvm.vector.reduce.smax.v8i8(<8 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i8 @llvm.vector.reduce.smax.v16i8(<16 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i8 @llvm.vector.reduce.smax.v32i8(<32 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i8 @llvm.vector.reduce.smax.v64i8(<64 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i8 @llvm.vector.reduce.smax.v128i8(<128 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i8 @llvm.vector.reduce.smax.v1i8(<1 x i8> undef)
   %V2   = call i8 @llvm.vector.reduce.smax.v2i8(<2 x i8> undef)
   %V4   = call i8 @llvm.vector.reduce.smax.v4i8(<4 x i8> undef)
   %V8   = call i8 @llvm.vector.reduce.smax.v8i8(<8 x i8> undef)
   %V16  = call i8 @llvm.vector.reduce.smax.v16i8(<16 x i8> undef)
   %V32  = call i8 @llvm.vector.reduce.smax.v32i8(<32 x i8> undef)
   %V64  = call i8 @llvm.vector.reduce.smax.v64i8(<64 x i8> undef)
   %V128 = call i8 @llvm.vector.reduce.smax.v128i8(<128 x i8> undef)
   ret i32 undef
 }
 
 define i32 @reduce_smax_i16(i32 %arg) {
 ; CHECK-LABEL: 'reduce_smax_i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i16 @llvm.vector.reduce.smax.v1i16(<1 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i16 @llvm.vector.reduce.smax.v2i16(<2 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i16 @llvm.vector.reduce.smax.v4i16(<4 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i16 @llvm.vector.reduce.smax.v8i16(<8 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i16 @llvm.vector.reduce.smax.v16i16(<16 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i16 @llvm.vector.reduce.smax.v32i16(<32 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i16 @llvm.vector.reduce.smax.v64i16(<64 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call i16 @llvm.vector.reduce.smax.v128i16(<128 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_smax_i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i16 @llvm.vector.reduce.smax.v1i16(<1 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i16 @llvm.vector.reduce.smax.v2i16(<2 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i16 @llvm.vector.reduce.smax.v4i16(<4 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i16 @llvm.vector.reduce.smax.v8i16(<8 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i16 @llvm.vector.reduce.smax.v16i16(<16 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i16 @llvm.vector.reduce.smax.v32i16(<32 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i16 @llvm.vector.reduce.smax.v64i16(<64 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i16 @llvm.vector.reduce.smax.v128i16(<128 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i16 @llvm.vector.reduce.smax.v1i16(<1 x i16> undef)
   %V2   = call i16 @llvm.vector.reduce.smax.v2i16(<2 x i16> undef)
   %V4   = call i16 @llvm.vector.reduce.smax.v4i16(<4 x i16> undef)
   %V8   = call i16 @llvm.vector.reduce.smax.v8i16(<8 x i16> undef)
   %V16  = call i16 @llvm.vector.reduce.smax.v16i16(<16 x i16> undef)
   %V32  = call i16 @llvm.vector.reduce.smax.v32i16(<32 x i16> undef)
   %V64  = call i16 @llvm.vector.reduce.smax.v64i16(<64 x i16> undef)
   %V128 = call i16 @llvm.vector.reduce.smax.v128i16(<128 x i16> undef)
   ret i32 undef
 }
 
 define i32 @reduce_smax_i32(i32 %arg) {
 ; CHECK-LABEL: 'reduce_smax_i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i32 @llvm.vector.reduce.smax.v1i32(<1 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i32 @llvm.vector.reduce.smax.v2i32(<2 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i32 @llvm.vector.reduce.smax.v4i32(<4 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i32 @llvm.vector.reduce.smax.v8i32(<8 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i32 @llvm.vector.reduce.smax.v16i32(<16 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i32 @llvm.vector.reduce.smax.v32i32(<32 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i32 @llvm.vector.reduce.smax.v64i32(<64 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V128 = call i32 @llvm.vector.reduce.smax.v128i32(<128 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_smax_i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i32 @llvm.vector.reduce.smax.v1i32(<1 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i32 @llvm.vector.reduce.smax.v2i32(<2 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i32 @llvm.vector.reduce.smax.v4i32(<4 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i32 @llvm.vector.reduce.smax.v8i32(<8 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i32 @llvm.vector.reduce.smax.v16i32(<16 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i32 @llvm.vector.reduce.smax.v32i32(<32 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i32 @llvm.vector.reduce.smax.v64i32(<64 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i32 @llvm.vector.reduce.smax.v128i32(<128 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i32 @llvm.vector.reduce.smax.v1i32(<1 x i32> undef)
   %V2   = call i32 @llvm.vector.reduce.smax.v2i32(<2 x i32> undef)
   %V4   = call i32 @llvm.vector.reduce.smax.v4i32(<4 x i32> undef)
   %V8   = call i32 @llvm.vector.reduce.smax.v8i32(<8 x i32> undef)
   %V16  = call i32 @llvm.vector.reduce.smax.v16i32(<16 x i32> undef)
   %V32  = call i32 @llvm.vector.reduce.smax.v32i32(<32 x i32> undef)
   %V64  = call i32 @llvm.vector.reduce.smax.v64i32(<64 x i32> undef)
   %V128 = call i32 @llvm.vector.reduce.smax.v128i32(<128 x i32> undef)
   ret i32 undef
 }
 
 define i32 @reduce_smax_i64(i32 %arg) {
 ; CHECK-LABEL: 'reduce_smax_i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.vector.reduce.smax.v1i64(<1 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i64 @llvm.vector.reduce.smax.v2i64(<2 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i64 @llvm.vector.reduce.smax.v4i64(<4 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i64 @llvm.vector.reduce.smax.v8i64(<8 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i64 @llvm.vector.reduce.smax.v16i64(<16 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i64 @llvm.vector.reduce.smax.v32i64(<32 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V64 = call i64 @llvm.vector.reduce.smax.v64i64(<64 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V128 = call i64 @llvm.vector.reduce.smax.v128i64(<128 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_smax_i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.vector.reduce.smax.v1i64(<1 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i64 @llvm.vector.reduce.smax.v2i64(<2 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i64 @llvm.vector.reduce.smax.v4i64(<4 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i64 @llvm.vector.reduce.smax.v8i64(<8 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i64 @llvm.vector.reduce.smax.v16i64(<16 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i64 @llvm.vector.reduce.smax.v32i64(<32 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i64 @llvm.vector.reduce.smax.v64i64(<64 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V128 = call i64 @llvm.vector.reduce.smax.v128i64(<128 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i64 @llvm.vector.reduce.smax.v1i64(<1 x i64> undef)
   %V2   = call i64 @llvm.vector.reduce.smax.v2i64(<2 x i64> undef)
   %V4   = call i64 @llvm.vector.reduce.smax.v4i64(<4 x i64> undef)
   %V8   = call i64 @llvm.vector.reduce.smax.v8i64(<8 x i64> undef)
   %V16  = call i64 @llvm.vector.reduce.smax.v16i64(<16 x i64> undef)
   %V32  = call i64 @llvm.vector.reduce.smax.v32i64(<32 x i64> undef)
   %V64  = call i64 @llvm.vector.reduce.smax.v64i64(<64 x i64> undef)
   %V128 = call i64 @llvm.vector.reduce.smax.v128i64(<128 x i64> undef)

llvm/test/Analysis/CostModel/RISCV/reduce-min.ll

 ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
 ; RUN: opt < %s -mtriple=riscv32 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s
 ; RUN: opt < %s -mtriple=riscv64 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s
+; RUN: opt < %s -mtriple=riscv32 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s --check-prefix=SIZE
+; RUN: opt < %s -mtriple=riscv64 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s --check-prefix=SIZE
 
 define i32 @reduce_umin_i1(i32 %arg) {
 ; CHECK-LABEL: 'reduce_umin_i1'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V1 = call i1 @llvm.vector.reduce.umin.v1i1(<1 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i1 @llvm.vector.reduce.umin.v2i1(<2 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i1 @llvm.vector.reduce.umin.v4i1(<4 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i1 @llvm.vector.reduce.umin.v8i1(<8 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i1 @llvm.vector.reduce.umin.v16i1(<16 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32 = call i1 @llvm.vector.reduce.umin.v32i1(<32 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i1 @llvm.vector.reduce.umin.v64i1(<64 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i1 @llvm.vector.reduce.umin.v128i1(<128 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_umin_i1'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V1 = call i1 @llvm.vector.reduce.umin.v1i1(<1 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i1 @llvm.vector.reduce.umin.v2i1(<2 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i1 @llvm.vector.reduce.umin.v4i1(<4 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i1 @llvm.vector.reduce.umin.v8i1(<8 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i1 @llvm.vector.reduce.umin.v16i1(<16 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32 = call i1 @llvm.vector.reduce.umin.v32i1(<32 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i1 @llvm.vector.reduce.umin.v64i1(<64 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i1 @llvm.vector.reduce.umin.v128i1(<128 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i1 @llvm.vector.reduce.umin.v1i1(<1 x i1> undef)
   %V2   = call i1 @llvm.vector.reduce.umin.v2i1(<2 x i1> undef)
   %V4   = call i1 @llvm.vector.reduce.umin.v4i1(<4 x i1> undef)
   %V8   = call i1 @llvm.vector.reduce.umin.v8i1(<8 x i1> undef)
   %V16  = call i1 @llvm.vector.reduce.umin.v16i1(<16 x i1> undef)
   %V32  = call i1 @llvm.vector.reduce.umin.v32i1(<32 x i1> undef)
   %V64  = call i1 @llvm.vector.reduce.umin.v64i1(<64 x i1> undef)
   %V128 = call i1 @llvm.vector.reduce.umin.v128i1(<128 x i1> undef)
   ret i32 undef
 }
 
 define i32 @reduce_umin_i8(i32 %arg) {
 ; CHECK-LABEL: 'reduce_umin_i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i8 @llvm.vector.reduce.umin.v1i8(<1 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i8 @llvm.vector.reduce.umin.v2i8(<2 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i8 @llvm.vector.reduce.umin.v4i8(<4 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i8 @llvm.vector.reduce.umin.v8i8(<8 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i8 @llvm.vector.reduce.umin.v16i8(<16 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i8 @llvm.vector.reduce.umin.v32i8(<32 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i8 @llvm.vector.reduce.umin.v64i8(<64 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call i8 @llvm.vector.reduce.umin.v128i8(<128 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_umin_i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i8 @llvm.vector.reduce.umin.v1i8(<1 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i8 @llvm.vector.reduce.umin.v2i8(<2 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i8 @llvm.vector.reduce.umin.v4i8(<4 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i8 @llvm.vector.reduce.umin.v8i8(<8 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i8 @llvm.vector.reduce.umin.v16i8(<16 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i8 @llvm.vector.reduce.umin.v32i8(<32 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i8 @llvm.vector.reduce.umin.v64i8(<64 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i8 @llvm.vector.reduce.umin.v128i8(<128 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i8 @llvm.vector.reduce.umin.v1i8(<1 x i8> undef)
   %V2   = call i8 @llvm.vector.reduce.umin.v2i8(<2 x i8> undef)
   %V4   = call i8 @llvm.vector.reduce.umin.v4i8(<4 x i8> undef)
   %V8   = call i8 @llvm.vector.reduce.umin.v8i8(<8 x i8> undef)
   %V16  = call i8 @llvm.vector.reduce.umin.v16i8(<16 x i8> undef)
   %V32  = call i8 @llvm.vector.reduce.umin.v32i8(<32 x i8> undef)
   %V64  = call i8 @llvm.vector.reduce.umin.v64i8(<64 x i8> undef)
   %V128 = call i8 @llvm.vector.reduce.umin.v128i8(<128 x i8> undef)
   ret i32 undef
 }
 
 define i32 @reduce_umin_i16(i32 %arg) {
 ; CHECK-LABEL: 'reduce_umin_i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i16 @llvm.vector.reduce.umin.v1i16(<1 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i16 @llvm.vector.reduce.umin.v2i16(<2 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i16 @llvm.vector.reduce.umin.v4i16(<4 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i16 @llvm.vector.reduce.umin.v8i16(<8 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i16 @llvm.vector.reduce.umin.v16i16(<16 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i16 @llvm.vector.reduce.umin.v32i16(<32 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i16 @llvm.vector.reduce.umin.v64i16(<64 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call i16 @llvm.vector.reduce.umin.v128i16(<128 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_umin_i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i16 @llvm.vector.reduce.umin.v1i16(<1 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i16 @llvm.vector.reduce.umin.v2i16(<2 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i16 @llvm.vector.reduce.umin.v4i16(<4 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i16 @llvm.vector.reduce.umin.v8i16(<8 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i16 @llvm.vector.reduce.umin.v16i16(<16 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i16 @llvm.vector.reduce.umin.v32i16(<32 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i16 @llvm.vector.reduce.umin.v64i16(<64 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i16 @llvm.vector.reduce.umin.v128i16(<128 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i16 @llvm.vector.reduce.umin.v1i16(<1 x i16> undef)
   %V2   = call i16 @llvm.vector.reduce.umin.v2i16(<2 x i16> undef)
   %V4   = call i16 @llvm.vector.reduce.umin.v4i16(<4 x i16> undef)
   %V8   = call i16 @llvm.vector.reduce.umin.v8i16(<8 x i16> undef)
   %V16  = call i16 @llvm.vector.reduce.umin.v16i16(<16 x i16> undef)
   %V32  = call i16 @llvm.vector.reduce.umin.v32i16(<32 x i16> undef)
   %V64  = call i16 @llvm.vector.reduce.umin.v64i16(<64 x i16> undef)
   %V128 = call i16 @llvm.vector.reduce.umin.v128i16(<128 x i16> undef)
   ret i32 undef
 }
 
 define i32 @reduce_umin_i32(i32 %arg) {
 ; CHECK-LABEL: 'reduce_umin_i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i32 @llvm.vector.reduce.umin.v1i32(<1 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i32 @llvm.vector.reduce.umin.v2i32(<2 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i32 @llvm.vector.reduce.umin.v4i32(<4 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i32 @llvm.vector.reduce.umin.v8i32(<8 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i32 @llvm.vector.reduce.umin.v16i32(<16 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i32 @llvm.vector.reduce.umin.v32i32(<32 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i32 @llvm.vector.reduce.umin.v64i32(<64 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V128 = call i32 @llvm.vector.reduce.umin.v128i32(<128 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_umin_i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i32 @llvm.vector.reduce.umin.v1i32(<1 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i32 @llvm.vector.reduce.umin.v2i32(<2 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i32 @llvm.vector.reduce.umin.v4i32(<4 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i32 @llvm.vector.reduce.umin.v8i32(<8 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i32 @llvm.vector.reduce.umin.v16i32(<16 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i32 @llvm.vector.reduce.umin.v32i32(<32 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i32 @llvm.vector.reduce.umin.v64i32(<64 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i32 @llvm.vector.reduce.umin.v128i32(<128 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i32 @llvm.vector.reduce.umin.v1i32(<1 x i32> undef)
   %V2   = call i32 @llvm.vector.reduce.umin.v2i32(<2 x i32> undef)
   %V4   = call i32 @llvm.vector.reduce.umin.v4i32(<4 x i32> undef)
   %V8   = call i32 @llvm.vector.reduce.umin.v8i32(<8 x i32> undef)
   %V16  = call i32 @llvm.vector.reduce.umin.v16i32(<16 x i32> undef)
   %V32  = call i32 @llvm.vector.reduce.umin.v32i32(<32 x i32> undef)
   %V64  = call i32 @llvm.vector.reduce.umin.v64i32(<64 x i32> undef)
   %V128 = call i32 @llvm.vector.reduce.umin.v128i32(<128 x i32> undef)
   ret i32 undef
 }
 
 define i32 @reduce_umin_i64(i32 %arg) {
 ; CHECK-LABEL: 'reduce_umin_i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.vector.reduce.umin.v1i64(<1 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i64 @llvm.vector.reduce.umin.v2i64(<2 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i64 @llvm.vector.reduce.umin.v4i64(<4 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i64 @llvm.vector.reduce.umin.v8i64(<8 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i64 @llvm.vector.reduce.umin.v16i64(<16 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i64 @llvm.vector.reduce.umin.v32i64(<32 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V64 = call i64 @llvm.vector.reduce.umin.v64i64(<64 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V128 = call i64 @llvm.vector.reduce.umin.v128i64(<128 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_umin_i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.vector.reduce.umin.v1i64(<1 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i64 @llvm.vector.reduce.umin.v2i64(<2 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i64 @llvm.vector.reduce.umin.v4i64(<4 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i64 @llvm.vector.reduce.umin.v8i64(<8 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i64 @llvm.vector.reduce.umin.v16i64(<16 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i64 @llvm.vector.reduce.umin.v32i64(<32 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i64 @llvm.vector.reduce.umin.v64i64(<64 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V128 = call i64 @llvm.vector.reduce.umin.v128i64(<128 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i64 @llvm.vector.reduce.umin.v1i64(<1 x i64> undef)
   %V2   = call i64 @llvm.vector.reduce.umin.v2i64(<2 x i64> undef)
   %V4   = call i64 @llvm.vector.reduce.umin.v4i64(<4 x i64> undef)
   %V8   = call i64 @llvm.vector.reduce.umin.v8i64(<8 x i64> undef)
   %V16  = call i64 @llvm.vector.reduce.umin.v16i64(<16 x i64> undef)
   %V32  = call i64 @llvm.vector.reduce.umin.v32i64(<32 x i64> undef)
   %V64  = call i64 @llvm.vector.reduce.umin.v64i64(<64 x i64> undef)
   %V128 = call i64 @llvm.vector.reduce.umin.v128i64(<128 x i64> undef)
   ret i32 undef
 }
 
