[AArch64, ARM] Add v8.1a architecture and generic cpu

[vsukharev]

Summary

Based on http://reviews.llvm.org/D8504

[t.p.northover]

This looks OK, with one small clarification. (Feel free to commit with that change unless you disagree).

Tim.

lib/Target/ARM/ARMAsmPrinter.cpp

669 ↗

(On Diff #22390)

I think this is a bit non-obvious. I grant that it's correct (and equivalent to !StringRef::startswith), but I think this looks too much like "if generic is in CPUString".

I'd suggest an explicit comparison against 0, which might make people at least consider whether string::find returns an index and what it returns for failure.

[vsukharev]

• CPUString.find("generic") is replaced with more understandable construction
• The first revision was based not on trunk, but on patches, that are not yet approved. To be able to commit, definitions of "V8_1" subtarget feature added from http://reviews.llvm.org/D8503 (AArch64) and http://reviews.llvm.org/D8498 (ARM). Now the second revision is based on ToT.

Tim, would you please approve the second change?

llvm/trunk/include/llvm/ADT/Triple.h

enum ArchType {
     hsail64,    // AMD HSAIL with 64-bit pointers
     spir,       // SPIR: standard portable IR for OpenCL 32-bit version
     spir64,     // SPIR: standard portable IR for OpenCL 64-bit version
     kalimba     // Kalimba: generic kalimba
   };
   enum SubArchType {
     NoSubArch,
 
+    ARMSubArch_v8_1a,
     ARMSubArch_v8,
     ARMSubArch_v7,
     ARMSubArch_v7em,
     ARMSubArch_v7m,
     ARMSubArch_v7s,
     ARMSubArch_v6,
     ARMSubArch_v6m,
     ARMSubArch_v6k,

llvm/trunk/include/llvm/Support/ARMBuildAttributes.h

enum CPUArch {
   v6       = 6,   // e.g. ARM1136J_S
   v6KZ     = 7,   // e.g. ARM1176JZ_S
   v6T2     = 8,   // e.g. ARM1156T2_S
   v6K      = 9,   // e.g. ARM1176JZ_S
   v7       = 10,  // e.g. Cortex A8, Cortex M3
   v6_M     = 11,  // e.g. Cortex M1
   v6S_M    = 12,  // v6_M with the System extensions
   v7E_M    = 13,  // v7_M with DSP extensions
-  v8       = 14   // v8, AArch32
+  v8       = 14,  // v8,v8.1a AArch32
 };
 
 enum CPUArchProfile {               // (=7), uleb128
   Not_Applicable          = 0,      // pre v7, or cross-profile code
   ApplicationProfile      = (0x41), // 'A' (e.g. for Cortex A8)
   RealTimeProfile         = (0x52), // 'R' (e.g. for Cortex R4)
   MicroControllerProfile  = (0x4D), // 'M' (e.g. for Cortex M3)
   SystemProfile           = (0x53)  // 'S' Application or real-time profile
enum {
   // Tag_WMMX_arch, (=11), uleb128
   AllowWMMXv1 = 1,  // The user permitted this entity to use WMMX v1
   AllowWMMXv2 = 2,  // The user permitted this entity to use WMMX v2
 
   // Tag_Advanced_SIMD_arch, (=12), uleb128
   AllowNeon = 1,      // SIMDv1 was permitted
   AllowNeon2 = 2,     // SIMDv2 was permitted (Half-precision FP, MAC operations)
   AllowNeonARMv8 = 3, // ARM v8-A SIMD was permitted
+  AllowNeonARMv8_1a = 4,// ARM v8.1-A SIMD was permitted (RDMA)
 
   // Tag_ABI_PCS_R9_use, (=14), uleb128
   R9IsGPR = 0,        // R9 used as v6 (just another callee-saved register)
   R9IsSB = 1,         // R9 used as a global static base rgister
   R9IsTLSPointer = 2, // R9 used as a thread local storage pointer
   R9Reserved = 3,     // R9 not used by code associated with attributed entity
 
