[llvm] b1ef919 - [ARM] Add CostKind to getMVEVectorCostFactor.
David Green via llvm-commits
llvm-commits at lists.llvm.org
Thu Feb 11 07:34:15 PST 2021
Author: David Green
Date: 2021-02-11T15:33:59Z
New Revision: b1ef919aada5217b7111433959399b73b3547293
URL: https://github.com/llvm/llvm-project/commit/b1ef919aada5217b7111433959399b73b3547293
DIFF: https://github.com/llvm/llvm-project/commit/b1ef919aada5217b7111433959399b73b3547293.diff
LOG: [ARM] Add CostKind to getMVEVectorCostFactor.
This adds the CostKind to getMVEVectorCostFactor, so that it can
automatically account for CodeSize costs, where it returns a cost of 1
not the MVEFactor used for Throughput/Latency. This helps simplify the
caller code and allows us to get the codesize cost more correct in more
cases.
Added:
Modified:
llvm/lib/Target/ARM/ARMSubtarget.h
llvm/lib/Target/ARM/ARMTargetTransformInfo.cpp
llvm/test/Analysis/CostModel/ARM/arith-ssat.ll
llvm/test/Analysis/CostModel/ARM/arith-usat.ll
llvm/test/Analysis/CostModel/ARM/cast.ll
llvm/test/Analysis/CostModel/ARM/cast_ldst.ll
Removed:
################################################################################
diff --git a/llvm/lib/Target/ARM/ARMSubtarget.h b/llvm/lib/Target/ARM/ARMSubtarget.h
index 1e1195ce9a99..44ef8593f6ff 100644
--- a/llvm/lib/Target/ARM/ARMSubtarget.h
+++ b/llvm/lib/Target/ARM/ARMSubtarget.h
@@ -20,6 +20,7 @@
#include "ARMISelLowering.h"
#include "ARMSelectionDAGInfo.h"
#include "llvm/ADT/Triple.h"
+#include "llvm/Analysis/TargetTransformInfo.h"
#include "llvm/CodeGen/GlobalISel/CallLowering.h"
#include "llvm/CodeGen/GlobalISel/InstructionSelector.h"
#include "llvm/CodeGen/GlobalISel/LegalizerInfo.h"
@@ -911,7 +912,12 @@ class ARMSubtarget : public ARMGenSubtargetInfo {
unsigned getPrefLoopLogAlignment() const { return PrefLoopLogAlignment; }
- unsigned getMVEVectorCostFactor() const { return MVEVectorCostFactor; }
+ unsigned
+ getMVEVectorCostFactor(TargetTransformInfo::TargetCostKind CostKind) const {
+ if (CostKind == TargetTransformInfo::TCK_CodeSize)
+ return 1;
+ return MVEVectorCostFactor;
+ }
bool ignoreCSRForAllocationOrder(const MachineFunction &MF,
unsigned PhysReg) const override;
diff --git a/llvm/lib/Target/ARM/ARMTargetTransformInfo.cpp b/llvm/lib/Target/ARM/ARMTargetTransformInfo.cpp
index de2c0607d2ed..f122cc59bb4f 100644
--- a/llvm/lib/Target/ARM/ARMTargetTransformInfo.cpp
+++ b/llvm/lib/Target/ARM/ARMTargetTransformInfo.cpp
@@ -426,7 +426,8 @@ int ARMTTIImpl::getCastInstrCost(unsigned Opcode, Type *Dst, Type *Src,
(Opcode == Instruction::FPExt || Opcode == Instruction::FPTrunc) &&
IsLegalFPType(SrcTy) && IsLegalFPType(DstTy)))
if (CCH == TTI::CastContextHint::Masked && DstTy.getSizeInBits() > 128)
- return 2 * DstTy.getVectorNumElements() * ST->getMVEVectorCostFactor();
+ return 2 * DstTy.getVectorNumElements() *
+ ST->getMVEVectorCostFactor(CostKind);
// The extend of other kinds of load is free
if (CCH == TTI::CastContextHint::Normal ||
@@ -470,7 +471,7 @@ int ARMTTIImpl::getCastInstrCost(unsigned Opcode, Type *Dst, Type *Src,
if (const auto *Entry =
ConvertCostTableLookup(MVELoadConversionTbl, ISD,
DstTy.getSimpleVT(), SrcTy.getSimpleVT()))
- return AdjustCost(Entry->Cost * ST->getMVEVectorCostFactor());
+ return Entry->Cost * ST->getMVEVectorCostFactor(CostKind);
}
static const TypeConversionCostTblEntry MVEFLoadConversionTbl[] = {
@@ -482,7 +483,7 @@ int ARMTTIImpl::getCastInstrCost(unsigned Opcode, Type *Dst, Type *Src,
if (const auto *Entry =
ConvertCostTableLookup(MVEFLoadConversionTbl, ISD,
DstTy.getSimpleVT(), SrcTy.getSimpleVT()))
- return AdjustCost(Entry->Cost * ST->getMVEVectorCostFactor());
+ return Entry->Cost * ST->getMVEVectorCostFactor(CostKind);
}
// The truncate of a store is free. This is the mirror of extends above.
