llvm-6502/lib/Target/PowerPC/PPC.td
Hal Finkel 94dc061e85 [PowerPC] Loosen ELFv1 PPC64 func descriptor loads for indirect calls
Function pointers under PPC64 ELFv1 (which is used on PPC64/Linux on the
POWER7, A2 and earlier cores) are really pointers to a function descriptor, a
structure with three pointers: the actual pointer to the code to which to jump,
the pointer to the TOC needed by the callee, and an environment pointer. We
used to chain these loads, and make them opaque to the rest of the optimizer,
so that they'd always occur directly before the call. This is not necessary,
and in fact, highly suboptimal on embedded cores. Once the function pointer is
known, the loads can be performed ahead of time; in fact, they can be hoisted
out of loops.

Now these function descriptors are almost always generated by the linker, and
thus the contents of the descriptors are invariant. As a result, by default,
we'll mark the associated loads as invariant (allowing them to be hoisted out
of loops). I've added a target feature to turn this off, however, just in case
someone needs that option (constructing an on-stack descriptor, casting it to a
function pointer, and then calling it cannot be well-defined C/C++ code, but I
can imagine some JIT-compilation system doing so).

Consider this simple test:
  $ cat call.c

  typedef void (*fp)();
  void bar(fp x) {
    for (int i = 0; i < 1600000000; ++i)
      x();
  }

  $ cat main.c

  typedef void (*fp)();
  void bar(fp x);
  void foo() {}
  int main() {
    bar(foo);
  }

On the PPC A2 (the BG/Q supercomputer), marking the function-descriptor loads
as invariant brings the execution time down to ~8 seconds from ~32 seconds with
the loads in the loop.

The difference on the POWER7 is smaller. Compiling with:

  gcc -std=c99 -O3 -mcpu=native call.c main.c : ~6 seconds [this is 4.8.2]

  clang -O3 -mcpu=native call.c main.c : ~5.3 seconds

  clang -O3 -mcpu=native call.c main.c -mno-invariant-function-descriptors : ~4 seconds
  (looks like we'd benefit from additional loop unrolling here, as a first
   guess, because this is faster with the extra loads)

The -mno-invariant-function-descriptors will be added to Clang shortly.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@226207 91177308-0d34-0410-b5e6-96231b3b80d8
2015-01-15 21:17:34 +00:00

