summaryrefslogtreecommitdiff
path: root/src/jit/hwintrinsiclistxarch.h
diff options
context:
space:
mode:
authorTanner Gooding <tagoo@outlook.com>2018-02-25 09:27:18 -0800
committerTanner Gooding <tagoo@outlook.com>2018-02-28 07:19:27 -0800
commitc23b4d6780095ac1e58f785e8f635d94120e1f64 (patch)
tree9e3673e005498bd34b080280a74cb70567dbf2a1 /src/jit/hwintrinsiclistxarch.h
parent7207e1b422de2d2ad451605d5753a7e7b1883a08 (diff)
downloadcoreclr-c23b4d6780095ac1e58f785e8f635d94120e1f64.tar.gz
coreclr-c23b4d6780095ac1e58f785e8f635d94120e1f64.tar.bz2
coreclr-c23b4d6780095ac1e58f785e8f635d94120e1f64.zip
Adding partial support for the SSE41 hardware intrinsics
Diffstat (limited to 'src/jit/hwintrinsiclistxarch.h')
-rw-r--r--src/jit/hwintrinsiclistxarch.h45
1 files changed, 40 insertions, 5 deletions
diff --git a/src/jit/hwintrinsiclistxarch.h b/src/jit/hwintrinsiclistxarch.h
index 0c28863f17..be9d031136 100644
--- a/src/jit/hwintrinsiclistxarch.h
+++ b/src/jit/hwintrinsiclistxarch.h
@@ -261,12 +261,47 @@ HARDWARE_INTRINSIC(SSSE3_MultiplyHighRoundScale, "MultiplyHi
HARDWARE_INTRINSIC(SSSE3_Shuffle, "Shuffle", SSSE3, -1, 16, 2, {INS_pshufb, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_NoFlag)
HARDWARE_INTRINSIC(SSSE3_Sign, "Sign", SSSE3, -1, 16, 2, {INS_psignb, INS_invalid, INS_psignw, INS_invalid, INS_psignd, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_NoFlag)
+// ************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************
+// Intrinsic ID Function name ISA ival SIMD size NumArg instructions Category Flags
+// {TYP_BYTE, TYP_UBYTE, TYP_SHORT, TYP_USHORT, TYP_INT, TYP_UINT, TYP_LONG, TYP_ULONG, TYP_FLOAT, TYP_DOUBLE}
+// ************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************
// SSE41 Intrinsics
-HARDWARE_INTRINSIC(SSE41_IsSupported, "get_IsSupported", SSE41, -1, 0, 0, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_IsSupportedProperty, HW_Flag_NoFlag)
-HARDWARE_INTRINSIC(SSE41_BlendVariable, "BlendVariable", SSE41, -1, 16, 3, {INS_pblendvb, INS_pblendvb, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_blendvps, INS_blendvpd}, HW_Category_SimpleSIMD, HW_Flag_NoFlag)
-HARDWARE_INTRINSIC(SSE41_CompareEqual, "CompareEqual", SSE41, -1, 16, 2, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_pcmpeqq, INS_pcmpeqq, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_Commutative)
-HARDWARE_INTRINSIC(SSE41_LoadAlignedVector128NonTemporal, "LoadAlignedVector128NonTemporal", SSE41, -1, 16, 1, {INS_movntdqa, INS_movntdqa, INS_movntdqa, INS_movntdqa, INS_movntdqa, INS_movntdqa, INS_movntdqa, INS_movntdqa, INS_invalid, INS_invalid}, HW_Category_MemoryLoad, HW_Flag_NoFlag)
-HARDWARE_INTRINSIC(SSE41_Multiply, "Multiply", SSE41, -1, 16, 2, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_pmuldq, INS_invalid, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_Commutative)
+HARDWARE_INTRINSIC(SSE41_IsSupported, "get_IsSupported", SSE41, -1, 0, 0, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_IsSupportedProperty, HW_Flag_NoFlag)
+HARDWARE_INTRINSIC(SSE41_Blend, "Blend", SSE41, -1, 16, 3, {INS_invalid, INS_invalid, INS_pblendw, INS_pblendw, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_blendps, INS_blendpd}, HW_Category_IMM, HW_Flag_FullRangeIMM)
+HARDWARE_INTRINSIC(SSE41_BlendVariable, "BlendVariable", SSE41, -1, 16, 3, {INS_pblendvb, INS_pblendvb, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_blendvps, INS_blendvpd}, HW_Category_SimpleSIMD, HW_Flag_NoFlag)
