[AMDGPU] gfx908 target

[kzhuravl]

LGTM

llvm/trunk/docs/AMDGPUUsage.rst

``gfx902``                  ``amdgcn``   APU   - xnack                   - Ryzen 3 2200G
                                                       [on]                    - Ryzen 5 2400G
      ``gfx904``                  ``amdgcn``   dGPU  - xnack                   *TBA*
                                                       [off]
                                                                               .. TODO
                                                                                  Add product
                                                                                  names.
      ``gfx906``                  ``amdgcn``   dGPU  - xnack                   - Radeon Instinct MI50
                                                       [off]                   - Radeon Instinct MI60
+     ``gfx908``                  ``amdgcn``   dGPU  - xnack                   *TBA*
+                                                      [off]
+                                                      sram-ecc
+                                                      [on]
      ``gfx909``                  ``amdgcn``   APU   - xnack                   *TBA* (Raven Ridge 2)
                                                       [on]
                                                                               .. TODO
                                                                                  Add product
                                                                                  names.
      **GCN GFX10** [AMD-GCN-GFX10]_
      -----------------------------------------------------------------------------------------------
      ``gfx1010``                 ``amdgcn``   dGPU  - xnack                   *TBA*
.. table:: AMDGPU ``EF_AMDGPU_MACH`` Values
      ``EF_AMDGPU_MACH_AMDGCN_GFX801``  0x028      ``gfx801``
      ``EF_AMDGPU_MACH_AMDGCN_GFX802``  0x029      ``gfx802``
      ``EF_AMDGPU_MACH_AMDGCN_GFX803``  0x02a      ``gfx803``
      ``EF_AMDGPU_MACH_AMDGCN_GFX810``  0x02b      ``gfx810``
      ``EF_AMDGPU_MACH_AMDGCN_GFX900``  0x02c      ``gfx900``
      ``EF_AMDGPU_MACH_AMDGCN_GFX902``  0x02d      ``gfx902``
      ``EF_AMDGPU_MACH_AMDGCN_GFX904``  0x02e      ``gfx904``
      ``EF_AMDGPU_MACH_AMDGCN_GFX906``  0x02f      ``gfx906``
-     *reserved*                        0x030      Reserved.
+     ``EF_AMDGPU_MACH_AMDGCN_GFX908``  0x030      ``gfx908``
      ``EF_AMDGPU_MACH_AMDGCN_GFX909``  0x031      ``gfx909``
      *reserved*                        0x032      Reserved.
      ``EF_AMDGPU_MACH_AMDGCN_GFX1010`` 0x033      ``gfx1010``
      ``EF_AMDGPU_MACH_AMDGCN_GFX1011`` 0x034      ``gfx1011``
      ``EF_AMDGPU_MACH_AMDGCN_GFX1012`` 0x035      ``gfx1012``
      ================================= ========== =============================
 
 Sections

llvm/trunk/include/llvm/BinaryFormat/ELF.h

enum : unsigned {
   EF_AMDGPU_MACH_AMDGCN_GFX802 = 0x029,
   EF_AMDGPU_MACH_AMDGCN_GFX803 = 0x02a,
   EF_AMDGPU_MACH_AMDGCN_GFX810 = 0x02b,
   // AMDGCN GFX9.
   EF_AMDGPU_MACH_AMDGCN_GFX900 = 0x02c,
   EF_AMDGPU_MACH_AMDGCN_GFX902 = 0x02d,
   EF_AMDGPU_MACH_AMDGCN_GFX904 = 0x02e,
   EF_AMDGPU_MACH_AMDGCN_GFX906 = 0x02f,
+  EF_AMDGPU_MACH_AMDGCN_GFX908 = 0x030,
   EF_AMDGPU_MACH_AMDGCN_GFX909 = 0x031,
   // AMDGCN GFX10.
   EF_AMDGPU_MACH_AMDGCN_GFX1010 = 0x033,
   EF_AMDGPU_MACH_AMDGCN_GFX1011 = 0x034,
   EF_AMDGPU_MACH_AMDGCN_GFX1012 = 0x035,
 
   // Reserved for AMDGCN-based processors.
   EF_AMDGPU_MACH_AMDGCN_RESERVED0 = 0x027,

llvm/trunk/include/llvm/Support/TargetParser.h

enum GPUKind : uint32_t {
   GK_GFX802 = 51,
   GK_GFX803 = 52,
   GK_GFX810 = 53,
 
   GK_GFX900 = 60,
   GK_GFX902 = 61,
   GK_GFX904 = 62,
   GK_GFX906 = 63,
+  GK_GFX908 = 64,
   GK_GFX909 = 65,
 
   GK_GFX1010 = 71,
   GK_GFX1011 = 72,
   GK_GFX1012 = 73,
 
   GK_AMDGCN_FIRST = GK_GFX600,
   GK_AMDGCN_LAST = GK_GFX1012,

llvm/trunk/lib/ObjectYAML/ELFYAML.cpp

case ELF::EM_AMDGPU:
     BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX801, EF_AMDGPU_MACH);
     BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX802, EF_AMDGPU_MACH);
     BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX803, EF_AMDGPU_MACH);
     BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX810, EF_AMDGPU_MACH);
     BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX900, EF_AMDGPU_MACH);
     BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX902, EF_AMDGPU_MACH);
     BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX904, EF_AMDGPU_MACH);
     BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX906, EF_AMDGPU_MACH);
+    BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX908, EF_AMDGPU_MACH);
     BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX909, EF_AMDGPU_MACH);
     BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX1010, EF_AMDGPU_MACH);
     BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX1011, EF_AMDGPU_MACH);
     BCaseMask(EF_AMDGPU_MACH_AMDGCN_GFX1012, EF_AMDGPU_MACH);
     BCase(EF_AMDGPU_XNACK);
     BCase(EF_AMDGPU_SRAM_ECC);
     break;
   case ELF::EM_X86_64:

