Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

ARM64-SVE: GatherVectorWithByteOffsets #103564

Merged
merged 2 commits into from
Jun 18, 2024
Merged

ARM64-SVE: GatherVectorWithByteOffsets #103564

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
abaa66a
ARM64-SVE: GatherVectorWithByteOffsets
a74nh Jun 7, 2024
699cac3
Spaces
a74nh Jun 18, 2024
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
1 change: 1 addition & 0 deletions src/coreclr/jit/gentree.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27476,6 +27476,7 @@ bool GenTreeHWIntrinsic::OperIsMemoryLoad(GenTree** pAddr) const
addr = Op(3);
break;

case NI_Sve_GatherVectorWithByteOffsets:
kunalspathak marked this conversation as resolved.
Show resolved Hide resolved
case NI_Sve_LoadVector:
case NI_Sve_LoadVectorNonTemporal:
case NI_Sve_LoadVector128AndReplicateToVector:
Expand Down
1 change: 1 addition & 0 deletions src/coreclr/jit/hwintrinsic.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1598,6 +1598,7 @@ GenTree* Compiler::impHWIntrinsic(NamedIntrinsic intrinsic,
case NI_Sve_GatherVectorUInt16ZeroExtend:
case NI_Sve_GatherVectorUInt32WithByteOffsetsZeroExtend:
case NI_Sve_GatherVectorUInt32ZeroExtend:
case NI_Sve_GatherVectorWithByteOffsets:
assert(varTypeIsSIMD(op3->TypeGet()));
retNode->AsHWIntrinsic()->SetAuxiliaryJitType(getBaseJitTypeOfSIMDType(sigReader.op3ClsHnd));
break;
Expand Down
21 changes: 21 additions & 0 deletions src/coreclr/jit/hwintrinsiccodegenarm64.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1922,6 +1922,27 @@ void CodeGen::genHWIntrinsic(GenTreeHWIntrinsic* node)
break;
}

case NI_Sve_GatherVectorWithByteOffsets:
{
assert(!varTypeIsSIMD(intrin.op2->gtType));
assert(intrin.numOperands == 3);
emitAttr baseSize = emitActualTypeSize(intrin.baseType);

if (baseSize == EA_4BYTE)
{
// Index is sign or zero extended to 64bits.
opt = varTypeIsUnsigned(node->GetAuxiliaryType()) ? INS_OPTS_SCALABLE_S_UXTW
: INS_OPTS_SCALABLE_S_SXTW;
}
else
{
assert(baseSize == EA_8BYTE);
}

GetEmitter()->emitIns_R_R_R_R(ins, emitSize, targetReg, op1Reg, op2Reg, op3Reg, opt);
break;
}