   // Tag_ABI_PCS_RW_data, (=15), uleb128

llvm/trunk/lib/Support/Triple.cpp

return StringSwitch<Triple::ArchType>(ArchName.substr(offset))
     .Cases("v3", "v3m", isThumb ? Triple::UnknownArch : arch)
     .Cases("v4", "v4t", arch)
     .Cases("v5", "v5e", "v5t", "v5te", "v5tej", arch)
     .Cases("v6", "v6j", "v6k", "v6m", "v6sm", arch)
     .Cases("v6t2", "v6z", "v6zk", arch)
     .Cases("v7", "v7a", "v7em", "v7l", arch)
     .Cases("v7m", "v7r", "v7s", arch)
     .Cases("v8", "v8a", arch)
+    .Cases("v8.1", "v8.1a", arch)
     .Default(Triple::UnknownArch);
 }
 
 static Triple::ArchType parseArch(StringRef ArchName) {
   Triple::ArchType ARMArch(parseARMArch(ArchName));
 
   return StringSwitch<Triple::ArchType>(ArchName)
     .Cases("i386", "i486", "i586", "i686", Triple::x86)
return StringSwitch<Triple::ObjectFormatType>(EnvironmentName)
     .Default(Triple::UnknownObjectFormat);
 }
 
 static Triple::SubArchType parseSubArch(StringRef SubArchName) {
   if (SubArchName.endswith("eb"))
     SubArchName = SubArchName.substr(0, SubArchName.size() - 2);
 
   return StringSwitch<Triple::SubArchType>(SubArchName)
+    .EndsWith("v8.1a", Triple::ARMSubArch_v8_1a)
     .EndsWith("v8", Triple::ARMSubArch_v8)
     .EndsWith("v8a", Triple::ARMSubArch_v8)
     .EndsWith("v7", Triple::ARMSubArch_v7)
     .EndsWith("v7a", Triple::ARMSubArch_v7)
     .EndsWith("v7em", Triple::ARMSubArch_v7em)
     .EndsWith("v7l", Triple::ARMSubArch_v7)
     .EndsWith("v7m", Triple::ARMSubArch_v7m)
     .EndsWith("v7r", Triple::ARMSubArch_v7)
result = llvm::StringSwitch<const char *>(MArch.substr(offset))
       .Case("v6t2", "arm1156t2-s")
       .Cases("v6m", "v6-m", "v6sm", "v6s-m", "cortex-m0")
       .Cases("v7", "v7a", "v7-a", "v7l", "v7-l", "cortex-a8")
       .Cases("v7s", "v7-s", "swift")
       .Cases("v7r", "v7-r", "cortex-r4")
       .Cases("v7m", "v7-m", "cortex-m3")
       .Cases("v7em", "v7e-m", "cortex-m4")
       .Cases("v8", "v8a", "v8-a", "cortex-a53")
+      .Cases("v8.1a", "v8.1-a", "generic-armv8.1-a")
       .Default(nullptr);
   else
     result = llvm::StringSwitch<const char *>(MArch)
       .Case("ep9312", "ep9312")
       .Case("iwmmxt", "iwmmxt")
       .Case("xscale", "xscale")
       .Default(nullptr);
 

llvm/trunk/lib/Target/AArch64/AArch64.td

def FeatureNEON : SubtargetFeature<"neon", "HasNEON", "true",
   "Enable Advanced SIMD instructions", [FeatureFPARMv8]>;
 
 def FeatureCrypto : SubtargetFeature<"crypto", "HasCrypto", "true",
   "Enable cryptographic instructions">;
 
 def FeatureCRC : SubtargetFeature<"crc", "HasCRC", "true",
   "Enable ARMv8 CRC-32 checksum instructions">;
 
+def FeatureV8_1a : SubtargetFeature<"v8.1a", "HasV8_1a", "true",
+  "Enable ARMv8.1a extensions", [FeatureCRC]>;
+
 /// Cyclone has register move instructions which are "free".
 def FeatureZCRegMove : SubtargetFeature<"zcm", "HasZeroCycleRegMove", "true",
                                         "Has zero-cycle register moves">;
 
 /// Cyclone has instructions which zero registers for "free".
 def FeatureZCZeroing : SubtargetFeature<"zcz", "HasZeroCycleZeroing", "true",
                                         "Has zero-cycle zeroing instructions">;
 
def ProcCyclone : SubtargetFeature<"cyclone", "ARMProcFamily", "Cyclone",
                                    FeatureCrypto,
                                    FeatureCRC,
                                    FeatureZCRegMove, FeatureZCZeroing]>;
 
 def : ProcessorModel<"generic", NoSchedModel, [FeatureFPARMv8,
                                               FeatureNEON,
                                               FeatureCRC]>;
 
+def : ProcessorModel<"generic-armv8.1-a", NoSchedModel, [FeatureV8_1a,
+                                                         FeatureNEON,
+                                                         FeatureCrypto]>;
+
 def : ProcessorModel<"cortex-a53", CortexA53Model, [ProcA53]>;
 def : ProcessorModel<"cortex-a57", CortexA57Model, [ProcA57]>;
 // FIXME: Cortex-A72 is currently modelled as an Cortex-A57.
 def : ProcessorModel<"cortex-a72", CortexA57Model, [ProcA57]>;
 def : ProcessorModel<"cyclone", CycloneModel, [ProcCyclone]>;
 
