Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion src/coreclr/jit/hwintrinsicarm64.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -2784,7 +2784,7 @@ GenTree* Compiler::impSpecialIntrinsic(NamedIntrinsic intrinsic,
op4 = getArgForHWIntrinsic(argType, argClass);

assert(HWIntrinsicInfo::isImmOp(intrinsic, op4));
op3 = addRangeCheckIfNeeded(intrinsic, op4, mustExpand, immLowerBound, immUpperBound);
op4 = addRangeCheckIfNeeded(intrinsic, op4, mustExpand, immLowerBound, immUpperBound);

argType = JITtype2varType(strip(info.compCompHnd->getArgType(sig, arg3, &argClass)));
op3 = getArgForHWIntrinsic(argType, argClass);
Expand Down
10 changes: 5 additions & 5 deletions src/coreclr/jit/hwintrinsiccodegenarm64.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -308,18 +308,18 @@ void CodeGen::genHWIntrinsic(GenTreeHWIntrinsic* node)
emitSize = emitActualTypeSize(intrin.baseType);
opt = INS_OPTS_NONE;
}
else if (HWIntrinsicInfo::IsScalable(intrin.id))
{
emitSize = EA_SCALABLE;
opt = emitter::optGetSveInsOpt(emitTypeSize(intrin.baseType));
}
else if (intrin.category == HW_Category_Special)
{
assert(intrin.id == NI_ArmBase_Yield);