 define i32 @reduce_smin_i1(i32 %arg) {
 ; CHECK-LABEL: 'reduce_smin_i1'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V1 = call i1 @llvm.vector.reduce.smin.v1i1(<1 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i1 @llvm.vector.reduce.smin.v2i1(<2 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i1 @llvm.vector.reduce.smin.v4i1(<4 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i1 @llvm.vector.reduce.smin.v8i1(<8 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i1 @llvm.vector.reduce.smin.v16i1(<16 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32 = call i1 @llvm.vector.reduce.smin.v32i1(<32 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i1 @llvm.vector.reduce.smin.v64i1(<64 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i1 @llvm.vector.reduce.smin.v128i1(<128 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_smin_i1'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V1 = call i1 @llvm.vector.reduce.smin.v1i1(<1 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i1 @llvm.vector.reduce.smin.v2i1(<2 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4 = call i1 @llvm.vector.reduce.smin.v4i1(<4 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8 = call i1 @llvm.vector.reduce.smin.v8i1(<8 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16 = call i1 @llvm.vector.reduce.smin.v16i1(<16 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V32 = call i1 @llvm.vector.reduce.smin.v32i1(<32 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i1 @llvm.vector.reduce.smin.v64i1(<64 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i1 @llvm.vector.reduce.smin.v128i1(<128 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i1 @llvm.vector.reduce.smin.v1i1(<1 x i1> undef)
   %V2   = call i1 @llvm.vector.reduce.smin.v2i1(<2 x i1> undef)
   %V4   = call i1 @llvm.vector.reduce.smin.v4i1(<4 x i1> undef)
   %V8   = call i1 @llvm.vector.reduce.smin.v8i1(<8 x i1> undef)
   %V16  = call i1 @llvm.vector.reduce.smin.v16i1(<16 x i1> undef)
   %V32  = call i1 @llvm.vector.reduce.smin.v32i1(<32 x i1> undef)
   %V64  = call i1 @llvm.vector.reduce.smin.v64i1(<64 x i1> undef)
   %V128 = call i1 @llvm.vector.reduce.smin.v128i1(<128 x i1> undef)
   ret i32 undef
 }
 
 define i32 @reduce_smin_i8(i32 %arg) {
 ; CHECK-LABEL: 'reduce_smin_i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i8 @llvm.vector.reduce.smin.v1i8(<1 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i8 @llvm.vector.reduce.smin.v2i8(<2 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i8 @llvm.vector.reduce.smin.v4i8(<4 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i8 @llvm.vector.reduce.smin.v8i8(<8 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i8 @llvm.vector.reduce.smin.v16i8(<16 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i8 @llvm.vector.reduce.smin.v32i8(<32 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i8 @llvm.vector.reduce.smin.v64i8(<64 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call i8 @llvm.vector.reduce.smin.v128i8(<128 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_smin_i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i8 @llvm.vector.reduce.smin.v1i8(<1 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i8 @llvm.vector.reduce.smin.v2i8(<2 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i8 @llvm.vector.reduce.smin.v4i8(<4 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i8 @llvm.vector.reduce.smin.v8i8(<8 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i8 @llvm.vector.reduce.smin.v16i8(<16 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i8 @llvm.vector.reduce.smin.v32i8(<32 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i8 @llvm.vector.reduce.smin.v64i8(<64 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i8 @llvm.vector.reduce.smin.v128i8(<128 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i8 @llvm.vector.reduce.smin.v1i8(<1 x i8> undef)
   %V2   = call i8 @llvm.vector.reduce.smin.v2i8(<2 x i8> undef)
   %V4   = call i8 @llvm.vector.reduce.smin.v4i8(<4 x i8> undef)
   %V8   = call i8 @llvm.vector.reduce.smin.v8i8(<8 x i8> undef)
   %V16  = call i8 @llvm.vector.reduce.smin.v16i8(<16 x i8> undef)
   %V32  = call i8 @llvm.vector.reduce.smin.v32i8(<32 x i8> undef)
   %V64  = call i8 @llvm.vector.reduce.smin.v64i8(<64 x i8> undef)
   %V128 = call i8 @llvm.vector.reduce.smin.v128i8(<128 x i8> undef)
   ret i32 undef
 }
 
 define i32 @reduce_smin_i16(i32 %arg) {
 ; CHECK-LABEL: 'reduce_smin_i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i16 @llvm.vector.reduce.smin.v1i16(<1 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i16 @llvm.vector.reduce.smin.v2i16(<2 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i16 @llvm.vector.reduce.smin.v4i16(<4 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i16 @llvm.vector.reduce.smin.v8i16(<8 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i16 @llvm.vector.reduce.smin.v16i16(<16 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i16 @llvm.vector.reduce.smin.v32i16(<32 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i16 @llvm.vector.reduce.smin.v64i16(<64 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call i16 @llvm.vector.reduce.smin.v128i16(<128 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_smin_i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i16 @llvm.vector.reduce.smin.v1i16(<1 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i16 @llvm.vector.reduce.smin.v2i16(<2 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i16 @llvm.vector.reduce.smin.v4i16(<4 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i16 @llvm.vector.reduce.smin.v8i16(<8 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i16 @llvm.vector.reduce.smin.v16i16(<16 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i16 @llvm.vector.reduce.smin.v32i16(<32 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i16 @llvm.vector.reduce.smin.v64i16(<64 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i16 @llvm.vector.reduce.smin.v128i16(<128 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i16 @llvm.vector.reduce.smin.v1i16(<1 x i16> undef)
   %V2   = call i16 @llvm.vector.reduce.smin.v2i16(<2 x i16> undef)
   %V4   = call i16 @llvm.vector.reduce.smin.v4i16(<4 x i16> undef)
   %V8   = call i16 @llvm.vector.reduce.smin.v8i16(<8 x i16> undef)
   %V16  = call i16 @llvm.vector.reduce.smin.v16i16(<16 x i16> undef)
   %V32  = call i16 @llvm.vector.reduce.smin.v32i16(<32 x i16> undef)
   %V64  = call i16 @llvm.vector.reduce.smin.v64i16(<64 x i16> undef)
   %V128 = call i16 @llvm.vector.reduce.smin.v128i16(<128 x i16> undef)
   ret i32 undef
 }
 
 define i32 @reduce_smin_i32(i32 %arg) {
 ; CHECK-LABEL: 'reduce_smin_i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i32 @llvm.vector.reduce.smin.v1i32(<1 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i32 @llvm.vector.reduce.smin.v2i32(<2 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i32 @llvm.vector.reduce.smin.v4i32(<4 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i32 @llvm.vector.reduce.smin.v8i32(<8 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i32 @llvm.vector.reduce.smin.v16i32(<16 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i32 @llvm.vector.reduce.smin.v32i32(<32 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i32 @llvm.vector.reduce.smin.v64i32(<64 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V128 = call i32 @llvm.vector.reduce.smin.v128i32(<128 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_smin_i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i32 @llvm.vector.reduce.smin.v1i32(<1 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i32 @llvm.vector.reduce.smin.v2i32(<2 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i32 @llvm.vector.reduce.smin.v4i32(<4 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i32 @llvm.vector.reduce.smin.v8i32(<8 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i32 @llvm.vector.reduce.smin.v16i32(<16 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i32 @llvm.vector.reduce.smin.v32i32(<32 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i32 @llvm.vector.reduce.smin.v64i32(<64 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i32 @llvm.vector.reduce.smin.v128i32(<128 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i32 @llvm.vector.reduce.smin.v1i32(<1 x i32> undef)
   %V2   = call i32 @llvm.vector.reduce.smin.v2i32(<2 x i32> undef)
   %V4   = call i32 @llvm.vector.reduce.smin.v4i32(<4 x i32> undef)
   %V8   = call i32 @llvm.vector.reduce.smin.v8i32(<8 x i32> undef)
   %V16  = call i32 @llvm.vector.reduce.smin.v16i32(<16 x i32> undef)
   %V32  = call i32 @llvm.vector.reduce.smin.v32i32(<32 x i32> undef)
   %V64  = call i32 @llvm.vector.reduce.smin.v64i32(<64 x i32> undef)
   %V128 = call i32 @llvm.vector.reduce.smin.v128i32(<128 x i32> undef)
   ret i32 undef
 }
 
 define i32 @reduce_smin_i64(i32 %arg) {
 ; CHECK-LABEL: 'reduce_smin_i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.vector.reduce.smin.v1i64(<1 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i64 @llvm.vector.reduce.smin.v2i64(<2 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i64 @llvm.vector.reduce.smin.v4i64(<4 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i64 @llvm.vector.reduce.smin.v8i64(<8 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i64 @llvm.vector.reduce.smin.v16i64(<16 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i64 @llvm.vector.reduce.smin.v32i64(<32 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V64 = call i64 @llvm.vector.reduce.smin.v64i64(<64 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V128 = call i64 @llvm.vector.reduce.smin.v128i64(<128 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_smin_i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.vector.reduce.smin.v1i64(<1 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i64 @llvm.vector.reduce.smin.v2i64(<2 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i64 @llvm.vector.reduce.smin.v4i64(<4 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i64 @llvm.vector.reduce.smin.v8i64(<8 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i64 @llvm.vector.reduce.smin.v16i64(<16 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i64 @llvm.vector.reduce.smin.v32i64(<32 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i64 @llvm.vector.reduce.smin.v64i64(<64 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V128 = call i64 @llvm.vector.reduce.smin.v128i64(<128 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i64 @llvm.vector.reduce.smin.v1i64(<1 x i64> undef)
   %V2   = call i64 @llvm.vector.reduce.smin.v2i64(<2 x i64> undef)
   %V4   = call i64 @llvm.vector.reduce.smin.v4i64(<4 x i64> undef)
   %V8   = call i64 @llvm.vector.reduce.smin.v8i64(<8 x i64> undef)
   %V16  = call i64 @llvm.vector.reduce.smin.v16i64(<16 x i64> undef)
   %V32  = call i64 @llvm.vector.reduce.smin.v32i64(<32 x i64> undef)
   %V64  = call i64 @llvm.vector.reduce.smin.v64i64(<64 x i64> undef)
   %V128 = call i64 @llvm.vector.reduce.smin.v128i64(<128 x i64> undef)

llvm/test/Analysis/CostModel/RISCV/reduce-or.ll

 ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
 ; RUN: opt < %s -mtriple=riscv32 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s
 ; RUN: opt < %s -mtriple=riscv64 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s
+; RUN: opt < %s -mtriple=riscv32 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s --check-prefix=SIZE
+; RUN: opt < %s -mtriple=riscv64 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s --check-prefix=SIZE
 
 define i32 @reduce_i1(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i1'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i1 @llvm.vector.reduce.or.v1i1(<1 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i1 @llvm.vector.reduce.or.v2i1(<2 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i1 @llvm.vector.reduce.or.v8i1(<8 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i1 @llvm.vector.reduce.or.v16i1(<16 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i1 @llvm.vector.reduce.or.v32i1(<32 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i1 @llvm.vector.reduce.or.v64i1(<64 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i1 @llvm.vector.reduce.or.v128i1(<128 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V256 = call i1 @llvm.vector.reduce.or.v256i1(<256 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V512 = call i1 @llvm.vector.reduce.or.v512i1(<512 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V1024 = call i1 @llvm.vector.reduce.or.v1024i1(<1024 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i1'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i1 @llvm.vector.reduce.or.v1i1(<1 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i1 @llvm.vector.reduce.or.v2i1(<2 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i1 @llvm.vector.reduce.or.v8i1(<8 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i1 @llvm.vector.reduce.or.v16i1(<16 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i1 @llvm.vector.reduce.or.v32i1(<32 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i1 @llvm.vector.reduce.or.v64i1(<64 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i1 @llvm.vector.reduce.or.v128i1(<128 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V256 = call i1 @llvm.vector.reduce.or.v256i1(<256 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V512 = call i1 @llvm.vector.reduce.or.v512i1(<512 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V1024 = call i1 @llvm.vector.reduce.or.v1024i1(<1024 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i1 @llvm.vector.reduce.or.v1i1(<1 x i1> undef)
   %V2   = call i1 @llvm.vector.reduce.or.v2i1(<2 x i1> undef)
   %V4   = call i1 @llvm.vector.reduce.or.v4i1(<4 x i1> undef)
   %V8   = call i1 @llvm.vector.reduce.or.v8i1(<8 x i1> undef)
   %V16  = call i1 @llvm.vector.reduce.or.v16i1(<16 x i1> undef)
   %V32  = call i1 @llvm.vector.reduce.or.v32i1(<32 x i1> undef)
   %V64  = call i1 @llvm.vector.reduce.or.v64i1(<64 x i1> undef)
   %V128 = call i1 @llvm.vector.reduce.or.v128i1(<128 x i1> undef)

 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i8 @llvm.vector.reduce.or.v4i8(<4 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i8 @llvm.vector.reduce.or.v8i8(<8 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i8 @llvm.vector.reduce.or.v16i8(<16 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i8 @llvm.vector.reduce.or.v32i8(<32 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i8 @llvm.vector.reduce.or.v64i8(<64 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call i8 @llvm.vector.reduce.or.v128i8(<128 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i8 @llvm.vector.reduce.or.v1i8(<1 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i8 @llvm.vector.reduce.or.v2i8(<2 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i8 @llvm.vector.reduce.or.v4i8(<4 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i8 @llvm.vector.reduce.or.v8i8(<8 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i8 @llvm.vector.reduce.or.v16i8(<16 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i8 @llvm.vector.reduce.or.v32i8(<32 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i8 @llvm.vector.reduce.or.v64i8(<64 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i8 @llvm.vector.reduce.or.v128i8(<128 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i8 @llvm.vector.reduce.or.v1i8(<1 x i8> undef)
   %V2   = call i8 @llvm.vector.reduce.or.v2i8(<2 x i8> undef)
   %V4   = call i8 @llvm.vector.reduce.or.v4i8(<4 x i8> undef)
   %V8   = call i8 @llvm.vector.reduce.or.v8i8(<8 x i8> undef)
   %V16  = call i8 @llvm.vector.reduce.or.v16i8(<16 x i8> undef)
   %V32  = call i8 @llvm.vector.reduce.or.v32i8(<32 x i8> undef)
   %V64  = call i8 @llvm.vector.reduce.or.v64i8(<64 x i8> undef)
   %V128 = call i8 @llvm.vector.reduce.or.v128i8(<128 x i8> undef)
   ret i32 undef
 }
 
 define i32 @reduce_i16(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i16 @llvm.vector.reduce.or.v1i16(<1 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i16 @llvm.vector.reduce.or.v2i16(<2 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i16 @llvm.vector.reduce.or.v4i16(<4 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i16 @llvm.vector.reduce.or.v8i16(<8 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i16 @llvm.vector.reduce.or.v16i16(<16 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i16 @llvm.vector.reduce.or.v32i16(<32 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i16 @llvm.vector.reduce.or.v64i16(<64 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call i16 @llvm.vector.reduce.or.v128i16(<128 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i16 @llvm.vector.reduce.or.v1i16(<1 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i16 @llvm.vector.reduce.or.v2i16(<2 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i16 @llvm.vector.reduce.or.v4i16(<4 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i16 @llvm.vector.reduce.or.v8i16(<8 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i16 @llvm.vector.reduce.or.v16i16(<16 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i16 @llvm.vector.reduce.or.v32i16(<32 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i16 @llvm.vector.reduce.or.v64i16(<64 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i16 @llvm.vector.reduce.or.v128i16(<128 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i16 @llvm.vector.reduce.or.v1i16(<1 x i16> undef)
   %V2   = call i16 @llvm.vector.reduce.or.v2i16(<2 x i16> undef)
   %V4   = call i16 @llvm.vector.reduce.or.v4i16(<4 x i16> undef)
   %V8   = call i16 @llvm.vector.reduce.or.v8i16(<8 x i16> undef)
   %V16  = call i16 @llvm.vector.reduce.or.v16i16(<16 x i16> undef)
   %V32  = call i16 @llvm.vector.reduce.or.v32i16(<32 x i16> undef)
   %V64  = call i16 @llvm.vector.reduce.or.v64i16(<64 x i16> undef)
   %V128 = call i16 @llvm.vector.reduce.or.v128i16(<128 x i16> undef)
   ret i32 undef
 }
 
