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.

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

Add BSF and BSR fallbacks for BitOperations methods #34550

Merged
merged 14 commits into from
Apr 29, 2020
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
10 changes: 5 additions & 5 deletions src/coreclr/src/inc/corinfo.h
Original file line number Diff line number Diff line change
Expand Up @@ -217,11 +217,11 @@ TODO: Talk about initializing strutures before use
#endif
#endif

SELECTANY const GUID JITEEVersionIdentifier = { /* bb6ea6c3-ce5a-4543-86b7-c9c88f9ec780 */
0xbb6ea6c3,
0xce5a,
0x4543,
{ 0x86, 0xb7, 0xc9, 0xc8, 0x8f, 0x9e, 0xc7, 0x80 }
SELECTANY const GUID JITEEVersionIdentifier = { /* 8b2226a2-ac30-4f5c-ae5c-926c792ecdb9 */
0x8b2226a2,
0xac30,
0x4f5c,
{ 0xae, 0x5c, 0x92, 0x6c, 0x79, 0x2e, 0xcd, 0xb9 }
};

//////////////////////////////////////////////////////////////////////////////////////////////////////////
Expand Down
122 changes: 72 additions & 50 deletions src/coreclr/src/inc/corinfoinstructionset.h
Original file line number Diff line number Diff line change
Expand Up @@ -32,58 +32,62 @@ enum CORINFO_InstructionSet
InstructionSet_Vector128=12,
#endif // TARGET_ARM64
#ifdef TARGET_AMD64
InstructionSet_SSE=1,
InstructionSet_SSE2=2,
InstructionSet_SSE3=3,
InstructionSet_SSSE3=4,
InstructionSet_SSE41=5,
InstructionSet_SSE42=6,
InstructionSet_AVX=7,
InstructionSet_AVX2=8,
InstructionSet_AES=9,
InstructionSet_BMI1=10,
InstructionSet_BMI2=11,
InstructionSet_FMA=12,
InstructionSet_LZCNT=13,
InstructionSet_PCLMULQDQ=14,
InstructionSet_POPCNT=15,
InstructionSet_Vector128=16,
InstructionSet_Vector256=17,
InstructionSet_BMI1_X64=18,
InstructionSet_BMI2_X64=19,
InstructionSet_LZCNT_X64=20,
InstructionSet_POPCNT_X64=21,
InstructionSet_SSE_X64=22,
InstructionSet_SSE2_X64=23,
InstructionSet_SSE41_X64=24,
InstructionSet_SSE42_X64=25,
InstructionSet_X86Base=1,
InstructionSet_SSE=2,
InstructionSet_SSE2=3,
InstructionSet_SSE3=4,
InstructionSet_SSSE3=5,
InstructionSet_SSE41=6,
InstructionSet_SSE42=7,
InstructionSet_AVX=8,
InstructionSet_AVX2=9,
InstructionSet_AES=10,
InstructionSet_BMI1=11,
InstructionSet_BMI2=12,
InstructionSet_FMA=13,
InstructionSet_LZCNT=14,
InstructionSet_PCLMULQDQ=15,
InstructionSet_POPCNT=16,
InstructionSet_Vector128=17,
InstructionSet_Vector256=18,
InstructionSet_X86Base_X64=19,
InstructionSet_BMI1_X64=20,
InstructionSet_BMI2_X64=21,
InstructionSet_LZCNT_X64=22,
InstructionSet_POPCNT_X64=23,
InstructionSet_SSE_X64=24,
InstructionSet_SSE2_X64=25,
InstructionSet_SSE41_X64=26,
InstructionSet_SSE42_X64=27,
#endif // TARGET_AMD64
#ifdef TARGET_X86
InstructionSet_SSE=1,
InstructionSet_SSE2=2,
InstructionSet_SSE3=3,
InstructionSet_SSSE3=4,
InstructionSet_SSE41=5,
InstructionSet_SSE42=6,
InstructionSet_AVX=7,
InstructionSet_AVX2=8,
InstructionSet_AES=9,
InstructionSet_BMI1=10,
InstructionSet_BMI2=11,
InstructionSet_FMA=12,
InstructionSet_LZCNT=13,
InstructionSet_PCLMULQDQ=14,
InstructionSet_POPCNT=15,
InstructionSet_Vector128=16,
InstructionSet_Vector256=17,
InstructionSet_BMI1_X64=18,
InstructionSet_BMI2_X64=19,
InstructionSet_LZCNT_X64=20,
InstructionSet_POPCNT_X64=21,
InstructionSet_SSE_X64=22,
InstructionSet_SSE2_X64=23,
InstructionSet_SSE41_X64=24,
InstructionSet_SSE42_X64=25,
InstructionSet_X86Base=1,
InstructionSet_SSE=2,
InstructionSet_SSE2=3,
InstructionSet_SSE3=4,
InstructionSet_SSSE3=5,
InstructionSet_SSE41=6,
InstructionSet_SSE42=7,
InstructionSet_AVX=8,
InstructionSet_AVX2=9,
InstructionSet_AES=10,
InstructionSet_BMI1=11,
InstructionSet_BMI2=12,
InstructionSet_FMA=13,
InstructionSet_LZCNT=14,
InstructionSet_PCLMULQDQ=15,
InstructionSet_POPCNT=16,
InstructionSet_Vector128=17,
InstructionSet_Vector256=18,
InstructionSet_X86Base_X64=19,
InstructionSet_BMI1_X64=20,
InstructionSet_BMI2_X64=21,
InstructionSet_LZCNT_X64=22,
InstructionSet_POPCNT_X64=23,
InstructionSet_SSE_X64=24,
InstructionSet_SSE2_X64=25,
InstructionSet_SSE41_X64=26,
InstructionSet_SSE42_X64=27,
#endif // TARGET_X86