 //===----------------------------------------------------------------------===//
 // Assembly parser

llvm/trunk/lib/Target/AArch64/AArch64InstrInfo.td

 def HasFPARMv8       : Predicate<"Subtarget->hasFPARMv8()">,
                                AssemblerPredicate<"FeatureFPARMv8", "fp-armv8">;
 def HasNEON          : Predicate<"Subtarget->hasNEON()">,
                                  AssemblerPredicate<"FeatureNEON", "neon">;
 def HasCrypto        : Predicate<"Subtarget->hasCrypto()">,
                                  AssemblerPredicate<"FeatureCrypto", "crypto">;
 def HasCRC           : Predicate<"Subtarget->hasCRC()">,
                                  AssemblerPredicate<"FeatureCRC", "crc">;
+def HasV8_1a         : Predicate<"Subtarget->hasV8_1a()">,
+                                 AssemblerPredicate<"FeatureV8_1a", "v8.1a">;
 def IsLE             : Predicate<"Subtarget->isLittleEndian()">;
 def IsBE             : Predicate<"!Subtarget->isLittleEndian()">;
 def IsCyclone        : Predicate<"Subtarget->isCyclone()">;
 
 //===----------------------------------------------------------------------===//
 // AArch64-specific DAG Nodes.
 //
 

llvm/trunk/lib/Target/AArch64/AArch64Subtarget.h

protected:
 
   /// ARMProcFamily - ARM processor family: Cortex-A53, Cortex-A57, and others.
   ARMProcFamilyEnum ARMProcFamily;
 
   bool HasFPARMv8;
   bool HasNEON;
   bool HasCrypto;
   bool HasCRC;
+  bool HasV8_1a;
 
   // HasZeroCycleRegMove - Has zero-cycle register mov instructions.
   bool HasZeroCycleRegMove;
 
   // HasZeroCycleZeroing - Has zero-cycle zeroing instructions.
   bool HasZeroCycleZeroing;
 
   bool IsLittle;
public:
   bool hasZeroCycleRegMove() const { return HasZeroCycleRegMove; }
 
   bool hasZeroCycleZeroing() const { return HasZeroCycleZeroing; }
 
   bool hasFPARMv8() const { return HasFPARMv8; }
   bool hasNEON() const { return HasNEON; }
   bool hasCrypto() const { return HasCrypto; }
   bool hasCRC() const { return HasCRC; }
+  bool hasV8_1a() const { return HasV8_1a; }
 
   bool isLittleEndian() const { return IsLittle; }
 
   bool isTargetDarwin() const { return TargetTriple.isOSDarwin(); }
   bool isTargetIOS() const { return TargetTriple.isiOS(); }
   bool isTargetLinux() const { return TargetTriple.isOSLinux(); }
   bool isTargetWindows() const { return TargetTriple.isOSWindows(); }
 

llvm/trunk/lib/Target/AArch64/AArch64Subtarget.cpp

 }
 
 AArch64Subtarget::AArch64Subtarget(const std::string &TT,
                                    const std::string &CPU,
                                    const std::string &FS,
                                    const TargetMachine &TM, bool LittleEndian)
     : AArch64GenSubtargetInfo(TT, CPU, FS), ARMProcFamily(Others),
       HasFPARMv8(false), HasNEON(false), HasCrypto(false), HasCRC(false),
-      HasZeroCycleRegMove(false), HasZeroCycleZeroing(false),
+      HasV8_1a(false), HasZeroCycleRegMove(false), HasZeroCycleZeroing(false),
       IsLittle(LittleEndian), CPUString(CPU), TargetTriple(TT), FrameLowering(),
       InstrInfo(initializeSubtargetDependencies(FS)),
       TSInfo(TM.getDataLayout()), TLInfo(TM, *this) {}
 
 /// ClassifyGlobalReference - Find the target operand flags that describe
 /// how a global value should be referenced for the current subtarget.
 unsigned char
 AArch64Subtarget::ClassifyGlobalReference(const GlobalValue *GV,

llvm/trunk/lib/Target/ARM/ARM.td

def HasV6T2Ops  : SubtargetFeature<"v6t2", "HasV6T2Ops", "true",
                                    [HasV6MOps, HasV6KOps, FeatureThumb2]>;
 def HasV7Ops    : SubtargetFeature<"v7", "HasV7Ops", "true",
                                    "Support ARM v7 instructions",
                                    [HasV6T2Ops, FeaturePerfMon]>;
 def HasV8Ops    : SubtargetFeature<"v8", "HasV8Ops", "true",
                                    "Support ARM v8 instructions",
                                    [HasV7Ops, FeatureVirtualization,
                                     FeatureMP]>;
+def FeatureV8_1a : SubtargetFeature<"v8.1a", "HasV8_1a", "true",
+                                   "Support ARM v8.1a instructions",
+                                   [HasV8Ops, FeatureAClass, FeatureCRC]>;
 
 //===----------------------------------------------------------------------===//
 // ARM Processors supported.
 //
 
 include "ARMSchedule.td"
 
 // ARM processor families.