 define i32 @reduce_i32(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i32 @llvm.vector.reduce.or.v1i32(<1 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i32 @llvm.vector.reduce.or.v2i32(<2 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i32 @llvm.vector.reduce.or.v4i32(<4 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i32 @llvm.vector.reduce.or.v8i32(<8 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i32 @llvm.vector.reduce.or.v16i32(<16 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i32 @llvm.vector.reduce.or.v32i32(<32 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i32 @llvm.vector.reduce.or.v64i32(<64 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V128 = call i32 @llvm.vector.reduce.or.v128i32(<128 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i32 @llvm.vector.reduce.or.v1i32(<1 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i32 @llvm.vector.reduce.or.v2i32(<2 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i32 @llvm.vector.reduce.or.v4i32(<4 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i32 @llvm.vector.reduce.or.v8i32(<8 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i32 @llvm.vector.reduce.or.v16i32(<16 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i32 @llvm.vector.reduce.or.v32i32(<32 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i32 @llvm.vector.reduce.or.v64i32(<64 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i32 @llvm.vector.reduce.or.v128i32(<128 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i32 @llvm.vector.reduce.or.v1i32(<1 x i32> undef)
   %V2   = call i32 @llvm.vector.reduce.or.v2i32(<2 x i32> undef)
   %V4   = call i32 @llvm.vector.reduce.or.v4i32(<4 x i32> undef)
   %V8   = call i32 @llvm.vector.reduce.or.v8i32(<8 x i32> undef)
   %V16  = call i32 @llvm.vector.reduce.or.v16i32(<16 x i32> undef)
   %V32  = call i32 @llvm.vector.reduce.or.v32i32(<32 x i32> undef)
   %V64  = call i32 @llvm.vector.reduce.or.v64i32(<64 x i32> undef)
   %V128 = call i32 @llvm.vector.reduce.or.v128i32(<128 x i32> undef)
   ret i32 undef
 }
 
 define i32 @reduce_i64(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.vector.reduce.or.v1i64(<1 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i64 @llvm.vector.reduce.or.v2i64(<2 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i64 @llvm.vector.reduce.or.v4i64(<4 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i64 @llvm.vector.reduce.or.v8i64(<8 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i64 @llvm.vector.reduce.or.v16i64(<16 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i64 @llvm.vector.reduce.or.v32i64(<32 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V64 = call i64 @llvm.vector.reduce.or.v64i64(<64 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V128 = call i64 @llvm.vector.reduce.or.v128i64(<128 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.vector.reduce.or.v1i64(<1 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i64 @llvm.vector.reduce.or.v2i64(<2 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i64 @llvm.vector.reduce.or.v4i64(<4 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i64 @llvm.vector.reduce.or.v8i64(<8 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i64 @llvm.vector.reduce.or.v16i64(<16 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i64 @llvm.vector.reduce.or.v32i64(<32 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i64 @llvm.vector.reduce.or.v64i64(<64 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V128 = call i64 @llvm.vector.reduce.or.v128i64(<128 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i64 @llvm.vector.reduce.or.v1i64(<1 x i64> undef)
   %V2   = call i64 @llvm.vector.reduce.or.v2i64(<2 x i64> undef)
   %V4   = call i64 @llvm.vector.reduce.or.v4i64(<4 x i64> undef)
   %V8   = call i64 @llvm.vector.reduce.or.v8i64(<8 x i64> undef)
   %V16  = call i64 @llvm.vector.reduce.or.v16i64(<16 x i64> undef)
   %V32  = call i64 @llvm.vector.reduce.or.v32i64(<32 x i64> undef)
   %V64  = call i64 @llvm.vector.reduce.or.v64i64(<64 x i64> undef)
   %V128 = call i64 @llvm.vector.reduce.or.v128i64(<128 x i64> undef)

llvm/test/Analysis/CostModel/RISCV/reduce-scalable-fp.ll

 ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
 ; RUN: opt < %s -mtriple=riscv32 -mattr=+d,+zfh,+experimental-zvfh,+v -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s
 ; RUN: opt < %s -mtriple=riscv64 -mattr=+d,+zfh,+experimental-zvfh,+v -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s
+; RUN: opt < %s -mtriple=riscv32 -mattr=+d,+zfh,+experimental-zvfh,+v -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s --check-prefix=SIZE
+; RUN: opt < %s -mtriple=riscv64 -mattr=+d,+zfh,+experimental-zvfh,+v -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s --check-prefix=SIZE
 
 declare half @llvm.vector.reduce.fadd.nxv1f16(half, <vscale x 1 x half>)
 
 define half @vreduce_fadd_nxv1f16(<vscale x 1 x half> %v, half %s) {
 ; CHECK-LABEL: 'vreduce_fadd_nxv1f16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call reassoc half @llvm.vector.reduce.fadd.nxv1f16(half %s, <vscale x 1 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_fadd_nxv1f16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call reassoc half @llvm.vector.reduce.fadd.nxv1f16(half %s, <vscale x 1 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call reassoc half @llvm.vector.reduce.fadd.nxv1f16(half %s, <vscale x 1 x half> %v)
   ret half %red
 }
 
 define half @vreduce_ord_fadd_nxv1f16(<vscale x 1 x half> %v, half %s) {
 ; CHECK-LABEL: 'vreduce_ord_fadd_nxv1f16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call half @llvm.vector.reduce.fadd.nxv1f16(half %s, <vscale x 1 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_ord_fadd_nxv1f16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call half @llvm.vector.reduce.fadd.nxv1f16(half %s, <vscale x 1 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call half @llvm.vector.reduce.fadd.nxv1f16(half %s, <vscale x 1 x half> %v)
   ret half %red
 }
 
 declare half @llvm.vector.reduce.fadd.nxv2f16(half, <vscale x 2 x half>)
 
 define half @vreduce_fadd_nxv2f16(<vscale x 2 x half> %v, half %s) {
 ; CHECK-LABEL: 'vreduce_fadd_nxv2f16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call reassoc half @llvm.vector.reduce.fadd.nxv2f16(half %s, <vscale x 2 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_fadd_nxv2f16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call reassoc half @llvm.vector.reduce.fadd.nxv2f16(half %s, <vscale x 2 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call reassoc half @llvm.vector.reduce.fadd.nxv2f16(half %s, <vscale x 2 x half> %v)
   ret half %red
 }
 
 define half @vreduce_ord_fadd_nxv2f16(<vscale x 2 x half> %v, half %s) {
 ; CHECK-LABEL: 'vreduce_ord_fadd_nxv2f16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %red = call half @llvm.vector.reduce.fadd.nxv2f16(half %s, <vscale x 2 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_ord_fadd_nxv2f16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call half @llvm.vector.reduce.fadd.nxv2f16(half %s, <vscale x 2 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call half @llvm.vector.reduce.fadd.nxv2f16(half %s, <vscale x 2 x half> %v)
   ret half %red
 }
 
 declare half @llvm.vector.reduce.fadd.nxv4f16(half, <vscale x 4 x half>)
 
 define half @vreduce_fadd_nxv4f16(<vscale x 4 x half> %v, half %s) {
 ; CHECK-LABEL: 'vreduce_fadd_nxv4f16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call reassoc half @llvm.vector.reduce.fadd.nxv4f16(half %s, <vscale x 4 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_fadd_nxv4f16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call reassoc half @llvm.vector.reduce.fadd.nxv4f16(half %s, <vscale x 4 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call reassoc half @llvm.vector.reduce.fadd.nxv4f16(half %s, <vscale x 4 x half> %v)
   ret half %red
 }
 
 define half @vreduce_ord_fadd_nxv4f16(<vscale x 4 x half> %v, half %s) {
 ; CHECK-LABEL: 'vreduce_ord_fadd_nxv4f16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %red = call half @llvm.vector.reduce.fadd.nxv4f16(half %s, <vscale x 4 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_ord_fadd_nxv4f16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call half @llvm.vector.reduce.fadd.nxv4f16(half %s, <vscale x 4 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call half @llvm.vector.reduce.fadd.nxv4f16(half %s, <vscale x 4 x half> %v)
   ret half %red
 }
 
 declare float @llvm.vector.reduce.fadd.nxv1f32(float, <vscale x 1 x float>)
 
 define float @vreduce_fadd_nxv1f32(<vscale x 1 x float> %v, float %s) {
 ; CHECK-LABEL: 'vreduce_fadd_nxv1f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call reassoc float @llvm.vector.reduce.fadd.nxv1f32(float %s, <vscale x 1 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fadd_nxv1f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call reassoc float @llvm.vector.reduce.fadd.nxv1f32(float %s, <vscale x 1 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call reassoc float @llvm.vector.reduce.fadd.nxv1f32(float %s, <vscale x 1 x float> %v)
   ret float %red
 }
 
 define float @vreduce_ord_fadd_nxv1f32(<vscale x 1 x float> %v, float %s) {
 ; CHECK-LABEL: 'vreduce_ord_fadd_nxv1f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call float @llvm.vector.reduce.fadd.nxv1f32(float %s, <vscale x 1 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_ord_fadd_nxv1f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call float @llvm.vector.reduce.fadd.nxv1f32(float %s, <vscale x 1 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call float @llvm.vector.reduce.fadd.nxv1f32(float %s, <vscale x 1 x float> %v)
   ret float %red
 }
 
 define float @vreduce_fwadd_nxv1f32(<vscale x 1 x half> %v, float %s) {
 ; CHECK-LABEL: 'vreduce_fwadd_nxv1f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 1 x half> %v to <vscale x 1 x float>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call reassoc float @llvm.vector.reduce.fadd.nxv1f32(float %s, <vscale x 1 x float> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fwadd_nxv1f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 1 x half> %v to <vscale x 1 x float>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call reassoc float @llvm.vector.reduce.fadd.nxv1f32(float %s, <vscale x 1 x float> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %e = fpext <vscale x 1 x half> %v to <vscale x 1 x float>
   %red = call reassoc float @llvm.vector.reduce.fadd.nxv1f32(float %s, <vscale x 1 x float> %e)
   ret float %red
 }
 
 define float @vreduce_ord_fwadd_nxv1f32(<vscale x 1 x half> %v, float %s) {
 ; CHECK-LABEL: 'vreduce_ord_fwadd_nxv1f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 1 x half> %v to <vscale x 1 x float>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call float @llvm.vector.reduce.fadd.nxv1f32(float %s, <vscale x 1 x float> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_ord_fwadd_nxv1f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 1 x half> %v to <vscale x 1 x float>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call float @llvm.vector.reduce.fadd.nxv1f32(float %s, <vscale x 1 x float> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %e = fpext <vscale x 1 x half> %v to <vscale x 1 x float>
   %red = call float @llvm.vector.reduce.fadd.nxv1f32(float %s, <vscale x 1 x float> %e)
   ret float %red
 }
 
 declare float @llvm.vector.reduce.fadd.nxv2f32(float, <vscale x 2 x float>)
 
 define float @vreduce_fadd_nxv2f32(<vscale x 2 x float> %v, float %s) {
 ; CHECK-LABEL: 'vreduce_fadd_nxv2f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call reassoc float @llvm.vector.reduce.fadd.nxv2f32(float %s, <vscale x 2 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fadd_nxv2f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call reassoc float @llvm.vector.reduce.fadd.nxv2f32(float %s, <vscale x 2 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call reassoc float @llvm.vector.reduce.fadd.nxv2f32(float %s, <vscale x 2 x float> %v)
   ret float %red
 }
 
 define float @vreduce_ord_fadd_nxv2f32(<vscale x 2 x float> %v, float %s) {
 ; CHECK-LABEL: 'vreduce_ord_fadd_nxv2f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %red = call float @llvm.vector.reduce.fadd.nxv2f32(float %s, <vscale x 2 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_ord_fadd_nxv2f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call float @llvm.vector.reduce.fadd.nxv2f32(float %s, <vscale x 2 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call float @llvm.vector.reduce.fadd.nxv2f32(float %s, <vscale x 2 x float> %v)
   ret float %red
 }
 
 define float @vreduce_fwadd_nxv2f32(<vscale x 2 x half> %v, float %s) {
 ; CHECK-LABEL: 'vreduce_fwadd_nxv2f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 2 x half> %v to <vscale x 2 x float>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call reassoc float @llvm.vector.reduce.fadd.nxv2f32(float %s, <vscale x 2 x float> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fwadd_nxv2f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 2 x half> %v to <vscale x 2 x float>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call reassoc float @llvm.vector.reduce.fadd.nxv2f32(float %s, <vscale x 2 x float> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %e = fpext <vscale x 2 x half> %v to <vscale x 2 x float>
   %red = call reassoc float @llvm.vector.reduce.fadd.nxv2f32(float %s, <vscale x 2 x float> %e)
   ret float %red
 }
 
 define float @vreduce_ord_fwadd_nxv2f32(<vscale x 2 x half> %v, float %s) {
 ; CHECK-LABEL: 'vreduce_ord_fwadd_nxv2f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 2 x half> %v to <vscale x 2 x float>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %red = call float @llvm.vector.reduce.fadd.nxv2f32(float %s, <vscale x 2 x float> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_ord_fwadd_nxv2f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 2 x half> %v to <vscale x 2 x float>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call float @llvm.vector.reduce.fadd.nxv2f32(float %s, <vscale x 2 x float> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %e = fpext <vscale x 2 x half> %v to <vscale x 2 x float>
   %red = call float @llvm.vector.reduce.fadd.nxv2f32(float %s, <vscale x 2 x float> %e)
   ret float %red
 }
 
 declare float @llvm.vector.reduce.fadd.nxv4f32(float, <vscale x 4 x float>)
 
 define float @vreduce_fadd_nxv4f32(<vscale x 4 x float> %v, float %s) {
 ; CHECK-LABEL: 'vreduce_fadd_nxv4f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call reassoc float @llvm.vector.reduce.fadd.nxv4f32(float %s, <vscale x 4 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fadd_nxv4f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call reassoc float @llvm.vector.reduce.fadd.nxv4f32(float %s, <vscale x 4 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call reassoc float @llvm.vector.reduce.fadd.nxv4f32(float %s, <vscale x 4 x float> %v)
   ret float %red
 }
 
 define float @vreduce_ord_fadd_nxv4f32(<vscale x 4 x float> %v, float %s) {
 ; CHECK-LABEL: 'vreduce_ord_fadd_nxv4f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %red = call float @llvm.vector.reduce.fadd.nxv4f32(float %s, <vscale x 4 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_ord_fadd_nxv4f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call float @llvm.vector.reduce.fadd.nxv4f32(float %s, <vscale x 4 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call float @llvm.vector.reduce.fadd.nxv4f32(float %s, <vscale x 4 x float> %v)
   ret float %red
 }
 
 define float @vreduce_fwadd_nxv4f32(<vscale x 4 x half> %v, float %s) {
 ; CHECK-LABEL: 'vreduce_fwadd_nxv4f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 4 x half> %v to <vscale x 4 x float>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call reassoc float @llvm.vector.reduce.fadd.nxv4f32(float %s, <vscale x 4 x float> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fwadd_nxv4f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 4 x half> %v to <vscale x 4 x float>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call reassoc float @llvm.vector.reduce.fadd.nxv4f32(float %s, <vscale x 4 x float> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %e = fpext <vscale x 4 x half> %v to <vscale x 4 x float>
   %red = call reassoc float @llvm.vector.reduce.fadd.nxv4f32(float %s, <vscale x 4 x float> %e)
   ret float %red
 }
 
 define float @vreduce_ord_fwadd_nxv4f32(<vscale x 4 x half> %v, float %s) {
 ; CHECK-LABEL: 'vreduce_ord_fwadd_nxv4f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 4 x half> %v to <vscale x 4 x float>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %red = call float @llvm.vector.reduce.fadd.nxv4f32(float %s, <vscale x 4 x float> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_ord_fwadd_nxv4f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 4 x half> %v to <vscale x 4 x float>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call float @llvm.vector.reduce.fadd.nxv4f32(float %s, <vscale x 4 x float> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %e = fpext <vscale x 4 x half> %v to <vscale x 4 x float>
   %red = call float @llvm.vector.reduce.fadd.nxv4f32(float %s, <vscale x 4 x float> %e)
   ret float %red
 }
 
 declare double @llvm.vector.reduce.fadd.nxv1f64(double, <vscale x 1 x double>)
 
 define double @vreduce_fadd_nxv1f64(<vscale x 1 x double> %v, double %s) {
 ; CHECK-LABEL: 'vreduce_fadd_nxv1f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call reassoc double @llvm.vector.reduce.fadd.nxv1f64(double %s, <vscale x 1 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fadd_nxv1f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call reassoc double @llvm.vector.reduce.fadd.nxv1f64(double %s, <vscale x 1 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call reassoc double @llvm.vector.reduce.fadd.nxv1f64(double %s, <vscale x 1 x double> %v)
   ret double %red
 }
 
 define double @vreduce_ord_fadd_nxv1f64(<vscale x 1 x double> %v, double %s) {
 ; CHECK-LABEL: 'vreduce_ord_fadd_nxv1f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call double @llvm.vector.reduce.fadd.nxv1f64(double %s, <vscale x 1 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_ord_fadd_nxv1f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call double @llvm.vector.reduce.fadd.nxv1f64(double %s, <vscale x 1 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call double @llvm.vector.reduce.fadd.nxv1f64(double %s, <vscale x 1 x double> %v)
   ret double %red
 }
 