@@ -499,7 +500,7 @@ int ARMTTIImpl::getCastInstrCost(unsigned Opcode, Type *Dst, Type *Src,
if (const auto *Entry =
ConvertCostTableLookup(MVEStoreConversionTbl, ISD,
SrcTy.getSimpleVT(), DstTy.getSimpleVT()))
- return AdjustCost(Entry->Cost * ST->getMVEVectorCostFactor());
+ return Entry->Cost * ST->getMVEVectorCostFactor(CostKind);
}
static const TypeConversionCostTblEntry MVEFStoreConversionTbl[] = {
@@ -510,7 +511,7 @@ int ARMTTIImpl::getCastInstrCost(unsigned Opcode, Type *Dst, Type *Src,
if (const auto *Entry =
ConvertCostTableLookup(MVEFStoreConversionTbl, ISD,
SrcTy.getSimpleVT(), DstTy.getSimpleVT()))
- return AdjustCost(Entry->Cost * ST->getMVEVectorCostFactor());
+ return Entry->Cost * ST->getMVEVectorCostFactor(CostKind);
}
}
@@ -734,7 +735,7 @@ int ARMTTIImpl::getCastInstrCost(unsigned Opcode, Type *Dst, Type *Src,
if (const auto *Entry = ConvertCostTableLookup(MVEVectorConversionTbl,
ISD, DstTy.getSimpleVT(),
SrcTy.getSimpleVT()))
- return AdjustCost(Entry->Cost * ST->getMVEVectorCostFactor());
+ return Entry->Cost * ST->getMVEVectorCostFactor(CostKind);
}
if (ISD == ISD::FP_ROUND || ISD == ISD::FP_EXTEND) {
@@ -784,7 +785,7 @@ int ARMTTIImpl::getCastInstrCost(unsigned Opcode, Type *Dst, Type *Src,
}
int BaseCost = ST->hasMVEIntegerOps() && Src->isVectorTy()
- ? ST->getMVEVectorCostFactor()
+ ? ST->getMVEVectorCostFactor(CostKind)
: 1;
return AdjustCost(
BaseCost * BaseT::getCastInstrCost(Opcode, Dst, Src, CCH, CostKind, I));
@@ -819,7 +820,7 @@ int ARMTTIImpl::getVectorInstrCost(unsigned Opcode, Type *ValTy,
// vector, to prevent vectorising where we end up just scalarising the
// result anyway.
return std::max(BaseT::getVectorInstrCost(Opcode, ValTy, Index),
- ST->getMVEVectorCostFactor()) *
+ ST->getMVEVectorCostFactor(TTI::TCK_RecipThroughput)) *
cast<FixedVectorType>(ValTy)->getNumElements() / 2;
}
@@ -881,9 +882,8 @@ int ARMTTIImpl::getCmpSelInstrCost(unsigned Opcode, Type *ValTy, Type *CondTy,
// Default to cheap (throughput/size of 1 instruction) but adjust throughput
// for "multiple beats" potentially needed by MVE instructions.
int BaseCost = 1;
- if (CostKind != TTI::TCK_CodeSize && ST->hasMVEIntegerOps() &&
- ValTy->isVectorTy())
- BaseCost = ST->getMVEVectorCostFactor();
+ if (ST->hasMVEIntegerOps() && ValTy->isVectorTy())
+ BaseCost = ST->getMVEVectorCostFactor(CostKind);
return BaseCost *
BaseT::getCmpSelInstrCost(Opcode, ValTy, CondTy, VecPred, CostKind, I);
@@ -1132,11 +1132,12 @@ int ARMTTIImpl::getShuffleCost(TTI::ShuffleKind Kind, VectorType *Tp,
if (const auto *Entry = CostTableLookup(MVEDupTbl, ISD::VECTOR_SHUFFLE,
LT.second))
- return LT.first * Entry->Cost * ST->getMVEVectorCostFactor();
+ return LT.first * Entry->Cost *
+ ST->getMVEVectorCostFactor(TTI::TCK_RecipThroughput);
}
}
int BaseCost = ST->hasMVEIntegerOps() && Tp->isVectorTy()
- ? ST->getMVEVectorCostFactor()
+ ? ST->getMVEVectorCostFactor(TTI::TCK_RecipThroughput)
: 1;
return BaseCost * BaseT::getShuffleCost(Kind, Tp, Index, SubTp);
}
@@ -1262,9 +1263,8 @@ int ARMTTIImpl::getArithmeticInstrCost(unsigned Opcode, Type *Ty,
// Default to cheap (throughput/size of 1 instruction) but adjust throughput
// for "multiple beats" potentially needed by MVE instructions.
int BaseCost = 1;
- if (CostKind != TTI::TCK_CodeSize && ST->hasMVEIntegerOps() &&
- Ty->isVectorTy())
- BaseCost = ST->getMVEVectorCostFactor();
+ if (ST->hasMVEIntegerOps() && Ty->isVectorTy())
+ BaseCost = ST->getMVEVectorCostFactor(CostKind);
// The rest of this mostly follows what is done in BaseT::getArithmeticInstrCost,
// without treating floats as more expensive that scalars or increasing the
@@ -1321,11 +1321,11 @@ int ARMTTIImpl::getMemoryOpCost(unsigned Opcode, Type *Src,
: cast<Instruction>(I->getOperand(0))->getOperand(0)->getType();
if (SrcVTy->getNumElements() == 4 && SrcVTy->getScalarType()->isHalfTy() &&
DstTy->getScalarType()->isFloatTy())
- return ST->getMVEVectorCostFactor();
+ return ST->getMVEVectorCostFactor(CostKind);
}
int BaseCost = ST->hasMVEIntegerOps() && Src->isVectorTy()
- ? ST->getMVEVectorCostFactor()
+ ? ST->getMVEVectorCostFactor(CostKind)
: 1;
return BaseCost * BaseT::getMemoryOpCost(Opcode, Src, Alignment, AddressSpace,
CostKind, I);
@@ -1337,9 +1337,9 @@ unsigned ARMTTIImpl::getMaskedMemoryOpCost(unsigned Opcode, Type *Src,
TTI::TargetCostKind CostKind) {
if (ST->hasMVEIntegerOps()) {
if (Opcode == Instruction::Load && isLegalMaskedLoad(Src, Alignment))
- return ST->getMVEVectorCostFactor();
+ return ST->getMVEVectorCostFactor(CostKind);
if (Opcode == Instruction::Store && isLegalMaskedStore(Src, Alignment))
- return ST->getMVEVectorCostFactor();
+ return ST->getMVEVectorCostFactor(CostKind);
}
if (!isa<FixedVectorType>(Src))
return BaseT::getMaskedMemoryOpCost(Opcode, Src, Alignment, AddressSpace,
@@ -1368,7 +1368,8 @@ int ARMTTIImpl::getInterleavedMemoryOpCost(
// vldN/vstN only support legal vector types of size 64 or 128 in bits.