 
 // Cyclone is very similar to swift
 def : ProcessorModel<"cyclone",     SwiftModel,
                                     [ProcSwift, HasV8Ops, HasV7Ops,
                                      FeatureCrypto, FeatureFPARMv8,
                                      FeatureDB,FeatureDSPThumb2,
                                      FeatureHasRAS, FeatureZCZeroing]>;
 
+// V8.1 Processors
+def : ProcNoItin<"generic-armv8.1-a", [HasV8Ops, FeatureV8_1a,
+                                       FeatureDB, FeatureFPARMv8,
+                                       FeatureNEON, FeatureDSPThumb2,
+                                       FeatureHWDiv, FeatureHWDivARM,
+                                       FeatureTrustZone, FeatureT2XtPk,
+                                       FeatureCrypto]>;
+
 //===----------------------------------------------------------------------===//
 // Register File Description
 //===----------------------------------------------------------------------===//
 
 include "ARMRegisterInfo.td"
 
 include "ARMCallingConv.td"
 

llvm/trunk/lib/Target/ARM/ARMAsmPrinter.cpp

else
       ArchFS = FS;
   }
   const ARMBaseTargetMachine &ATM =
       static_cast<const ARMBaseTargetMachine &>(TM);
   const ARMSubtarget STI(TT, CPU, ArchFS, ATM, ATM.isLittleEndian());
 
   std::string CPUString = STI.getCPUString();
 
-  if (CPUString != "generic") {
+  if (CPUString.find("generic") != 0) { //CPUString doesn't start with "generic"
     // FIXME: remove krait check when GNU tools support krait cpu
     if (STI.isKrait()) {
       ATS.emitTextAttribute(ARMBuildAttrs::CPU_name, "cortex-a9");
       // We consider krait as a "cortex-a9" + hwdiv CPU
       // Enable hwdiv through ".arch_extension idiv"
       if (STI.hasDivide() || STI.hasDivideInARMMode())
         ATS.emitArchExtension(ARM::HWDIV);
     } else
if (STI.hasFPARMv8()) {
         ATS.emitFPU(ARM::NEON_FP_ARMV8);
     } else if (STI.hasVFP4())
       ATS.emitFPU(ARM::NEON_VFPV4);
     else
       ATS.emitFPU(ARM::NEON);
     // Emit Tag_Advanced_SIMD_arch for ARMv8 architecture
     if (STI.hasV8Ops())
       ATS.emitAttribute(ARMBuildAttrs::Advanced_SIMD_arch,
+                        STI.hasV8_1a() ? ARMBuildAttrs::AllowNeonARMv8_1a:
                                          ARMBuildAttrs::AllowNeonARMv8);
   } else {
     if (STI.hasFPARMv8())
       // FPv5 and FP-ARMv8 have the same instructions, so are modeled as one
       // FPU, but there are two different names for it depending on the CPU.
       ATS.emitFPU(STI.hasD16() ? ARM::FPV5_D16 : ARM::FP_ARMV8);
     else if (STI.hasVFP4())
       ATS.emitFPU(STI.hasD16() ? ARM::VFPV4_D16 : ARM::VFPV4);
     else if (STI.hasVFP3())

llvm/trunk/lib/Target/ARM/ARMSubtarget.h

protected:
   bool HasTrustZone;
 
   /// HasCrypto - if true, processor supports Cryptography extensions
   bool HasCrypto;
 
   /// HasCRC - if true, processor supports CRC instructions
   bool HasCRC;
 
+  /// HasV8_1a - if true, the processor has V8.1a: PAN and RDMA extensions
+  bool HasV8_1a;
+
   /// If true, the instructions "vmov.i32 d0, #0" and "vmov.i32 q0, #0" are
   /// particularly effective at zeroing a VFP register.
   bool HasZeroCycleZeroing;
 