case NI_Sve_ReverseElement:
// Use non-predicated version explicitly
GetEmitter()->emitIns_R_R(ins, emitSize, targetReg, op1Reg, opt, INS_SCALABLE_OPTS_UNPREDICATED);
Expand Down
1 change: 1 addition & 0 deletions src/coreclr/jit/hwintrinsiclistarm64sve.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -86,6 +86,7 @@ HARDWARE_INTRINSIC(Sve, GatherVectorUInt16WithByteOffsetsZeroExtend,
HARDWARE_INTRINSIC(Sve, GatherVectorUInt16ZeroExtend, -1, -1, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_ld1h, INS_sve_ld1h, INS_sve_ld1h, INS_sve_ld1h, INS_invalid, INS_invalid}, HW_Category_MemoryLoad, HW_Flag_Scalable|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation)
HARDWARE_INTRINSIC(Sve, GatherVectorUInt32WithByteOffsetsZeroExtend, -1, -1, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_ld1w, INS_sve_ld1w, INS_sve_ld1w, INS_sve_ld1w, INS_invalid, INS_invalid}, HW_Category_MemoryLoad, HW_Flag_Scalable|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation)
HARDWARE_INTRINSIC(Sve, GatherVectorUInt32ZeroExtend, -1, -1, false, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_sve_ld1w, INS_sve_ld1w, INS_sve_ld1w, INS_sve_ld1w, INS_invalid, INS_invalid}, HW_Category_MemoryLoad, HW_Flag_Scalable|HW_Flag_SpecialCodeGen|HW_Flag_ExplicitMaskedOperation|HW_Flag_LowMaskedOperation)
HARDWARE_INTRINSIC(Sve, GatherVectorWithByteOffsets, -1, 3, true, {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, GetActiveElementCount, -1, 2, true, {INS_sve_cntp, INS_sve_cntp, INS_sve_cntp, INS_sve_cntp, INS_sve_cntp, INS_sve_cntp, INS_sve_cntp, INS_sve_cntp, INS_sve_cntp, INS_sve_cntp}, HW_Category_SIMD, HW_Flag_Scalable|HW_Flag_BaseTypeFromFirstArg|HW_Flag_ExplicitMaskedOperation)
HARDWARE_INTRINSIC(Sve, LeadingSignCount, -1, -1, false, {INS_sve_cls, INS_invalid, INS_sve_cls, INS_invalid, INS_sve_cls, INS_invalid, INS_sve_cls, INS_invalid, INS_invalid, INS_invalid}, HW_Category_SIMD, HW_Flag_Scalable|HW_Flag_BaseTypeFromFirstArg|HW_Flag_EmbeddedMaskedOperation|HW_Flag_LowMaskedOperation)
HARDWARE_INTRINSIC(Sve, LeadingZeroCount, -1, -1, false, {INS_sve_clz, INS_sve_clz, INS_sve_clz, INS_sve_clz, INS_sve_clz, INS_sve_clz, INS_sve_clz, INS_sve_clz, INS_invalid, INS_invalid}, HW_Category_SIMD, HW_Flag_Scalable|HW_Flag_BaseTypeFromFirstArg|HW_Flag_EmbeddedMaskedOperation|HW_Flag_LowMaskedOperation)
Expand Down
75 changes: 75 additions & 0 deletions ...ries/System.Private.CoreLib/src/System/Runtime/Intrinsics/Arm/Sve.PlatformNotSupported.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2227,6 +2227,81 @@ internal Arm64() { }
public static unsafe Vector<ulong> GatherVectorUInt32ZeroExtend(Vector<ulong> mask, uint* address, Vector<ulong> indices) { throw new PlatformNotSupportedException(); }


/// Unextended load

/// <summary>
/// svfloat64_t svld1_gather_[s64]offset[_f64](svbool_t pg, const float64_t *base, svint64_t offsets)
/// LD1D Zresult.D, Pg/Z, [Xbase, Zoffsets.D]
/// </summary>
public static unsafe Vector<double> GatherVectorWithByteOffsets(Vector<double> mask, double* address, Vector<long> offsets) { throw new PlatformNotSupportedException(); }

/// <summary>
/// svfloat64_t svld1_gather_[u64]offset[_f64](svbool_t pg, const float64_t *base, svuint64_t offsets)
/// LD1D Zresult.D, Pg/Z, [Xbase, Zoffsets.D]
/// </summary>
public static unsafe Vector<double> GatherVectorWithByteOffsets(Vector<double> mask, double* address, Vector<ulong> offsets) { throw new PlatformNotSupportedException(); }

/// <summary>
/// svint32_t svld1_gather_[s32]offset[_s32](svbool_t pg, const int32_t *base, svint32_t offsets)
/// LD1W Zresult.S, Pg/Z, [Xbase, Zoffsets.S, SXTW]
/// </summary>
public static unsafe Vector<int> GatherVectorWithByteOffsets(Vector<int> mask, int* address, Vector<int> offsets) { throw new PlatformNotSupportedException(); }

/// <summary>
/// svint32_t svld1_gather_[u32]offset[_s32](svbool_t pg, const int32_t *base, svuint32_t offsets)
/// LD1W Zresult.S, Pg/Z, [Xbase, Zoffsets.S, UXTW]
/// </summary>
public static unsafe Vector<int> GatherVectorWithByteOffsets(Vector<int> mask, int* address, Vector<uint> offsets) { throw new PlatformNotSupportedException(); }

/// <summary>
/// svint64_t svld1_gather_[s64]offset[_s64](svbool_t pg, const int64_t *base, svint64_t offsets)
/// LD1D Zresult.D, Pg/Z, [Xbase, Zoffsets.D]
/// </summary>
public static unsafe Vector<long> GatherVectorWithByteOffsets(Vector<long> mask, long* address, Vector<long> offsets) { throw new PlatformNotSupportedException(); }