llvm/trunk/lib/Support/TargetParser.cpp

constexpr GPUInfo R600GPUs[26] = {
   {{"caicos"},  {"caicos"},  GK_CAICOS,  FEATURE_NONE },
   {{"aruba"},   {"cayman"},  GK_CAYMAN,  FEATURE_FMA  },
   {{"cayman"},  {"cayman"},  GK_CAYMAN,  FEATURE_FMA  },
   {{"turks"},   {"turks"},   GK_TURKS,   FEATURE_NONE }
 };
 
 // This table should be sorted by the value of GPUKind
 // Don't bother listing the implicitly true features
-constexpr GPUInfo AMDGCNGPUs[36] = {
+constexpr GPUInfo AMDGCNGPUs[37] = {
   // Name         Canonical    Kind        Features
   //              Name
   {{"gfx600"},    {"gfx600"},  GK_GFX600,  FEATURE_FAST_FMA_F32},
   {{"tahiti"},    {"gfx600"},  GK_GFX600,  FEATURE_FAST_FMA_F32},
   {{"gfx601"},    {"gfx601"},  GK_GFX601,  FEATURE_NONE},
   {{"hainan"},    {"gfx601"},  GK_GFX601,  FEATURE_NONE},
   {{"oland"},     {"gfx601"},  GK_GFX601,  FEATURE_NONE},
   {{"pitcairn"},  {"gfx601"},  GK_GFX601,  FEATURE_NONE},
constexpr GPUInfo AMDGCNGPUs[37] = {
   {{"polaris10"}, {"gfx803"},  GK_GFX803,  FEATURE_FAST_DENORMAL_F32},
   {{"polaris11"}, {"gfx803"},  GK_GFX803,  FEATURE_FAST_DENORMAL_F32},
   {{"gfx810"},    {"gfx810"},  GK_GFX810,  FEATURE_FAST_DENORMAL_F32},
   {{"stoney"},    {"gfx810"},  GK_GFX810,  FEATURE_FAST_DENORMAL_F32},
   {{"gfx900"},    {"gfx900"},  GK_GFX900,  FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32},
   {{"gfx902"},    {"gfx902"},  GK_GFX902,  FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32},
   {{"gfx904"},    {"gfx904"},  GK_GFX904,  FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32},
   {{"gfx906"},    {"gfx906"},  GK_GFX906,  FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32},
+  {{"gfx908"},    {"gfx908"},  GK_GFX908,  FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32},
   {{"gfx909"},    {"gfx909"},  GK_GFX909,  FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32},
   {{"gfx1010"},   {"gfx1010"}, GK_GFX1010, FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32},
   {{"gfx1011"},   {"gfx1011"}, GK_GFX1011, FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32},
   {{"gfx1012"},   {"gfx1012"}, GK_GFX1012, FEATURE_FAST_FMA_F32|FEATURE_FAST_DENORMAL_F32},
 };
 
 const GPUInfo *getArchEntry(AMDGPU::GPUKind AK, ArrayRef<GPUInfo> Table) {
   GPUInfo Search = { {""}, {""}, AK, AMDGPU::FEATURE_NONE };
AMDGPU::IsaVersion AMDGPU::getIsaVersion(StringRef GPU) {
   case GK_GFX801:  return {8, 0, 1};
   case GK_GFX802:  return {8, 0, 2};
   case GK_GFX803:  return {8, 0, 3};
   case GK_GFX810:  return {8, 1, 0};
   case GK_GFX900:  return {9, 0, 0};
   case GK_GFX902:  return {9, 0, 2};
   case GK_GFX904:  return {9, 0, 4};
   case GK_GFX906:  return {9, 0, 6};
+  case GK_GFX908:  return {9, 0, 8};
   case GK_GFX909:  return {9, 0, 9};
   case GK_GFX1010: return {10, 1, 0};
   case GK_GFX1011: return {10, 1, 1};
   case GK_GFX1012: return {10, 1, 2};
   default:         return {0, 0, 0};
   }
 }

llvm/trunk/lib/Target/AMDGPU/AMDGPU.td

 >;
 
 def FeatureDot2Insts : SubtargetFeature<"dot2-insts",
   "HasDot2Insts",
   "true",
   "Has v_dot2_f32_f16, v_dot2_i32_i16, v_dot2_u32_u16, v_dot4_u32_u8, v_dot8_u32_u4 instructions"
 >;
 