   /// AllowsUnalignedMem - If true, the subtarget allows unaligned memory
   /// accesses for some types.  For details, see
   /// ARMTargetLowering::allowsMisalignedMemoryAccesses().
   bool AllowsUnalignedMem;
public:
 
   bool hasVFP2() const { return HasVFPv2; }
   bool hasVFP3() const { return HasVFPv3; }
   bool hasVFP4() const { return HasVFPv4; }
   bool hasFPARMv8() const { return HasFPARMv8; }
   bool hasNEON() const { return HasNEON;  }
   bool hasCrypto() const { return HasCrypto; }
   bool hasCRC() const { return HasCRC; }
+  bool hasV8_1a() const { return HasV8_1a; }
   bool hasVirtualization() const { return HasVirtualization; }
   bool useNEONForSinglePrecisionFP() const {
     return hasNEON() && UseNEONForSinglePrecisionFP;
   }
 
   bool hasDivide() const { return HasHardwareDivide; }
   bool hasDivideInARMMode() const { return HasHardwareDivideInARM; }
   bool hasT2ExtractPack() const { return HasT2ExtractPack; }

llvm/trunk/lib/Target/ARM/ARMSubtarget.cpp

void ARMSubtarget::initializeEnvironment() {
   HasRAS = false;
   HasMPExtension = false;
   HasVirtualization = false;
   FPOnlySP = false;
   HasPerfMon = false;
   HasTrustZone = false;
   HasCrypto = false;
   HasCRC = false;
+  HasV8_1a = false;
   HasZeroCycleZeroing = false;
   AllowsUnalignedMem = false;
   Thumb2DSP = false;
   UseNaClTrap = false;
   UnsafeFPMath = false;
 }
 
 void ARMSubtarget::initSubtargetFeatures(StringRef CPU, StringRef FS) {

llvm/trunk/lib/Target/ARM/AsmParser/ARMAsmParser.cpp

bool hasARM() const {
     return !(STI.getFeatureBits() & ARM::FeatureNoARM);
   }
   bool hasThumb2DSP() const {
     return STI.getFeatureBits() & ARM::FeatureDSPThumb2;
   }
   bool hasD16() const {
     return STI.getFeatureBits() & ARM::FeatureD16;
   }
+  bool hasV8_1a() const {
+    return STI.getFeatureBits() & ARM::FeatureV8_1a;
+  }
 
   void SwitchMode() {
     uint64_t FB = ComputeAvailableFeatures(STI.ToggleFeature(ARM::ModeThumb));
     setAvailableFeatures(FB);
   }
   bool isMClass() const {
     return STI.getFeatureBits() & ARM::FeatureMClass;
   }

llvm/trunk/lib/Target/ARM/MCTargetDesc/ARMArchName.def

 ARM_ARCH_NAME("armv7-a", ARMV7A,  "7-A",     v7)
 ARM_ARCH_ALIAS("armv7a", ARMV7A)
 ARM_ARCH_NAME("armv7-r", ARMV7R,  "7-R",     v7)
 ARM_ARCH_ALIAS("armv7r", ARMV7R)
 ARM_ARCH_NAME("armv7-m", ARMV7M,  "7-M",     v7)
 ARM_ARCH_ALIAS("armv7m", ARMV7M)
 ARM_ARCH_NAME("armv8-a", ARMV8A,  "8-A",     v8)
 ARM_ARCH_ALIAS("armv8a", ARMV8A)
+ARM_ARCH_NAME("armv8.1-a", ARMV8_1A, "8.1-A", v8)
+ARM_ARCH_ALIAS("armv8.1a", ARMV8_1A)
 ARM_ARCH_NAME("iwmmxt",  IWMMXT,  "iwmmxt",  v5TE)
 ARM_ARCH_NAME("iwmmxt2", IWMMXT2, "iwmmxt2", v5TE)
 
 #undef ARM_ARCH_NAME
 #undef ARM_ARCH_ALIAS

llvm/trunk/lib/Target/ARM/MCTargetDesc/ARMELFStreamer.cpp

case ARM::ARMV7R:
     break;
 
   case ARM::ARMV7M:
     setAttributeItem(CPU_arch_profile, MicroControllerProfile, false);
     setAttributeItem(THUMB_ISA_use, AllowThumb32, false);
     break;
 
   case ARM::ARMV8A:
+  case ARM::ARMV8_1A:
     setAttributeItem(CPU_arch_profile, ApplicationProfile, false);
     setAttributeItem(ARM_ISA_use, Allowed, false);
     setAttributeItem(THUMB_ISA_use, AllowThumb32, false);
     setAttributeItem(MPextension_use, Allowed, false);
     setAttributeItem(Virtualization_use, AllowTZVirtualization, false);
     break;
 
   case ARM::IWMMXT:
setAttributeItem(ARMBuildAttrs::Advanced_SIMD_arch,
                      /* OverwriteExisting= */ false);
     break;
 