 define double @vreduce_fwadd_nxv1f64(<vscale x 1 x float> %v, double %s) {
 ; CHECK-LABEL: 'vreduce_fwadd_nxv1f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 1 x float> %v to <vscale x 1 x double>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call reassoc double @llvm.vector.reduce.fadd.nxv1f64(double %s, <vscale x 1 x double> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fwadd_nxv1f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 1 x float> %v to <vscale x 1 x double>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call reassoc double @llvm.vector.reduce.fadd.nxv1f64(double %s, <vscale x 1 x double> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %e = fpext <vscale x 1 x float> %v to <vscale x 1 x double>
   %red = call reassoc double @llvm.vector.reduce.fadd.nxv1f64(double %s, <vscale x 1 x double> %e)
   ret double %red
 }
 
 define double @vreduce_ord_fwadd_nxv1f64(<vscale x 1 x float> %v, double %s) {
 ; CHECK-LABEL: 'vreduce_ord_fwadd_nxv1f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 1 x float> %v to <vscale x 1 x double>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call double @llvm.vector.reduce.fadd.nxv1f64(double %s, <vscale x 1 x double> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_ord_fwadd_nxv1f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 1 x float> %v to <vscale x 1 x double>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call double @llvm.vector.reduce.fadd.nxv1f64(double %s, <vscale x 1 x double> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %e = fpext <vscale x 1 x float> %v to <vscale x 1 x double>
   %red = call double @llvm.vector.reduce.fadd.nxv1f64(double %s, <vscale x 1 x double> %e)
   ret double %red
 }
 
 declare double @llvm.vector.reduce.fadd.nxv2f64(double, <vscale x 2 x double>)
 
 define double @vreduce_fadd_nxv2f64(<vscale x 2 x double> %v, double %s) {
 ; CHECK-LABEL: 'vreduce_fadd_nxv2f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call reassoc double @llvm.vector.reduce.fadd.nxv2f64(double %s, <vscale x 2 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fadd_nxv2f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call reassoc double @llvm.vector.reduce.fadd.nxv2f64(double %s, <vscale x 2 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call reassoc double @llvm.vector.reduce.fadd.nxv2f64(double %s, <vscale x 2 x double> %v)
   ret double %red
 }
 
 define double @vreduce_ord_fadd_nxv2f64(<vscale x 2 x double> %v, double %s) {
 ; CHECK-LABEL: 'vreduce_ord_fadd_nxv2f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %red = call double @llvm.vector.reduce.fadd.nxv2f64(double %s, <vscale x 2 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_ord_fadd_nxv2f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call double @llvm.vector.reduce.fadd.nxv2f64(double %s, <vscale x 2 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call double @llvm.vector.reduce.fadd.nxv2f64(double %s, <vscale x 2 x double> %v)
   ret double %red
 }
 
 define double @vreduce_fwadd_nxv2f64(<vscale x 2 x float> %v, double %s) {
 ; CHECK-LABEL: 'vreduce_fwadd_nxv2f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 2 x float> %v to <vscale x 2 x double>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call reassoc double @llvm.vector.reduce.fadd.nxv2f64(double %s, <vscale x 2 x double> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fwadd_nxv2f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 2 x float> %v to <vscale x 2 x double>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call reassoc double @llvm.vector.reduce.fadd.nxv2f64(double %s, <vscale x 2 x double> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %e = fpext <vscale x 2 x float> %v to <vscale x 2 x double>
   %red = call reassoc double @llvm.vector.reduce.fadd.nxv2f64(double %s, <vscale x 2 x double> %e)
   ret double %red
 }
 
 define double @vreduce_ord_fwadd_nxv2f64(<vscale x 2 x float> %v, double %s) {
 ; CHECK-LABEL: 'vreduce_ord_fwadd_nxv2f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 2 x float> %v to <vscale x 2 x double>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %red = call double @llvm.vector.reduce.fadd.nxv2f64(double %s, <vscale x 2 x double> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_ord_fwadd_nxv2f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 2 x float> %v to <vscale x 2 x double>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call double @llvm.vector.reduce.fadd.nxv2f64(double %s, <vscale x 2 x double> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %e = fpext <vscale x 2 x float> %v to <vscale x 2 x double>
   %red = call double @llvm.vector.reduce.fadd.nxv2f64(double %s, <vscale x 2 x double> %e)
   ret double %red
 }
 
 declare double @llvm.vector.reduce.fadd.nxv4f64(double, <vscale x 4 x double>)
 
 define double @vreduce_fadd_nxv4f64(<vscale x 4 x double> %v, double %s) {
 ; CHECK-LABEL: 'vreduce_fadd_nxv4f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call reassoc double @llvm.vector.reduce.fadd.nxv4f64(double %s, <vscale x 4 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fadd_nxv4f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call reassoc double @llvm.vector.reduce.fadd.nxv4f64(double %s, <vscale x 4 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call reassoc double @llvm.vector.reduce.fadd.nxv4f64(double %s, <vscale x 4 x double> %v)
   ret double %red
 }
 
 define double @vreduce_ord_fadd_nxv4f64(<vscale x 4 x double> %v, double %s) {
 ; CHECK-LABEL: 'vreduce_ord_fadd_nxv4f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %red = call double @llvm.vector.reduce.fadd.nxv4f64(double %s, <vscale x 4 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_ord_fadd_nxv4f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call double @llvm.vector.reduce.fadd.nxv4f64(double %s, <vscale x 4 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call double @llvm.vector.reduce.fadd.nxv4f64(double %s, <vscale x 4 x double> %v)
   ret double %red
 }
 
 define double @vreduce_fwadd_nxv4f64(<vscale x 4 x float> %v, double %s) {
 ; CHECK-LABEL: 'vreduce_fwadd_nxv4f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 4 x float> %v to <vscale x 4 x double>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call reassoc double @llvm.vector.reduce.fadd.nxv4f64(double %s, <vscale x 4 x double> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fwadd_nxv4f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 4 x float> %v to <vscale x 4 x double>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call reassoc double @llvm.vector.reduce.fadd.nxv4f64(double %s, <vscale x 4 x double> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %e = fpext <vscale x 4 x float> %v to <vscale x 4 x double>
   %red = call reassoc double @llvm.vector.reduce.fadd.nxv4f64(double %s, <vscale x 4 x double> %e)
   ret double %red
 }
 
 define double @vreduce_ord_fwadd_nxv4f64(<vscale x 4 x float> %v, double %s) {
 ; CHECK-LABEL: 'vreduce_ord_fwadd_nxv4f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 4 x float> %v to <vscale x 4 x double>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %red = call double @llvm.vector.reduce.fadd.nxv4f64(double %s, <vscale x 4 x double> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_ord_fwadd_nxv4f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = fpext <vscale x 4 x float> %v to <vscale x 4 x double>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call double @llvm.vector.reduce.fadd.nxv4f64(double %s, <vscale x 4 x double> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %e = fpext <vscale x 4 x float> %v to <vscale x 4 x double>
   %red = call double @llvm.vector.reduce.fadd.nxv4f64(double %s, <vscale x 4 x double> %e)
   ret double %red
 }
 
 declare half @llvm.vector.reduce.fmin.nxv1f16(<vscale x 1 x half>)
 
 define half @vreduce_fmin_nxv1f16(<vscale x 1 x half> %v) {
 ; CHECK-LABEL: 'vreduce_fmin_nxv1f16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call half @llvm.vector.reduce.fmin.nxv1f16(<vscale x 1 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv1f16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call half @llvm.vector.reduce.fmin.nxv1f16(<vscale x 1 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call half @llvm.vector.reduce.fmin.nxv1f16(<vscale x 1 x half> %v)
   ret half %red
 }
 
 define half @vreduce_fmin_nxv1f16_nonans(<vscale x 1 x half> %v) #0 {
 ; CHECK-LABEL: 'vreduce_fmin_nxv1f16_nonans'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call nnan half @llvm.vector.reduce.fmin.nxv1f16(<vscale x 1 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv1f16_nonans'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call nnan half @llvm.vector.reduce.fmin.nxv1f16(<vscale x 1 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call nnan half @llvm.vector.reduce.fmin.nxv1f16(<vscale x 1 x half> %v)
   ret half %red
 }
 
 define half @vreduce_fmin_nxv1f16_nonans_noinfs(<vscale x 1 x half> %v) #1 {
 ; CHECK-LABEL: 'vreduce_fmin_nxv1f16_nonans_noinfs'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call nnan ninf half @llvm.vector.reduce.fmin.nxv1f16(<vscale x 1 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv1f16_nonans_noinfs'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call nnan ninf half @llvm.vector.reduce.fmin.nxv1f16(<vscale x 1 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call nnan ninf half @llvm.vector.reduce.fmin.nxv1f16(<vscale x 1 x half> %v)
   ret half %red
 }
 
 declare half @llvm.vector.reduce.fmin.nxv2f16(<vscale x 2 x half>)
 
 define half @vreduce_fmin_nxv2f16(<vscale x 2 x half> %v) {
 ; CHECK-LABEL: 'vreduce_fmin_nxv2f16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call half @llvm.vector.reduce.fmin.nxv2f16(<vscale x 2 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv2f16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call half @llvm.vector.reduce.fmin.nxv2f16(<vscale x 2 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call half @llvm.vector.reduce.fmin.nxv2f16(<vscale x 2 x half> %v)
   ret half %red
 }
 
 declare half @llvm.vector.reduce.fmin.nxv4f16(<vscale x 4 x half>)
 
 define half @vreduce_fmin_nxv4f16(<vscale x 4 x half> %v) {
 ; CHECK-LABEL: 'vreduce_fmin_nxv4f16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call half @llvm.vector.reduce.fmin.nxv4f16(<vscale x 4 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv4f16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call half @llvm.vector.reduce.fmin.nxv4f16(<vscale x 4 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call half @llvm.vector.reduce.fmin.nxv4f16(<vscale x 4 x half> %v)
   ret half %red
 }
 
 declare half @llvm.vector.reduce.fmin.nxv64f16(<vscale x 64 x half>)
 
 define half @vreduce_fmin_nxv64f16(<vscale x 64 x half> %v) {
 ; CHECK-LABEL: 'vreduce_fmin_nxv64f16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %red = call half @llvm.vector.reduce.fmin.nxv64f16(<vscale x 64 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv64f16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call half @llvm.vector.reduce.fmin.nxv64f16(<vscale x 64 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call half @llvm.vector.reduce.fmin.nxv64f16(<vscale x 64 x half> %v)
   ret half %red
 }
 
 declare float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float>)
 
 define float @vreduce_fmin_nxv1f32(<vscale x 1 x float> %v) {
 ; CHECK-LABEL: 'vreduce_fmin_nxv1f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv1f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> %v)
   ret float %red
 }
 
 define float @vreduce_fmin_nxv1f32_nonans(<vscale x 1 x float> %v) {
 ; CHECK-LABEL: 'vreduce_fmin_nxv1f32_nonans'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call nnan float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv1f32_nonans'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call nnan float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call nnan float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> %v)
   ret float %red
 }
 
 define float @vreduce_fmin_nxv1f32_nonans_noinfs(<vscale x 1 x float> %v) {
 ; CHECK-LABEL: 'vreduce_fmin_nxv1f32_nonans_noinfs'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call nnan ninf float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv1f32_nonans_noinfs'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call nnan ninf float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call nnan ninf float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> %v)
   ret float %red
 }
 
 declare float @llvm.vector.reduce.fmin.nxv2f32(<vscale x 2 x float>)
 
 define float @vreduce_fmin_nxv2f32(<vscale x 2 x float> %v) {
 ; CHECK-LABEL: 'vreduce_fmin_nxv2f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call float @llvm.vector.reduce.fmin.nxv2f32(<vscale x 2 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv2f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call float @llvm.vector.reduce.fmin.nxv2f32(<vscale x 2 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call float @llvm.vector.reduce.fmin.nxv2f32(<vscale x 2 x float> %v)
   ret float %red
 }
 
 declare float @llvm.vector.reduce.fmin.nxv4f32(<vscale x 4 x float>)
 
 define float @vreduce_fmin_nxv4f32(<vscale x 4 x float> %v) {
 ; CHECK-LABEL: 'vreduce_fmin_nxv4f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call float @llvm.vector.reduce.fmin.nxv4f32(<vscale x 4 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv4f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call float @llvm.vector.reduce.fmin.nxv4f32(<vscale x 4 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call float @llvm.vector.reduce.fmin.nxv4f32(<vscale x 4 x float> %v)
   ret float %red
 }
 
 declare float @llvm.vector.reduce.fmin.nxv32f32(<vscale x 32 x float>)
 
 define float @vreduce_fmin_nxv32f32(<vscale x 32 x float> %v) {
 ; CHECK-LABEL: 'vreduce_fmin_nxv32f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %red = call float @llvm.vector.reduce.fmin.nxv32f32(<vscale x 32 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv32f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call float @llvm.vector.reduce.fmin.nxv32f32(<vscale x 32 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call float @llvm.vector.reduce.fmin.nxv32f32(<vscale x 32 x float> %v)
   ret float %red
 }
 
 declare double @llvm.vector.reduce.fmin.nxv1f64(<vscale x 1 x double>)
 
 define double @vreduce_fmin_nxv1f64(<vscale x 1 x double> %v) {
 ; CHECK-LABEL: 'vreduce_fmin_nxv1f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call double @llvm.vector.reduce.fmin.nxv1f64(<vscale x 1 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv1f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call double @llvm.vector.reduce.fmin.nxv1f64(<vscale x 1 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call double @llvm.vector.reduce.fmin.nxv1f64(<vscale x 1 x double> %v)
   ret double %red
 }
 
 define double @vreduce_fmin_nxv1f64_nonans(<vscale x 1 x double> %v) {
 ; CHECK-LABEL: 'vreduce_fmin_nxv1f64_nonans'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call nnan double @llvm.vector.reduce.fmin.nxv1f64(<vscale x 1 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv1f64_nonans'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call nnan double @llvm.vector.reduce.fmin.nxv1f64(<vscale x 1 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call nnan double @llvm.vector.reduce.fmin.nxv1f64(<vscale x 1 x double> %v)
   ret double %red
 }
 
 define double @vreduce_fmin_nxv1f64_nonans_noinfs(<vscale x 1 x double> %v) {
 ; CHECK-LABEL: 'vreduce_fmin_nxv1f64_nonans_noinfs'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call nnan ninf double @llvm.vector.reduce.fmin.nxv1f64(<vscale x 1 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv1f64_nonans_noinfs'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call nnan ninf double @llvm.vector.reduce.fmin.nxv1f64(<vscale x 1 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call nnan ninf double @llvm.vector.reduce.fmin.nxv1f64(<vscale x 1 x double> %v)
   ret double %red
 }
 
 declare double @llvm.vector.reduce.fmin.nxv2f64(<vscale x 2 x double>)
 
 define double @vreduce_fmin_nxv2f64(<vscale x 2 x double> %v) {
 ; CHECK-LABEL: 'vreduce_fmin_nxv2f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call double @llvm.vector.reduce.fmin.nxv2f64(<vscale x 2 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv2f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call double @llvm.vector.reduce.fmin.nxv2f64(<vscale x 2 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call double @llvm.vector.reduce.fmin.nxv2f64(<vscale x 2 x double> %v)
   ret double %red
 }
 
 declare double @llvm.vector.reduce.fmin.nxv4f64(<vscale x 4 x double>)
 
 define double @vreduce_fmin_nxv4f64(<vscale x 4 x double> %v) {
 ; CHECK-LABEL: 'vreduce_fmin_nxv4f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call double @llvm.vector.reduce.fmin.nxv4f64(<vscale x 4 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv4f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call double @llvm.vector.reduce.fmin.nxv4f64(<vscale x 4 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call double @llvm.vector.reduce.fmin.nxv4f64(<vscale x 4 x double> %v)
   ret double %red
 }
 
 declare double @llvm.vector.reduce.fmin.nxv16f64(<vscale x 16 x double>)
 
 define double @vreduce_fmin_nxv16f64(<vscale x 16 x double> %v) {
 ; CHECK-LABEL: 'vreduce_fmin_nxv16f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %red = call double @llvm.vector.reduce.fmin.nxv16f64(<vscale x 16 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fmin_nxv16f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call double @llvm.vector.reduce.fmin.nxv16f64(<vscale x 16 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call double @llvm.vector.reduce.fmin.nxv16f64(<vscale x 16 x double> %v)
   ret double %red
 }
 
 declare half @llvm.vector.reduce.fmax.nxv1f16(<vscale x 1 x half>)
 
 define half @vreduce_fmax_nxv1f16(<vscale x 1 x half> %v) {
 ; CHECK-LABEL: 'vreduce_fmax_nxv1f16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call half @llvm.vector.reduce.fmax.nxv1f16(<vscale x 1 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv1f16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call half @llvm.vector.reduce.fmax.nxv1f16(<vscale x 1 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call half @llvm.vector.reduce.fmax.nxv1f16(<vscale x 1 x half> %v)
   ret half %red
 }
 