+def FeatureDot3Insts : SubtargetFeature<"dot3-insts",
+  "HasDot3Insts",
+  "true",
+  "Has v_dot8c_i32_i4 instruction"
+>;
+
+def FeatureDot4Insts : SubtargetFeature<"dot4-insts",
+  "HasDot4Insts",
+  "true",
+  "Has v_dot2c_i32_i16 instruction"
+>;
+
 def FeatureDot5Insts : SubtargetFeature<"dot5-insts",
   "HasDot5Insts",
   "true",
   "Has v_dot2c_f32_f16 instruction"
 >;
 
 def FeatureDot6Insts : SubtargetFeature<"dot6-insts",
   "HasDot6Insts",
   "true",
   "Has v_dot4c_i32_i8 instruction"
 >;
 
+def FeatureMAIInsts : SubtargetFeature<"mai-insts",
+  "HasMAIInsts",
+  "true",
+  "Has mAI instructions"
+>;
+
+def FeaturePkFmacF16Inst : SubtargetFeature<"pk-fmac-f16-inst",
+  "HasPkFmacF16Inst",
+  "true",
+  "Has v_pk_fmac_f16 instruction"
+>;
+
+def FeatureAtomicFaddInsts : SubtargetFeature<"atomic-fadd-insts",
+  "HasAtomicFaddInsts",
+  "true",
+  "Has buffer_atomic_add_f32, buffer_atomic_pk_add_f16, global_atomic_add_f32, "
+  "global_atomic_pk_add_f16 instructions"
+>;
+
 def FeatureDoesNotSupportSRAMECC : SubtargetFeature<"no-sram-ecc-support",
   "DoesNotSupportSRAMECC",
   "true",
   "Hardware does not support SRAM ECC"
 >;
 
 def FeatureSRAMECC : SubtargetFeature<"sram-ecc",
   "EnableSRAMECC",
def FeatureISAVersion9_0_6 : FeatureSet<
    FeatureFmaMixInsts,
    FeatureLDSBankCount32,
    FeatureDLInsts,
    FeatureDot1Insts,
    FeatureDot2Insts,
    FeatureDoesNotSupportXNACK,
    FeatureCodeObjectV3]>;
 
+def FeatureISAVersion9_0_8 : FeatureSet<
+  [FeatureGFX9,
+   HalfRate64Ops,
+   FeatureFmaMixInsts,
+   FeatureLDSBankCount32,
+   FeatureDLInsts,
+   FeatureDot1Insts,
+   FeatureDot2Insts,
+   FeatureDot3Insts,
+   FeatureDot4Insts,
+   FeatureDot5Insts,
+   FeatureDot6Insts,
+   FeatureMAIInsts,
+   FeaturePkFmacF16Inst,
+   FeatureAtomicFaddInsts,
+   FeatureSRAMECC,
+   FeatureCodeObjectV3]>;
+
 def FeatureISAVersion9_0_9 : FeatureSet<
   [FeatureGFX9,
    FeatureMadMixInsts,
    FeatureLDSBankCount32,
    FeatureXNACK,
    FeatureCodeObjectV3]>;
 
 // TODO: Organize more features into groups.
def HasDLInsts : Predicate<"Subtarget->hasDLInsts()">,
   AssemblerPredicate<"FeatureDLInsts">;
 
 def HasDot1Insts : Predicate<"Subtarget->hasDot1Insts()">,
   AssemblerPredicate<"FeatureDot1Insts">;
 
 def HasDot2Insts : Predicate<"Subtarget->hasDot2Insts()">,
   AssemblerPredicate<"FeatureDot2Insts">;
 
+def HasDot3Insts : Predicate<"Subtarget->hasDot3Insts()">,
+  AssemblerPredicate<"FeatureDot3Insts">;
+
+def HasDot4Insts : Predicate<"Subtarget->hasDot4Insts()">,
+  AssemblerPredicate<"FeatureDot4Insts">;
+
 def HasDot5Insts : Predicate<"Subtarget->hasDot5Insts()">,
   AssemblerPredicate<"FeatureDot5Insts">;
 
 def HasDot6Insts : Predicate<"Subtarget->hasDot6Insts()">,
   AssemblerPredicate<"FeatureDot6Insts">;
 
+def HasMAIInsts : Predicate<"Subtarget->hasMAIInsts()">,
+  AssemblerPredicate<"FeatureMAIInsts">;
+
+def HasPkFmacF16Inst : Predicate<"Subtarget->hasPkFmacF16Inst()">,
+  AssemblerPredicate<"FeaturePkFmacF16Inst">;
+
+def HasAtomicFaddInsts : Predicate<"Subtarget->hasAtomicFaddInsts()">,
+  AssemblerPredicate<"FeatureAtomicFaddInsts">;
+
 def HasOffset3fBug : Predicate<"!Subtarget->hasOffset3fBug()">,
   AssemblerPredicate<"FeatureOffset3fBug">;
 
 def EnableLateCFGStructurize : Predicate<
   "EnableLateStructurizeCFG">;
 