   case ARM::NEON_FP_ARMV8:
   case ARM::CRYPTO_NEON_FP_ARMV8:
     setAttributeItem(ARMBuildAttrs::FP_arch,
                      ARMBuildAttrs::AllowFPARMv8A,
                      /* OverwriteExisting= */ false);
-    setAttributeItem(ARMBuildAttrs::Advanced_SIMD_arch,
-                     ARMBuildAttrs::AllowNeonARMv8,
-                     /* OverwriteExisting= */ false);
+    // 'Advanced_SIMD_arch' must be emitted not here, but within
+    // ARMAsmPrinter::emitAttributes(), depending on hasV8Ops() and hasV8_1a()
     break;
 
   case ARM::SOFTVFP:
     break;
 
   default:
     report_fatal_error("Unknown FPU: " + Twine(FPU));
     break;

llvm/trunk/lib/Target/ARM/MCTargetDesc/ARMMCTargetDesc.cpp

if (NoCPU)
       //      FeatureMP, FeatureHWDiv, FeatureHWDivARM, FeatureTrustZone,
       //      FeatureT2XtPk, FeatureCrypto, FeatureCRC
       ARMArchFeature = "+v8,+db,+fp-armv8,+neon,+t2dsp,+mp,+hwdiv,+hwdiv-arm,"
                        "+trustzone,+t2xtpk,+crypto,+crc";
     else
       // Use CPU to figure out the exact features
       ARMArchFeature = "+v8";
     break;
+  case Triple::ARMSubArch_v8_1a:
+    if (NoCPU)
+      // v8.1a: FeatureDB, FeatureFPARMv8, FeatureNEON, FeatureDSPThumb2,
+      //      FeatureMP, FeatureHWDiv, FeatureHWDivARM, FeatureTrustZone,
+      //      FeatureT2XtPk, FeatureCrypto, FeatureCRC, FeatureV8_1a
+      ARMArchFeature = "+v8.1a,+db,+fp-armv8,+neon,+t2dsp,+mp,+hwdiv,+hwdiv-arm,"
+                       "+trustzone,+t2xtpk,+crypto,+crc";
+    else
+      // Use CPU to figure out the exact features
+      ARMArchFeature = "+v8.1a";
+    break;
   case Triple::ARMSubArch_v7m:
     isThumb = true;
     if (NoCPU)
       // v7m: FeatureNoARM, FeatureDB, FeatureHWDiv, FeatureMClass
       ARMArchFeature = "+v7,+noarm,+db,+hwdiv,+mclass";
     else
       // Use CPU to figure out the exact features.
       ARMArchFeature = "+v7";