/// <summary>
/// svint64_t svld1_gather_[u64]offset[_s64](svbool_t pg, const int64_t *base, svuint64_t offsets)
/// LD1D Zresult.D, Pg/Z, [Xbase, Zoffsets.D]
/// </summary>
public static unsafe Vector<long> GatherVectorWithByteOffsets(Vector<long> mask, long* address, Vector<ulong> offsets) { throw new PlatformNotSupportedException(); }

/// <summary>
/// svfloat32_t svld1_gather_[s32]offset[_f32](svbool_t pg, const float32_t *base, svint32_t offsets)
/// LD1W Zresult.S, Pg/Z, [Xbase, Zoffsets.S, SXTW]
/// </summary>
public static unsafe Vector<float> GatherVectorWithByteOffsets(Vector<float> mask, float* address, Vector<int> offsets) { throw new PlatformNotSupportedException(); }

/// <summary>
/// svfloat32_t svld1_gather_[u32]offset[_f32](svbool_t pg, const float32_t *base, svuint32_t offsets)
/// LD1W Zresult.S, Pg/Z, [Xbase, Zoffsets.S, UXTW]
/// </summary>
public static unsafe Vector<float> GatherVectorWithByteOffsets(Vector<float> mask, float* address, Vector<uint> offsets) { throw new PlatformNotSupportedException(); }

/// <summary>
/// svuint32_t svld1_gather_[s32]offset[_u32](svbool_t pg, const uint32_t *base, svint32_t offsets)
/// LD1W Zresult.S, Pg/Z, [Xbase, Zoffsets.S, SXTW]
/// </summary>
public static unsafe Vector<uint> GatherVectorWithByteOffsets(Vector<uint> mask, uint* address, Vector<int> offsets) { throw new PlatformNotSupportedException(); }

/// <summary>
/// svuint32_t svld1_gather_[u32]offset[_u32](svbool_t pg, const uint32_t *base, svuint32_t offsets)
/// LD1W Zresult.S, Pg/Z, [Xbase, Zoffsets.S, UXTW]
/// </summary>
public static unsafe Vector<uint> GatherVectorWithByteOffsets(Vector<uint> mask, uint* address, Vector<uint> offsets) { throw new PlatformNotSupportedException(); }

/// <summary>
/// svuint64_t svld1_gather_[s64]offset[_u64](svbool_t pg, const uint64_t *base, svint64_t offsets)
/// LD1D Zresult.D, Pg/Z, [Xbase, Zoffsets.D]
/// </summary>
public static unsafe Vector<ulong> GatherVectorWithByteOffsets(Vector<ulong> mask, ulong* address, Vector<long> offsets) { throw new PlatformNotSupportedException(); }

/// <summary>
/// svuint64_t svld1_gather_[u64]offset[_u64](svbool_t pg, const uint64_t *base, svuint64_t offsets)
/// LD1D Zresult.D, Pg/Z, [Xbase, Zoffsets.D]
/// </summary>
public static unsafe Vector<ulong> GatherVectorWithByteOffsets(Vector<ulong> mask, ulong* address, Vector<ulong> offsets) { throw new PlatformNotSupportedException(); }


/// Count set predicate bits

/// <summary>
Expand Down
75 changes: 75 additions & 0 deletions src/libraries/System.Private.CoreLib/src/System/Runtime/Intrinsics/Arm/Sve.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2283,6 +2283,81 @@ internal Arm64() { }
public static unsafe Vector<ulong> GatherVectorUInt32ZeroExtend(Vector<ulong> mask, uint* address, Vector<ulong> indices) => GatherVectorUInt32ZeroExtend(mask, address, indices);


/// Unextended load

/// <summary>
/// svfloat64_t svld1_gather_[s64]offset[_f64](svbool_t pg, const float64_t *base, svint64_t offsets)
/// LD1D Zresult.D, Pg/Z, [Xbase, Zoffsets.D]
/// </summary>
public static unsafe Vector<double> GatherVectorWithByteOffsets(Vector<double> mask, double* address, Vector<long> offsets) => GatherVectorWithByteOffsets(mask, address, offsets);