 // Include AMDGPU TD files
 include "SISchedule.td"
 include "GCNProcessors.td"
 include "AMDGPUInstrInfo.td"
 include "AMDGPURegisterInfo.td"
 include "AMDGPURegisterBanks.td"
 include "AMDGPUInstructions.td"
 include "SIInstrInfo.td"
 include "AMDGPUCallingConv.td"
 include "AMDGPUSearchableTables.td"

llvm/trunk/lib/Target/AMDGPU/AMDGPUSubtarget.h

protected:
   bool HasSDWAOutModsVOPC;
   bool HasDPP;
   bool HasDPP8;
   bool HasR128A16;
   bool HasNSAEncoding;
   bool HasDLInsts;
   bool HasDot1Insts;
   bool HasDot2Insts;
+  bool HasDot3Insts;
+  bool HasDot4Insts;
   bool HasDot5Insts;
   bool HasDot6Insts;
+  bool HasMAIInsts;
+  bool HasPkFmacF16Inst;
+  bool HasAtomicFaddInsts;
   bool EnableSRAMECC;
   bool DoesNotSupportSRAMECC;
   bool HasNoSdstCMPX;
   bool HasVscnt;
   bool HasRegisterBanking;
   bool HasVOP3Literal;
   bool HasNoDataDepHazard;
   bool FlatAddressSpace;
public:
   bool hasDot1Insts() const {
     return HasDot1Insts;
   }
 
   bool hasDot2Insts() const {
     return HasDot2Insts;
   }
 
+  bool hasDot3Insts() const {
+    return HasDot3Insts;
+  }
+
+  bool hasDot4Insts() const {
+    return HasDot4Insts;
+  }
+
   bool hasDot5Insts() const {
     return HasDot5Insts;
   }
 
   bool hasDot6Insts() const {
     return HasDot6Insts;
   }
 
+  bool hasMAIInsts() const {
+    return HasMAIInsts;
+  }
+
+  bool hasPkFmacF16Inst() const {
+    return HasPkFmacF16Inst;
+  }
+
+  bool hasAtomicFaddInsts() const {
+    return HasAtomicFaddInsts;
+  }
+
   bool isSRAMECCEnabled() const {
     return EnableSRAMECC;
   }
 
   bool hasNoSdstCMPX() const {
     return HasNoSdstCMPX;
   }
 

llvm/trunk/lib/Target/AMDGPU/AMDGPUSubtarget.cpp

GCNSubtarget::GCNSubtarget(const Triple &TT, StringRef GPU, StringRef FS,
     HasSDWAOutModsVOPC(false),
     HasDPP(false),
     HasDPP8(false),
     HasR128A16(false),
     HasNSAEncoding(false),
     HasDLInsts(false),
     HasDot1Insts(false),
     HasDot2Insts(false),
+    HasDot3Insts(false),
+    HasDot4Insts(false),
     HasDot5Insts(false),
     HasDot6Insts(false),
+    HasMAIInsts(false),
+    HasPkFmacF16Inst(false),
+    HasAtomicFaddInsts(false),
     EnableSRAMECC(false),
     DoesNotSupportSRAMECC(false),
     HasNoSdstCMPX(false),
     HasVscnt(false),
     HasRegisterBanking(false),
     HasVOP3Literal(false),
     HasNoDataDepHazard(false),
     FlatAddressSpace(false),

llvm/trunk/lib/Target/AMDGPU/GCNProcessors.td

 def : ProcessorModel<"gfx904", SIQuarterSpeedModel,
   FeatureISAVersion9_0_4.Features
 >;
 
 def : ProcessorModel<"gfx906", SIQuarterSpeedModel,
   FeatureISAVersion9_0_6.Features
 >;
 
+def : ProcessorModel<"gfx908", SIQuarterSpeedModel,
+  FeatureISAVersion9_0_8.Features
+>;
+
 def : ProcessorModel<"gfx909", SIQuarterSpeedModel,
   FeatureISAVersion9_0_9.Features
 >;
 
 //===----------------------------------------------------------------------===//
 // GCN GFX10.
 //===----------------------------------------------------------------------===//
 

llvm/trunk/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp

StringRef AMDGPUTargetStreamer::getArchNameFromElfMach(unsigned ElfMach) {
   case ELF::EF_AMDGPU_MACH_AMDGCN_GFX801:  AK = GK_GFX801;  break;
   case ELF::EF_AMDGPU_MACH_AMDGCN_GFX802:  AK = GK_GFX802;  break;
   case ELF::EF_AMDGPU_MACH_AMDGCN_GFX803:  AK = GK_GFX803;  break;
   case ELF::EF_AMDGPU_MACH_AMDGCN_GFX810:  AK = GK_GFX810;  break;
   case ELF::EF_AMDGPU_MACH_AMDGCN_GFX900:  AK = GK_GFX900;  break;
   case ELF::EF_AMDGPU_MACH_AMDGCN_GFX902:  AK = GK_GFX902;  break;
   case ELF::EF_AMDGPU_MACH_AMDGCN_GFX904:  AK = GK_GFX904;  break;
   case ELF::EF_AMDGPU_MACH_AMDGCN_GFX906:  AK = GK_GFX906;  break;
+  case ELF::EF_AMDGPU_MACH_AMDGCN_GFX908:  AK = GK_GFX908;  break;
   case ELF::EF_AMDGPU_MACH_AMDGCN_GFX909:  AK = GK_GFX909;  break;
   case ELF::EF_AMDGPU_MACH_AMDGCN_GFX1010: AK = GK_GFX1010; break;
   case ELF::EF_AMDGPU_MACH_AMDGCN_GFX1011: AK = GK_GFX1011; break;
   case ELF::EF_AMDGPU_MACH_AMDGCN_GFX1012: AK = GK_GFX1012; break;
   case ELF::EF_AMDGPU_MACH_NONE:           AK = GK_NONE;    break;
   }
 
