-//=====---- X86Subtarget.h - Define Subtarget for the X86 -----*- C++ -*--====//
+//===-- X86Subtarget.h - Define Subtarget for the X86 ----------*- C++ -*--===//
//
// The LLVM Compiler Infrastructure
//
#define X86SUBTARGET_H
#include "llvm/ADT/Triple.h"
+#include "llvm/IR/CallingConv.h"
#include "llvm/Target/TargetSubtargetInfo.h"
-#include "llvm/CallingConv.h"
#include <string>
#define GET_SUBTARGETINFO_HEADER
class X86Subtarget : public X86GenSubtargetInfo {
protected:
enum X86SSEEnum {
- NoMMXSSE, MMX, SSE1, SSE2, SSE3, SSSE3, SSE41, SSE42
+ NoMMXSSE, MMX, SSE1, SSE2, SSE3, SSSE3, SSE41, SSE42, AVX, AVX2
};
enum X863DNowEnum {
NoThreeDNow, ThreeDNow, ThreeDNowA
};
+ enum X86ProcFamilyEnum {
+ Others, IntelAtom
+ };
+
+ /// X86ProcFamily - X86 processor family: Intel Atom, and others
+ X86ProcFamilyEnum X86ProcFamily;
+
/// PICStyle - Which PIC style to use
///
PICStyles::Style PICStyle;
/// HasSSE4A - True if the processor supports SSE4A instructions.
bool HasSSE4A;
- /// HasAVX - Target has AVX instructions
- bool HasAVX;
-
- /// HasAVX2 - Target has AVX2 instructions
- bool HasAVX2;
-
/// HasAES - Target has AES instructions
bool HasAES;
- /// HasCLMUL - Target has carry-less multiplication
- bool HasCLMUL;
+ /// HasPCLMUL - Target has carry-less multiplication
+ bool HasPCLMUL;
- /// HasFMA3 - Target has 3-operand fused multiply-add
- bool HasFMA3;
+ /// HasFMA - Target has 3-operand fused multiply-add
+ bool HasFMA;
/// HasFMA4 - Target has 4-operand fused multiply-add
bool HasFMA4;
/// HasBMI2 - Processor has BMI2 instructions.
bool HasBMI2;
+ /// HasRTM - Processor has RTM instructions.
+ bool HasRTM;
+
+ /// HasADX - Processor has ADX instructions.
+ bool HasADX;
+
/// IsBTMemSlow - True if BT (bit test) of memory instructions are slow.
bool IsBTMemSlow;
/// this is true for most x86-64 chips, but not the first AMD chips.
bool HasCmpxchg16b;
+ /// UseLeaForSP - True if the LEA instruction should be used for adjusting
+ /// the stack pointer. This is an optimization for Intel Atom processors.
+ bool UseLeaForSP;
+
+ /// HasSlowDivide - True if smaller divides are significantly faster than
+ /// full divides and should be used when possible.
+ bool HasSlowDivide;
+
+ /// PostRAScheduler - True if using post-register-allocation scheduler.
+ bool PostRAScheduler;
+
+ /// PadShortFunctions - True if the short functions should be padded to prevent
+ /// a stall when returning too early.
+ bool PadShortFunctions;
+
/// stackAlignment - The minimum alignment known to hold of the stack frame on
/// entry to the function and which must be maintained by every function.
unsigned stackAlignment;
/// TargetTriple - What processor and OS we're targeting.
Triple TargetTriple;
+ /// Instruction itineraries for scheduling
+ InstrItineraryData InstrItins;
+
private:
+ /// StackAlignOverride - Override the stack alignment.
+ unsigned StackAlignOverride;
+
/// In64BitMode - True if compiling for 64-bit, false for 32-bit.
bool In64BitMode;
public:
-
/// This constructor initializes the data members to match that
/// of the specified triple.
///
/// instruction.
void AutoDetectSubtargetFeatures();
- bool is64Bit() const { return In64BitMode; }
+ /// \brief Reset the features for the X86 target.
+ virtual void resetSubtargetFeatures(const MachineFunction *MF);
+private:
+ void initializeEnvironment();
+ void resetSubtargetFeatures(StringRef CPU, StringRef FS);
+public:
+ /// Is this x86_64? (disregarding specific ABI / programming model)
+ bool is64Bit() const {
+ return In64BitMode;
+ }
+
+ /// Is this x86_64 with the ILP32 programming model (x32 ABI)?
+ bool isTarget64BitILP32() const {
+ return In64BitMode && (TargetTriple.getEnvironment() == Triple::GNUX32);
+ }
+
+ /// Is this x86_64 with the LP64 programming model (standard AMD64, no x32)?