// Accesses having vector types that are a multiple of 128 bits can be
// matched to more than one vldN/vstN instruction.
- int BaseCost = ST->hasMVEIntegerOps() ? ST->getMVEVectorCostFactor() : 1;
+ int BaseCost =
+ ST->hasMVEIntegerOps() ? ST->getMVEVectorCostFactor(CostKind) : 1;
if (NumElts % Factor == 0 &&
TLI->isLegalInterleavedAccessType(Factor, SubVecTy, Alignment, DL))
return Factor * BaseCost * TLI->getNumInterleavedAccesses(SubVecTy, DL);
@@ -1413,7 +1414,8 @@ unsigned ARMTTIImpl::getGatherScatterOpCost(unsigned Opcode, Type *DataTy,
// multiplied by the number of elements being loaded. This is possibly very
// conservative, but even so we still end up vectorising loops because the
// cost per iteration for many loops is lower than for scalar loops.
- unsigned VectorCost = NumElems * LT.first * ST->getMVEVectorCostFactor();
+ unsigned VectorCost =
+ NumElems * LT.first * ST->getMVEVectorCostFactor(CostKind);
// The scalarization cost should be a lot higher. We use the number of vector
// elements plus the scalarization overhead.
unsigned ScalarCost = NumElems * LT.first +
@@ -1506,7 +1508,7 @@ int ARMTTIImpl::getArithmeticReductionCost(unsigned Opcode, VectorType *ValTy,
{ISD::ADD, MVT::v4i32, 1},
};
if (const auto *Entry = CostTableLookup(CostTblAdd, ISD, LT.second))
- return Entry->Cost * ST->getMVEVectorCostFactor() * LT.first;
+ return Entry->Cost * ST->getMVEVectorCostFactor(CostKind) * LT.first;
return BaseT::getArithmeticReductionCost(Opcode, ValTy, IsPairwiseForm,
CostKind);
@@ -1524,7 +1526,7 @@ ARMTTIImpl::getExtendedAddReductionCost(bool IsMLA, bool IsUnsigned,
(LT.second == MVT::v8i16 &&
ResVT.getSizeInBits() <= (IsMLA ? 64 : 32)) ||
(LT.second == MVT::v4i32 && ResVT.getSizeInBits() <= 64))
- return ST->getMVEVectorCostFactor() * LT.first;
+ return ST->getMVEVectorCostFactor(CostKind) * LT.first;
}
return BaseT::getExtendedAddReductionCost(IsMLA, IsUnsigned, ResTy, ValTy,
@@ -1566,7 +1568,7 @@ int ARMTTIImpl::getIntrinsicInstrCost(const IntrinsicCostAttributes &ICA,
ICA.getReturnType()->getScalarSizeInBits()
? 1
: 4;
- return LT.first * ST->getMVEVectorCostFactor() * Instrs;
+ return LT.first * ST->getMVEVectorCostFactor(CostKind) * Instrs;
}
break;
}
diff --git a/llvm/test/Analysis/CostModel/ARM/arith-ssat.ll b/llvm/test/Analysis/CostModel/ARM/arith-ssat.ll
index 9cf5a9b8ab19..251392af24c7 100644
--- a/llvm/test/Analysis/CostModel/ARM/arith-ssat.ll
+++ b/llvm/test/Analysis/CostModel/ARM/arith-ssat.ll
@@ -165,22 +165,22 @@ define i32 @add(i32 %arg) {
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 153 for instruction: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 17 for instruction: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 19 for instruction: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 17 for instruction: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 19 for instruction: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 17 for instruction: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 19 for instruction: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
;
%I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)
@@ -371,22 +371,22 @@ define i32 @sub(i32 %arg) {
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 153 for instruction: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 17 for instruction: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 19 for instruction: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 17 for instruction: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 19 for instruction: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 17 for instruction: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 19 for instruction: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
;
%I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)
diff --git a/llvm/test/Analysis/CostModel/ARM/arith-usat.ll b/llvm/test/Analysis/CostModel/ARM/arith-usat.ll
index 9670807db662..342956cb83e8 100644
--- a/llvm/test/Analysis/CostModel/ARM/arith-usat.ll
+++ b/llvm/test/Analysis/CostModel/ARM/arith-usat.ll
@@ -165,22 +165,22 @@ define i32 @add(i32 %arg) {
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 146 for instruction: %V8I64 = call <8 x i64> @llvm.uadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.uadd.sat.i32(i32 undef, i32 undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V2I32 = call <2 x i32> @llvm.uadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.uadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.uadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.uadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.uadd.sat.i16(i16 undef, i16 undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V2I16 = call <2 x i16> @llvm.uadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V4I16 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4I16 = call <4 x i16> @llvm.uadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.uadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.uadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.uadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.uadd.sat.i8(i8 undef, i8 undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V2I8 = call <2 x i8> @llvm.uadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V4I8 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V8I8 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = call <4 x i8> @llvm.uadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8I8 = call <8 x i8> @llvm.uadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.uadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.uadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.uadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
;
%I64 = call i64 @llvm.uadd.sat.i64(i64 undef, i64 undef)
@@ -371,22 +371,22 @@ define i32 @sub(i32 %arg) {