   StringRef GPUName = getArchNameAMDGCN(AK);
unsigned AMDGPUTargetStreamer::getElfMach(StringRef GPU) {
   case GK_GFX801:  return ELF::EF_AMDGPU_MACH_AMDGCN_GFX801;
   case GK_GFX802:  return ELF::EF_AMDGPU_MACH_AMDGCN_GFX802;
   case GK_GFX803:  return ELF::EF_AMDGPU_MACH_AMDGCN_GFX803;
   case GK_GFX810:  return ELF::EF_AMDGPU_MACH_AMDGCN_GFX810;
   case GK_GFX900:  return ELF::EF_AMDGPU_MACH_AMDGCN_GFX900;
   case GK_GFX902:  return ELF::EF_AMDGPU_MACH_AMDGCN_GFX902;
   case GK_GFX904:  return ELF::EF_AMDGPU_MACH_AMDGCN_GFX904;
   case GK_GFX906:  return ELF::EF_AMDGPU_MACH_AMDGCN_GFX906;
+  case GK_GFX908:  return ELF::EF_AMDGPU_MACH_AMDGCN_GFX908;
   case GK_GFX909:  return ELF::EF_AMDGPU_MACH_AMDGCN_GFX909;
   case GK_GFX1010: return ELF::EF_AMDGPU_MACH_AMDGCN_GFX1010;
   case GK_GFX1011: return ELF::EF_AMDGPU_MACH_AMDGCN_GFX1011;
   case GK_GFX1012: return ELF::EF_AMDGPU_MACH_AMDGCN_GFX1012;
   case GK_NONE:    return ELF::EF_AMDGPU_MACH_NONE;
   }
 
   llvm_unreachable("unknown GPU");

llvm/trunk/test/CodeGen/AMDGPU/elf-header-flags-mach.ll

 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=polaris10 < %s | llvm-readobj -file-headers - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX803 %s
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=polaris11 < %s | llvm-readobj -file-headers - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX803 %s
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx810 < %s | llvm-readobj -file-headers - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX810 %s
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=stoney < %s | llvm-readobj -file-headers - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX810 %s
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx900 < %s | llvm-readobj -file-headers - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX900 %s
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx902 < %s | llvm-readobj -file-headers - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX902 %s
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx904 < %s | llvm-readobj -file-headers - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX904 %s
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx906 < %s | llvm-readobj -file-headers - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX906 %s
+; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx908 < %s | llvm-readobj -file-headers - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX908 %s
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx909 < %s | llvm-readobj -file-headers - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX909 %s
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx1010 < %s | llvm-readobj -file-headers - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX1010 %s
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx1011 < %s | llvm-readobj -file-headers - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX1011 %s
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx1012 < %s | llvm-readobj -file-headers - | FileCheck --check-prefixes=ALL,ARCH-GCN,GFX1012 %s
 
 ; ARCH-R600: Arch: r600
 ; ARCH-GCN:  Arch: amdgcn
 

 ; GFX803:        EF_AMDGPU_MACH_AMDGCN_GFX803 (0x2A)
 ; GFX810:        EF_AMDGPU_MACH_AMDGCN_GFX810 (0x2B)
 ; GFX810-NEXT:   EF_AMDGPU_XNACK              (0x100)
 ; GFX900:        EF_AMDGPU_MACH_AMDGCN_GFX900 (0x2C)
 ; GFX902:        EF_AMDGPU_MACH_AMDGCN_GFX902 (0x2D)
 ; GFX902-NEXT:   EF_AMDGPU_XNACK              (0x100)
 ; GFX904:        EF_AMDGPU_MACH_AMDGCN_GFX904 (0x2E)
 ; GFX906:        EF_AMDGPU_MACH_AMDGCN_GFX906 (0x2F)
+; GFX908:        EF_AMDGPU_MACH_AMDGCN_GFX908 (0x30)
+; GFX908-NEXT:   EF_AMDGPU_SRAM_ECC           (0x200)
 ; GFX909:        EF_AMDGPU_MACH_AMDGCN_GFX909 (0x31)
 ; GFX1010:       EF_AMDGPU_MACH_AMDGCN_GFX1010 (0x33)
 ; GFX1011:       EF_AMDGPU_MACH_AMDGCN_GFX1011 (0x34)
 ; GFX1012:       EF_AMDGPU_MACH_AMDGCN_GFX1012 (0x35)
 ; ALL:         ]
 
 define amdgpu_kernel void @elf_header() {
   ret void
 }

llvm/trunk/test/CodeGen/AMDGPU/elf-header-flags-sram-ecc.ll

 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx902 < %s | llvm-readobj --file-headers - | FileCheck --check-prefixes=NO-SRAM-ECC-GFX902 %s
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx902 -mattr=-sram-ecc < %s | llvm-readobj --file-headers - | FileCheck --check-prefixes=NO-SRAM-ECC-GFX902 %s
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx902 -mattr=+sram-ecc < %s | llvm-readobj --file-headers - | FileCheck --check-prefixes=SRAM-ECC-GFX902 %s
 