 define half @vreduce_fmax_nxv1f16_nonans(<vscale x 1 x half> %v) #0 {
 ; CHECK-LABEL: 'vreduce_fmax_nxv1f16_nonans'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call nnan half @llvm.vector.reduce.fmax.nxv1f16(<vscale x 1 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv1f16_nonans'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call nnan half @llvm.vector.reduce.fmax.nxv1f16(<vscale x 1 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call nnan half @llvm.vector.reduce.fmax.nxv1f16(<vscale x 1 x half> %v)
   ret half %red
 }
 
 define half @vreduce_fmax_nxv1f16_nonans_noinfs(<vscale x 1 x half> %v) #1 {
 ; CHECK-LABEL: 'vreduce_fmax_nxv1f16_nonans_noinfs'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call nnan ninf half @llvm.vector.reduce.fmax.nxv1f16(<vscale x 1 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv1f16_nonans_noinfs'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call nnan ninf half @llvm.vector.reduce.fmax.nxv1f16(<vscale x 1 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call nnan ninf half @llvm.vector.reduce.fmax.nxv1f16(<vscale x 1 x half> %v)
   ret half %red
 }
 
 declare half @llvm.vector.reduce.fmax.nxv2f16(<vscale x 2 x half>)
 
 define half @vreduce_fmax_nxv2f16(<vscale x 2 x half> %v) {
 ; CHECK-LABEL: 'vreduce_fmax_nxv2f16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call half @llvm.vector.reduce.fmax.nxv2f16(<vscale x 2 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv2f16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call half @llvm.vector.reduce.fmax.nxv2f16(<vscale x 2 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call half @llvm.vector.reduce.fmax.nxv2f16(<vscale x 2 x half> %v)
   ret half %red
 }
 
 declare half @llvm.vector.reduce.fmax.nxv4f16(<vscale x 4 x half>)
 
 define half @vreduce_fmax_nxv4f16(<vscale x 4 x half> %v) {
 ; CHECK-LABEL: 'vreduce_fmax_nxv4f16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call half @llvm.vector.reduce.fmax.nxv4f16(<vscale x 4 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv4f16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call half @llvm.vector.reduce.fmax.nxv4f16(<vscale x 4 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call half @llvm.vector.reduce.fmax.nxv4f16(<vscale x 4 x half> %v)
   ret half %red
 }
 
 declare half @llvm.vector.reduce.fmax.nxv64f16(<vscale x 64 x half>)
 
 define half @vreduce_fmax_nxv64f16(<vscale x 64 x half> %v) {
 ; CHECK-LABEL: 'vreduce_fmax_nxv64f16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %red = call half @llvm.vector.reduce.fmax.nxv64f16(<vscale x 64 x half> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv64f16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call half @llvm.vector.reduce.fmax.nxv64f16(<vscale x 64 x half> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret half %red
+;
   %red = call half @llvm.vector.reduce.fmax.nxv64f16(<vscale x 64 x half> %v)
   ret half %red
 }
 
 declare float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float>)
 
 define float @vreduce_fmax_nxv1f32(<vscale x 1 x float> %v) {
 ; CHECK-LABEL: 'vreduce_fmax_nxv1f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv1f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> %v)
   ret float %red
 }
 
 define float @vreduce_fmax_nxv1f32_nonans(<vscale x 1 x float> %v) {
 ; CHECK-LABEL: 'vreduce_fmax_nxv1f32_nonans'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call nnan float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv1f32_nonans'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call nnan float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call nnan float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> %v)
   ret float %red
 }
 
 define float @vreduce_fmax_nxv1f32_nonans_noinfs(<vscale x 1 x float> %v) {
 ; CHECK-LABEL: 'vreduce_fmax_nxv1f32_nonans_noinfs'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call nnan ninf float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv1f32_nonans_noinfs'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call nnan ninf float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call nnan ninf float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> %v)
   ret float %red
 }
 
 declare float @llvm.vector.reduce.fmax.nxv2f32(<vscale x 2 x float>)
 
 define float @vreduce_fmax_nxv2f32(<vscale x 2 x float> %v) {
 ; CHECK-LABEL: 'vreduce_fmax_nxv2f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call float @llvm.vector.reduce.fmax.nxv2f32(<vscale x 2 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv2f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call float @llvm.vector.reduce.fmax.nxv2f32(<vscale x 2 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call float @llvm.vector.reduce.fmax.nxv2f32(<vscale x 2 x float> %v)
   ret float %red
 }
 
 declare float @llvm.vector.reduce.fmax.nxv4f32(<vscale x 4 x float>)
 
 define float @vreduce_fmax_nxv4f32(<vscale x 4 x float> %v) {
 ; CHECK-LABEL: 'vreduce_fmax_nxv4f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call float @llvm.vector.reduce.fmax.nxv4f32(<vscale x 4 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv4f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call float @llvm.vector.reduce.fmax.nxv4f32(<vscale x 4 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call float @llvm.vector.reduce.fmax.nxv4f32(<vscale x 4 x float> %v)
   ret float %red
 }
 
 declare float @llvm.vector.reduce.fmax.nxv32f32(<vscale x 32 x float>)
 
 define float @vreduce_fmax_nxv32f32(<vscale x 32 x float> %v) {
 ; CHECK-LABEL: 'vreduce_fmax_nxv32f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %red = call float @llvm.vector.reduce.fmax.nxv32f32(<vscale x 32 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv32f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call float @llvm.vector.reduce.fmax.nxv32f32(<vscale x 32 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call float @llvm.vector.reduce.fmax.nxv32f32(<vscale x 32 x float> %v)
   ret float %red
 }
 
 declare double @llvm.vector.reduce.fmax.nxv1f64(<vscale x 1 x double>)
 
 define double @vreduce_fmax_nxv1f64(<vscale x 1 x double> %v) {
 ; CHECK-LABEL: 'vreduce_fmax_nxv1f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call double @llvm.vector.reduce.fmax.nxv1f64(<vscale x 1 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv1f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call double @llvm.vector.reduce.fmax.nxv1f64(<vscale x 1 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call double @llvm.vector.reduce.fmax.nxv1f64(<vscale x 1 x double> %v)
   ret double %red
 }
 
 define double @vreduce_fmax_nxv1f64_nonans(<vscale x 1 x double> %v) {
 ; CHECK-LABEL: 'vreduce_fmax_nxv1f64_nonans'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call nnan double @llvm.vector.reduce.fmax.nxv1f64(<vscale x 1 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv1f64_nonans'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call nnan double @llvm.vector.reduce.fmax.nxv1f64(<vscale x 1 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call nnan double @llvm.vector.reduce.fmax.nxv1f64(<vscale x 1 x double> %v)
   ret double %red
 }
 
 define double @vreduce_fmax_nxv1f64_nonans_noinfs(<vscale x 1 x double> %v) {
 ; CHECK-LABEL: 'vreduce_fmax_nxv1f64_nonans_noinfs'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call nnan ninf double @llvm.vector.reduce.fmax.nxv1f64(<vscale x 1 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv1f64_nonans_noinfs'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call nnan ninf double @llvm.vector.reduce.fmax.nxv1f64(<vscale x 1 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call nnan ninf double @llvm.vector.reduce.fmax.nxv1f64(<vscale x 1 x double> %v)
   ret double %red
 }
 
 declare double @llvm.vector.reduce.fmax.nxv2f64(<vscale x 2 x double>)
 
 define double @vreduce_fmax_nxv2f64(<vscale x 2 x double> %v) {
 ; CHECK-LABEL: 'vreduce_fmax_nxv2f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call double @llvm.vector.reduce.fmax.nxv2f64(<vscale x 2 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv2f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call double @llvm.vector.reduce.fmax.nxv2f64(<vscale x 2 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call double @llvm.vector.reduce.fmax.nxv2f64(<vscale x 2 x double> %v)
   ret double %red
 }
 
 declare double @llvm.vector.reduce.fmax.nxv4f64(<vscale x 4 x double>)
 
 define double @vreduce_fmax_nxv4f64(<vscale x 4 x double> %v) {
 ; CHECK-LABEL: 'vreduce_fmax_nxv4f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call double @llvm.vector.reduce.fmax.nxv4f64(<vscale x 4 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv4f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call double @llvm.vector.reduce.fmax.nxv4f64(<vscale x 4 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call double @llvm.vector.reduce.fmax.nxv4f64(<vscale x 4 x double> %v)
   ret double %red
 }
 
 declare double @llvm.vector.reduce.fmax.nxv16f64(<vscale x 16 x double>)
 
 define double @vreduce_fmax_nxv16f64(<vscale x 16 x double> %v) {
 ; CHECK-LABEL: 'vreduce_fmax_nxv16f64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %red = call double @llvm.vector.reduce.fmax.nxv16f64(<vscale x 16 x double> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
 ;
+; SIZE-LABEL: 'vreduce_fmax_nxv16f64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call double @llvm.vector.reduce.fmax.nxv16f64(<vscale x 16 x double> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret double %red
+;
   %red = call double @llvm.vector.reduce.fmax.nxv16f64(<vscale x 16 x double> %v)
   ret double %red
 }
 
 define float @vreduce_nsz_fadd_nxv1f32(<vscale x 1 x float> %v, float %s) {
 ; CHECK-LABEL: 'vreduce_nsz_fadd_nxv1f32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call reassoc nsz float @llvm.vector.reduce.fadd.nxv1f32(float %s, <vscale x 1 x float> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
 ;
+; SIZE-LABEL: 'vreduce_nsz_fadd_nxv1f32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call reassoc nsz float @llvm.vector.reduce.fadd.nxv1f32(float %s, <vscale x 1 x float> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret float %red
+;
   %red = call reassoc nsz float @llvm.vector.reduce.fadd.nxv1f32(float %s, <vscale x 1 x float> %v)
   ret float %red
 }

llvm/test/Analysis/CostModel/RISCV/reduce-scalable-int.ll

 ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
 ; RUN: opt < %s -mtriple=riscv32 -mattr=+v -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s
 ; RUN: opt < %s -mtriple=riscv64 -mattr=+v -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s
+; RUN: opt < %s -mtriple=riscv32 -mattr=+v -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s --check-prefix=SIZE
+; RUN: opt < %s -mtriple=riscv64 -mattr=+v -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s --check-prefix=SIZE
 
 declare i8 @llvm.vector.reduce.add.nxv1i8(<vscale x 1 x i8>)
 
 define signext i8 @vreduce_add_nxv1i8(<vscale x 1 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_add_nxv1i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i8 @llvm.vector.reduce.add.nxv1i8(<vscale x 1 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_add_nxv1i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.add.nxv1i8(<vscale x 1 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.add.nxv1i8(<vscale x 1 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.umax.nxv1i8(<vscale x 1 x i8>)
 
 define signext i8 @vreduce_umax_nxv1i8(<vscale x 1 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_umax_nxv1i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i8 @llvm.vector.reduce.umax.nxv1i8(<vscale x 1 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_umax_nxv1i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.umax.nxv1i8(<vscale x 1 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.umax.nxv1i8(<vscale x 1 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.smax.nxv1i8(<vscale x 1 x i8>)
 
 define signext i8 @vreduce_smax_nxv1i8(<vscale x 1 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_smax_nxv1i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i8 @llvm.vector.reduce.smax.nxv1i8(<vscale x 1 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_smax_nxv1i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.smax.nxv1i8(<vscale x 1 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.smax.nxv1i8(<vscale x 1 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.umin.nxv1i8(<vscale x 1 x i8>)
 
 define signext i8 @vreduce_umin_nxv1i8(<vscale x 1 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_umin_nxv1i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i8 @llvm.vector.reduce.umin.nxv1i8(<vscale x 1 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_umin_nxv1i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.umin.nxv1i8(<vscale x 1 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.umin.nxv1i8(<vscale x 1 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.smin.nxv1i8(<vscale x 1 x i8>)
 
 define signext i8 @vreduce_smin_nxv1i8(<vscale x 1 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_smin_nxv1i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i8 @llvm.vector.reduce.smin.nxv1i8(<vscale x 1 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_smin_nxv1i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.smin.nxv1i8(<vscale x 1 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.smin.nxv1i8(<vscale x 1 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.and.nxv1i8(<vscale x 1 x i8>)
 
 define signext i8 @vreduce_and_nxv1i8(<vscale x 1 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_and_nxv1i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i8 @llvm.vector.reduce.and.nxv1i8(<vscale x 1 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_and_nxv1i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.and.nxv1i8(<vscale x 1 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.and.nxv1i8(<vscale x 1 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.or.nxv1i8(<vscale x 1 x i8>)
 
 define signext i8 @vreduce_or_nxv1i8(<vscale x 1 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_or_nxv1i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i8 @llvm.vector.reduce.or.nxv1i8(<vscale x 1 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_or_nxv1i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.or.nxv1i8(<vscale x 1 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.or.nxv1i8(<vscale x 1 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.xor.nxv1i8(<vscale x 1 x i8>)
 
 define signext i8 @vreduce_xor_nxv1i8(<vscale x 1 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_xor_nxv1i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i8 @llvm.vector.reduce.xor.nxv1i8(<vscale x 1 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_xor_nxv1i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.xor.nxv1i8(<vscale x 1 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.xor.nxv1i8(<vscale x 1 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.add.nxv2i8(<vscale x 2 x i8>)
 
 define signext i8 @vreduce_add_nxv2i8(<vscale x 2 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_add_nxv2i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i8 @llvm.vector.reduce.add.nxv2i8(<vscale x 2 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_add_nxv2i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.add.nxv2i8(<vscale x 2 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.add.nxv2i8(<vscale x 2 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.umax.nxv2i8(<vscale x 2 x i8>)
 
 define signext i8 @vreduce_umax_nxv2i8(<vscale x 2 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_umax_nxv2i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i8 @llvm.vector.reduce.umax.nxv2i8(<vscale x 2 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_umax_nxv2i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.umax.nxv2i8(<vscale x 2 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.umax.nxv2i8(<vscale x 2 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.smax.nxv2i8(<vscale x 2 x i8>)
 
 define signext i8 @vreduce_smax_nxv2i8(<vscale x 2 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_smax_nxv2i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i8 @llvm.vector.reduce.smax.nxv2i8(<vscale x 2 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_smax_nxv2i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.smax.nxv2i8(<vscale x 2 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.smax.nxv2i8(<vscale x 2 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.umin.nxv2i8(<vscale x 2 x i8>)
 
 define signext i8 @vreduce_umin_nxv2i8(<vscale x 2 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_umin_nxv2i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i8 @llvm.vector.reduce.umin.nxv2i8(<vscale x 2 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_umin_nxv2i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.umin.nxv2i8(<vscale x 2 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.umin.nxv2i8(<vscale x 2 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.smin.nxv2i8(<vscale x 2 x i8>)
 
 define signext i8 @vreduce_smin_nxv2i8(<vscale x 2 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_smin_nxv2i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i8 @llvm.vector.reduce.smin.nxv2i8(<vscale x 2 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_smin_nxv2i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.smin.nxv2i8(<vscale x 2 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.smin.nxv2i8(<vscale x 2 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.and.nxv2i8(<vscale x 2 x i8>)
 
 define signext i8 @vreduce_and_nxv2i8(<vscale x 2 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_and_nxv2i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i8 @llvm.vector.reduce.and.nxv2i8(<vscale x 2 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_and_nxv2i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.and.nxv2i8(<vscale x 2 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.and.nxv2i8(<vscale x 2 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.or.nxv2i8(<vscale x 2 x i8>)
 
 define signext i8 @vreduce_or_nxv2i8(<vscale x 2 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_or_nxv2i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i8 @llvm.vector.reduce.or.nxv2i8(<vscale x 2 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_or_nxv2i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.or.nxv2i8(<vscale x 2 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.or.nxv2i8(<vscale x 2 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.xor.nxv2i8(<vscale x 2 x i8>)
 
 define signext i8 @vreduce_xor_nxv2i8(<vscale x 2 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_xor_nxv2i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i8 @llvm.vector.reduce.xor.nxv2i8(<vscale x 2 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_xor_nxv2i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.xor.nxv2i8(<vscale x 2 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.xor.nxv2i8(<vscale x 2 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.add.nxv4i8(<vscale x 4 x i8>)
 
 define signext i8 @vreduce_add_nxv4i8(<vscale x 4 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_add_nxv4i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i8 @llvm.vector.reduce.add.nxv4i8(<vscale x 4 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_add_nxv4i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.add.nxv4i8(<vscale x 4 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.add.nxv4i8(<vscale x 4 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.umax.nxv4i8(<vscale x 4 x i8>)
 
 define signext i8 @vreduce_umax_nxv4i8(<vscale x 4 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_umax_nxv4i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i8 @llvm.vector.reduce.umax.nxv4i8(<vscale x 4 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_umax_nxv4i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.umax.nxv4i8(<vscale x 4 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.umax.nxv4i8(<vscale x 4 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.smax.nxv4i8(<vscale x 4 x i8>)
 
 define signext i8 @vreduce_smax_nxv4i8(<vscale x 4 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_smax_nxv4i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i8 @llvm.vector.reduce.smax.nxv4i8(<vscale x 4 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_smax_nxv4i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.smax.nxv4i8(<vscale x 4 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.smax.nxv4i8(<vscale x 4 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.umin.nxv4i8(<vscale x 4 x i8>)
 