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 146 for instruction: %V8I64 = call <8 x i64> @llvm.usub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.usub.sat.i32(i32 undef, i32 undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V2I32 = call <2 x i32> @llvm.usub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4I32 = call <4 x i32> @llvm.usub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8I32 = call <8 x i32> @llvm.usub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16I32 = call <16 x i32> @llvm.usub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %I16 = call i16 @llvm.usub.sat.i16(i16 undef, i16 undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V2I16 = call <2 x i16> @llvm.usub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V4I16 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4I16 = call <4 x i16> @llvm.usub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8I16 = call <8 x i16> @llvm.usub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = call <16 x i16> @llvm.usub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = call <32 x i16> @llvm.usub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %I8 = call i8 @llvm.usub.sat.i8(i8 undef, i8 undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V2I8 = call <2 x i8> @llvm.usub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V4I8 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V8I8 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
-; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = call <4 x i8> @llvm.usub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8I8 = call <8 x i8> @llvm.usub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V16I8 = call <16 x i8> @llvm.usub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V32I8 = call <32 x i8> @llvm.usub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)
+; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V64I8 = call <64 x i8> @llvm.usub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)
; MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
;
%I64 = call i64 @llvm.usub.sat.i64(i64 undef, i64 undef)
diff --git a/llvm/test/Analysis/CostModel/ARM/cast.ll b/llvm/test/Analysis/CostModel/ARM/cast.ll
index 6741f92f6e8e..f8377ad3f0c1 100644
--- a/llvm/test/Analysis/CostModel/ARM/cast.ll
+++ b/llvm/test/Analysis/CostModel/ARM/cast.ll
@@ -2072,14 +2072,14 @@ define i32 @casts() {
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %r67 = uitofp i64 undef to half
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %r68 = sitofp i64 undef to double
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %r69 = uitofp i64 undef to double
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %q70 = sext <4 x i8> undef to <4 x i32>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %q70 = sext <4 x i8> undef to <4 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %q71 = sext <8 x i8> undef to <8 x i16>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %s70 = sext <4 x i8> undef to <4 x i32>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %s70 = sext <4 x i8> undef to <4 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %r70 = sext <8 x i8> undef to <8 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %r71 = sext <16 x i8> undef to <16 x i32>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %q72 = zext <4 x i8> undef to <4 x i32>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %q72 = zext <4 x i8> undef to <4 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %q73 = zext <8 x i8> undef to <8 x i16>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %s72 = zext <4 x i8> undef to <4 x i32>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %s72 = zext <4 x i8> undef to <4 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %r72 = zext <8 x i8> undef to <8 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %r73 = zext <16 x i8> undef to <16 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %rext_0 = sext <8 x i8> undef to <8 x i64>
@@ -2088,12 +2088,12 @@ define i32 @casts() {
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %rext_3 = zext <8 x i16> undef to <8 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %rext_4 = sext <4 x i16> undef to <4 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %rext_5 = zext <4 x i16> undef to <4 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %rext_6 = sext <2 x i8> undef to <2 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %rext_7 = zext <2 x i8> undef to <2 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %rext_8 = sext <2 x i16> undef to <2 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %rext_9 = zext <2 x i16> undef to <2 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %rext_a = sext <2 x i32> undef to <2 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %rext_b = zext <2 x i32> undef to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 10 for instruction: %rext_6 = sext <2 x i8> undef to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %rext_7 = zext <2 x i8> undef to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 10 for instruction: %rext_8 = sext <2 x i16> undef to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %rext_9 = zext <2 x i16> undef to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %rext_a = sext <2 x i32> undef to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %rext_b = zext <2 x i32> undef to <2 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %tv4i32i8 = trunc <4 x i32> undef to <4 x i8>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %tv4i32i16 = trunc <4 x i32> undef to <4 x i16>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %tv4i16i8 = trunc <4 x i16> undef to <4 x i8>
diff --git a/llvm/test/Analysis/CostModel/ARM/cast_ldst.ll b/llvm/test/Analysis/CostModel/ARM/cast_ldst.ll
index 8b0e7095914f..fd4e1a6c2e6b 100644
--- a/llvm/test/Analysis/CostModel/ARM/cast_ldst.ll
+++ b/llvm/test/Analysis/CostModel/ARM/cast_ldst.ll
@@ -374,8 +374,8 @@ define i32 @load_extends() {
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v2816u = zext <2 x i8> %loadv2i8 to <2 x i16>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v2832s = sext <2 x i8> %loadv2i8 to <2 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v2832u = zext <2 x i8> %loadv2i8 to <2 x i32>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v2864s = sext <2 x i8> %loadv2i8 to <2 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v2864u = zext <2 x i8> %loadv2i8 to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 10 for instruction: %v2864s = sext <2 x i8> %loadv2i8 to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %v2864u = zext <2 x i8> %loadv2i8 to <2 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v4816s = sext <4 x i8> %loadv4i8 to <4 x i16>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v4816u = zext <4 x i8> %loadv4i8 to <4 x i16>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %v4832s = sext <4 x i8> %loadv4i8 to <4 x i32>