379 lines
19 KiB
TableGen

//===-- PPC.td - Describe the PowerPC Target Machine -------*- tablegen -*-===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This is the top level entry point for the PowerPC target.
//
//===----------------------------------------------------------------------===//
// Get the target-independent interfaces which we are implementing.
//
include "llvm/Target/Target.td"
//===----------------------------------------------------------------------===//
// PowerPC Subtarget features.
//
//===----------------------------------------------------------------------===//
// CPU Directives //
//===----------------------------------------------------------------------===//
def Directive440 : SubtargetFeature<"", "DarwinDirective", "PPC::DIR_440", "">;
def Directive601 : SubtargetFeature<"", "DarwinDirective", "PPC::DIR_601", "">;
def Directive602 : SubtargetFeature<"", "DarwinDirective", "PPC::DIR_602", "">;
def Directive603 : SubtargetFeature<"", "DarwinDirective", "PPC::DIR_603", "">;
def Directive604 : SubtargetFeature<"", "DarwinDirective", "PPC::DIR_603", "">;
def Directive620 : SubtargetFeature<"", "DarwinDirective", "PPC::DIR_603", "">;
def Directive7400: SubtargetFeature<"", "DarwinDirective", "PPC::DIR_7400", "">;
def Directive750 : SubtargetFeature<"", "DarwinDirective", "PPC::DIR_750", "">;
def Directive970 : SubtargetFeature<"", "DarwinDirective", "PPC::DIR_970", "">;
def Directive32 : SubtargetFeature<"", "DarwinDirective", "PPC::DIR_32", "">;
def Directive64 : SubtargetFeature<"", "DarwinDirective", "PPC::DIR_64", "">;
def DirectiveA2 : SubtargetFeature<"", "DarwinDirective", "PPC::DIR_A2", "">;
def DirectiveE500mc : SubtargetFeature<"", "DarwinDirective",
"PPC::DIR_E500mc", "">;
def DirectiveE5500 : SubtargetFeature<"", "DarwinDirective",
"PPC::DIR_E5500", "">;
def DirectivePwr3: SubtargetFeature<"", "DarwinDirective", "PPC::DIR_PWR3", "">;
def DirectivePwr4: SubtargetFeature<"", "DarwinDirective", "PPC::DIR_PWR4", "">;
def DirectivePwr5: SubtargetFeature<"", "DarwinDirective", "PPC::DIR_PWR5", "">;
def DirectivePwr5x: SubtargetFeature<"", "DarwinDirective", "PPC::DIR_PWR5X", "">;
def DirectivePwr6: SubtargetFeature<"", "DarwinDirective", "PPC::DIR_PWR6", "">;
def DirectivePwr6x: SubtargetFeature<"", "DarwinDirective", "PPC::DIR_PWR6X", "">;
def DirectivePwr7: SubtargetFeature<"", "DarwinDirective", "PPC::DIR_PWR7", "">;
def DirectivePwr8: SubtargetFeature<"", "DarwinDirective", "PPC::DIR_PWR8", "">;
def Feature64Bit : SubtargetFeature<"64bit","Has64BitSupport", "true",
"Enable 64-bit instructions">;
def Feature64BitRegs : SubtargetFeature<"64bitregs","Use64BitRegs", "true",
"Enable 64-bit registers usage for ppc32 [beta]">;
def FeatureCRBits : SubtargetFeature<"crbits", "UseCRBits", "true",
"Use condition-register bits individually">;
def FeatureAltivec : SubtargetFeature<"altivec","HasAltivec", "true",
"Enable Altivec instructions">;
def FeatureSPE : SubtargetFeature<"spe","HasSPE", "true",
"Enable SPE instructions">;
def FeatureMFOCRF : SubtargetFeature<"mfocrf","HasMFOCRF", "true",
"Enable the MFOCRF instruction">;
def FeatureFSqrt : SubtargetFeature<"fsqrt","HasFSQRT", "true",
"Enable the fsqrt instruction">;
def FeatureFCPSGN : SubtargetFeature<"fcpsgn", "HasFCPSGN", "true",
"Enable the fcpsgn instruction">;
def FeatureFRE : SubtargetFeature<"fre", "HasFRE", "true",
"Enable the fre instruction">;
def FeatureFRES : SubtargetFeature<"fres", "HasFRES", "true",
"Enable the fres instruction">;
def FeatureFRSQRTE : SubtargetFeature<"frsqrte", "HasFRSQRTE", "true",
"Enable the frsqrte instruction">;
def FeatureFRSQRTES : SubtargetFeature<"frsqrtes", "HasFRSQRTES", "true",
"Enable the frsqrtes instruction">;
def FeatureRecipPrec : SubtargetFeature<"recipprec", "HasRecipPrec", "true",
"Assume higher precision reciprocal estimates">;
def FeatureSTFIWX : SubtargetFeature<"stfiwx","HasSTFIWX", "true",