 define signext i8 @vreduce_umin_nxv4i8(<vscale x 4 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_umin_nxv4i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i8 @llvm.vector.reduce.umin.nxv4i8(<vscale x 4 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_umin_nxv4i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.umin.nxv4i8(<vscale x 4 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.umin.nxv4i8(<vscale x 4 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.smin.nxv4i8(<vscale x 4 x i8>)
 
 define signext i8 @vreduce_smin_nxv4i8(<vscale x 4 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_smin_nxv4i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i8 @llvm.vector.reduce.smin.nxv4i8(<vscale x 4 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_smin_nxv4i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.smin.nxv4i8(<vscale x 4 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.smin.nxv4i8(<vscale x 4 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.and.nxv4i8(<vscale x 4 x i8>)
 
 define signext i8 @vreduce_and_nxv4i8(<vscale x 4 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_and_nxv4i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i8 @llvm.vector.reduce.and.nxv4i8(<vscale x 4 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_and_nxv4i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.and.nxv4i8(<vscale x 4 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.and.nxv4i8(<vscale x 4 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.or.nxv4i8(<vscale x 4 x i8>)
 
 define signext i8 @vreduce_or_nxv4i8(<vscale x 4 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_or_nxv4i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i8 @llvm.vector.reduce.or.nxv4i8(<vscale x 4 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_or_nxv4i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.or.nxv4i8(<vscale x 4 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.or.nxv4i8(<vscale x 4 x i8> %v)
   ret i8 %red
 }
 
 declare i8 @llvm.vector.reduce.xor.nxv4i8(<vscale x 4 x i8>)
 
 define signext i8 @vreduce_xor_nxv4i8(<vscale x 4 x i8> %v) {
 ; CHECK-LABEL: 'vreduce_xor_nxv4i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i8 @llvm.vector.reduce.xor.nxv4i8(<vscale x 4 x i8> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
 ;
+; SIZE-LABEL: 'vreduce_xor_nxv4i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i8 @llvm.vector.reduce.xor.nxv4i8(<vscale x 4 x i8> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i8 %red
+;
   %red = call i8 @llvm.vector.reduce.xor.nxv4i8(<vscale x 4 x i8> %v)
   ret i8 %red
 }
 
 declare i16 @llvm.vector.reduce.add.nxv1i16(<vscale x 1 x i16>)
 
 define signext i16 @vreduce_add_nxv1i16(<vscale x 1 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_add_nxv1i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv1i16(<vscale x 1 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_add_nxv1i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv1i16(<vscale x 1 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.add.nxv1i16(<vscale x 1 x i16> %v)
   ret i16 %red
 }
 
 define signext i16 @vwreduce_add_nxv1i8(<vscale x 1 x i8> %v) {
 ; CHECK-LABEL: 'vwreduce_add_nxv1i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 1 x i8> %v to <vscale x 1 x i16>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv1i16(<vscale x 1 x i16> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vwreduce_add_nxv1i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 1 x i8> %v to <vscale x 1 x i16>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv1i16(<vscale x 1 x i16> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %e = sext <vscale x 1 x i8> %v to <vscale x 1 x i16>
   %red = call i16 @llvm.vector.reduce.add.nxv1i16(<vscale x 1 x i16> %e)
   ret i16 %red
 }
 
 define signext i16 @vwreduce_uadd_nxv1i8(<vscale x 1 x i8> %v) {
 ; CHECK-LABEL: 'vwreduce_uadd_nxv1i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 1 x i8> %v to <vscale x 1 x i16>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv1i16(<vscale x 1 x i16> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vwreduce_uadd_nxv1i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 1 x i8> %v to <vscale x 1 x i16>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv1i16(<vscale x 1 x i16> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %e = sext <vscale x 1 x i8> %v to <vscale x 1 x i16>
   %red = call i16 @llvm.vector.reduce.add.nxv1i16(<vscale x 1 x i16> %e)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.umax.nxv1i16(<vscale x 1 x i16>)
 
 define signext i16 @vreduce_umax_nxv1i16(<vscale x 1 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_umax_nxv1i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i16 @llvm.vector.reduce.umax.nxv1i16(<vscale x 1 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_umax_nxv1i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.umax.nxv1i16(<vscale x 1 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.umax.nxv1i16(<vscale x 1 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.smax.nxv1i16(<vscale x 1 x i16>)
 
 define signext i16 @vreduce_smax_nxv1i16(<vscale x 1 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_smax_nxv1i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i16 @llvm.vector.reduce.smax.nxv1i16(<vscale x 1 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_smax_nxv1i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.smax.nxv1i16(<vscale x 1 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.smax.nxv1i16(<vscale x 1 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.umin.nxv1i16(<vscale x 1 x i16>)
 
 define signext i16 @vreduce_umin_nxv1i16(<vscale x 1 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_umin_nxv1i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i16 @llvm.vector.reduce.umin.nxv1i16(<vscale x 1 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_umin_nxv1i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.umin.nxv1i16(<vscale x 1 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.umin.nxv1i16(<vscale x 1 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.smin.nxv1i16(<vscale x 1 x i16>)
 
 define signext i16 @vreduce_smin_nxv1i16(<vscale x 1 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_smin_nxv1i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i16 @llvm.vector.reduce.smin.nxv1i16(<vscale x 1 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_smin_nxv1i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.smin.nxv1i16(<vscale x 1 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.smin.nxv1i16(<vscale x 1 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.and.nxv1i16(<vscale x 1 x i16>)
 
 define signext i16 @vreduce_and_nxv1i16(<vscale x 1 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_and_nxv1i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i16 @llvm.vector.reduce.and.nxv1i16(<vscale x 1 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_and_nxv1i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.and.nxv1i16(<vscale x 1 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.and.nxv1i16(<vscale x 1 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.or.nxv1i16(<vscale x 1 x i16>)
 
 define signext i16 @vreduce_or_nxv1i16(<vscale x 1 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_or_nxv1i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i16 @llvm.vector.reduce.or.nxv1i16(<vscale x 1 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_or_nxv1i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.or.nxv1i16(<vscale x 1 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.or.nxv1i16(<vscale x 1 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.xor.nxv1i16(<vscale x 1 x i16>)
 
 define signext i16 @vreduce_xor_nxv1i16(<vscale x 1 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_xor_nxv1i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i16 @llvm.vector.reduce.xor.nxv1i16(<vscale x 1 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_xor_nxv1i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.xor.nxv1i16(<vscale x 1 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.xor.nxv1i16(<vscale x 1 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.add.nxv2i16(<vscale x 2 x i16>)
 
 define signext i16 @vreduce_add_nxv2i16(<vscale x 2 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_add_nxv2i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv2i16(<vscale x 2 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_add_nxv2i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv2i16(<vscale x 2 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.add.nxv2i16(<vscale x 2 x i16> %v)
   ret i16 %red
 }
 
 define signext i16 @vwreduce_add_nxv2i8(<vscale x 2 x i8> %v) {
 ; CHECK-LABEL: 'vwreduce_add_nxv2i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 2 x i8> %v to <vscale x 2 x i16>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv2i16(<vscale x 2 x i16> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vwreduce_add_nxv2i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 2 x i8> %v to <vscale x 2 x i16>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv2i16(<vscale x 2 x i16> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %e = sext <vscale x 2 x i8> %v to <vscale x 2 x i16>
   %red = call i16 @llvm.vector.reduce.add.nxv2i16(<vscale x 2 x i16> %e)
   ret i16 %red
 }
 
 define signext i16 @vwreduce_uadd_nxv2i8(<vscale x 2 x i8> %v) {
 ; CHECK-LABEL: 'vwreduce_uadd_nxv2i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 2 x i8> %v to <vscale x 2 x i16>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv2i16(<vscale x 2 x i16> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vwreduce_uadd_nxv2i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 2 x i8> %v to <vscale x 2 x i16>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv2i16(<vscale x 2 x i16> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %e = sext <vscale x 2 x i8> %v to <vscale x 2 x i16>
   %red = call i16 @llvm.vector.reduce.add.nxv2i16(<vscale x 2 x i16> %e)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.umax.nxv2i16(<vscale x 2 x i16>)
 
 define signext i16 @vreduce_umax_nxv2i16(<vscale x 2 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_umax_nxv2i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i16 @llvm.vector.reduce.umax.nxv2i16(<vscale x 2 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_umax_nxv2i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.umax.nxv2i16(<vscale x 2 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.umax.nxv2i16(<vscale x 2 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.smax.nxv2i16(<vscale x 2 x i16>)
 
 define signext i16 @vreduce_smax_nxv2i16(<vscale x 2 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_smax_nxv2i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i16 @llvm.vector.reduce.smax.nxv2i16(<vscale x 2 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_smax_nxv2i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.smax.nxv2i16(<vscale x 2 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.smax.nxv2i16(<vscale x 2 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.umin.nxv2i16(<vscale x 2 x i16>)
 
 define signext i16 @vreduce_umin_nxv2i16(<vscale x 2 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_umin_nxv2i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i16 @llvm.vector.reduce.umin.nxv2i16(<vscale x 2 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_umin_nxv2i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.umin.nxv2i16(<vscale x 2 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.umin.nxv2i16(<vscale x 2 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.smin.nxv2i16(<vscale x 2 x i16>)
 
 define signext i16 @vreduce_smin_nxv2i16(<vscale x 2 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_smin_nxv2i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i16 @llvm.vector.reduce.smin.nxv2i16(<vscale x 2 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_smin_nxv2i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.smin.nxv2i16(<vscale x 2 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.smin.nxv2i16(<vscale x 2 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.and.nxv2i16(<vscale x 2 x i16>)
 
 define signext i16 @vreduce_and_nxv2i16(<vscale x 2 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_and_nxv2i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i16 @llvm.vector.reduce.and.nxv2i16(<vscale x 2 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_and_nxv2i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.and.nxv2i16(<vscale x 2 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.and.nxv2i16(<vscale x 2 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.or.nxv2i16(<vscale x 2 x i16>)
 
 define signext i16 @vreduce_or_nxv2i16(<vscale x 2 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_or_nxv2i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i16 @llvm.vector.reduce.or.nxv2i16(<vscale x 2 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_or_nxv2i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.or.nxv2i16(<vscale x 2 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.or.nxv2i16(<vscale x 2 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.xor.nxv2i16(<vscale x 2 x i16>)
 
 define signext i16 @vreduce_xor_nxv2i16(<vscale x 2 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_xor_nxv2i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i16 @llvm.vector.reduce.xor.nxv2i16(<vscale x 2 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_xor_nxv2i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.xor.nxv2i16(<vscale x 2 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.xor.nxv2i16(<vscale x 2 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.add.nxv4i16(<vscale x 4 x i16>)
 
 define signext i16 @vreduce_add_nxv4i16(<vscale x 4 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_add_nxv4i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv4i16(<vscale x 4 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_add_nxv4i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv4i16(<vscale x 4 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.add.nxv4i16(<vscale x 4 x i16> %v)
   ret i16 %red
 }
 
 define signext i16 @vwreduce_add_nxv4i8(<vscale x 4 x i8> %v) {
 ; CHECK-LABEL: 'vwreduce_add_nxv4i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 4 x i8> %v to <vscale x 4 x i16>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv4i16(<vscale x 4 x i16> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vwreduce_add_nxv4i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 4 x i8> %v to <vscale x 4 x i16>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv4i16(<vscale x 4 x i16> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %e = sext <vscale x 4 x i8> %v to <vscale x 4 x i16>
   %red = call i16 @llvm.vector.reduce.add.nxv4i16(<vscale x 4 x i16> %e)
   ret i16 %red
 }
 
 define signext i16 @vwreduce_uadd_nxv4i8(<vscale x 4 x i8> %v) {
 ; CHECK-LABEL: 'vwreduce_uadd_nxv4i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 4 x i8> %v to <vscale x 4 x i16>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv4i16(<vscale x 4 x i16> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vwreduce_uadd_nxv4i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 4 x i8> %v to <vscale x 4 x i16>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.add.nxv4i16(<vscale x 4 x i16> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %e = sext <vscale x 4 x i8> %v to <vscale x 4 x i16>
   %red = call i16 @llvm.vector.reduce.add.nxv4i16(<vscale x 4 x i16> %e)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.umax.nxv4i16(<vscale x 4 x i16>)
 
 define signext i16 @vreduce_umax_nxv4i16(<vscale x 4 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_umax_nxv4i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i16 @llvm.vector.reduce.umax.nxv4i16(<vscale x 4 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_umax_nxv4i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.umax.nxv4i16(<vscale x 4 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.umax.nxv4i16(<vscale x 4 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.smax.nxv4i16(<vscale x 4 x i16>)
 
 define signext i16 @vreduce_smax_nxv4i16(<vscale x 4 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_smax_nxv4i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i16 @llvm.vector.reduce.smax.nxv4i16(<vscale x 4 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_smax_nxv4i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.smax.nxv4i16(<vscale x 4 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.smax.nxv4i16(<vscale x 4 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.umin.nxv4i16(<vscale x 4 x i16>)
 
 define signext i16 @vreduce_umin_nxv4i16(<vscale x 4 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_umin_nxv4i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i16 @llvm.vector.reduce.umin.nxv4i16(<vscale x 4 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_umin_nxv4i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.umin.nxv4i16(<vscale x 4 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.umin.nxv4i16(<vscale x 4 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.smin.nxv4i16(<vscale x 4 x i16>)
 
 define signext i16 @vreduce_smin_nxv4i16(<vscale x 4 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_smin_nxv4i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i16 @llvm.vector.reduce.smin.nxv4i16(<vscale x 4 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_smin_nxv4i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.smin.nxv4i16(<vscale x 4 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.smin.nxv4i16(<vscale x 4 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.and.nxv4i16(<vscale x 4 x i16>)
 
 define signext i16 @vreduce_and_nxv4i16(<vscale x 4 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_and_nxv4i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i16 @llvm.vector.reduce.and.nxv4i16(<vscale x 4 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_and_nxv4i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.and.nxv4i16(<vscale x 4 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.and.nxv4i16(<vscale x 4 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.or.nxv4i16(<vscale x 4 x i16>)
 
 define signext i16 @vreduce_or_nxv4i16(<vscale x 4 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_or_nxv4i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i16 @llvm.vector.reduce.or.nxv4i16(<vscale x 4 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_or_nxv4i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.or.nxv4i16(<vscale x 4 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.or.nxv4i16(<vscale x 4 x i16> %v)
   ret i16 %red
 }
 
 declare i16 @llvm.vector.reduce.xor.nxv4i16(<vscale x 4 x i16>)
 
 define signext i16 @vreduce_xor_nxv4i16(<vscale x 4 x i16> %v) {
 ; CHECK-LABEL: 'vreduce_xor_nxv4i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i16 @llvm.vector.reduce.xor.nxv4i16(<vscale x 4 x i16> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
 ;
+; SIZE-LABEL: 'vreduce_xor_nxv4i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i16 @llvm.vector.reduce.xor.nxv4i16(<vscale x 4 x i16> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i16 %red
+;
   %red = call i16 @llvm.vector.reduce.xor.nxv4i16(<vscale x 4 x i16> %v)
   ret i16 %red
 }
 
 declare i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32>)
 
 define signext i32 @vreduce_add_nxv1i32(<vscale x 1 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_add_nxv1i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_add_nxv1i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> %v)
   ret i32 %red
 }
 
 define signext i32 @vwreduce_add_nxv1i16(<vscale x 1 x i16> %v) {
 ; CHECK-LABEL: 'vwreduce_add_nxv1i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 1 x i16> %v to <vscale x 1 x i32>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vwreduce_add_nxv1i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 1 x i16> %v to <vscale x 1 x i32>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %e = sext <vscale x 1 x i16> %v to <vscale x 1 x i32>
   %red = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> %e)
   ret i32 %red
 }
 
 define signext i32 @vwreduce_uadd_nxv1i16(<vscale x 1 x i16> %v) {
 ; CHECK-LABEL: 'vwreduce_uadd_nxv1i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = zext <vscale x 1 x i16> %v to <vscale x 1 x i32>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vwreduce_uadd_nxv1i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = zext <vscale x 1 x i16> %v to <vscale x 1 x i32>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %e = zext <vscale x 1 x i16> %v to <vscale x 1 x i32>
   %red = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> %e)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.umax.nxv1i32(<vscale x 1 x i32>)
 
 define signext i32 @vreduce_umax_nxv1i32(<vscale x 1 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_umax_nxv1i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i32 @llvm.vector.reduce.umax.nxv1i32(<vscale x 1 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_umax_nxv1i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.umax.nxv1i32(<vscale x 1 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.umax.nxv1i32(<vscale x 1 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.smax.nxv1i32(<vscale x 1 x i32>)
 
 define signext i32 @vreduce_smax_nxv1i32(<vscale x 1 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_smax_nxv1i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i32 @llvm.vector.reduce.smax.nxv1i32(<vscale x 1 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_smax_nxv1i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.smax.nxv1i32(<vscale x 1 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.smax.nxv1i32(<vscale x 1 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.umin.nxv1i32(<vscale x 1 x i32>)
 