@@ -390,14 +390,14 @@ define i32 @load_extends() {
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v8864u = zext <8 x i8> %loadv8i8 to <8 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v16816s = sext <16 x i8> %loadv16i8 to <16 x i16>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v16816u = zext <16 x i8> %loadv16i8 to <16 x i16>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v16832s = sext <16 x i8> %loadv16i8 to <16 x i32>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v16832u = zext <16 x i8> %loadv16i8 to <16 x i32>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %v16832s = sext <16 x i8> %loadv16i8 to <16 x i32>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %v16832u = zext <16 x i8> %loadv16i8 to <16 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v16864s = sext <16 x i8> %loadv16i8 to <16 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v16864u = zext <16 x i8> %loadv16i8 to <16 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v21632s = sext <2 x i16> %loadv2i16 to <2 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v21632u = zext <2 x i16> %loadv2i16 to <2 x i32>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v21664s = sext <2 x i16> %loadv2i16 to <2 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v21664u = zext <2 x i16> %loadv2i16 to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 10 for instruction: %v21664s = sext <2 x i16> %loadv2i16 to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %v21664u = zext <2 x i16> %loadv2i16 to <2 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %v41632s = sext <4 x i16> %loadv4i16 to <4 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %v41632u = zext <4 x i16> %loadv4i16 to <4 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v41664s = sext <4 x i16> %loadv4i16 to <4 x i64>
@@ -406,8 +406,8 @@ define i32 @load_extends() {
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v81632u = zext <8 x i16> %loadv8i16 to <8 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v81664s = sext <8 x i16> %loadv8i16 to <8 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v81664u = zext <8 x i16> %loadv8i16 to <8 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v23264s = sext <2 x i32> %loadv2i32 to <2 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v23264u = zext <2 x i32> %loadv2i32 to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %v23264s = sext <2 x i32> %loadv2i32 to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %v23264u = zext <2 x i32> %loadv2i32 to <2 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v43264s = sext <4 x i32> %loadv4i32 to <4 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v43264u = zext <4 x i32> %loadv4i32 to <4 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
@@ -978,7 +978,7 @@ define i32 @store_trunc() {
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v8832 = trunc <8 x i32> undef to <8 x i8>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v8864 = trunc <8 x i64> undef to <8 x i8>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v16816 = trunc <16 x i16> undef to <16 x i8>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v16832 = trunc <16 x i32> undef to <16 x i8>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %v16832 = trunc <16 x i32> undef to <16 x i8>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v16864 = trunc <16 x i64> undef to <16 x i8>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %v21632 = trunc <2 x i32> undef to <2 x i16>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %v21664 = trunc <2 x i64> undef to <2 x i16>
@@ -1397,7 +1397,7 @@ define i32 @load_fpextends() {
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 10 for instruction: %r3 = fpext float %loadf32 to double
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %v1 = fpext <2 x half> %loadv2f16 to <2 x float>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v2 = fpext <4 x half> %loadv4f16 to <4 x float>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v3 = fpext <8 x half> %loadv8f16 to <8 x float>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %v3 = fpext <8 x half> %loadv8f16 to <8 x float>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %v4 = fpext <16 x half> %loadv16f16 to <16 x float>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 20 for instruction: %v5 = fpext <2 x half> %loadv2f16 to <2 x double>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 40 for instruction: %v6 = fpext <4 x half> %loadv4f16 to <4 x double>
@@ -1660,7 +1660,7 @@ define i32 @load_fptrunc() {
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 20 for instruction: %v21664 = fptrunc <2 x double> undef to <2 x half>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v41632 = fptrunc <4 x float> undef to <4 x half>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 40 for instruction: %v41664 = fptrunc <4 x double> undef to <4 x half>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v81632 = fptrunc <8 x float> undef to <8 x half>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %v81632 = fptrunc <8 x float> undef to <8 x half>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 80 for instruction: %v81664 = fptrunc <8 x double> undef to <8 x half>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 20 for instruction: %v23264 = fptrunc <2 x double> undef to <2 x float>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 40 for instruction: %v43264 = fptrunc <4 x double> undef to <4 x float>
@@ -2044,54 +2044,54 @@ define i32 @maskedload_extends() {
;
; CHECK-MVE-SIZE-LABEL: 'maskedload_extends'