"Enable the stfiwx instruction">;
def FeatureLFIWAX : SubtargetFeature<"lfiwax","HasLFIWAX", "true",
"Enable the lfiwax instruction">;
def FeatureFPRND : SubtargetFeature<"fprnd", "HasFPRND", "true",
"Enable the fri[mnpz] instructions">;
def FeatureFPCVT : SubtargetFeature<"fpcvt", "HasFPCVT", "true",
"Enable fc[ft]* (unsigned and single-precision) and lfiwzx instructions">;
def FeatureISEL : SubtargetFeature<"isel","HasISEL", "true",
"Enable the isel instruction">;
def FeaturePOPCNTD : SubtargetFeature<"popcntd","HasPOPCNTD", "true",
"Enable the popcnt[dw] instructions">;
def FeatureLDBRX : SubtargetFeature<"ldbrx","HasLDBRX", "true",
"Enable the ldbrx instruction">;
def FeatureCMPB : SubtargetFeature<"cmpb", "HasCMPB", "true",
"Enable the cmpb instruction">;
def FeatureICBT : SubtargetFeature<"icbt","HasICBT", "true",
"Enable icbt instruction">;
def FeatureBookE : SubtargetFeature<"booke", "IsBookE", "true",
"Enable Book E instructions",
[FeatureICBT]>;
def FeatureMSYNC : SubtargetFeature<"msync", "HasOnlyMSYNC", "true",
"Has only the msync instruction instead of sync",
[FeatureBookE]>;
def FeatureE500 : SubtargetFeature<"e500", "IsE500", "true",
"Enable E500/E500mc instructions">;
def FeaturePPC4xx : SubtargetFeature<"ppc4xx", "IsPPC4xx", "true",
"Enable PPC 4xx instructions">;
def FeaturePPC6xx : SubtargetFeature<"ppc6xx", "IsPPC6xx", "true",
"Enable PPC 6xx instructions">;
def FeatureQPX : SubtargetFeature<"qpx","HasQPX", "true",
"Enable QPX instructions">;
def FeatureVSX : SubtargetFeature<"vsx","HasVSX", "true",
"Enable VSX instructions",
[FeatureAltivec]>;
def FeatureP8Vector : SubtargetFeature<"power8-vector", "HasP8Vector", "true",
"Enable POWER8 vector instructions",
[FeatureVSX, FeatureAltivec]>;
def FeatureInvariantFunctionDescriptors :
SubtargetFeature<"invariant-function-descriptors",
"HasInvariantFunctionDescriptors", "true",
"Assume function descriptors are invariant">;
def DeprecatedMFTB : SubtargetFeature<"", "DeprecatedMFTB", "true",
"Treat mftb as deprecated">;
def DeprecatedDST : SubtargetFeature<"", "DeprecatedDST", "true",
"Treat vector data stream cache control instructions as deprecated">;
// Note: Future features to add when support is extended to more
// recent ISA levels:
//
// DFP p6, p6x, p7 decimal floating-point instructions
// POPCNTB p5 through p7 popcntb and related instructions
//===----------------------------------------------------------------------===//
// ABI Selection //
//===----------------------------------------------------------------------===//
def FeatureELFv1 : SubtargetFeature<"elfv1", "TargetABI", "PPC_ABI_ELFv1",
"Use the ELFv1 ABI">;
def FeatureELFv2 : SubtargetFeature<"elfv2", "TargetABI", "PPC_ABI_ELFv2",
"Use the ELFv2 ABI">;
//===----------------------------------------------------------------------===//
// Classes used for relation maps.
//===----------------------------------------------------------------------===//
// RecFormRel - Filter class used to relate non-record-form instructions with
// their record-form variants.
class RecFormRel;
// AltVSXFMARel - Filter class used to relate the primary addend-killing VSX
// FMA instruction forms with their corresponding factor-killing forms.
class AltVSXFMARel {
bit IsVSXFMAAlt = 0;
}
//===----------------------------------------------------------------------===//
// Relation Map Definitions.
//===----------------------------------------------------------------------===//
def getRecordFormOpcode : InstrMapping {
let FilterClass = "RecFormRel";
// Instructions with the same BaseName and Interpretation64Bit values
// form a row.
let RowFields = ["BaseName", "Interpretation64Bit"];
// Instructions with the same RC value form a column.
let ColFields = ["RC"];
// The key column are the non-record-form instructions.
let KeyCol = ["0"];
// Value columns RC=1
let ValueCols = [["1"]];
}