 define signext i32 @vreduce_umin_nxv1i32(<vscale x 1 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_umin_nxv1i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i32 @llvm.vector.reduce.umin.nxv1i32(<vscale x 1 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_umin_nxv1i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.umin.nxv1i32(<vscale x 1 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.umin.nxv1i32(<vscale x 1 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.smin.nxv1i32(<vscale x 1 x i32>)
 
 define signext i32 @vreduce_smin_nxv1i32(<vscale x 1 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_smin_nxv1i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i32 @llvm.vector.reduce.smin.nxv1i32(<vscale x 1 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_smin_nxv1i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.smin.nxv1i32(<vscale x 1 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.smin.nxv1i32(<vscale x 1 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.and.nxv1i32(<vscale x 1 x i32>)
 
 define signext i32 @vreduce_and_nxv1i32(<vscale x 1 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_and_nxv1i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i32 @llvm.vector.reduce.and.nxv1i32(<vscale x 1 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_and_nxv1i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.and.nxv1i32(<vscale x 1 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.and.nxv1i32(<vscale x 1 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.or.nxv1i32(<vscale x 1 x i32>)
 
 define signext i32 @vreduce_or_nxv1i32(<vscale x 1 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_or_nxv1i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i32 @llvm.vector.reduce.or.nxv1i32(<vscale x 1 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_or_nxv1i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.or.nxv1i32(<vscale x 1 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.or.nxv1i32(<vscale x 1 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.xor.nxv1i32(<vscale x 1 x i32>)
 
 define signext i32 @vreduce_xor_nxv1i32(<vscale x 1 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_xor_nxv1i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i32 @llvm.vector.reduce.xor.nxv1i32(<vscale x 1 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_xor_nxv1i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.xor.nxv1i32(<vscale x 1 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.xor.nxv1i32(<vscale x 1 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.add.nxv2i32(<vscale x 2 x i32>)
 
 define signext i32 @vreduce_add_nxv2i32(<vscale x 2 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_add_nxv2i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv2i32(<vscale x 2 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_add_nxv2i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv2i32(<vscale x 2 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.add.nxv2i32(<vscale x 2 x i32> %v)
   ret i32 %red
 }
 
 define signext i32 @vwreduce_add_nxv2i16(<vscale x 2 x i16> %v) {
 ; CHECK-LABEL: 'vwreduce_add_nxv2i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 2 x i16> %v to <vscale x 2 x i32>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv2i32(<vscale x 2 x i32> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vwreduce_add_nxv2i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 2 x i16> %v to <vscale x 2 x i32>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv2i32(<vscale x 2 x i32> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %e = sext <vscale x 2 x i16> %v to <vscale x 2 x i32>
   %red = call i32 @llvm.vector.reduce.add.nxv2i32(<vscale x 2 x i32> %e)
   ret i32 %red
 }
 
 define signext i32 @vwreduce_uadd_nxv2i16(<vscale x 2 x i16> %v) {
 ; CHECK-LABEL: 'vwreduce_uadd_nxv2i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = zext <vscale x 2 x i16> %v to <vscale x 2 x i32>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv2i32(<vscale x 2 x i32> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vwreduce_uadd_nxv2i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = zext <vscale x 2 x i16> %v to <vscale x 2 x i32>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv2i32(<vscale x 2 x i32> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %e = zext <vscale x 2 x i16> %v to <vscale x 2 x i32>
   %red = call i32 @llvm.vector.reduce.add.nxv2i32(<vscale x 2 x i32> %e)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.umax.nxv2i32(<vscale x 2 x i32>)
 
 define signext i32 @vreduce_umax_nxv2i32(<vscale x 2 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_umax_nxv2i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i32 @llvm.vector.reduce.umax.nxv2i32(<vscale x 2 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_umax_nxv2i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.umax.nxv2i32(<vscale x 2 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.umax.nxv2i32(<vscale x 2 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.smax.nxv2i32(<vscale x 2 x i32>)
 
 define signext i32 @vreduce_smax_nxv2i32(<vscale x 2 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_smax_nxv2i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i32 @llvm.vector.reduce.smax.nxv2i32(<vscale x 2 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_smax_nxv2i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.smax.nxv2i32(<vscale x 2 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.smax.nxv2i32(<vscale x 2 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.umin.nxv2i32(<vscale x 2 x i32>)
 
 define signext i32 @vreduce_umin_nxv2i32(<vscale x 2 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_umin_nxv2i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i32 @llvm.vector.reduce.umin.nxv2i32(<vscale x 2 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_umin_nxv2i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.umin.nxv2i32(<vscale x 2 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.umin.nxv2i32(<vscale x 2 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.smin.nxv2i32(<vscale x 2 x i32>)
 
 define signext i32 @vreduce_smin_nxv2i32(<vscale x 2 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_smin_nxv2i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i32 @llvm.vector.reduce.smin.nxv2i32(<vscale x 2 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_smin_nxv2i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.smin.nxv2i32(<vscale x 2 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.smin.nxv2i32(<vscale x 2 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.and.nxv2i32(<vscale x 2 x i32>)
 
 define signext i32 @vreduce_and_nxv2i32(<vscale x 2 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_and_nxv2i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i32 @llvm.vector.reduce.and.nxv2i32(<vscale x 2 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_and_nxv2i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.and.nxv2i32(<vscale x 2 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.and.nxv2i32(<vscale x 2 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.or.nxv2i32(<vscale x 2 x i32>)
 
 define signext i32 @vreduce_or_nxv2i32(<vscale x 2 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_or_nxv2i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i32 @llvm.vector.reduce.or.nxv2i32(<vscale x 2 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_or_nxv2i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.or.nxv2i32(<vscale x 2 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.or.nxv2i32(<vscale x 2 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.xor.nxv2i32(<vscale x 2 x i32>)
 
 define signext i32 @vreduce_xor_nxv2i32(<vscale x 2 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_xor_nxv2i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i32 @llvm.vector.reduce.xor.nxv2i32(<vscale x 2 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_xor_nxv2i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.xor.nxv2i32(<vscale x 2 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.xor.nxv2i32(<vscale x 2 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.add.nxv4i32(<vscale x 4 x i32>)
 
 define signext i32 @vreduce_add_nxv4i32(<vscale x 4 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_add_nxv4i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv4i32(<vscale x 4 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_add_nxv4i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv4i32(<vscale x 4 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.add.nxv4i32(<vscale x 4 x i32> %v)
   ret i32 %red
 }
 
 define signext i32 @vwreduce_add_nxv4i16(<vscale x 4 x i16> %v) {
 ; CHECK-LABEL: 'vwreduce_add_nxv4i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 4 x i16> %v to <vscale x 4 x i32>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv4i32(<vscale x 4 x i32> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vwreduce_add_nxv4i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 4 x i16> %v to <vscale x 4 x i32>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv4i32(<vscale x 4 x i32> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %e = sext <vscale x 4 x i16> %v to <vscale x 4 x i32>
   %red = call i32 @llvm.vector.reduce.add.nxv4i32(<vscale x 4 x i32> %e)
   ret i32 %red
 }
 
 define signext i32 @vwreduce_uadd_nxv4i16(<vscale x 4 x i16> %v) {
 ; CHECK-LABEL: 'vwreduce_uadd_nxv4i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = zext <vscale x 4 x i16> %v to <vscale x 4 x i32>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv4i32(<vscale x 4 x i32> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vwreduce_uadd_nxv4i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = zext <vscale x 4 x i16> %v to <vscale x 4 x i32>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.add.nxv4i32(<vscale x 4 x i32> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %e = zext <vscale x 4 x i16> %v to <vscale x 4 x i32>
   %red = call i32 @llvm.vector.reduce.add.nxv4i32(<vscale x 4 x i32> %e)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.umax.nxv4i32(<vscale x 4 x i32>)
 
 define signext i32 @vreduce_umax_nxv4i32(<vscale x 4 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_umax_nxv4i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i32 @llvm.vector.reduce.umax.nxv4i32(<vscale x 4 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_umax_nxv4i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.umax.nxv4i32(<vscale x 4 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.umax.nxv4i32(<vscale x 4 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.smax.nxv4i32(<vscale x 4 x i32>)
 
 define signext i32 @vreduce_smax_nxv4i32(<vscale x 4 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_smax_nxv4i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i32 @llvm.vector.reduce.smax.nxv4i32(<vscale x 4 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_smax_nxv4i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.smax.nxv4i32(<vscale x 4 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.smax.nxv4i32(<vscale x 4 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.umin.nxv4i32(<vscale x 4 x i32>)
 
 define signext i32 @vreduce_umin_nxv4i32(<vscale x 4 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_umin_nxv4i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i32 @llvm.vector.reduce.umin.nxv4i32(<vscale x 4 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_umin_nxv4i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.umin.nxv4i32(<vscale x 4 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.umin.nxv4i32(<vscale x 4 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.smin.nxv4i32(<vscale x 4 x i32>)
 
 define signext i32 @vreduce_smin_nxv4i32(<vscale x 4 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_smin_nxv4i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i32 @llvm.vector.reduce.smin.nxv4i32(<vscale x 4 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_smin_nxv4i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.smin.nxv4i32(<vscale x 4 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.smin.nxv4i32(<vscale x 4 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.and.nxv4i32(<vscale x 4 x i32>)
 
 define signext i32 @vreduce_and_nxv4i32(<vscale x 4 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_and_nxv4i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i32 @llvm.vector.reduce.and.nxv4i32(<vscale x 4 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_and_nxv4i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.and.nxv4i32(<vscale x 4 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.and.nxv4i32(<vscale x 4 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.or.nxv4i32(<vscale x 4 x i32>)
 
 define signext i32 @vreduce_or_nxv4i32(<vscale x 4 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_or_nxv4i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i32 @llvm.vector.reduce.or.nxv4i32(<vscale x 4 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_or_nxv4i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.or.nxv4i32(<vscale x 4 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.or.nxv4i32(<vscale x 4 x i32> %v)
   ret i32 %red
 }
 
 declare i32 @llvm.vector.reduce.xor.nxv4i32(<vscale x 4 x i32>)
 
 define signext i32 @vreduce_xor_nxv4i32(<vscale x 4 x i32> %v) {
 ; CHECK-LABEL: 'vreduce_xor_nxv4i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i32 @llvm.vector.reduce.xor.nxv4i32(<vscale x 4 x i32> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
 ;
+; SIZE-LABEL: 'vreduce_xor_nxv4i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i32 @llvm.vector.reduce.xor.nxv4i32(<vscale x 4 x i32> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 %red
+;
   %red = call i32 @llvm.vector.reduce.xor.nxv4i32(<vscale x 4 x i32> %v)
   ret i32 %red
 }
 
 declare i64 @llvm.vector.reduce.add.nxv1i64(<vscale x 1 x i64>)
 
 define i64 @vreduce_add_nxv1i64(<vscale x 1 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_add_nxv1i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv1i64(<vscale x 1 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_add_nxv1i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv1i64(<vscale x 1 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.add.nxv1i64(<vscale x 1 x i64> %v)
   ret i64 %red
 }
 
 define i64 @vwreduce_add_nxv1i32(<vscale x 1 x i32> %v) {
 ; CHECK-LABEL: 'vwreduce_add_nxv1i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 1 x i32> %v to <vscale x 1 x i64>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv1i64(<vscale x 1 x i64> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vwreduce_add_nxv1i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 1 x i32> %v to <vscale x 1 x i64>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv1i64(<vscale x 1 x i64> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %e = sext <vscale x 1 x i32> %v to <vscale x 1 x i64>
   %red = call i64 @llvm.vector.reduce.add.nxv1i64(<vscale x 1 x i64> %e)
   ret i64 %red
 }
 
 define i64 @vwreduce_uadd_nxv1i32(<vscale x 1 x i32> %v) {
 ; CHECK-LABEL: 'vwreduce_uadd_nxv1i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = zext <vscale x 1 x i32> %v to <vscale x 1 x i64>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv1i64(<vscale x 1 x i64> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vwreduce_uadd_nxv1i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = zext <vscale x 1 x i32> %v to <vscale x 1 x i64>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv1i64(<vscale x 1 x i64> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %e = zext <vscale x 1 x i32> %v to <vscale x 1 x i64>
   %red = call i64 @llvm.vector.reduce.add.nxv1i64(<vscale x 1 x i64> %e)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.umax.nxv1i64(<vscale x 1 x i64>)
 
 define i64 @vreduce_umax_nxv1i64(<vscale x 1 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_umax_nxv1i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i64 @llvm.vector.reduce.umax.nxv1i64(<vscale x 1 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_umax_nxv1i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.umax.nxv1i64(<vscale x 1 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.umax.nxv1i64(<vscale x 1 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.smax.nxv1i64(<vscale x 1 x i64>)
 
 define i64 @vreduce_smax_nxv1i64(<vscale x 1 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_smax_nxv1i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i64 @llvm.vector.reduce.smax.nxv1i64(<vscale x 1 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_smax_nxv1i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.smax.nxv1i64(<vscale x 1 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.smax.nxv1i64(<vscale x 1 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.umin.nxv1i64(<vscale x 1 x i64>)
 
 define i64 @vreduce_umin_nxv1i64(<vscale x 1 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_umin_nxv1i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i64 @llvm.vector.reduce.umin.nxv1i64(<vscale x 1 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_umin_nxv1i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.umin.nxv1i64(<vscale x 1 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.umin.nxv1i64(<vscale x 1 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.smin.nxv1i64(<vscale x 1 x i64>)
 
 define i64 @vreduce_smin_nxv1i64(<vscale x 1 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_smin_nxv1i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i64 @llvm.vector.reduce.smin.nxv1i64(<vscale x 1 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_smin_nxv1i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.smin.nxv1i64(<vscale x 1 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.smin.nxv1i64(<vscale x 1 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.and.nxv1i64(<vscale x 1 x i64>)
 
 define i64 @vreduce_and_nxv1i64(<vscale x 1 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_and_nxv1i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i64 @llvm.vector.reduce.and.nxv1i64(<vscale x 1 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_and_nxv1i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.and.nxv1i64(<vscale x 1 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.and.nxv1i64(<vscale x 1 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.or.nxv1i64(<vscale x 1 x i64>)
 
 define i64 @vreduce_or_nxv1i64(<vscale x 1 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_or_nxv1i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i64 @llvm.vector.reduce.or.nxv1i64(<vscale x 1 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_or_nxv1i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.or.nxv1i64(<vscale x 1 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.or.nxv1i64(<vscale x 1 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.xor.nxv1i64(<vscale x 1 x i64>)
 
 define i64 @vreduce_xor_nxv1i64(<vscale x 1 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_xor_nxv1i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %red = call i64 @llvm.vector.reduce.xor.nxv1i64(<vscale x 1 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_xor_nxv1i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.xor.nxv1i64(<vscale x 1 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.xor.nxv1i64(<vscale x 1 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.add.nxv2i64(<vscale x 2 x i64>)
 
 define i64 @vreduce_add_nxv2i64(<vscale x 2 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_add_nxv2i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv2i64(<vscale x 2 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_add_nxv2i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv2i64(<vscale x 2 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.add.nxv2i64(<vscale x 2 x i64> %v)
   ret i64 %red
 }
 
 define i64 @vwreduce_add_nxv2i32(<vscale x 2 x i32> %v) {
 ; CHECK-LABEL: 'vwreduce_add_nxv2i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 2 x i32> %v to <vscale x 2 x i64>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv2i64(<vscale x 2 x i64> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vwreduce_add_nxv2i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 2 x i32> %v to <vscale x 2 x i64>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv2i64(<vscale x 2 x i64> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %e = sext <vscale x 2 x i32> %v to <vscale x 2 x i64>
   %red = call i64 @llvm.vector.reduce.add.nxv2i64(<vscale x 2 x i64> %e)
   ret i64 %red
 }
 
 define i64 @vwreduce_uadd_nxv2i32(<vscale x 2 x i32> %v) {
 ; CHECK-LABEL: 'vwreduce_uadd_nxv2i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = zext <vscale x 2 x i32> %v to <vscale x 2 x i64>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv2i64(<vscale x 2 x i64> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vwreduce_uadd_nxv2i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = zext <vscale x 2 x i32> %v to <vscale x 2 x i64>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv2i64(<vscale x 2 x i64> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %e = zext <vscale x 2 x i32> %v to <vscale x 2 x i64>
   %red = call i64 @llvm.vector.reduce.add.nxv2i64(<vscale x 2 x i64> %e)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.umax.nxv2i64(<vscale x 2 x i64>)
 
 define i64 @vreduce_umax_nxv2i64(<vscale x 2 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_umax_nxv2i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i64 @llvm.vector.reduce.umax.nxv2i64(<vscale x 2 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_umax_nxv2i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.umax.nxv2i64(<vscale x 2 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.umax.nxv2i64(<vscale x 2 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.smax.nxv2i64(<vscale x 2 x i64>)
 
 define i64 @vreduce_smax_nxv2i64(<vscale x 2 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_smax_nxv2i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i64 @llvm.vector.reduce.smax.nxv2i64(<vscale x 2 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_smax_nxv2i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.smax.nxv2i64(<vscale x 2 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.smax.nxv2i64(<vscale x 2 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.umin.nxv2i64(<vscale x 2 x i64>)
 