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %loadv2i8 = call <2 x i8> @llvm.masked.load.v2i8.p0v2i8(<2 x i8>* undef, i32 1, <2 x i1> undef, <2 x i8> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %loadv4i8 = call <4 x i8> @llvm.masked.load.v4i8.p0v4i8(<4 x i8>* undef, i32 1, <4 x i1> undef, <4 x i8> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %loadv8i8 = call <8 x i8> @llvm.masked.load.v8i8.p0v8i8(<8 x i8>* undef, i32 1, <8 x i1> undef, <8 x i8> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %loadv16i8 = call <16 x i8> @llvm.masked.load.v16i8.p0v16i8(<16 x i8>* undef, i32 1, <16 x i1> undef, <16 x i8> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %loadv4i8 = call <4 x i8> @llvm.masked.load.v4i8.p0v4i8(<4 x i8>* undef, i32 1, <4 x i1> undef, <4 x i8> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %loadv8i8 = call <8 x i8> @llvm.masked.load.v8i8.p0v8i8(<8 x i8>* undef, i32 1, <8 x i1> undef, <8 x i8> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %loadv16i8 = call <16 x i8> @llvm.masked.load.v16i8.p0v16i8(<16 x i8>* undef, i32 1, <16 x i1> undef, <16 x i8> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %loadv2i16 = call <2 x i16> @llvm.masked.load.v2i16.p0v2i16(<2 x i16>* undef, i32 2, <2 x i1> undef, <2 x i16> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %loadv4i16 = call <4 x i16> @llvm.masked.load.v4i16.p0v4i16(<4 x i16>* undef, i32 2, <4 x i1> undef, <4 x i16> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %loadv8i16 = call <8 x i16> @llvm.masked.load.v8i16.p0v8i16(<8 x i16>* undef, i32 2, <8 x i1> undef, <8 x i16> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %loadv4i16 = call <4 x i16> @llvm.masked.load.v4i16.p0v4i16(<4 x i16>* undef, i32 2, <4 x i1> undef, <4 x i16> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %loadv8i16 = call <8 x i16> @llvm.masked.load.v8i16.p0v8i16(<8 x i16>* undef, i32 2, <8 x i1> undef, <8 x i16> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %loadv2i32 = call <2 x i32> @llvm.masked.load.v2i32.p0v2i32(<2 x i32>* undef, i32 4, <2 x i1> undef, <2 x i32> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %loadv4i32 = call <4 x i32> @llvm.masked.load.v4i32.p0v4i32(<4 x i32>* undef, i32 4, <4 x i1> undef, <4 x i32> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %loadv4i32 = call <4 x i32> @llvm.masked.load.v4i32.p0v4i32(<4 x i32>* undef, i32 4, <4 x i1> undef, <4 x i32> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v2816s = sext <2 x i8> %loadv2i8 to <2 x i16>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v2816u = zext <2 x i8> %loadv2i8 to <2 x i16>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v2832s = sext <2 x i8> %loadv2i8 to <2 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v2832u = zext <2 x i8> %loadv2i8 to <2 x i32>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v2864s = sext <2 x i8> %loadv2i8 to <2 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v2864u = zext <2 x i8> %loadv2i8 to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 10 for instruction: %v2864s = sext <2 x i8> %loadv2i8 to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %v2864u = zext <2 x i8> %loadv2i8 to <2 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v4816s = sext <4 x i8> %loadv4i8 to <4 x i16>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v4816u = zext <4 x i8> %loadv4i8 to <4 x i16>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %v4832s = sext <4 x i8> %loadv4i8 to <4 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %v4832u = zext <4 x i8> %loadv4i8 to <4 x i32>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %v4864s = sext <4 x i8> %loadv4i8 to <4 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %v4864u = zext <4 x i8> %loadv4i8 to <4 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %v4864s = sext <4 x i8> %loadv4i8 to <4 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %v4864u = zext <4 x i8> %loadv4i8 to <4 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %v8816s = sext <8 x i8> %loadv8i8 to <8 x i16>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %v8816u = zext <8 x i8> %loadv8i8 to <8 x i16>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %v8832s = sext <8 x i8> %loadv8i8 to <8 x i32>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %v8832u = zext <8 x i8> %loadv8i8 to <8 x i32>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %v8864s = sext <8 x i8> %loadv8i8 to <8 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %v8864u = zext <8 x i8> %loadv8i8 to <8 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 64 for instruction: %v16816s = sext <16 x i8> %loadv16i8 to <16 x i16>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 64 for instruction: %v16816u = zext <16 x i8> %loadv16i8 to <16 x i16>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 64 for instruction: %v16832s = sext <16 x i8> %loadv16i8 to <16 x i32>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 64 for instruction: %v16832u = zext <16 x i8> %loadv16i8 to <16 x i32>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 64 for instruction: %v16864s = sext <16 x i8> %loadv16i8 to <16 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 64 for instruction: %v16864u = zext <16 x i8> %loadv16i8 to <16 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %v8832s = sext <8 x i8> %loadv8i8 to <8 x i32>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %v8832u = zext <8 x i8> %loadv8i8 to <8 x i32>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %v8864s = sext <8 x i8> %loadv8i8 to <8 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %v8864u = zext <8 x i8> %loadv8i8 to <8 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %v16816s = sext <16 x i8> %loadv16i8 to <16 x i16>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %v16816u = zext <16 x i8> %loadv16i8 to <16 x i16>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %v16832s = sext <16 x i8> %loadv16i8 to <16 x i32>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %v16832u = zext <16 x i8> %loadv16i8 to <16 x i32>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %v16864s = sext <16 x i8> %loadv16i8 to <16 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %v16864u = zext <16 x i8> %loadv16i8 to <16 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v21632s = sext <2 x i16> %loadv2i16 to <2 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v21632u = zext <2 x i16> %loadv2i16 to <2 x i32>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v21664s = sext <2 x i16> %loadv2i16 to <2 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v21664u = zext <2 x i16> %loadv2i16 to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 10 for instruction: %v21664s = sext <2 x i16> %loadv2i16 to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %v21664u = zext <2 x i16> %loadv2i16 