emitSize = EA_UNKNOWN;
opt = INS_OPTS_NONE;
}
else if (HWIntrinsicInfo::IsScalable(intrin.id))
{
emitSize = EA_SCALABLE;
opt = emitter::optGetSveInsOpt(emitTypeSize(intrin.baseType));
}
else
{
emitSize = emitActualTypeSize(Compiler::getSIMDTypeForSize(node->GetSimdSize()));
Expand Down
16 changes: 8 additions & 8 deletions src/coreclr/jit/hwintrinsiclistarm64sve.h
Original file line number Diff line number Diff line change
Expand Up @@ -84,10 +84,10 @@ HARDWARE_INTRINSIC(Sve, FusedMultiplyAddNegated,
HARDWARE_INTRINSIC(Sve, FusedMultiplySubtract, -1, -1, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_fmls, INS_sve_fmls}, HW_Category_SIMD, HW_Flag_Scalable|HW_Flag_EmbeddedMaskedOperation|HW_Flag_HasRMWSemantics|HW_Flag_LowMaskedOperation|HW_Flag_FmaIntrinsic|HW_Flag_SpecialCodeGen)
HARDWARE_INTRINSIC(Sve, FusedMultiplySubtractBySelectedScalar, -1, 4, true, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_fmls, INS_sve_fmls}, HW_Category_SIMDByIndexedElement, HW_Flag_Scalable|HW_Flag_HasImmediateOperand|HW_Flag_HasRMWSemantics|HW_Flag_FmaIntrinsic|HW_Flag_LowVectorOperation)
HARDWARE_INTRINSIC(Sve, FusedMultiplySubtractNegated, -1, -1, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_fnmls, INS_sve_fnmls}, HW_Category_SIMD, HW_Flag_Scalable|HW_Flag_EmbeddedMaskedOperation|HW_Flag_HasRMWSemantics|HW_Flag_LowMaskedOperation|HW_Flag_FmaIntrinsic|HW_Flag_SpecialCodeGen)
HARDWARE_INTRINSIC(Sve, GatherPrefetch16Bit, -1, -1, false, {INS_invalid, INS_invalid, INS_sve_prfh, INS_sve_prfh, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_Special, HW_Flag_Scalable|HW_Flag_BaseTypeFromFirstArg|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation|HW_Flag_HasImmediateOperand|HW_Flag_HasEnumOperand)
HARDWARE_INTRINSIC(Sve, GatherPrefetch32Bit, -1, -1, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_prfw, INS_sve_prfw, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_Special, HW_Flag_Scalable|HW_Flag_BaseTypeFromFirstArg|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation|HW_Flag_HasImmediateOperand|HW_Flag_HasEnumOperand)
HARDWARE_INTRINSIC(Sve, GatherPrefetch64Bit, -1, -1, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_prfd, INS_sve_prfd, INS_invalid, INS_invalid}, HW_Category_Special, HW_Flag_Scalable|HW_Flag_BaseTypeFromFirstArg|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation|HW_Flag_HasImmediateOperand|HW_Flag_HasEnumOperand)
HARDWARE_INTRINSIC(Sve, GatherPrefetch8Bit, -1, -1, false, {INS_sve_prfb, INS_sve_prfb, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_Special, HW_Flag_Scalable|HW_Flag_BaseTypeFromFirstArg|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation|HW_Flag_HasImmediateOperand|HW_Flag_HasEnumOperand)
HARDWARE_INTRINSIC(Sve, GatherPrefetch16Bit, -1, -1, false, {INS_invalid, INS_invalid, INS_sve_prfh, INS_sve_prfh, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_Special, HW_Flag_Scalable|HW_Flag_BaseTypeFromFirstArg|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation|HW_Flag_HasImmediateOperand|HW_Flag_HasEnumOperand|HW_Flag_SpecialSideEffect_Other)
HARDWARE_INTRINSIC(Sve, GatherPrefetch32Bit, -1, -1, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_prfw, INS_sve_prfw, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_Special, HW_Flag_Scalable|HW_Flag_BaseTypeFromFirstArg|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation|HW_Flag_HasImmediateOperand|HW_Flag_HasEnumOperand|HW_Flag_SpecialSideEffect_Other)
HARDWARE_INTRINSIC(Sve, GatherPrefetch64Bit, -1, -1, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_prfd, INS_sve_prfd, INS_invalid, INS_invalid}, HW_Category_Special, HW_Flag_Scalable|HW_Flag_BaseTypeFromFirstArg|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation|HW_Flag_HasImmediateOperand|HW_Flag_HasEnumOperand|HW_Flag_SpecialSideEffect_Other)
HARDWARE_INTRINSIC(Sve, GatherPrefetch8Bit, -1, -1, false, {INS_sve_prfb, INS_sve_prfb, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_Special, HW_Flag_Scalable|HW_Flag_BaseTypeFromFirstArg|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation|HW_Flag_HasImmediateOperand|HW_Flag_HasEnumOperand|HW_Flag_SpecialSideEffect_Other)
HARDWARE_INTRINSIC(Sve, GatherVector, -1, -1, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_ld1w, INS_sve_ld1w, INS_sve_ld1d, INS_sve_ld1d, INS_sve_ld1w, INS_sve_ld1d}, HW_Category_MemoryLoad, HW_Flag_Scalable|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation)
HARDWARE_INTRINSIC(Sve, GatherVectorByteZeroExtend, -1, -1, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_ld1b, INS_sve_ld1b, INS_sve_ld1b, INS_sve_ld1b, INS_invalid, INS_invalid}, HW_Category_MemoryLoad, HW_Flag_Scalable|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation)
HARDWARE_INTRINSIC(Sve, GatherVectorInt16SignExtend, -1, -1, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_ld1sh, INS_sve_ld1sh, INS_sve_ld1sh, INS_sve_ld1sh, INS_invalid, INS_invalid}, HW_Category_MemoryLoad, HW_Flag_Scalable|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation)
Expand Down Expand Up @@ -177,10 +177,10 @@ HARDWARE_INTRINSIC(Sve, Not,
HARDWARE_INTRINSIC(Sve, Or, -1, -1, false, {INS_sve_orr, INS_sve_orr, INS_sve_orr, INS_sve_orr, INS_sve_orr, INS_sve_orr, INS_sve_orr, INS_sve_orr, INS_invalid, INS_invalid}, HW_Category_SIMD, HW_Flag_Scalable|HW_Flag_OptionalEmbeddedMaskedOperation|HW_Flag_HasRMWSemantics|HW_Flag_LowMaskedOperation)
HARDWARE_INTRINSIC(Sve, OrAcross, -1, -1, false, {INS_sve_orv, INS_sve_orv, INS_sve_orv, INS_sve_orv, INS_sve_orv, INS_sve_orv, INS_sve_orv, INS_sve_orv, INS_invalid, INS_invalid}, HW_Category_SIMD, HW_Flag_Scalable|HW_Flag_EmbeddedMaskedOperation|HW_Flag_LowMaskedOperation)
HARDWARE_INTRINSIC(Sve, PopCount, -1, -1, false, {INS_sve_cnt, INS_sve_cnt, INS_sve_cnt, INS_sve_cnt, INS_sve_cnt, INS_sve_cnt, INS_sve_cnt, INS_sve_cnt, INS_sve_cnt, INS_sve_cnt}, HW_Category_SIMD, HW_Flag_Scalable|HW_Flag_BaseTypeFromFirstArg|HW_Flag_EmbeddedMaskedOperation|HW_Flag_LowMaskedOperation)
HARDWARE_INTRINSIC(Sve, PrefetchBytes, -1, 3, false, {INS_invalid, INS_sve_prfb, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_Special, HW_Flag_Scalable|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation|HW_Flag_BaseTypeFromFirstArg|HW_Flag_HasImmediateOperand|HW_Flag_HasEnumOperand)
HARDWARE_INTRINSIC(Sve, PrefetchInt16, -1, 3, false, {INS_invalid, INS_invalid, INS_invalid, INS_sve_prfh, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_Special, HW_Flag_Scalable|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation|HW_Flag_BaseTypeFromFirstArg|HW_Flag_HasImmediateOperand|HW_Flag_HasEnumOperand)
HARDWARE_INTRINSIC(Sve, PrefetchInt32, -1, 3, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_prfw, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_Special, HW_Flag_Scalable|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation|HW_Flag_BaseTypeFromFirstArg|HW_Flag_HasImmediateOperand|HW_Flag_HasEnumOperand)
HARDWARE_INTRINSIC(Sve, PrefetchInt64, -1, 3, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_prfd, INS_invalid, INS_invalid}, HW_Category_Special, HW_Flag_Scalable|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation|HW_Flag_BaseTypeFromFirstArg|HW_Flag_HasImmediateOperand|HW_Flag_HasEnumOperand)
HARDWARE_INTRINSIC(Sve, PrefetchBytes, -1, 3, false, {INS_invalid, INS_sve_prfb, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_Special, HW_Flag_Scalable|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation|HW_Flag_BaseTypeFromFirstArg|HW_Flag_HasImmediateOperand|HW_Flag_HasEnumOperand|HW_Flag_SpecialSideEffect_Other)
HARDWARE_INTRINSIC(Sve, PrefetchInt16, -1, 3, false, {INS_invalid, INS_invalid, INS_invalid, INS_sve_prfh, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_Special, HW_Flag_Scalable|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation|HW_Flag_BaseTypeFromFirstArg|HW_Flag_HasImmediateOperand|HW_Flag_HasEnumOperand|HW_Flag_SpecialSideEffect_Other)
HARDWARE_INTRINSIC(Sve, PrefetchInt32, -1, 3, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_prfw, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_Special, HW_Flag_Scalable|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation|HW_Flag_BaseTypeFromFirstArg|HW_Flag_HasImmediateOperand|HW_Flag_HasEnumOperand|HW_Flag_SpecialSideEffect_Other)
HARDWARE_INTRINSIC(Sve, PrefetchInt64, -1, 3, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_prfd, INS_invalid, INS_invalid}, HW_Category_Special, HW_Flag_Scalable|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation|HW_Flag_BaseTypeFromFirstArg|HW_Flag_HasImmediateOperand|HW_Flag_HasEnumOperand|HW_Flag_SpecialSideEffect_Other)
HARDWARE_INTRINSIC(Sve, ReciprocalEstimate, -1, 1, true, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_frecpe, INS_sve_frecpe}, HW_Category_SIMD, HW_Flag_Scalable)
HARDWARE_INTRINSIC(Sve, ReciprocalExponent, -1, -1, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_frecpx, INS_sve_frecpx}, HW_Category_SIMD, HW_Flag_Scalable|HW_Flag_EmbeddedMaskedOperation|HW_Flag_LowMaskedOperation)
HARDWARE_INTRINSIC(Sve, ReciprocalSqrtEstimate, -1, 1, true, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_frsqrte, INS_sve_frsqrte}, HW_Category_SIMD, HW_Flag_Scalable)
Expand Down
18 changes: 10 additions & 8 deletions src/coreclr/jit/lsraarm64.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -2003,18 +2003,10 @@ int LinearScan::BuildHWIntrinsic(GenTreeHWIntrinsic* intrinsicTree, int* pDstCou
case NI_Sve_LoadVectorNonTemporal:
case NI_Sve_LoadVector128AndReplicateToVector:
case NI_Sve_StoreAndZip:
case NI_Sve_PrefetchBytes:
case NI_Sve_PrefetchInt16:
case NI_Sve_PrefetchInt32:
case NI_Sve_PrefetchInt64:
assert(intrinsicTree->OperIsMemoryLoadOrStore());
srcCount += BuildAddrUses(intrin.op2);
break;

