[AMDGPU] Runtime metadata fixes:

  - Verify that runtime metadata is actually valid runtime metadata when assembling, otherwise we could accept the following when assembling, but ocl runtime will reject it:
    .amdgpu_runtime_metadata
    { amd.MDVersion: [ 2, 1 ], amd.RandomUnknownKey, amd.IsaInfo: ...
  - Make IsaInfo optional, and always emit it.

Differential Revision: https://reviews.llvm.org/D30349

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296324 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/AMDGPU/AsmParser/AMDGPUAsmParser.cpp b/lib/Target/AMDGPU/AsmParser/AMDGPUAsmParser.cpp
index 0f0e0ab68c9ccdd8b72fe409a054bb5949888ba4..b2135d518d105a8df5a916748095b9f648cba519 100644 (file)
--- a/lib/Target/AMDGPU/AsmParser/AMDGPUAsmParser.cpp
+++ b/lib/Target/AMDGPU/AsmParser/AMDGPUAsmParser.cpp
@@ -789,12 +789,12 @@ public:
       : MCTargetAsmParser(Options, STI), MII(MII), Parser(_Parser) {
     MCAsmParserExtension::Initialize(Parser);
 
-    if (getSTI().getFeatureBits().none()) {
+    if (getFeatureBits().none()) {
       // Set default features.
       copySTI().ToggleFeature("SOUTHERN_ISLANDS");
     }
 
-    setAvailableFeatures(ComputeAvailableFeatures(getSTI().getFeatureBits()));
+    setAvailableFeatures(ComputeAvailableFeatures(getFeatureBits()));
 
     {
       // TODO: make those pre-defined variables read-only.
@@ -802,7 +802,7 @@ public:
       // MCSymbol::isRedefinable is intended for another purpose, and
       // AsmParser::parseDirectiveSet() cannot be specialized for specific target.
       AMDGPU::IsaInfo::IsaVersion ISA =
-          AMDGPU::IsaInfo::getIsaVersion(getSTI().getFeatureBits());
+          AMDGPU::IsaInfo::getIsaVersion(getFeatureBits());
       MCContext &Ctx = getContext();
       MCSymbol *Sym =
           Ctx.getOrCreateSymbol(Twine(".option.machine_version_major"));
@@ -828,7 +828,7 @@ public:
   }
 
   bool hasInv2PiInlineImm() const {
-    return getSTI().getFeatureBits()[AMDGPU::FeatureInv2PiInlineImm];
+    return getFeatureBits()[AMDGPU::FeatureInv2PiInlineImm];
   }
 
   bool hasSGPR102_SGPR103() const {
@@ -850,6 +850,10 @@ public:
     return &MII;
   }
 
+  const FeatureBitset &getFeatureBits() const {
+    return getSTI().getFeatureBits();
+  }
+
   void setForcedEncodingSize(unsigned Size) { ForcedEncodingSize = Size; }
   void setForcedDPP(bool ForceDPP_) { ForcedDPP = ForceDPP_; }
   void setForcedSDWA(bool ForceSDWA_) { ForcedSDWA = ForceSDWA_; }
@@ -1870,7 +1874,7 @@ bool AMDGPUAsmParser::ParseDirectiveHSACodeObjectISA() {
   // targeted GPU.
   if (getLexer().is(AsmToken::EndOfStatement)) {
     AMDGPU::IsaInfo::IsaVersion ISA =
-        AMDGPU::IsaInfo::getIsaVersion(getSTI().getFeatureBits());
+        AMDGPU::IsaInfo::getIsaVersion(getFeatureBits());
     getTargetStreamer().EmitDirectiveHSACodeObjectISA(ISA.Major, ISA.Minor,
                                                       ISA.Stepping,
                                                       "AMD", "AMDGPU");
@@ -1947,7 +1951,8 @@ bool AMDGPUAsmParser::ParseDirectiveRuntimeMetadata() {
 
   MS.flush();
 