};
Expand Down Expand Up @@ -139,6 +143,8 @@ struct CORINFO_InstructionSetFlags
AddInstructionSet(InstructionSet_Crc32_Arm64);
#endif // TARGET_ARM64
#ifdef TARGET_AMD64
if (HasInstructionSet(InstructionSet_X86Base))
AddInstructionSet(InstructionSet_X86Base_X64);
if (HasInstructionSet(InstructionSet_SSE))
AddInstructionSet(InstructionSet_SSE_X64);
if (HasInstructionSet(InstructionSet_SSE2))
Expand Down Expand Up @@ -204,6 +210,10 @@ inline CORINFO_InstructionSetFlags EnsureInstructionSetFlagsAreValid(CORINFO_Ins
resultflags.RemoveInstructionSet(InstructionSet_Sha256);
#endif // TARGET_ARM64
#ifdef TARGET_AMD64
if (resultflags.HasInstructionSet(InstructionSet_X86Base) && !resultflags.HasInstructionSet(InstructionSet_X86Base_X64))
resultflags.RemoveInstructionSet(InstructionSet_X86Base);
if (resultflags.HasInstructionSet(InstructionSet_X86Base_X64) && !resultflags.HasInstructionSet(InstructionSet_X86Base))
resultflags.RemoveInstructionSet(InstructionSet_X86Base_X64);
if (resultflags.HasInstructionSet(InstructionSet_SSE) && !resultflags.HasInstructionSet(InstructionSet_SSE_X64))
resultflags.RemoveInstructionSet(InstructionSet_SSE);
if (resultflags.HasInstructionSet(InstructionSet_SSE_X64) && !resultflags.HasInstructionSet(InstructionSet_SSE))
Expand Down Expand Up @@ -236,6 +246,8 @@ inline CORINFO_InstructionSetFlags EnsureInstructionSetFlagsAreValid(CORINFO_Ins
resultflags.RemoveInstructionSet(InstructionSet_POPCNT);
if (resultflags.HasInstructionSet(InstructionSet_POPCNT_X64) && !resultflags.HasInstructionSet(InstructionSet_POPCNT))
resultflags.RemoveInstructionSet(InstructionSet_POPCNT_X64);
if (resultflags.HasInstructionSet(InstructionSet_SSE) && !resultflags.HasInstructionSet(InstructionSet_X86Base))
resultflags.RemoveInstructionSet(InstructionSet_SSE);
if (resultflags.HasInstructionSet(InstructionSet_SSE2) && !resultflags.HasInstructionSet(InstructionSet_SSE))
resultflags.RemoveInstructionSet(InstructionSet_SSE2);
if (resultflags.HasInstructionSet(InstructionSet_SSE3) && !resultflags.HasInstructionSet(InstructionSet_SSE2))
Expand Down Expand Up @@ -264,6 +276,8 @@ inline CORINFO_InstructionSetFlags EnsureInstructionSetFlagsAreValid(CORINFO_Ins
resultflags.RemoveInstructionSet(InstructionSet_POPCNT);
#endif // TARGET_AMD64
#ifdef TARGET_X86
if (resultflags.HasInstructionSet(InstructionSet_SSE) && !resultflags.HasInstructionSet(InstructionSet_X86Base))