to <2 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %v41632s = sext <4 x i16> %loadv4i16 to <4 x i32>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %v41632u = zext <4 x i16> %loadv4i16 to <4 x i32>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %v41664s = sext <4 x i16> %loadv4i16 to <4 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %v41664u = zext <4 x i16> %loadv4i16 to <4 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %v81632s = sext <8 x i16> %loadv8i16 to <8 x i32>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %v81632u = zext <8 x i16> %loadv8i16 to <8 x i32>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %v81664s = sext <8 x i16> %loadv8i16 to <8 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %v81664u = zext <8 x i16> %loadv8i16 to <8 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v23264s = sext <2 x i32> %loadv2i32 to <2 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v23264u = zext <2 x i32> %loadv2i32 to <2 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %v43264s = sext <4 x i32> %loadv4i32 to <4 x i64>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %v43264u = zext <4 x i32> %loadv4i32 to <4 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %v41664s = sext <4 x i16> %loadv4i16 to <4 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %v41664u = zext <4 x i16> %loadv4i16 to <4 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %v81632s = sext <8 x i16> %loadv8i16 to <8 x i32>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %v81632u = zext <8 x i16> %loadv8i16 to <8 x i32>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %v81664s = sext <8 x i16> %loadv8i16 to <8 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %v81664u = zext <8 x i16> %loadv8i16 to <8 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %v23264s = sext <2 x i32> %loadv2i32 to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %v23264u = zext <2 x i32> %loadv2i32 to <2 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %v43264s = sext <4 x i32> %loadv4i32 to <4 x i64>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %v43264u = zext <4 x i32> %loadv4i32 to <4 x i64>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
;
; CHECK-V8M-MAIN-SIZE-LABEL: 'maskedload_extends'
@@ -2533,7 +2533,7 @@ define i32 @maskedstore_trunc() {
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v8832 = trunc <8 x i32> undef to <8 x i8>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v8864 = trunc <8 x i64> undef to <8 x i8>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v16816 = trunc <16 x i16> undef to <16 x i8>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v16832 = trunc <16 x i32> undef to <16 x i8>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %v16832 = trunc <16 x i32> undef to <16 x i8>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v16864 = trunc <16 x i64> undef to <16 x i8>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %v21632 = trunc <2 x i32> undef to <2 x i16>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %v21664 = trunc <2 x i64> undef to <2 x i16>
@@ -2546,23 +2546,23 @@ define i32 @maskedstore_trunc() {
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: call void @llvm.masked.store.v2i8.p0v2i8(<2 x i8> %v2816, <2 x i8>* undef, i32 1, <2 x i1> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: call void @llvm.masked.store.v2i8.p0v2i8(<2 x i8> %v2832, <2 x i8>* undef, i32 1, <2 x i1> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: call void @llvm.masked.store.v2i8.p0v2i8(<2 x i8> %v2864, <2 x i8>* undef, i32 1, <2 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v4i8.p0v4i8(<4 x i8> %v4816, <4 x i8>* undef, i32 1, <4 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v4i8.p0v4i8(<4 x i8> %v4832, <4 x i8>* undef, i32 1, <4 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v4i8.p0v4i8(<4 x i8> %v4864, <4 x i8>* undef, i32 1, <4 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v8i8.p0v8i8(<8 x i8> %v8816, <8 x i8>* undef, i32 1, <8 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v8i8.p0v8i8(<8 x i8> %v8832, <8 x i8>* undef, i32 1, <8 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v8i8.p0v8i8(<8 x i8> %v8864, <8 x i8>* undef, i32 1, <8 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v16i8.p0v16i8(<16 x i8> %v16816, <16 x i8>* undef, i32 1, <16 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v16i8.p0v16i8(<16 x i8> %v16832, <16 x i8>* undef, i32 1, <16 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v16i8.p0v16i8(<16 x i8> %v16864, <16 x i8>* undef, i32 1, <16 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v4i8.p0v4i8(<4 x i8> %v4816, <4 x i8>* undef, i32 1, <4 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v4i8.p0v4i8(<4 x i8> %v4832, <4 x i8>* undef, i32 1, <4 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v4i8.p0v4i8(<4 x i8> %v4864, <4 x i8>* undef, i32 1, <4 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v8i8.p0v8i8(<8 x i8> %v8816, <8 x i8>* undef, i32 1, <8 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v8i8.p0v8i8(<8 x i8> %v8832, <8 x i8>* undef, i32 1, <8 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v8i8.p0v8i8(<8 x i8> %v8864, <8 x i8>* undef, i32 1, <8 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v16i8.p0v16i8(<16 x i8> %v16816, <16 x i8>* undef, i32 1, <16 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v16i8.p0v16i8(<16 x i8> %v16832, <16 x i8>* undef, i32 1, <16 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v16i8.p0v16i8(<16 x i8> %v16864, <16 x i8>* undef, i32 1, <16 x i1> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: call void @llvm.masked.store.v2i16.p0v2i16(<2 x i16> %v21632, <2 x i16>* undef, i32 2, <2 x i1> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: call void @llvm.masked.store.v2i16.p0v2i16(<2 x i16> %v21664, <2 x i16>* undef, i32 2, <2 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v4i16.p0v4i16(<4 x i16> %v41632, <4 x i16>* undef, i32 2, <4 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v4i16.p0v4i16(<4 x i16> %v41664, <4 x i16>* undef, i32 2, <4 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v8i16.p0v8i16(<8 x i16> %v81632, <8 x i16>* undef, i32 2, <8 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v8i16.p0v8i16(<8 x i16> %v81664, <8 x i16>* undef, i32 2, <8 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v4i16.p0v4i16(<4 x i16> %v41632, <4 x i16>* undef, i32 2, <4 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v4i16.p0v4i16(<4 x i16> %v41664, <4 x i16>* undef, i32 2, <4 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v8i16.p0v8i16(<8 x i16> %v81632, <8 x i16>* undef, i32 2, <8 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v8i16.p0v8i16(<8 x i16> %v81664, <8 x i16>* undef, i32 2, <8 x i1> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: call void @llvm.masked.store.v2i32.p0v2i32(<2 x i32> %v23264, <2 x i32>* undef, i32 4, <2 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v4i32.p0v4i32(<4 x i32> %v43264, <4 x i32>* undef, i32 4, <4 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v4i32.p0v4i32(<4 x i32> %v43264, <4 x i32>* undef, i32 4, <4 x i1> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