-  getTargetStreamer().EmitRuntimeMetadata(Metadata);
+  if (getTargetStreamer().EmitRuntimeMetadata(getFeatureBits(), Metadata))
+    return Error(getParser().getTok().getLoc(), "invalid runtime metadata");
 
   return false;
 }
@@ -1965,7 +1970,7 @@ bool AMDGPUAsmParser::ParseAMDKernelCodeTValue(StringRef ID,
 
 bool AMDGPUAsmParser::ParseDirectiveAMDKernelCodeT() {
   amd_kernel_code_t Header;
-  AMDGPU::initDefaultAMDKernelCodeT(Header, getSTI().getFeatureBits());
+  AMDGPU::initDefaultAMDKernelCodeT(Header, getFeatureBits());
 
   while (true) {
     // Lex EndOfStatement.  This is in a while loop, because lexing a comment
@@ -2459,7 +2464,7 @@ bool AMDGPUAsmParser::parseCnt(int64_t &IntVal) {
     Parser.Lex();
 
   AMDGPU::IsaInfo::IsaVersion ISA =
-      AMDGPU::IsaInfo::getIsaVersion(getSTI().getFeatureBits());
+      AMDGPU::IsaInfo::getIsaVersion(getFeatureBits());
   if (CntName == "vmcnt")
     IntVal = encodeVmcnt(ISA, IntVal, CntVal);
   else if (CntName == "expcnt")
@@ -2475,7 +2480,7 @@ bool AMDGPUAsmParser::parseCnt(int64_t &IntVal) {
 OperandMatchResultTy
 AMDGPUAsmParser::parseSWaitCntOps(OperandVector &Operands) {
   AMDGPU::IsaInfo::IsaVersion ISA =
-      AMDGPU::IsaInfo::getIsaVersion(getSTI().getFeatureBits());
+      AMDGPU::IsaInfo::getIsaVersion(getFeatureBits());
   int64_t Waitcnt = getWaitcntBitMask(ISA);
   SMLoc S = Parser.getTok().getLoc();
 

diff --git a/lib/Target/AMDGPU/MCTargetDesc/AMDGPURuntimeMD.cpp b/lib/Target/AMDGPU/MCTargetDesc/AMDGPURuntimeMD.cpp
index 9bd8eddb60a8734af4668f8762dd6cc41d26d3ac..0e87d750f5df20c6d42f9f0c2c7885430734ccba 100644 (file)
--- a/lib/Target/AMDGPU/MCTargetDesc/AMDGPURuntimeMD.cpp
+++ b/lib/Target/AMDGPU/MCTargetDesc/AMDGPURuntimeMD.cpp
@@ -38,6 +38,7 @@
 #include <vector>
 
 using namespace llvm;
+using namespace llvm::AMDGPU::IsaInfo;
 using namespace ::AMDGPU::RuntimeMD;
 
 static cl::opt<bool>
@@ -88,7 +89,7 @@ template <> struct MappingTraits<Kernel::Metadata> {
         INVALID_KERNEL_INDEX);
     YamlIO.mapOptional(KeyName::NoPartialWorkGroups, K.NoPartialWorkGroups,
         uint8_t(0));
-    YamlIO.mapRequired(KeyName::Args, K.Args);
+    YamlIO.mapOptional(KeyName::Args, K.Args);
   }
   static const bool flow = true;
 };
@@ -116,7 +117,7 @@ template <> struct MappingTraits<IsaInfo::Metadata> {
 template <> struct MappingTraits<Program::Metadata> {
   static void mapping(IO &YamlIO, Program::Metadata &Prog) {
     YamlIO.mapRequired(KeyName::MDVersion, Prog.MDVersionSeq);
-    YamlIO.mapRequired(KeyName::IsaInfo, Prog.IsaInfo);
+    YamlIO.mapOptional(KeyName::IsaInfo, Prog.IsaInfo);
     YamlIO.mapOptional(KeyName::PrintfInfo, Prog.PrintfInfo);
     YamlIO.mapOptional(KeyName::Kernels, Prog.Kernels);
   }
@@ -375,6 +376,20 @@ static Kernel::Metadata getRuntimeMDForKernel(const Function &F) {
   return Kernel;
 }
 