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx906 < %s | llvm-readobj --file-headers - | FileCheck --check-prefixes=NO-SRAM-ECC-GFX906 %s
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx906 -mattr=-sram-ecc < %s | llvm-readobj --file-headers - | FileCheck --check-prefixes=NO-SRAM-ECC-GFX906 %s
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx906 -mattr=+sram-ecc < %s | llvm-readobj --file-headers - | FileCheck --check-prefixes=SRAM-ECC-GFX906 %s
 ; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx906 -mattr=+sram-ecc,+xnack < %s | llvm-readobj --file-headers - | FileCheck --check-prefixes=SRAM-ECC-XNACK-GFX906 %s
 
+; RUN: llc -filetype=obj -march=amdgcn -mcpu=gfx908 < %s | llvm-readobj -file-headers - | FileCheck --check-prefix=SRAM-ECC-GFX908 %s
+
 ; NO-SRAM-ECC-GFX902:      Flags [
 ; NO-SRAM-ECC-GFX902-NEXT:   EF_AMDGPU_MACH_AMDGCN_GFX902 (0x2D)
 ; NO-SRAM-ECC-GFX902-NEXT:   EF_AMDGPU_XNACK              (0x100)
 ; NO-SRAM-ECC-GFX902-NEXT: ]
 
 ; SRAM-ECC-GFX902:      Flags [
 ; SRAM-ECC-GFX902-NEXT:   EF_AMDGPU_MACH_AMDGCN_GFX902 (0x2D)
 ; SRAM-ECC-GFX902-NEXT:   EF_AMDGPU_SRAM_ECC           (0x200)

 ; SRAM-ECC-GFX906-NEXT: ]
 
 ; SRAM-ECC-XNACK-GFX906:      Flags [
 ; SRAM-ECC-XNACK-GFX906-NEXT:   EF_AMDGPU_MACH_AMDGCN_GFX906 (0x2F)
 ; SRAM-ECC-XNACK-GFX906-NEXT:   EF_AMDGPU_SRAM_ECC           (0x200)
 ; SRAM-ECC-XNACK-GFX906-NEXT:   EF_AMDGPU_XNACK              (0x100)
 ; SRAM-ECC-XNACK-GFX906-NEXT: ]
 
+; SRAM-ECC-GFX908: Flags [ (0x230)
+; SRAM-ECC-GFX908:    EF_AMDGPU_MACH_AMDGCN_GFX908 (0x30)
+; SRAM-ECC-GFX908:    EF_AMDGPU_SRAM_ECC (0x200)
+; SRAM-ECC-GFX908:  ]
+
 define amdgpu_kernel void @elf_header() {
   ret void
 }

llvm/trunk/test/CodeGen/AMDGPU/hsa-note-no-func.ll

 ; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx801 -mattr=-code-object-v3 | FileCheck --check-prefix=HSA --check-prefix=HSA-VI801 %s
 ; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx802 -mattr=-code-object-v3 | FileCheck --check-prefix=HSA --check-prefix=HSA-VI802 %s
 ; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx803 -mattr=-code-object-v3 | FileCheck --check-prefix=HSA --check-prefix=HSA-VI803 %s
 ; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx810 -mattr=-code-object-v3 | FileCheck --check-prefix=HSA --check-prefix=HSA-VI810 %s
 ; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx900 -mattr=-code-object-v3 | FileCheck --check-prefix=HSA --check-prefix=HSA-GFX900 %s
 ; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx902 -mattr=-code-object-v3 | FileCheck --check-prefix=HSA --check-prefix=HSA-GFX902 %s
 ; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx904 -mattr=-code-object-v3 | FileCheck --check-prefix=HSA --check-prefix=HSA-GFX904 %s
 ; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx906 -mattr=-code-object-v3 | FileCheck --check-prefix=HSA --check-prefix=HSA-GFX906 %s
+; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx908 -mattr=-code-object-v3 | FileCheck --check-prefix=HSA --check-prefix=HSA-GFX908 %s
 ; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx909 -mattr=-code-object-v3 | FileCheck --check-prefix=HSA --check-prefix=HSA-GFX909 %s
 ; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx1010 -mattr=-code-object-v3 | FileCheck --check-prefix=HSA --check-prefix=HSA-GFX1010 %s
 ; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx1011 -mattr=-code-object-v3 | FileCheck --check-prefix=HSA --check-prefix=HSA-GFX1011 %s
 ; RUN: llc < %s -mtriple=amdgcn--amdhsa -mcpu=gfx1012 -mattr=-code-object-v3 | FileCheck --check-prefix=HSA --check-prefix=HSA-GFX1012 %s
 