+HARDWARE_INTRINSIC(SSE41_Ceiling, "Ceiling", SSE41, 10, 16, 1, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_roundps, INS_roundpd}, HW_Category_SimpleSIMD, HW_Flag_NoFlag)
+HARDWARE_INTRINSIC(SSE41_CeilingScalar, "CeilingScalar", SSE41, 10, 16, -1, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_roundss, INS_roundsd}, HW_Category_Special, HW_Flag_CopyUpperBits)
+HARDWARE_INTRINSIC(SSE41_CompareEqual, "CompareEqual", SSE41, -1, 16, 2, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_pcmpeqq, INS_pcmpeqq, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_Commutative)
+HARDWARE_INTRINSIC(SSE41_ConvertToVector128Int16, "ConvertToVector128Int16", SSE41, -1, 16, 1, {INS_pmovsxbw, INS_pmovzxbw, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_NoContainment|HW_Flag_BaseTypeFromArg)
+HARDWARE_INTRINSIC(SSE41_ConvertToVector128Int32, "ConvertToVector128Int32", SSE41, -1, 16, 1, {INS_pmovsxbd, INS_pmovzxbd, INS_pmovsxwd, INS_pmovzxwd, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_NoContainment|HW_Flag_BaseTypeFromArg)
+HARDWARE_INTRINSIC(SSE41_ConvertToVector128Int64, "ConvertToVector128Int64", SSE41, -1, 16, 1, {INS_pmovsxbq, INS_pmovzxbq, INS_pmovsxwq, INS_pmovzxwq, INS_pmovsxdq, INS_pmovzxdq, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_NoContainment|HW_Flag_BaseTypeFromArg)
+HARDWARE_INTRINSIC(SSE41_DotProduct, "DotProduct", SSE41, -1, 16, 3, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_dpps, INS_dppd}, HW_Category_IMM, HW_Flag_FullRangeIMM)
+HARDWARE_INTRINSIC(SSE41_Floor, "Floor", SSE41, 9, 16, 1, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_roundps, INS_roundpd}, HW_Category_SimpleSIMD, HW_Flag_NoFlag)
+HARDWARE_INTRINSIC(SSE41_FloorScalar, "FloorScalar", SSE41, 9, 16, -1, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_roundss, INS_roundsd}, HW_Category_Special, HW_Flag_CopyUpperBits)
+HARDWARE_INTRINSIC(SSE41_LoadAlignedVector128NonTemporal, "LoadAlignedVector128NonTemporal", SSE41, -1, 16, 1, {INS_movntdqa, INS_movntdqa, INS_movntdqa, INS_movntdqa, INS_movntdqa, INS_movntdqa, INS_movntdqa, INS_movntdqa, INS_invalid, INS_invalid}, HW_Category_MemoryLoad, HW_Flag_NoFlag)
+HARDWARE_INTRINSIC(SSE41_Max, "Max", SSE41, -1, 16, 2, {INS_pmaxsb, INS_invalid, INS_invalid, INS_pmaxuw, INS_pmaxsd, INS_pmaxud, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_NoFlag)
+HARDWARE_INTRINSIC(SSE41_Min, "Min", SSE41, -1, 16, 2, {INS_pminsb, INS_invalid, INS_invalid, INS_pminuw, INS_pminsd, INS_pminud, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_NoFlag)
+HARDWARE_INTRINSIC(SSE41_MinHorizontal, "MinHorizontal", SSE41, -1, 16, 1, {INS_invalid, INS_invalid, INS_invalid, INS_phminposuw, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_NoFlag)
+HARDWARE_INTRINSIC(SSE41_MultipleSumAbsoluteDifferences, "MultipleSumAbsoluteDifferences", SSE41, -1, 16, 3, {INS_invalid, INS_invalid, INS_invalid, INS_mpsadbw, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_IMM, HW_Flag_FullRangeIMM)
+HARDWARE_INTRINSIC(SSE41_Multiply, "Multiply", SSE41, -1, 16, 2, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_pmuldq, INS_invalid, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_Commutative)
+HARDWARE_INTRINSIC(SSE41_MultiplyLow, "MultiplyLow", SSE41, -1, 16, 2, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_pmulld, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_NoFlag)