saucecontrol marked this conversation as resolved.
Show resolved Hide resolved
resultflags.RemoveInstructionSet(InstructionSet_SSE);
if (resultflags.HasInstructionSet(InstructionSet_SSE2) && !resultflags.HasInstructionSet(InstructionSet_SSE))
resultflags.RemoveInstructionSet(InstructionSet_SSE2);
if (resultflags.HasInstructionSet(InstructionSet_SSE3) && !resultflags.HasInstructionSet(InstructionSet_SSE2))
Expand Down Expand Up @@ -332,6 +346,10 @@ inline const char *InstructionSetToString(CORINFO_InstructionSet instructionSet)
return "Vector128";
#endif // TARGET_ARM64
#ifdef TARGET_AMD64
case InstructionSet_X86Base :
return "X86Base";
case InstructionSet_X86Base_X64 :
return "X86Base_X64";
case InstructionSet_SSE :
return "SSE";
case InstructionSet_SSE_X64 :
Expand Down Expand Up @@ -384,6 +402,8 @@ inline const char *InstructionSetToString(CORINFO_InstructionSet instructionSet)
return "Vector256";
#endif // TARGET_AMD64
#ifdef TARGET_X86
case InstructionSet_X86Base :
return "X86Base";
case InstructionSet_SSE :
return "SSE";
case InstructionSet_SSE2 :
Expand Down Expand Up @@ -447,6 +467,7 @@ inline CORINFO_InstructionSet InstructionSetFromR2RInstructionSet(ReadyToRunInst
case READYTORUN_INSTRUCTION_Atomics: return InstructionSet_Atomics;
#endif // TARGET_ARM64
#ifdef TARGET_AMD64
case READYTORUN_INSTRUCTION_X86Base: return InstructionSet_X86Base;
case READYTORUN_INSTRUCTION_Sse: return InstructionSet_SSE;
case READYTORUN_INSTRUCTION_Sse2: return InstructionSet_SSE2;
case READYTORUN_INSTRUCTION_Sse3: return InstructionSet_SSE3;
Expand All @@ -464,6 +485,7 @@ inline CORINFO_InstructionSet InstructionSetFromR2RInstructionSet(ReadyToRunInst
case READYTORUN_INSTRUCTION_Popcnt: return InstructionSet_POPCNT;
#endif // TARGET_AMD64
#ifdef TARGET_X86
case READYTORUN_INSTRUCTION_X86Base: return InstructionSet_X86Base;
case READYTORUN_INSTRUCTION_Sse: return InstructionSet_SSE;
case READYTORUN_INSTRUCTION_Sse2: return InstructionSet_SSE2;
case READYTORUN_INSTRUCTION_Sse3: return InstructionSet_SSE3;
Expand Down
1 change: 1 addition & 0 deletions src/coreclr/src/inc/readytoruninstructionset.h
Original file line number Diff line number Diff line change
Expand Up @@ -32,6 +32,7 @@ enum ReadyToRunInstructionSet
READYTORUN_INSTRUCTION_Sha1=19,
READYTORUN_INSTRUCTION_Sha256=20,
READYTORUN_INSTRUCTION_Atomics=21,
READYTORUN_INSTRUCTION_X86Base=22,