def getNonRecordFormOpcode : InstrMapping {
let FilterClass = "RecFormRel";
// Instructions with the same BaseName and Interpretation64Bit values
// form a row.
let RowFields = ["BaseName", "Interpretation64Bit"];
// Instructions with the same RC value form a column.
let ColFields = ["RC"];
// The key column are the record-form instructions.
let KeyCol = ["1"];
// Value columns are RC=0
let ValueCols = [["0"]];
}
def getAltVSXFMAOpcode : InstrMapping {
let FilterClass = "AltVSXFMARel";
// Instructions with the same BaseName and Interpretation64Bit values
// form a row.
let RowFields = ["BaseName"];
// Instructions with the same RC value form a column.
let ColFields = ["IsVSXFMAAlt"];
// The key column are the (default) addend-killing instructions.
let KeyCol = ["0"];
// Value columns IsVSXFMAAlt=1
let ValueCols = [["1"]];
}
//===----------------------------------------------------------------------===//
// Register File Description
//===----------------------------------------------------------------------===//
include "PPCRegisterInfo.td"
include "PPCSchedule.td"
include "PPCInstrInfo.td"
//===----------------------------------------------------------------------===//
// PowerPC processors supported.
//
def : Processor<"generic", G3Itineraries, [Directive32]>;
def : ProcessorModel<"440", PPC440Model, [Directive440, FeatureISEL,
FeatureFRES, FeatureFRSQRTE,
FeatureICBT, FeatureBookE,
FeatureMSYNC, DeprecatedMFTB]>;
def : ProcessorModel<"450", PPC440Model, [Directive440, FeatureISEL,
FeatureFRES, FeatureFRSQRTE,
FeatureICBT, FeatureBookE,
FeatureMSYNC, DeprecatedMFTB]>;
def : Processor<"601", G3Itineraries, [Directive601]>;
def : Processor<"602", G3Itineraries, [Directive602]>;
def : Processor<"603", G3Itineraries, [Directive603,
FeatureFRES, FeatureFRSQRTE]>;
def : Processor<"603e", G3Itineraries, [Directive603,
FeatureFRES, FeatureFRSQRTE]>;
def : Processor<"603ev", G3Itineraries, [Directive603,
FeatureFRES, FeatureFRSQRTE]>;
def : Processor<"604", G3Itineraries, [Directive604,
FeatureFRES, FeatureFRSQRTE]>;
def : Processor<"604e", G3Itineraries, [Directive604,
FeatureFRES, FeatureFRSQRTE]>;
def : Processor<"620", G3Itineraries, [Directive620,
FeatureFRES, FeatureFRSQRTE]>;
def : Processor<"750", G4Itineraries, [Directive750,
FeatureFRES, FeatureFRSQRTE]>;
def : Processor<"g3", G3Itineraries, [Directive750,
FeatureFRES, FeatureFRSQRTE]>;
def : Processor<"7400", G4Itineraries, [Directive7400, FeatureAltivec,
FeatureFRES, FeatureFRSQRTE]>;
def : Processor<"g4", G4Itineraries, [Directive7400, FeatureAltivec,
FeatureFRES, FeatureFRSQRTE]>;
def : Processor<"7450", G4PlusItineraries, [Directive7400, FeatureAltivec,
FeatureFRES, FeatureFRSQRTE]>;
def : Processor<"g4+", G4PlusItineraries, [Directive7400, FeatureAltivec,
FeatureFRES, FeatureFRSQRTE]>;
def : ProcessorModel<"970", G5Model,
[Directive970, FeatureAltivec,
FeatureMFOCRF, FeatureFSqrt,
FeatureFRES, FeatureFRSQRTE, FeatureSTFIWX,
Feature64Bit /*, Feature64BitRegs */]>;
def : ProcessorModel<"g5", G5Model,
[Directive970, FeatureAltivec,
FeatureMFOCRF, FeatureFSqrt, FeatureSTFIWX,
FeatureFRES, FeatureFRSQRTE,
Feature64Bit /*, Feature64BitRegs */,
DeprecatedMFTB, DeprecatedDST]>;
def : ProcessorModel<"e500mc", PPCE500mcModel,
[DirectiveE500mc, FeatureMFOCRF,
FeatureSTFIWX, FeatureICBT, FeatureBookE,
FeatureISEL, DeprecatedMFTB]>;
def : ProcessorModel<"e5500", PPCE5500Model,
[DirectiveE5500, FeatureMFOCRF, Feature64Bit,
FeatureSTFIWX, FeatureICBT, FeatureBookE,
FeatureISEL, DeprecatedMFTB]>;
def : ProcessorModel<"a2", PPCA2Model,
[DirectiveA2, FeatureICBT, FeatureBookE, FeatureMFOCRF,
FeatureFCPSGN, FeatureFSqrt, FeatureFRE, FeatureFRES,
FeatureFRSQRTE, FeatureFRSQRTES, FeatureRecipPrec,
FeatureSTFIWX, FeatureLFIWAX,
FeatureFPRND, FeatureFPCVT, FeatureISEL,
FeaturePOPCNTD, FeatureCMPB, FeatureLDBRX, Feature64Bit
/*, Feature64BitRegs */, DeprecatedMFTB]>;
def : ProcessorModel<"a2q", PPCA2Model,