 ; HSA: .hsa_code_object_version 2,1
 ; HSA-SI600: .hsa_code_object_isa 6,0,0,"AMD","AMDGPU"
 ; HSA-SI601: .hsa_code_object_isa 6,0,1,"AMD","AMDGPU"
 ; HSA-CI700: .hsa_code_object_isa 7,0,0,"AMD","AMDGPU"
 ; HSA-CI701: .hsa_code_object_isa 7,0,1,"AMD","AMDGPU"
 ; HSA-CI702: .hsa_code_object_isa 7,0,2,"AMD","AMDGPU"
 ; HSA-CI703: .hsa_code_object_isa 7,0,3,"AMD","AMDGPU"
 ; HSA-CI704: .hsa_code_object_isa 7,0,4,"AMD","AMDGPU"
 ; HSA-VI801: .hsa_code_object_isa 8,0,1,"AMD","AMDGPU"
 ; HSA-VI802: .hsa_code_object_isa 8,0,2,"AMD","AMDGPU"
 ; HSA-VI803: .hsa_code_object_isa 8,0,3,"AMD","AMDGPU"
 ; HSA-VI810: .hsa_code_object_isa 8,1,0,"AMD","AMDGPU"
 ; HSA-GFX900: .hsa_code_object_isa 9,0,0,"AMD","AMDGPU"
 ; HSA-GFX902: .hsa_code_object_isa 9,0,2,"AMD","AMDGPU"
 ; HSA-GFX904: .hsa_code_object_isa 9,0,4,"AMD","AMDGPU"
 ; HSA-GFX906: .hsa_code_object_isa 9,0,6,"AMD","AMDGPU"
+; HSA-GFX908: .hsa_code_object_isa 9,0,8,"AMD","AMDGPU"
 ; HSA-GFX909: .hsa_code_object_isa 9,0,9,"AMD","AMDGPU"
 ; HSA-GFX1010: .hsa_code_object_isa 10,1,0,"AMD","AMDGPU"
 ; HSA-GFX1011: .hsa_code_object_isa 10,1,1,"AMD","AMDGPU"
 ; HSA-GFX1012: .hsa_code_object_isa 10,1,2,"AMD","AMDGPU"

llvm/trunk/test/MC/AMDGPU/dl-insts.s

 // RUN: llvm-mc -arch=amdgcn -mcpu=gfx906 -show-encoding %s | FileCheck %s
+// RUN: llvm-mc -arch=amdgcn -mcpu=gfx908 -show-encoding %s | FileCheck %s
 
 //
 // VOP2 Instructions.
 //
 
 // CHECK: encoding: [0x01,0x05,0x0a,0x76]
 v_fmac_f32 v5, v1, v2
 // CHECK: encoding: [0x01,0x05,0xfe,0x77]

llvm/trunk/test/MC/Disassembler/AMDGPU/dl-insts.txt

 # RUN: llvm-mc -arch=amdgcn -mcpu=gfx906 -disassemble -show-encoding < %s | FileCheck %s
+# RUN: llvm-mc -arch=amdgcn -mcpu=gfx908 -disassemble -show-encoding < %s | FileCheck %s
 
 # CHECK: v_fmac_f32_e32 v5, v1, v2 ; encoding: [0x01,0x05,0x0a,0x76]
 0x01,0x05,0x0a,0x76
 
 # CHECK: v_fmac_f32_e32 v255, v1, v2 ; encoding: [0x01,0x05,0xfe,0x77]
 0x01,0x05,0xfe,0x77
 
 # CHECK: v_fmac_f32_e32 v5, v255, v2 ; encoding: [0xff,0x05,0x0a,0x76]

llvm/trunk/test/Object/AMDGPU/elf-header-flags-mach.yaml

 # RUN: obj2yaml %t.o.29 | FileCheck --check-prefixes=YAML-GFX902 %s
 # RUN: yaml2obj -docnum=30 %s > %t.o.30
 # RUN: llvm-readobj -S --file-headers %t.o.30 | FileCheck --check-prefixes=ELF-ALL,ELF-GFX904 %s
 # RUN: obj2yaml %t.o.30 | FileCheck --check-prefixes=YAML-GFX904 %s
 # RUN: yaml2obj -docnum=31 %s > %t.o.31
 # RUN: llvm-readobj -S --file-headers %t.o.31 | FileCheck --check-prefixes=ELF-ALL,ELF-GFX906 %s
 # RUN: obj2yaml %t.o.31 | FileCheck --check-prefixes=YAML-GFX906 %s
 # RUN: yaml2obj -docnum=32 %s > %t.o.32
-# RUN: llvm-readobj -S --file-headers %t.o.32 | FileCheck --check-prefixes=ELF-ALL,ELF-GFX909 %s
-# RUN: obj2yaml %t.o.32 | FileCheck --check-prefixes=YAML-GFX909 %s
+# RUN: llvm-readobj -s -file-headers %t.o.32 | FileCheck --check-prefixes=ELF-ALL,ELF-GFX908 %s
+# RUN: obj2yaml %t.o.32 | FileCheck --check-prefixes=YAML-GFX908 %s
 # RUN: yaml2obj -docnum=33 %s > %t.o.33
-# RUN: llvm-readobj -s -file-headers %t.o.33 | FileCheck --check-prefixes=ELF-ALL,ELF-GFX1010 %s
-# RUN: obj2yaml %t.o.33 | FileCheck --check-prefixes=YAML-GFX1010 %s
+# RUN: llvm-readobj -s -file-headers %t.o.33 | FileCheck --check-prefixes=ELF-ALL,ELF-GFX909 %s
+# RUN: obj2yaml %t.o.33 | FileCheck --check-prefixes=YAML-GFX909 %s
+# RUN: yaml2obj -docnum=34 %s > %t.o.34
+# RUN: llvm-readobj -s -file-headers %t.o.34 | FileCheck --check-prefixes=ELF-ALL,ELF-GFX1010 %s
+# RUN: obj2yaml %t.o.34 | FileCheck --check-prefixes=YAML-GFX1010 %s
 