};

Expand Down
1 change: 1 addition & 0 deletions src/coreclr/src/jit/codegen.h
Original file line number Diff line number Diff line change
Expand Up @@ -1028,6 +1028,7 @@ XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
void genHWIntrinsic_R_R_R_RM(
instruction ins, emitAttr attr, regNumber targetReg, regNumber op1Reg, regNumber op2Reg, GenTree* op3);
void genBaseIntrinsic(GenTreeHWIntrinsic* node);
void genX86BaseIntrinsic(GenTreeHWIntrinsic* node);
saucecontrol marked this conversation as resolved.
Show resolved Hide resolved
void genSSEIntrinsic(GenTreeHWIntrinsic* node);
void genSSE2Intrinsic(GenTreeHWIntrinsic* node);
void genSSE41Intrinsic(GenTreeHWIntrinsic* node);
Expand Down
9 changes: 4 additions & 5 deletions src/coreclr/src/jit/compiler.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -2161,7 +2161,7 @@ const char* Compiler::compLocalVarName(unsigned varNum, unsigned offs)
void Compiler::compSetProcessor()
{
//
// NOTE: This function needs to be kept in sync with EEJitManager::SetCpuInfo() in vm\codemap.cpp
// NOTE: This function needs to be kept in sync with EEJitManager::SetCpuInfo() in vm\codeman.cpp
//

const JitFlags& jitFlags = *opts.jitFlags;
Expand Down Expand Up @@ -2195,13 +2195,14 @@ void Compiler::compSetProcessor()

#endif // TARGET_X86

// The VM will set the ISA flags depending on actual hardware support.
// We then select which ISAs to leave enabled based on the JIT config.
// The exception to this is the dummy Vector64/128/256 ISAs, which must be added explicitly.
CORINFO_InstructionSetFlags instructionSetFlags = jitFlags.GetInstructionSetFlags();
opts.compSupportsISA = 0;
opts.compSupportsISAReported = 0;

#ifdef TARGET_XARCH
bool avxSupported = false;

if (JitConfig.EnableHWIntrinsic())
{
// Dummy ISAs for simplifying the JIT code
Expand Down Expand Up @@ -2315,8 +2316,6 @@ void Compiler::compSetProcessor()
if (JitConfig.EnableHWIntrinsic())
{
// Dummy ISAs for simplifying the JIT code
instructionSetFlags.AddInstructionSet(InstructionSet_ArmBase);
instructionSetFlags.AddInstructionSet(InstructionSet_ArmBase_Arm64);
instructionSetFlags.AddInstructionSet(InstructionSet_Vector64);
instructionSetFlags.AddInstructionSet(InstructionSet_Vector128);
}
Expand Down
13 changes: 9 additions & 4 deletions src/coreclr/src/jit/emitxarch.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -9448,7 +9448,8 @@ BYTE* emitter::emitOutputAM(BYTE* dst, instrDesc* id, code_t code, CnsVal* addc)

// Use the large version if this is not a byte. This trick will not
// work in case of SSE2 and AVX instructions.
if ((size != EA_1BYTE) && (ins != INS_imul) && !IsSSEInstruction(ins) && !IsAVXInstruction(ins))
if ((size != EA_1BYTE) && (ins != INS_imul) && (ins != INS_bsf) && (ins != INS_bsr) && !IsSSEInstruction(ins) &&
!IsAVXInstruction(ins))
{
code++;
}
Expand Down Expand Up @@ -10214,8 +10215,9 @@ BYTE* emitter::emitOutputSV(BYTE* dst, instrDesc* id, code_t code, CnsVal* addc)
}