+HARDWARE_INTRINSIC(SSE41_PackUnsignedSaturate, "PackUnsignedSaturate", SSE41, -1, 16, 2, {INS_invalid, INS_invalid, INS_invalid, INS_packusdw, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_NoFlag)
+HARDWARE_INTRINSIC(SSE41_RoundCurrentDirection, "RoundCurrentDirection", SSE41, 4, 16, 1, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_roundps, INS_roundpd}, HW_Category_SimpleSIMD, HW_Flag_NoFlag)
+HARDWARE_INTRINSIC(SSE41_RoundCurrentDirectionScalar, "RoundCurrentDirectionScalar", SSE41, 4, 16, -1, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_roundss, INS_roundsd}, HW_Category_Special, HW_Flag_CopyUpperBits)
+HARDWARE_INTRINSIC(SSE41_RoundToNearestInteger, "RoundToNearestInteger", SSE41, 8, 16, 1, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_roundps, INS_roundpd}, HW_Category_SimpleSIMD, HW_Flag_NoFlag)
+HARDWARE_INTRINSIC(SSE41_RoundToNearestIntegerScalar, "RoundToNearestIntegerScalar", SSE41, 8, 16, -1, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_roundss, INS_roundsd}, HW_Category_Special, HW_Flag_CopyUpperBits)
+HARDWARE_INTRINSIC(SSE41_RoundToNegativeInfinity, "RoundToNegativeInfinity", SSE41, 9, 16, 1, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_roundps, INS_roundpd}, HW_Category_SimpleSIMD, HW_Flag_NoFlag)
+HARDWARE_INTRINSIC(SSE41_RoundToNegativeInfinityScalar, "RoundToNegativeInfinityScalar", SSE41, 9, 16, -1, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_roundss, INS_roundsd}, HW_Category_Special, HW_Flag_CopyUpperBits)
+HARDWARE_INTRINSIC(SSE41_RoundToPositiveInfinity, "RoundToPositiveInfinity", SSE41, 10, 16, 1, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_roundps, INS_roundpd}, HW_Category_SimpleSIMD, HW_Flag_NoFlag)
+HARDWARE_INTRINSIC(SSE41_RoundToPositiveInfinityScalar, "RoundToPositiveInfinityScalar", SSE41, 10, 16, -1, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_roundss, INS_roundsd}, HW_Category_Special, HW_Flag_CopyUpperBits)
+HARDWARE_INTRINSIC(SSE41_RoundToZero, "RoundToZero", SSE41, 11, 16, 1, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_roundps, INS_roundpd}, HW_Category_SimpleSIMD, HW_Flag_NoFlag)
+HARDWARE_INTRINSIC(SSE41_RoundToZeroScalar, "RoundToZeroScalar", SSE41, 11, 16, -1, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_roundss, INS_roundsd}, HW_Category_Special, HW_Flag_CopyUpperBits)
+HARDWARE_INTRINSIC(SSE41_TestAllOnes, "TestAllOnes", SSE41, -1, 16, 1, {INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_MultiIns|HW_Flag_NoContainment)
+HARDWARE_INTRINSIC(SSE41_TestAllZeros, "TestAllZeros", SSE41, -1, 16, 2, {INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_MultiIns|HW_Flag_NoContainment)
+HARDWARE_INTRINSIC(SSE41_TestC, "TestC", SSE41, -1, 16, 2, {INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_MultiIns|HW_Flag_NoContainment)
+HARDWARE_INTRINSIC(SSE41_TestMixOnesZeros, "TestMixOnesZeros", SSE41, -1, 16, 2, {INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_MultiIns|HW_Flag_NoContainment)
+HARDWARE_INTRINSIC(SSE41_TestNotZAndNotC, "TestNotZAndNotC", SSE41, -1, 16, 2, {INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_MultiIns|HW_Flag_NoContainment)
+HARDWARE_INTRINSIC(SSE41_TestZ, "TestZ", SSE41, -1, 16, 2, {INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_ptest, INS_invalid, INS_invalid}, HW_Category_SimpleSIMD, HW_Flag_MultiIns|HW_Flag_NoContainment)
// SSE42 Intrinsics
HARDWARE_INTRINSIC(SSE42_IsSupported, "get_IsSupported", SSE42, -1, 0, 0, {INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid, INS_invalid}, HW_Category_IsSupportedProperty, HW_Flag_NoFlag)