 # ELF-ALL:     Flags [
 # ELF-R600:      EF_AMDGPU_MACH_R600_R600     (0x1)
 # ELF-R630:      EF_AMDGPU_MACH_R600_R630     (0x2)
 # ELF-RS880:     EF_AMDGPU_MACH_R600_RS880    (0x3)
 # ELF-RV670:     EF_AMDGPU_MACH_R600_RV670    (0x4)
 # ELF-RV710:     EF_AMDGPU_MACH_R600_RV710    (0x5)
 # ELF-RV730:     EF_AMDGPU_MACH_R600_RV730    (0x6)

 # ELF-GFX801:    EF_AMDGPU_MACH_AMDGCN_GFX801 (0x28)
 # ELF-GFX802:    EF_AMDGPU_MACH_AMDGCN_GFX802 (0x29)
 # ELF-GFX803:    EF_AMDGPU_MACH_AMDGCN_GFX803 (0x2A)
 # ELF-GFX810:    EF_AMDGPU_MACH_AMDGCN_GFX810 (0x2B)
 # ELF-GFX900:    EF_AMDGPU_MACH_AMDGCN_GFX900 (0x2C)
 # ELF-GFX902:    EF_AMDGPU_MACH_AMDGCN_GFX902 (0x2D)
 # ELF-GFX904:    EF_AMDGPU_MACH_AMDGCN_GFX904 (0x2E)
 # ELF-GFX906:    EF_AMDGPU_MACH_AMDGCN_GFX906 (0x2F)
+# ELF-GFX908:    EF_AMDGPU_MACH_AMDGCN_GFX908 (0x30)
 # ELF-GFX909:    EF_AMDGPU_MACH_AMDGCN_GFX909 (0x31)
 # ELF-GFX1010:   EF_AMDGPU_MACH_AMDGCN_GFX1010 (0x33)
 # ELF-ALL:     ]
 
 # YAML-R600:    Flags: [ EF_AMDGPU_MACH_R600_R600 ]
 # YAML-R630:    Flags: [ EF_AMDGPU_MACH_R600_R630 ]
 # YAML-RS880:   Flags: [ EF_AMDGPU_MACH_R600_RS880 ]
 # YAML-RV670:   Flags: [ EF_AMDGPU_MACH_R600_RV670 ]

 # YAML-GFX801:  Flags: [ EF_AMDGPU_MACH_AMDGCN_GFX801 ]
 # YAML-GFX802:  Flags: [ EF_AMDGPU_MACH_AMDGCN_GFX802 ]
 # YAML-GFX803:  Flags: [ EF_AMDGPU_MACH_AMDGCN_GFX803 ]
 # YAML-GFX810:  Flags: [ EF_AMDGPU_MACH_AMDGCN_GFX810 ]
 # YAML-GFX900:  Flags: [ EF_AMDGPU_MACH_AMDGCN_GFX900 ]
 # YAML-GFX902:  Flags: [ EF_AMDGPU_MACH_AMDGCN_GFX902 ]
 # YAML-GFX904:  Flags: [ EF_AMDGPU_MACH_AMDGCN_GFX904 ]
 # YAML-GFX906:  Flags: [ EF_AMDGPU_MACH_AMDGCN_GFX906 ]
+# YAML-GFX908:  Flags: [ EF_AMDGPU_MACH_AMDGCN_GFX908 ]
 # YAML-GFX909:  Flags: [ EF_AMDGPU_MACH_AMDGCN_GFX909 ]
 # YAML-GFX1010: Flags: [ EF_AMDGPU_MACH_AMDGCN_GFX1010 ]
 
 # Doc1
 --- !ELF
 FileHeader:
   Class:   ELFCLASS32
   Data:    ELFDATA2LSB

 # Doc32
 --- !ELF
 FileHeader:
   Class:   ELFCLASS64
   Data:    ELFDATA2LSB
   OSABI:   ELFOSABI_NONE
   Type:    ET_REL
   Machine: EM_AMDGPU
-  Flags:   [ EF_AMDGPU_MACH_AMDGCN_GFX909 ]
+  Flags:   [ EF_AMDGPU_MACH_AMDGCN_GFX908 ]
 ...
 
 # Doc33
 --- !ELF
 FileHeader:
   Class:   ELFCLASS64
   Data:    ELFDATA2LSB
   OSABI:   ELFOSABI_NONE
   Type:    ET_REL
   Machine: EM_AMDGPU
+  Flags:   [ EF_AMDGPU_MACH_AMDGCN_GFX909 ]
+...
+
+# Doc34
+--- !ELF
+FileHeader:
+  Class:   ELFCLASS64
+  Data:    ELFDATA2LSB
+  OSABI:   ELFOSABI_NONE
+  Type:    ET_REL
+  Machine: EM_AMDGPU
   Flags:   [ EF_AMDGPU_MACH_AMDGCN_GFX1010 ]
 ...
+

llvm/trunk/tools/llvm-readobj/ELFDumper.cpp

static const EnumEntry<unsigned> ElfHeaderAMDGPUFlags[] = {
   LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX801),
   LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX802),
   LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX803),
   LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX810),
   LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX900),
   LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX902),
   LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX904),
   LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX906),
+  LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX908),
   LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX909),
   LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX1010),
   LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX1011),
   LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_MACH_AMDGCN_GFX1012),
   LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_XNACK),
   LLVM_READOBJ_ENUM_ENT(ELF, EF_AMDGPU_SRAM_ECC)
 };