/// <summary>
/// svfloat64_t svld1_gather_[u64]offset[_f64](svbool_t pg, const float64_t *base, svuint64_t offsets)
/// LD1D Zresult.D, Pg/Z, [Xbase, Zoffsets.D]
/// </summary>
public static unsafe Vector<double> GatherVectorWithByteOffsets(Vector<double> mask, double* address, Vector<ulong> offsets) => GatherVectorWithByteOffsets(mask, address, offsets);

/// <summary>
/// svint32_t svld1_gather_[s32]offset[_s32](svbool_t pg, const int32_t *base, svint32_t offsets)
/// LD1W Zresult.S, Pg/Z, [Xbase, Zoffsets.S, SXTW]
/// </summary>
public static unsafe Vector<int> GatherVectorWithByteOffsets(Vector<int> mask, int* address, Vector<int> offsets) => GatherVectorWithByteOffsets(mask, address, offsets);

/// <summary>
/// svint32_t svld1_gather_[u32]offset[_s32](svbool_t pg, const int32_t *base, svuint32_t offsets)
/// LD1W Zresult.S, Pg/Z, [Xbase, Zoffsets.S, UXTW]
/// </summary>
public static unsafe Vector<int> GatherVectorWithByteOffsets(Vector<int> mask, int* address, Vector<uint> offsets) => GatherVectorWithByteOffsets(mask, address, offsets);

/// <summary>
/// svint64_t svld1_gather_[s64]offset[_s64](svbool_t pg, const int64_t *base, svint64_t offsets)
/// LD1D Zresult.D, Pg/Z, [Xbase, Zoffsets.D]
/// </summary>
public static unsafe Vector<long> GatherVectorWithByteOffsets(Vector<long> mask, long* address, Vector<long> offsets) => GatherVectorWithByteOffsets(mask, address, offsets);

/// <summary>
/// svint64_t svld1_gather_[u64]offset[_s64](svbool_t pg, const int64_t *base, svuint64_t offsets)
/// LD1D Zresult.D, Pg/Z, [Xbase, Zoffsets.D]
/// </summary>
public static unsafe Vector<long> GatherVectorWithByteOffsets(Vector<long> mask, long* address, Vector<ulong> offsets) => GatherVectorWithByteOffsets(mask, address, offsets);

/// <summary>
/// svfloat32_t svld1_gather_[s32]offset[_f32](svbool_t pg, const float32_t *base, svint32_t offsets)
/// LD1W Zresult.S, Pg/Z, [Xbase, Zoffsets.S, SXTW]
/// </summary>
public static unsafe Vector<float> GatherVectorWithByteOffsets(Vector<float> mask, float* address, Vector<int> offsets) => GatherVectorWithByteOffsets(mask, address, offsets);

/// <summary>
/// svfloat32_t svld1_gather_[u32]offset[_f32](svbool_t pg, const float32_t *base, svuint32_t offsets)
/// LD1W Zresult.S, Pg/Z, [Xbase, Zoffsets.S, UXTW]
/// </summary>
public static unsafe Vector<float> GatherVectorWithByteOffsets(Vector<float> mask, float* address, Vector<uint> offsets) => GatherVectorWithByteOffsets(mask, address, offsets);

/// <summary>
/// svuint32_t svld1_gather_[s32]offset[_u32](svbool_t pg, const uint32_t *base, svint32_t offsets)
/// LD1W Zresult.S, Pg/Z, [Xbase, Zoffsets.S, SXTW]
/// </summary>
public static unsafe Vector<uint> GatherVectorWithByteOffsets(Vector<uint> mask, uint* address, Vector<int> offsets) => GatherVectorWithByteOffsets(mask, address, offsets);

/// <summary>
/// svuint32_t svld1_gather_[u32]offset[_u32](svbool_t pg, const uint32_t *base, svuint32_t offsets)
/// LD1W Zresult.S, Pg/Z, [Xbase, Zoffsets.S, UXTW]
/// </summary>
public static unsafe Vector<uint> GatherVectorWithByteOffsets(Vector<uint> mask, uint* address, Vector<uint> offsets) => GatherVectorWithByteOffsets(mask, address, offsets);

/// <summary>
/// svuint64_t svld1_gather_[s64]offset[_u64](svbool_t pg, const uint64_t *base, svint64_t offsets)
/// LD1D Zresult.D, Pg/Z, [Xbase, Zoffsets.D]
/// </summary>
public static unsafe Vector<ulong> GatherVectorWithByteOffsets(Vector<ulong> mask, ulong* address, Vector<long> offsets) => GatherVectorWithByteOffsets(mask, address, offsets);