// Use the large version if this is not a byte
if ((size != EA_1BYTE) && (ins != INS_imul) && (!insIsCMOV(ins)) && !IsSSEInstruction(ins) &&
!IsAVXInstruction(ins))
// TODO-XArch-Cleanup Can the need for the 'w' size bit be encoded in the instruction flags?
if ((size != EA_1BYTE) && (ins != INS_imul) && (ins != INS_bsf) && (ins != INS_bsr) && (!insIsCMOV(ins)) &&
!IsSSEInstruction(ins) && !IsAVXInstruction(ins))
{
code |= 0x1;
}
Expand Down Expand Up @@ -11248,7 +11250,8 @@ BYTE* emitter::emitOutputRR(BYTE* dst, instrDesc* id)
#endif // TARGET_AMD64
}
#ifdef FEATURE_HW_INTRINSICS
else if ((ins == INS_crc32) || (ins == INS_lzcnt) || (ins == INS_popcnt) || (ins == INS_tzcnt))
else if ((ins == INS_bsf) || (ins == INS_bsr) || (ins == INS_crc32) || (ins == INS_lzcnt) || (ins == INS_popcnt) ||
(ins == INS_tzcnt))
{
code = insEncodeRMreg(ins, code);
if ((ins == INS_crc32) && (size > EA_1BYTE))
Expand Down Expand Up @@ -14826,6 +14829,8 @@ emitter::insExecutionCharacteristics emitter::getInsExecutionCharacteristics(ins
result.insLatency += PERFSCORE_LATENCY_2C;
break;

case INS_bsf:
case INS_bsr:
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It would be nice if we had a way to model Intel vs AMD here.

Instruction Intel AMD
BSF 1.00 4.00
BSR 1.00 5.00
POPCNT 1.00 0.50
LZCNT 1.00 0.50
TZCNT 1.00 0.50

case INS_pextrb:
case INS_pextrd:
case INS_pextrw:
Expand Down
38 changes: 38 additions & 0 deletions src/coreclr/src/jit/hwintrinsiccodegenxarch.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -359,6 +359,10 @@ void CodeGen::genHWIntrinsic(GenTreeHWIntrinsic* node)
case InstructionSet_Vector256:
genBaseIntrinsic(node);
break;
case InstructionSet_X86Base:
case InstructionSet_X86Base_X64:
genX86BaseIntrinsic(node);
break;
case InstructionSet_SSE:
case InstructionSet_SSE_X64:
genSSEIntrinsic(node);
Expand Down Expand Up @@ -1249,6 +1253,40 @@ void CodeGen::genBaseIntrinsic(GenTreeHWIntrinsic* node)
genProduceReg(node);
}

//------------------------------------------------------------------------
// genX86BaseIntrinsic: Generates the code for an X86 base hardware intrinsic node
//
// Arguments:
// node - The hardware intrinsic node
//
void CodeGen::genX86BaseIntrinsic(GenTreeHWIntrinsic* node)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Not needed for this PR, but we allowed HW_Category_Scalar intrinsics to be table driven in importation.
It would be nice to do the same here in codegen...

{
NamedIntrinsic intrinsicId = node->gtHWIntrinsicId;

switch (intrinsicId)
{
case NI_X86Base_BitScanForward:
case NI_X86Base_BitScanReverse:
case NI_X86Base_X64_BitScanForward:
case NI_X86Base_X64_BitScanReverse:
{
GenTree* op1 = node->gtGetOp1();
regNumber targetReg = node->GetRegNum();
var_types targetType = node->TypeGet();
instruction ins = HWIntrinsicInfo::lookupIns(intrinsicId, targetType);

genConsumeOperands(node);
genHWIntrinsic_R_RM(node, ins, emitTypeSize(targetType), targetReg, op1);
genProduceReg(node);
break;
}

default:
unreached();
break;
}
}

//------------------------------------------------------------------------
// genSSEIntrinsic: Generates the code for an SSE hardware intrinsic node
//
Expand Down
Loading