 define i64 @vreduce_umin_nxv2i64(<vscale x 2 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_umin_nxv2i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i64 @llvm.vector.reduce.umin.nxv2i64(<vscale x 2 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_umin_nxv2i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.umin.nxv2i64(<vscale x 2 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.umin.nxv2i64(<vscale x 2 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.smin.nxv2i64(<vscale x 2 x i64>)
 
 define i64 @vreduce_smin_nxv2i64(<vscale x 2 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_smin_nxv2i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i64 @llvm.vector.reduce.smin.nxv2i64(<vscale x 2 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_smin_nxv2i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.smin.nxv2i64(<vscale x 2 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.smin.nxv2i64(<vscale x 2 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.and.nxv2i64(<vscale x 2 x i64>)
 
 define i64 @vreduce_and_nxv2i64(<vscale x 2 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_and_nxv2i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i64 @llvm.vector.reduce.and.nxv2i64(<vscale x 2 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_and_nxv2i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.and.nxv2i64(<vscale x 2 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.and.nxv2i64(<vscale x 2 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.or.nxv2i64(<vscale x 2 x i64>)
 
 define i64 @vreduce_or_nxv2i64(<vscale x 2 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_or_nxv2i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i64 @llvm.vector.reduce.or.nxv2i64(<vscale x 2 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_or_nxv2i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.or.nxv2i64(<vscale x 2 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.or.nxv2i64(<vscale x 2 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.xor.nxv2i64(<vscale x 2 x i64>)
 
 define i64 @vreduce_xor_nxv2i64(<vscale x 2 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_xor_nxv2i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %red = call i64 @llvm.vector.reduce.xor.nxv2i64(<vscale x 2 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_xor_nxv2i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.xor.nxv2i64(<vscale x 2 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.xor.nxv2i64(<vscale x 2 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.add.nxv4i64(<vscale x 4 x i64>)
 
 define i64 @vreduce_add_nxv4i64(<vscale x 4 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_add_nxv4i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv4i64(<vscale x 4 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_add_nxv4i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv4i64(<vscale x 4 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.add.nxv4i64(<vscale x 4 x i64> %v)
   ret i64 %red
 }
 
 define i64 @vwreduce_add_nxv4i32(<vscale x 4 x i32> %v) {
 ; CHECK-LABEL: 'vwreduce_add_nxv4i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 4 x i32> %v to <vscale x 4 x i64>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv4i64(<vscale x 4 x i64> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vwreduce_add_nxv4i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = sext <vscale x 4 x i32> %v to <vscale x 4 x i64>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv4i64(<vscale x 4 x i64> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %e = sext <vscale x 4 x i32> %v to <vscale x 4 x i64>
   %red = call i64 @llvm.vector.reduce.add.nxv4i64(<vscale x 4 x i64> %e)
   ret i64 %red
 }
 
 define i64 @vwreduce_uadd_nxv4i32(<vscale x 4 x i32> %v) {
 ; CHECK-LABEL: 'vwreduce_uadd_nxv4i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = zext <vscale x 4 x i32> %v to <vscale x 4 x i64>
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv4i64(<vscale x 4 x i64> %e)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vwreduce_uadd_nxv4i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %e = zext <vscale x 4 x i32> %v to <vscale x 4 x i64>
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.add.nxv4i64(<vscale x 4 x i64> %e)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %e = zext <vscale x 4 x i32> %v to <vscale x 4 x i64>
   %red = call i64 @llvm.vector.reduce.add.nxv4i64(<vscale x 4 x i64> %e)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.umax.nxv4i64(<vscale x 4 x i64>)
 
 define i64 @vreduce_umax_nxv4i64(<vscale x 4 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_umax_nxv4i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i64 @llvm.vector.reduce.umax.nxv4i64(<vscale x 4 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_umax_nxv4i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.umax.nxv4i64(<vscale x 4 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.umax.nxv4i64(<vscale x 4 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.smax.nxv4i64(<vscale x 4 x i64>)
 
 define i64 @vreduce_smax_nxv4i64(<vscale x 4 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_smax_nxv4i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i64 @llvm.vector.reduce.smax.nxv4i64(<vscale x 4 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_smax_nxv4i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.smax.nxv4i64(<vscale x 4 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.smax.nxv4i64(<vscale x 4 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.umin.nxv4i64(<vscale x 4 x i64>)
 
 define i64 @vreduce_umin_nxv4i64(<vscale x 4 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_umin_nxv4i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i64 @llvm.vector.reduce.umin.nxv4i64(<vscale x 4 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_umin_nxv4i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.umin.nxv4i64(<vscale x 4 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.umin.nxv4i64(<vscale x 4 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.smin.nxv4i64(<vscale x 4 x i64>)
 
 define i64 @vreduce_smin_nxv4i64(<vscale x 4 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_smin_nxv4i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i64 @llvm.vector.reduce.smin.nxv4i64(<vscale x 4 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_smin_nxv4i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.smin.nxv4i64(<vscale x 4 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.smin.nxv4i64(<vscale x 4 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.and.nxv4i64(<vscale x 4 x i64>)
 
 define i64 @vreduce_and_nxv4i64(<vscale x 4 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_and_nxv4i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i64 @llvm.vector.reduce.and.nxv4i64(<vscale x 4 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_and_nxv4i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.and.nxv4i64(<vscale x 4 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.and.nxv4i64(<vscale x 4 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.or.nxv4i64(<vscale x 4 x i64>)
 
 define i64 @vreduce_or_nxv4i64(<vscale x 4 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_or_nxv4i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i64 @llvm.vector.reduce.or.nxv4i64(<vscale x 4 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_or_nxv4i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.or.nxv4i64(<vscale x 4 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.or.nxv4i64(<vscale x 4 x i64> %v)
   ret i64 %red
 }
 
 declare i64 @llvm.vector.reduce.xor.nxv4i64(<vscale x 4 x i64>)
 
 define i64 @vreduce_xor_nxv4i64(<vscale x 4 x i64> %v) {
 ; CHECK-LABEL: 'vreduce_xor_nxv4i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %red = call i64 @llvm.vector.reduce.xor.nxv4i64(<vscale x 4 x i64> %v)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
 ;
+; SIZE-LABEL: 'vreduce_xor_nxv4i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %red = call i64 @llvm.vector.reduce.xor.nxv4i64(<vscale x 4 x i64> %v)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i64 %red
+;
   %red = call i64 @llvm.vector.reduce.xor.nxv4i64(<vscale x 4 x i64> %v)
   ret i64 %red
 }

llvm/test/Analysis/CostModel/RISCV/reduce-xor.ll

 ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
 ; RUN: opt < %s -mtriple=riscv32 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s
 ; RUN: opt < %s -mtriple=riscv64 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=throughput 2>&1 -disable-output | FileCheck %s
+; RUN: opt < %s -mtriple=riscv32 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s --check-prefix=SIZE
+; RUN: opt < %s -mtriple=riscv64 -mattr=+v -riscv-v-vector-bits-min=256 -passes="print<cost-model>" -cost-kind=code-size 2>&1 -disable-output | FileCheck %s --check-prefix=SIZE
 
 define i32 @reduce_i1(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i1'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i1 @llvm.vector.reduce.xor.v1i1(<1 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i1 @llvm.vector.reduce.xor.v2i1(<2 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i1 @llvm.vector.reduce.xor.v4i1(<4 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i1 @llvm.vector.reduce.xor.v8i1(<8 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i1 @llvm.vector.reduce.xor.v16i1(<16 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i1 @llvm.vector.reduce.xor.v32i1(<32 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i1 @llvm.vector.reduce.xor.v64i1(<64 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i1 @llvm.vector.reduce.xor.v128i1(<128 x i1> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i1'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i1 @llvm.vector.reduce.xor.v1i1(<1 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i1 @llvm.vector.reduce.xor.v2i1(<2 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i1 @llvm.vector.reduce.xor.v4i1(<4 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i1 @llvm.vector.reduce.xor.v8i1(<8 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i1 @llvm.vector.reduce.xor.v16i1(<16 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i1 @llvm.vector.reduce.xor.v32i1(<32 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i1 @llvm.vector.reduce.xor.v64i1(<64 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i1 @llvm.vector.reduce.xor.v128i1(<128 x i1> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i1 @llvm.vector.reduce.xor.v1i1(<1 x i1> undef)
   %V2   = call i1 @llvm.vector.reduce.xor.v2i1(<2 x i1> undef)
   %V4   = call i1 @llvm.vector.reduce.xor.v4i1(<4 x i1> undef)
   %V8   = call i1 @llvm.vector.reduce.xor.v8i1(<8 x i1> undef)
   %V16  = call i1 @llvm.vector.reduce.xor.v16i1(<16 x i1> undef)
   %V32  = call i1 @llvm.vector.reduce.xor.v32i1(<32 x i1> undef)
   %V64  = call i1 @llvm.vector.reduce.xor.v64i1(<64 x i1> undef)
   %V128 = call i1 @llvm.vector.reduce.xor.v128i1(<128 x i1> undef)
   ret i32 undef
 }
 
 define i32 @reduce_i8(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i8'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i8 @llvm.vector.reduce.xor.v1i8(<1 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i8 @llvm.vector.reduce.xor.v2i8(<2 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i8 @llvm.vector.reduce.xor.v4i8(<4 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i8 @llvm.vector.reduce.xor.v8i8(<8 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i8 @llvm.vector.reduce.xor.v16i8(<16 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i8 @llvm.vector.reduce.xor.v32i8(<32 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i8 @llvm.vector.reduce.xor.v64i8(<64 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call i8 @llvm.vector.reduce.xor.v128i8(<128 x i8> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i8'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i8 @llvm.vector.reduce.xor.v1i8(<1 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i8 @llvm.vector.reduce.xor.v2i8(<2 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i8 @llvm.vector.reduce.xor.v4i8(<4 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i8 @llvm.vector.reduce.xor.v8i8(<8 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i8 @llvm.vector.reduce.xor.v16i8(<16 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i8 @llvm.vector.reduce.xor.v32i8(<32 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i8 @llvm.vector.reduce.xor.v64i8(<64 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i8 @llvm.vector.reduce.xor.v128i8(<128 x i8> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i8 @llvm.vector.reduce.xor.v1i8(<1 x i8> undef)
   %V2   = call i8 @llvm.vector.reduce.xor.v2i8(<2 x i8> undef)
   %V4   = call i8 @llvm.vector.reduce.xor.v4i8(<4 x i8> undef)
   %V8   = call i8 @llvm.vector.reduce.xor.v8i8(<8 x i8> undef)
   %V16  = call i8 @llvm.vector.reduce.xor.v16i8(<16 x i8> undef)
   %V32  = call i8 @llvm.vector.reduce.xor.v32i8(<32 x i8> undef)
   %V64  = call i8 @llvm.vector.reduce.xor.v64i8(<64 x i8> undef)
   %V128 = call i8 @llvm.vector.reduce.xor.v128i8(<128 x i8> undef)
   ret i32 undef
 }
 
 define i32 @reduce_i16(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i16'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i16 @llvm.vector.reduce.xor.v1i16(<1 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i16 @llvm.vector.reduce.xor.v2i16(<2 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i16 @llvm.vector.reduce.xor.v4i16(<4 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i16 @llvm.vector.reduce.xor.v8i16(<8 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i16 @llvm.vector.reduce.xor.v16i16(<16 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i16 @llvm.vector.reduce.xor.v32i16(<32 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i16 @llvm.vector.reduce.xor.v64i16(<64 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V128 = call i16 @llvm.vector.reduce.xor.v128i16(<128 x i16> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i16'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i16 @llvm.vector.reduce.xor.v1i16(<1 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i16 @llvm.vector.reduce.xor.v2i16(<2 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i16 @llvm.vector.reduce.xor.v4i16(<4 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i16 @llvm.vector.reduce.xor.v8i16(<8 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i16 @llvm.vector.reduce.xor.v16i16(<16 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i16 @llvm.vector.reduce.xor.v32i16(<32 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i16 @llvm.vector.reduce.xor.v64i16(<64 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V128 = call i16 @llvm.vector.reduce.xor.v128i16(<128 x i16> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i16 @llvm.vector.reduce.xor.v1i16(<1 x i16> undef)
   %V2   = call i16 @llvm.vector.reduce.xor.v2i16(<2 x i16> undef)
   %V4   = call i16 @llvm.vector.reduce.xor.v4i16(<4 x i16> undef)
   %V8   = call i16 @llvm.vector.reduce.xor.v8i16(<8 x i16> undef)
   %V16  = call i16 @llvm.vector.reduce.xor.v16i16(<16 x i16> undef)
   %V32  = call i16 @llvm.vector.reduce.xor.v32i16(<32 x i16> undef)
   %V64  = call i16 @llvm.vector.reduce.xor.v64i16(<64 x i16> undef)
   %V128 = call i16 @llvm.vector.reduce.xor.v128i16(<128 x i16> undef)
   ret i32 undef
 }
 
 define i32 @reduce_i32(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i32'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i32 @llvm.vector.reduce.xor.v1i32(<1 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i32 @llvm.vector.reduce.xor.v2i32(<2 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i32 @llvm.vector.reduce.xor.v4i32(<4 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i32 @llvm.vector.reduce.xor.v8i32(<8 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i32 @llvm.vector.reduce.xor.v16i32(<16 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i32 @llvm.vector.reduce.xor.v32i32(<32 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V64 = call i32 @llvm.vector.reduce.xor.v64i32(<64 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V128 = call i32 @llvm.vector.reduce.xor.v128i32(<128 x i32> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i32'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i32 @llvm.vector.reduce.xor.v1i32(<1 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i32 @llvm.vector.reduce.xor.v2i32(<2 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i32 @llvm.vector.reduce.xor.v4i32(<4 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i32 @llvm.vector.reduce.xor.v8i32(<8 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i32 @llvm.vector.reduce.xor.v16i32(<16 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i32 @llvm.vector.reduce.xor.v32i32(<32 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V64 = call i32 @llvm.vector.reduce.xor.v64i32(<64 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V128 = call i32 @llvm.vector.reduce.xor.v128i32(<128 x i32> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i32 @llvm.vector.reduce.xor.v1i32(<1 x i32> undef)
   %V2   = call i32 @llvm.vector.reduce.xor.v2i32(<2 x i32> undef)
   %V4   = call i32 @llvm.vector.reduce.xor.v4i32(<4 x i32> undef)
   %V8   = call i32 @llvm.vector.reduce.xor.v8i32(<8 x i32> undef)
   %V16  = call i32 @llvm.vector.reduce.xor.v16i32(<16 x i32> undef)
   %V32  = call i32 @llvm.vector.reduce.xor.v32i32(<32 x i32> undef)
   %V64  = call i32 @llvm.vector.reduce.xor.v64i32(<64 x i32> undef)
   %V128 = call i32 @llvm.vector.reduce.xor.v128i32(<128 x i32> undef)
   ret i32 undef
 }
 
 define i32 @reduce_i64(i32 %arg) {
 ; CHECK-LABEL: 'reduce_i64'
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.vector.reduce.xor.v1i64(<1 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2 = call i64 @llvm.vector.reduce.xor.v2i64(<2 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4 = call i64 @llvm.vector.reduce.xor.v4i64(<4 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8 = call i64 @llvm.vector.reduce.xor.v8i64(<8 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16 = call i64 @llvm.vector.reduce.xor.v16i64(<16 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V32 = call i64 @llvm.vector.reduce.xor.v32i64(<32 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V64 = call i64 @llvm.vector.reduce.xor.v64i64(<64 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V128 = call i64 @llvm.vector.reduce.xor.v128i64(<128 x i64> undef)
 ; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
 ;
+; SIZE-LABEL: 'reduce_i64'
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V1 = call i64 @llvm.vector.reduce.xor.v1i64(<1 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2 = call i64 @llvm.vector.reduce.xor.v2i64(<2 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4 = call i64 @llvm.vector.reduce.xor.v4i64(<4 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8 = call i64 @llvm.vector.reduce.xor.v8i64(<8 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16 = call i64 @llvm.vector.reduce.xor.v16i64(<16 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32 = call i64 @llvm.vector.reduce.xor.v32i64(<32 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V64 = call i64 @llvm.vector.reduce.xor.v64i64(<64 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V128 = call i64 @llvm.vector.reduce.xor.v128i64(<128 x i64> undef)
+; SIZE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
+;
   %V1   = call i64 @llvm.vector.reduce.xor.v1i64(<1 x i64> undef)
   %V2   = call i64 @llvm.vector.reduce.xor.v2i64(<2 x i64> undef)
   %V4   = call i64 @llvm.vector.reduce.xor.v4i64(<4 x i64> undef)
   %V8   = call i64 @llvm.vector.reduce.xor.v8i64(<8 x i64> undef)
   %V16  = call i64 @llvm.vector.reduce.xor.v16i64(<16 x i64> undef)
   %V32  = call i64 @llvm.vector.reduce.xor.v32i64(<32 x i64> undef)
   %V64  = call i64 @llvm.vector.reduce.xor.v64i64(<64 x i64> undef)
   %V128 = call i64 @llvm.vector.reduce.xor.v128i64(<128 x i64> undef)