/// <summary>
/// svuint64_t svld1_gather_[u64]offset[_u64](svbool_t pg, const uint64_t *base, svuint64_t offsets)
/// LD1D Zresult.D, Pg/Z, [Xbase, Zoffsets.D]
/// </summary>
public static unsafe Vector<ulong> GatherVectorWithByteOffsets(Vector<ulong> mask, ulong* address, Vector<ulong> offsets) => GatherVectorWithByteOffsets(mask, address, offsets);


/// Count set predicate bits

/// <summary>
Expand Down
13 changes: 13 additions & 0 deletions src/libraries/System.Runtime.Intrinsics/ref/System.Runtime.Intrinsics.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4517,6 +4517,19 @@ internal Arm64() { }
public static unsafe System.Numerics.Vector<ulong> GatherVectorUInt32ZeroExtend(System.Numerics.Vector<ulong> mask, uint* address, System.Numerics.Vector<long> indices) { throw null; }
public static System.Numerics.Vector<ulong> GatherVectorUInt32ZeroExtend(System.Numerics.Vector<ulong> mask, System.Numerics.Vector<ulong> addresses) { throw null; }
public static unsafe System.Numerics.Vector<ulong> GatherVectorUInt32ZeroExtend(System.Numerics.Vector<ulong> mask, uint* address, System.Numerics.Vector<ulong> indices) { throw null; }
public static unsafe System.Numerics.Vector<double> GatherVectorWithByteOffsets(System.Numerics.Vector<double> mask, double* address, System.Numerics.Vector<long> offsets) { throw null; }
a74nh marked this conversation as resolved.
Show resolved Hide resolved
public static unsafe System.Numerics.Vector<double> GatherVectorWithByteOffsets(System.Numerics.Vector<double> mask, double* address, System.Numerics.Vector<ulong> offsets) { throw null; }

public static unsafe System.Numerics.Vector<int> GatherVectorWithByteOffsets(System.Numerics.Vector<int> mask, int* address, System.Numerics.Vector<int> offsets) { throw null; }
public static unsafe System.Numerics.Vector<int> GatherVectorWithByteOffsets(System.Numerics.Vector<int> mask, int* address, System.Numerics.Vector<uint> offsets) { throw null; }
public static unsafe System.Numerics.Vector<long> GatherVectorWithByteOffsets(System.Numerics.Vector<long> mask, long* address, System.Numerics.Vector<long> offsets) { throw null; }
public static unsafe System.Numerics.Vector<long> GatherVectorWithByteOffsets(System.Numerics.Vector<long> mask, long* address, System.Numerics.Vector<ulong> offsets) { throw null; }
public static unsafe System.Numerics.Vector<float> GatherVectorWithByteOffsets(System.Numerics.Vector<float> mask, float* address, System.Numerics.Vector<int> offsets) { throw null; }
public static unsafe System.Numerics.Vector<float> GatherVectorWithByteOffsets(System.Numerics.Vector<float> mask, float* address, System.Numerics.Vector<uint> offsets) { throw null; }
public static unsafe System.Numerics.Vector<uint> GatherVectorWithByteOffsets(System.Numerics.Vector<uint> mask, uint* address, System.Numerics.Vector<int> offsets) { throw null; }
public static unsafe System.Numerics.Vector<uint> GatherVectorWithByteOffsets(System.Numerics.Vector<uint> mask, uint* address, System.Numerics.Vector<uint> offsets) { throw null; }
public static unsafe System.Numerics.Vector<ulong> GatherVectorWithByteOffsets(System.Numerics.Vector<ulong> mask, ulong* address, System.Numerics.Vector<long> offsets) { throw null; }
public static unsafe System.Numerics.Vector<ulong> GatherVectorWithByteOffsets(System.Numerics.Vector<ulong> mask, ulong* address, System.Numerics.Vector<ulong> offsets) { throw null; }

public static ulong GetActiveElementCount(System.Numerics.Vector<byte> mask, System.Numerics.Vector<byte> from) { throw null; }
public static ulong GetActiveElementCount(System.Numerics.Vector<double> mask, System.Numerics.Vector<double> from) { throw null; }
Expand Down
Loading
Loading