[DirectiveA2, FeatureICBT, FeatureBookE, FeatureMFOCRF,
FeatureFCPSGN, FeatureFSqrt, FeatureFRE, FeatureFRES,
FeatureFRSQRTE, FeatureFRSQRTES, FeatureRecipPrec,
FeatureSTFIWX, FeatureLFIWAX,
FeatureFPRND, FeatureFPCVT, FeatureISEL,
FeaturePOPCNTD, FeatureCMPB, FeatureLDBRX, Feature64Bit
/*, Feature64BitRegs */, FeatureQPX, DeprecatedMFTB]>;
def : ProcessorModel<"pwr3", G5Model,
[DirectivePwr3, FeatureAltivec,
FeatureFRES, FeatureFRSQRTE, FeatureMFOCRF,
FeatureSTFIWX, Feature64Bit]>;
def : ProcessorModel<"pwr4", G5Model,
[DirectivePwr4, FeatureAltivec, FeatureMFOCRF,
FeatureFSqrt, FeatureFRES, FeatureFRSQRTE,
FeatureSTFIWX, Feature64Bit]>;
def : ProcessorModel<"pwr5", G5Model,
[DirectivePwr5, FeatureAltivec, FeatureMFOCRF,
FeatureFSqrt, FeatureFRE, FeatureFRES,
FeatureFRSQRTE, FeatureFRSQRTES,
FeatureSTFIWX, Feature64Bit,
DeprecatedMFTB, DeprecatedDST]>;
def : ProcessorModel<"pwr5x", G5Model,
[DirectivePwr5x, FeatureAltivec, FeatureMFOCRF,
FeatureFSqrt, FeatureFRE, FeatureFRES,
FeatureFRSQRTE, FeatureFRSQRTES,
FeatureSTFIWX, FeatureFPRND, Feature64Bit,
DeprecatedMFTB, DeprecatedDST]>;
def : ProcessorModel<"pwr6", G5Model,
[DirectivePwr6, FeatureAltivec,
FeatureMFOCRF, FeatureFCPSGN, FeatureFSqrt, FeatureFRE,
FeatureFRES, FeatureFRSQRTE, FeatureFRSQRTES,
FeatureRecipPrec, FeatureSTFIWX, FeatureLFIWAX, FeatureCMPB,
FeatureFPRND, Feature64Bit /*, Feature64BitRegs */,
DeprecatedMFTB, DeprecatedDST]>;
def : ProcessorModel<"pwr6x", G5Model,
[DirectivePwr5x, FeatureAltivec, FeatureMFOCRF,
FeatureFCPSGN, FeatureFSqrt, FeatureFRE, FeatureFRES,
FeatureFRSQRTE, FeatureFRSQRTES, FeatureRecipPrec,
FeatureSTFIWX, FeatureLFIWAX, FeatureCMPB,
FeatureFPRND, Feature64Bit,
DeprecatedMFTB, DeprecatedDST]>;
def : ProcessorModel<"pwr7", P7Model,
[DirectivePwr7, FeatureAltivec, FeatureVSX,
FeatureMFOCRF, FeatureFCPSGN, FeatureFSqrt, FeatureFRE,
FeatureFRES, FeatureFRSQRTE, FeatureFRSQRTES,
FeatureRecipPrec, FeatureSTFIWX, FeatureLFIWAX,
FeatureFPRND, FeatureFPCVT, FeatureISEL,
FeaturePOPCNTD, FeatureCMPB, FeatureLDBRX,
Feature64Bit /*, Feature64BitRegs */,
DeprecatedMFTB, DeprecatedDST]>;
def : ProcessorModel<"pwr8", P8Model,
[DirectivePwr8, FeatureAltivec, FeatureVSX, FeatureP8Vector,
FeatureMFOCRF, FeatureFCPSGN, FeatureFSqrt, FeatureFRE,
FeatureFRES, FeatureFRSQRTE, FeatureFRSQRTES,
FeatureRecipPrec, FeatureSTFIWX, FeatureLFIWAX,
FeatureFPRND, FeatureFPCVT, FeatureISEL,
FeaturePOPCNTD, FeatureCMPB, FeatureLDBRX,
Feature64Bit /*, Feature64BitRegs */, FeatureICBT,
DeprecatedMFTB, DeprecatedDST]>;
def : Processor<"ppc", G3Itineraries, [Directive32]>;
def : ProcessorModel<"ppc64", G5Model,
[Directive64, FeatureAltivec,
FeatureMFOCRF, FeatureFSqrt, FeatureFRES,
FeatureFRSQRTE, FeatureSTFIWX,
Feature64Bit /*, Feature64BitRegs */]>;
def : ProcessorModel<"ppc64le", G5Model,
[Directive64, FeatureAltivec,
FeatureMFOCRF, FeatureFSqrt, FeatureFRES,
FeatureFRSQRTE, FeatureSTFIWX,
Feature64Bit /*, Feature64BitRegs */]>;
//===----------------------------------------------------------------------===//
// Calling Conventions
//===----------------------------------------------------------------------===//
include "PPCCallingConv.td"
def PPCInstrInfo : InstrInfo {
let isLittleEndianEncoding = 1;
// FIXME: Unset this when no longer needed!
let decodePositionallyEncodedOperands = 1;
let noNamedPositionallyEncodedOperands = 1;
}
def PPCAsmParser : AsmParser {
let ShouldEmitMatchRegisterName = 0;
}
def PPCAsmParserVariant : AsmParserVariant {
int Variant = 0;
// We do not use hard coded registers in asm strings. However, some
// InstAlias definitions use immediate literals. Set RegisterPrefix
// so that those are not misinterpreted as registers.
string RegisterPrefix = "%";
}
def PPC : Target {
// Information about the instructions.
let InstructionSet = PPCInstrInfo;
let AssemblyParsers = [PPCAsmParser];
let AssemblyParserVariants = [PPCAsmParserVariant];
}