llvm/trunk/test/CodeGen/ARM/build-attributes.ll

 ; RUN: llc < %s -mtriple=armv8-linux-gnueabi -mcpu=cortex-a53  -enable-unsafe-fp-math -disable-fp-elim -enable-no-infs-fp-math -enable-no-nans-fp-math -fp-contract=fast | FileCheck %s --check-prefix=CORTEX-A53-FAST
 ; RUN: llc < %s -mtriple=armv8-linux-gnueabi -mcpu=cortex-a53 -enable-sign-dependent-rounding-fp-math | FileCheck %s --check-prefix=DYN-ROUNDING
 ; RUN: llc < %s -mtriple=armv8-linux-gnueabi -mcpu=cortex-a57 | FileCheck %s --check-prefix=CORTEX-A57
 ; RUN: llc < %s -mtriple=armv8-linux-gnueabi -mcpu=cortex-a57  -enable-unsafe-fp-math -disable-fp-elim -enable-no-infs-fp-math -enable-no-nans-fp-math -fp-contract=fast | FileCheck %s --check-prefix=CORTEX-A57-FAST
 ; RUN: llc < %s -mtriple=armv8-linux-gnueabi -mcpu=cortex-a57 -enable-sign-dependent-rounding-fp-math | FileCheck %s --check-prefix=DYN-ROUNDING
 ; RUN: llc < %s -mtriple=armv8-linux-gnueabi -mcpu=cortex-a72 | FileCheck %s --check-prefix=CORTEX-A72
 ; RUN: llc < %s -mtriple=armv8-linux-gnueabi -mcpu=cortex-a72  -enable-unsafe-fp-math -disable-fp-elim -enable-no-infs-fp-math -enable-no-nans-fp-math -fp-contract=fast | FileCheck %s --check-prefix=CORTEX-A72-FAST
 ; RUN: llc < %s -mtriple=armv8-linux-gnueabi -mcpu=cortex-a72 -enable-sign-dependent-rounding-fp-math | FileCheck %s --check-prefix=DYN-ROUNDING
+; RUN: llc < %s -mtriple=armv8.1a-linux-gnueabi | FileCheck %s --check-prefix=GENERIC-ARMV8_1-A
+; RUN: llc < %s -mtriple=armv8.1a-linux-gnueabi  -enable-unsafe-fp-math -disable-fp-elim -enable-no-infs-fp-math -enable-no-nans-fp-math -fp-contract=fast | FileCheck %s --check-prefix=GENERIC-ARMV8_1-A-FAST
+; RUN: llc < %s -mtriple=armv8.1a-linux-gnueabi -enable-sign-dependent-rounding-fp-math | FileCheck %s --check-prefix=DYN-ROUNDING
 ; RUN: llc < %s -mtriple=armv7-none-linux-gnueabi -mcpu=cortex-a7 | FileCheck %s  --check-prefix=CORTEX-A7-CHECK
 ; RUN: llc < %s -mtriple=armv7-none-linux-gnueabi -mcpu=cortex-a7  -enable-unsafe-fp-math -disable-fp-elim -enable-no-infs-fp-math -enable-no-nans-fp-math -fp-contract=fast | FileCheck %s  --check-prefix=CORTEX-A7-CHECK-FAST
 ; RUN: llc < %s -mtriple=armv7-none-linux-gnueabi -mcpu=cortex-a7 -mattr=-vfp2,-vfp3,-vfp4,-neon | FileCheck %s --check-prefix=CORTEX-A7-NOFPU
 ; RUN: llc < %s -mtriple=armv7-none-linux-gnueabi -mcpu=cortex-a7 -mattr=-vfp2,-vfp3,-vfp4,-neon  -enable-unsafe-fp-math -disable-fp-elim -enable-no-infs-fp-math -enable-no-nans-fp-math -fp-contract=fast | FileCheck %s --check-prefix=CORTEX-A7-NOFPU-FAST
 ; RUN: llc < %s -mtriple=armv7-none-linux-gnueabi -mcpu=cortex-a7 -mattr=+vfp4,-neon | FileCheck %s --check-prefix=CORTEX-A7-FPUV4
 ; RUN: llc < %s -mtriple=armv7-none-linux-gnueabi -mcpu=cortex-a7 -enable-sign-dependent-rounding-fp-math | FileCheck %s --check-prefix=DYN-ROUNDING
 ; RUN: llc < %s -mtriple=armv7-none-linux-gnueabi -mcpu=cortex-a7 -mattr=+vfp4,-neon  -enable-unsafe-fp-math -disable-fp-elim -enable-no-infs-fp-math -enable-no-nans-fp-math -fp-contract=fast | FileCheck %s --check-prefix=CORTEX-A7-FPUV4-FAST
 ; RUN: llc < %s -mtriple=armv7-none-linux-gnueabi -mcpu=cortex-a7 -mattr=+vfp4,,+d16,-neon | FileCheck %s --check-prefix=CORTEX-A7-FPUV4
 ; RUN: llc < %s -mtriple=arm-none-linux-gnueabi -relocation-model=pic | FileCheck %s --check-prefix=RELOC-PIC
 ; RUN: llc < %s -mtriple=arm-none-linux-gnueabi -relocation-model=static | FileCheck %s --check-prefix=RELOC-OTHER
 ; RUN: llc < %s -mtriple=arm-none-linux-gnueabi -relocation-model=default | FileCheck %s --check-prefix=RELOC-OTHER
 ; RUN: llc < %s -mtriple=arm-none-linux-gnueabi -relocation-model=dynamic-no-pic | FileCheck %s --check-prefix=RELOC-OTHER
 ; RUN: llc < %s -mtriple=arm-none-linux-gnueabi | FileCheck %s --check-prefix=RELOC-OTHER
 ; RUN: llc < %s -mtriple=arm-none-linux-gnueabi | FileCheck %s --check-prefix=PCS-R9-USE
 ; RUN: llc < %s -mtriple=arm-none-linux-gnueabi -arm-reserve-r9 | FileCheck %s --check-prefix=PCS-R9-RESERVE
 