+ bool isTarget64BitLP64() const {
+ return In64BitMode && (TargetTriple.getEnvironment() != Triple::GNUX32);
+ }
PICStyles::Style getPICStyle() const { return PICStyle; }
void setPICStyle(PICStyles::Style Style) { PICStyle = Style; }
bool hasSSSE3() const { return X86SSELevel >= SSSE3; }
bool hasSSE41() const { return X86SSELevel >= SSE41; }
bool hasSSE42() const { return X86SSELevel >= SSE42; }
+ bool hasAVX() const { return X86SSELevel >= AVX; }
+ bool hasAVX2() const { return X86SSELevel >= AVX2; }
+ bool hasFp256() const { return hasAVX(); }
+ bool hasInt256() const { return hasAVX2(); }
bool hasSSE4A() const { return HasSSE4A; }
bool has3DNow() const { return X863DNowLevel >= ThreeDNow; }
bool has3DNowA() const { return X863DNowLevel >= ThreeDNowA; }
bool hasPOPCNT() const { return HasPOPCNT; }
- bool hasAVX() const { return HasAVX; }
- bool hasAVX2() const { return HasAVX2; }
- bool hasXMM() const { return hasSSE1() || hasAVX(); }
- bool hasXMMInt() const { return hasSSE2() || hasAVX(); }
- bool hasSSE1orAVX() const { return hasSSE1() || hasAVX(); }
- bool hasSSE3orAVX() const { return hasSSE3() || hasAVX(); }
- bool hasSSSE3orAVX() const { return hasSSSE3() || hasAVX(); }
- bool hasSSE41orAVX() const { return hasSSE41() || hasAVX(); }
- bool hasSSE42orAVX() const { return hasSSE42() || hasAVX(); }
bool hasAES() const { return HasAES; }
- bool hasCLMUL() const { return HasCLMUL; }
- bool hasFMA3() const { return HasFMA3; }
- bool hasFMA4() const { return HasFMA4; }
+ bool hasPCLMUL() const { return HasPCLMUL; }
+ bool hasFMA() const { return HasFMA; }
+ // FIXME: Favor FMA when both are enabled. Is this the right thing to do?
+ bool hasFMA4() const { return HasFMA4 && !HasFMA; }
bool hasXOP() const { return HasXOP; }
bool hasMOVBE() const { return HasMOVBE; }
bool hasRDRAND() const { return HasRDRAND; }
bool hasLZCNT() const { return HasLZCNT; }
bool hasBMI() const { return HasBMI; }
bool hasBMI2() const { return HasBMI2; }
+ bool hasRTM() const { return HasRTM; }
+ bool hasADX() const { return HasADX; }
bool isBTMemSlow() const { return IsBTMemSlow; }
bool isUnalignedMemAccessFast() const { return IsUAMemFast; }
bool hasVectorUAMem() const { return HasVectorUAMem; }
bool hasCmpxchg16b() const { return HasCmpxchg16b; }
+ bool useLeaForSP() const { return UseLeaForSP; }
+ bool hasSlowDivide() const { return HasSlowDivide; }
+ bool padShortFunctions() const { return PadShortFunctions; }
+
+ bool isAtom() const { return X86ProcFamily == IntelAtom; }
const Triple &getTargetTriple() const { return TargetTriple; }
bool isTargetSolaris() const {
return TargetTriple.getOS() == Triple::Solaris;
}
-
- // ELF is a reasonably sane default and the only other X86 targets we
- // support are Darwin and Windows. Just use "not those".
bool isTargetELF() const {
- return !isTargetDarwin() && !isTargetWindows() && !isTargetCygMing();
+ return (TargetTriple.getEnvironment() == Triple::ELF ||
+ TargetTriple.isOSBinFormatELF());
}
bool isTargetLinux() const { return TargetTriple.getOS() == Triple::Linux; }
bool isTargetNaCl() const {
- return TargetTriple.getOS() == Triple::NativeClient;
+ return TargetTriple.getOS() == Triple::NaCl;
}
bool isTargetNaCl32() const { return isTargetNaCl() && !is64Bit(); }
bool isTargetNaCl64() const { return isTargetNaCl() && is64Bit(); }
-
bool isTargetWindows() const { return TargetTriple.getOS() == Triple::Win32; }
bool isTargetMingw() const { return TargetTriple.getOS() == Triple::MinGW32; }
bool isTargetCygwin() const { return TargetTriple.getOS() == Triple::Cygwin; }
- bool isTargetCygMing() const {
- return isTargetMingw() || isTargetCygwin();
- }
-
- /// isTargetCOFF - Return true if this is any COFF/Windows target variant.
+ bool isTargetCygMing() const { return TargetTriple.isOSCygMing(); }
bool isTargetCOFF() const {
- return isTargetMingw() || isTargetCygwin() || isTargetWindows();
+ return (TargetTriple.getEnvironment() != Triple::ELF &&
+ TargetTriple.isOSBinFormatCOFF());
}
+ bool isTargetEnvMacho() const { return TargetTriple.isEnvironmentMachO(); }
bool isTargetWin64() const {
// FIXME: x86_64-cygwin has not been released yet.
- return In64BitMode && (isTargetCygMing() || isTargetWindows());
- }
-
- bool isTargetEnvMacho() const {
- return isTargetDarwin() || (TargetTriple.getEnvironment() == Triple::MachO);
+ return In64BitMode && TargetTriple.isOSWindows();
}
bool isTargetWin32() const {
+ // FIXME: Cygwin is included for isTargetWin64 -- should it be included
+ // here too?
return !In64BitMode && (isTargetMingw() || isTargetWindows());
}
/// memset with zero passed as the second argument. Otherwise it
/// returns null.
const char *getBZeroEntry() const;
+
+ /// This function returns true if the target has sincos() routine in its
+ /// compiler runtime or math libraries.
+ bool hasSinCos() const;
+
+ /// enablePostRAScheduler - run for Atom optimization.
+ bool enablePostRAScheduler(CodeGenOpt::Level OptLevel,
+ TargetSubtargetInfo::AntiDepBreakMode& Mode,
+ RegClassVector& CriticalPathRCs) const;
+
+ bool postRAScheduler() const { return PostRAScheduler; }
- /// getSpecialAddressLatency - For targets where it is beneficial to
- /// backschedule instructions that compute addresses, return a value
- /// indicating the number of scheduling cycles of backscheduling that
- /// should be attempted.
- unsigned getSpecialAddressLatency() const;
+ /// getInstrItins = Return the instruction itineraries based on the
+ /// subtarget selection.
+ const InstrItineraryData &getInstrItineraryData() const { return InstrItins; }
};
} // End llvm namespace
projects / oota-llvm.git / blobdiff