case NI_Sve_GatherPrefetch8Bit:
case NI_Sve_GatherPrefetch16Bit:
case NI_Sve_GatherPrefetch32Bit:
case NI_Sve_GatherPrefetch64Bit:
case NI_Sve_GatherVector:
case NI_Sve_GatherVectorByteZeroExtend:
case NI_Sve_GatherVectorInt16SignExtend:
Expand All @@ -2027,6 +2019,16 @@ int LinearScan::BuildHWIntrinsic(GenTreeHWIntrinsic* intrinsicTree, int* pDstCou
case NI_Sve_GatherVectorUInt32WithByteOffsetsZeroExtend:
case NI_Sve_GatherVectorUInt32ZeroExtend:
assert(intrinsicTree->OperIsMemoryLoadOrStore());
FALLTHROUGH;

case NI_Sve_PrefetchBytes:
case NI_Sve_PrefetchInt16:
case NI_Sve_PrefetchInt32:
case NI_Sve_PrefetchInt64:
case NI_Sve_GatherPrefetch8Bit:
case NI_Sve_GatherPrefetch16Bit:
case NI_Sve_GatherPrefetch32Bit:
case NI_Sve_GatherPrefetch64Bit:
if (!varTypeIsSIMD(intrin.op2->gtType))
{
srcCount += BuildAddrUses(intrin.op2);
Expand Down