+; ARMv8.1a (AArch32)
+; RUN: llc < %s -mtriple=armv8.1a-none-linux-gnueabi -arm-no-strict-align | FileCheck %s --check-prefix=NO-STRICT-ALIGN
+; RUN: llc < %s -mtriple=armv8.1a-none-linux-gnueabi -arm-strict-align | FileCheck %s --check-prefix=STRICT-ALIGN
+; RUN: llc < %s -mtriple=armv8.1a-none-linux-gnueabi | FileCheck %s --check-prefix=NO-STRICT-ALIGN
 ; ARMv8a (AArch32)
 ; RUN: llc < %s -mtriple=armv8-none-linux-gnueabi -mcpu=cortex-a57 -arm-no-strict-align | FileCheck %s --check-prefix=NO-STRICT-ALIGN
 ; RUN: llc < %s -mtriple=armv8-none-linux-gnueabi -mcpu=cortex-a57 -arm-strict-align | FileCheck %s --check-prefix=STRICT-ALIGN
 ; RUN: llc < %s -mtriple=armv8-none-linux-gnueabi -mcpu=cortex-a57 | FileCheck %s --check-prefix=NO-STRICT-ALIGN
 ; RUN: llc < %s -mtriple=armv8-none-linux-gnueabi -mcpu=cortex-a72 -arm-no-strict-align | FileCheck %s --check-prefix=NO-STRICT-ALIGN
 ; RUN: llc < %s -mtriple=armv8-none-linux-gnueabi -mcpu=cortex-a72 -arm-strict-align | FileCheck %s --check-prefix=STRICT-ALIGN
 ; RUN: llc < %s -mtriple=armv8-none-linux-gnueabi -mcpu=cortex-a72 | FileCheck %s --check-prefix=NO-STRICT-ALIGN
 ; ARMv7a

 
 ; CORTEX-A72-FAST-NOT:   .eabi_attribute 19
 ;; The A72 has the ARMv8 FP unit, which always flushes preserving sign.
 ; CORTEX-A72-FAST:  .eabi_attribute 20, 2
 ; CORTEX-A72-FAST-NOT:  .eabi_attribute 21
 ; CORTEX-A72-FAST-NOT:  .eabi_attribute 22
 ; CORTEX-A72-FAST:  .eabi_attribute 23, 1
 
+; GENERIC-ARMV8_1-A:  .eabi_attribute 6, 14
+; GENERIC-ARMV8_1-A:  .eabi_attribute 7, 65
+; GENERIC-ARMV8_1-A:  .eabi_attribute 8, 1
+; GENERIC-ARMV8_1-A:  .eabi_attribute 9, 2
+; GENERIC-ARMV8_1-A:  .fpu crypto-neon-fp-armv8
+; GENERIC-ARMV8_1-A:  .eabi_attribute 12, 4
+; GENERIC-ARMV8_1-A-NOT:   .eabi_attribute 19
+;; We default to IEEE 754 compliance
+; GENERIC-ARMV8_1-A:  .eabi_attribute 20, 1
+; GENERIC-ARMV8_1-A:  .eabi_attribute 21, 1
+; GENERIC-ARMV8_1-A-NOT:  .eabi_attribute 22
+; GENERIC-ARMV8_1-A:  .eabi_attribute 23, 3
+; GENERIC-ARMV8_1-A:  .eabi_attribute 24, 1
+; GENERIC-ARMV8_1-A:  .eabi_attribute 25, 1
+; GENERIC-ARMV8_1-A-NOT:  .eabi_attribute 27
+; GENERIC-ARMV8_1-A-NOT:  .eabi_attribute 28
+; GENERIC-ARMV8_1-A:  .eabi_attribute 36, 1
+; GENERIC-ARMV8_1-A:  .eabi_attribute 38, 1
+; GENERIC-ARMV8_1-A:  .eabi_attribute 42, 1
+; GENERIC-ARMV8_1-A-NOT:  .eabi_attribute 44
+; GENERIC-ARMV8_1-A:  .eabi_attribute 68, 3
+
+; GENERIC-ARMV8_1-A-FAST-NOT:   .eabi_attribute 19
+;; GENERIC-ARMV8_1-A has the ARMv8 FP unit, which always flushes preserving sign.
+; GENERIC-ARMV8_1-A-FAST:  .eabi_attribute 20, 2
+; GENERIC-ARMV8_1-A-FAST-NOT:  .eabi_attribute 21
+; GENERIC-ARMV8_1-A-FAST-NOT:  .eabi_attribute 22
+; GENERIC-ARMV8_1-A-FAST:  .eabi_attribute 23, 1
+
 ; RELOC-PIC:  .eabi_attribute 15, 1
 ; RELOC-PIC:  .eabi_attribute 16, 1
 ; RELOC-PIC:  .eabi_attribute 17, 2
 ; RELOC-OTHER:  .eabi_attribute 17, 1
 
 ; PCS-R9-USE:  .eabi_attribute 14, 0
 ; PCS-R9-RESERVE:  .eabi_attribute 14, 3
 
 define i32 @f(i64 %z) {
     ret i32 0
 }