;
; CHECK-V8M-MAIN-SIZE-LABEL: 'maskedstore_trunc'
@@ -2860,15 +2860,15 @@ define i32 @maskedload_fpextends() {
; CHECK-MVE-SIZE-LABEL: 'maskedload_fpextends'
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %loadv2f16 = call <2 x half> @llvm.masked.load.v2f16.p0v2f16(<2 x half>* undef, i32 2, <2 x i1> undef, <2 x half> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %loadv4f16 = call <4 x half> @llvm.masked.load.v4f16.p0v4f16(<4 x half>* undef, i32 2, <4 x i1> undef, <4 x half> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %loadv8f16 = call <8 x half> @llvm.masked.load.v8f16.p0v8f16(<8 x half>* undef, i32 2, <8 x i1> undef, <8 x half> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %loadv8f16 = call <8 x half> @llvm.masked.load.v8f16.p0v8f16(<8 x half>* undef, i32 2, <8 x i1> undef, <8 x half> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 128 for instruction: %loadv16f16 = call <16 x half> @llvm.masked.load.v16f16.p0v16f16(<16 x half>* undef, i32 2, <16 x i1> undef, <16 x half> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %loadv2f32 = call <2 x float> @llvm.masked.load.v2f32.p0v2f32(<2 x float>* undef, i32 4, <2 x i1> undef, <2 x float> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %loadv4f32 = call <4 x float> @llvm.masked.load.v4f32.p0v4f32(<4 x float>* undef, i32 4, <4 x i1> undef, <4 x float> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %loadv4f32 = call <4 x float> @llvm.masked.load.v4f32.p0v4f32(<4 x float>* undef, i32 4, <4 x i1> undef, <4 x float> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 64 for instruction: %loadv8f32 = call <8 x float> @llvm.masked.load.v8f32.p0v8f32(<8 x float>* undef, i32 4, <8 x i1> undef, <8 x float> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %v1 = fpext <2 x half> %loadv2f16 to <2 x float>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v2 = fpext <4 x half> %loadv4f16 to <4 x float>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %v3 = fpext <8 x half> %loadv8f16 to <8 x float>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 64 for instruction: %v4 = fpext <16 x half> %loadv16f16 to <16 x float>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %v3 = fpext <8 x half> %loadv8f16 to <8 x float>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %v4 = fpext <16 x half> %loadv16f16 to <16 x float>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 20 for instruction: %v5 = fpext <2 x half> %loadv2f16 to <2 x double>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 40 for instruction: %v6 = fpext <4 x half> %loadv4f16 to <4 x double>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 80 for instruction: %v7 = fpext <8 x half> %loadv8f16 to <8 x double>
@@ -3076,7 +3076,7 @@ define i32 @maskedload_fptrunc() {
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 20 for instruction: %v21664 = fptrunc <2 x double> undef to <2 x half>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v41632 = fptrunc <4 x float> undef to <4 x half>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 40 for instruction: %v41664 = fptrunc <4 x double> undef to <4 x half>
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %v81632 = fptrunc <8 x float> undef to <8 x half>
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %v81632 = fptrunc <8 x float> undef to <8 x half>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 80 for instruction: %v81664 = fptrunc <8 x double> undef to <8 x half>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 20 for instruction: %v23264 = fptrunc <2 x double> undef to <2 x float>
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 40 for instruction: %v43264 = fptrunc <4 x double> undef to <4 x float>
@@ -3084,10 +3084,10 @@ define i32 @maskedload_fptrunc() {
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: call void @llvm.masked.store.v2f16.p0v2f16(<2 x half> %v21664, <2 x half>* undef, i32 2, <2 x i1> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: call void @llvm.masked.store.v4f16.p0v4f16(<4 x half> %v41632, <4 x half>* undef, i32 2, <4 x i1> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 32 for instruction: call void @llvm.masked.store.v4f16.p0v4f16(<4 x half> %v41664, <4 x half>* undef, i32 2, <4 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v8f16.p0v8f16(<8 x half> %v81632, <8 x half>* undef, i32 2, <8 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v8f16.p0v8f16(<8 x half> %v81664, <8 x half>* undef, i32 2, <8 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v8f16.p0v8f16(<8 x half> %v81632, <8 x half>* undef, i32 2, <8 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v8f16.p0v8f16(<8 x half> %v81664, <8 x half>* undef, i32 2, <8 x i1> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 16 for instruction: call void @llvm.masked.store.v2f32.p0v2f32(<2 x float> %v23264, <2 x float>* undef, i32 4, <2 x i1> undef)
-; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: call void @llvm.masked.store.v4f32.p0v4f32(<4 x float> %v43264, <4 x float>* undef, i32 4, <4 x i1> undef)
+; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: call void @llvm.masked.store.v4f32.p0v4f32(<4 x float> %v43264, <4 x float>* undef, i32 4, <4 x i1> undef)
; CHECK-MVE-SIZE-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef
;
; CHECK-V8M-MAIN-SIZE-LABEL: 'maskedload_fptrunc'
More information about the llvm-commits
mailing list