+static void getIsaInfo(const FeatureBitset &Features, IsaInfo::Metadata &IIM) {
+  IIM.WavefrontSize = getWavefrontSize(Features);
+  IIM.LocalMemorySize = getLocalMemorySize(Features);
+  IIM.EUsPerCU = getEUsPerCU(Features);
+  IIM.MaxWavesPerEU = getMaxWavesPerEU(Features);
+  IIM.MaxFlatWorkGroupSize = getMaxFlatWorkGroupSize(Features);
+  IIM.SGPRAllocGranule = getSGPRAllocGranule(Features);
+  IIM.TotalNumSGPRs = getTotalNumSGPRs(Features);
+  IIM.AddressableNumSGPRs = getAddressableNumSGPRs(Features);
+  IIM.VGPRAllocGranule = getVGPRAllocGranule(Features);
+  IIM.TotalNumVGPRs = getTotalNumVGPRs(Features);
+  IIM.AddressableNumVGPRs = getAddressableNumVGPRs(Features);
+}
+
 Program::Metadata::Metadata(const std::string &YAML) {
   yaml::Input Input(YAML);
   Input >> *this;
@@ -411,18 +426,7 @@ std::string llvm::getRuntimeMDYAMLString(const FeatureBitset &Features,
   Prog.MDVersionSeq.push_back(MDVersion);
   Prog.MDVersionSeq.push_back(MDRevision);
 
-  IsaInfo::Metadata &IIM = Prog.IsaInfo;
-  IIM.WavefrontSize = AMDGPU::IsaInfo::getWavefrontSize(Features);
-  IIM.LocalMemorySize = AMDGPU::IsaInfo::getLocalMemorySize(Features);
-  IIM.EUsPerCU = AMDGPU::IsaInfo::getEUsPerCU(Features);
-  IIM.MaxWavesPerEU = AMDGPU::IsaInfo::getMaxWavesPerEU(Features);
-  IIM.MaxFlatWorkGroupSize = AMDGPU::IsaInfo::getMaxFlatWorkGroupSize(Features);
-  IIM.SGPRAllocGranule = AMDGPU::IsaInfo::getSGPRAllocGranule(Features);
-  IIM.TotalNumSGPRs = AMDGPU::IsaInfo::getTotalNumSGPRs(Features);
-  IIM.AddressableNumSGPRs = AMDGPU::IsaInfo::getAddressableNumSGPRs(Features);
-  IIM.VGPRAllocGranule = AMDGPU::IsaInfo::getVGPRAllocGranule(Features);
-  IIM.TotalNumVGPRs = AMDGPU::IsaInfo::getTotalNumVGPRs(Features);
-  IIM.AddressableNumVGPRs = AMDGPU::IsaInfo::getAddressableNumVGPRs(Features);
+  getIsaInfo(Features, Prog.IsaInfo);
 
   // Set PrintfInfo.
   if (auto MD = M.getNamedMetadata("llvm.printf.fmts")) {
@@ -451,3 +455,16 @@ std::string llvm::getRuntimeMDYAMLString(const FeatureBitset &Features,
 
   return YAML;
 }
+
+ErrorOr<std::string> llvm::getRuntimeMDYAMLString(const FeatureBitset &Features,
+                                                  StringRef YAML) {
+  Program::Metadata Prog;
+  yaml::Input Input(YAML);
+  Input >> Prog;
+
+  getIsaInfo(Features, Prog.IsaInfo);
+
+  if (Input.error())
+    return Input.error();
+  return Prog.toYAML();
+}

diff --git a/lib/Target/AMDGPU/MCTargetDesc/AMDGPURuntimeMD.h b/lib/Target/AMDGPU/MCTargetDesc/AMDGPURuntimeMD.h
index 8a73b3d642626c87fe55a90aee9d97449cbb7f14..2c1acdca797c4fbfe7e5df8c070ef3f2baf2854c 100644 (file)
--- a/lib/Target/AMDGPU/MCTargetDesc/AMDGPURuntimeMD.h
+++ b/lib/Target/AMDGPU/MCTargetDesc/AMDGPURuntimeMD.h
@@ -14,15 +14,20 @@
 #ifndef LLVM_LIB_TARGET_AMDGPU_MCTARGETDESC_AMDGPURUNTIMEMD_H
 #define LLVM_LIB_TARGET_AMDGPU_MCTARGETDESC_AMDGPURUNTIMEMD_H
 
+#include "llvm/Support/ErrorOr.h"
 #include <string>
 
 namespace llvm {
 class FeatureBitset;
 class Module;
 
-// Get runtime metadata as YAML string.
+/// \returns Runtime metadata as YAML string.
 std::string getRuntimeMDYAMLString(const FeatureBitset &Features,
                                    const Module &M);
 
+/// \returns \p YAML if \p YAML is valid runtime metadata, error otherwise.
+ErrorOr<std::string> getRuntimeMDYAMLString(const FeatureBitset &Features,
+                                            StringRef YAML);
+
 }
 #endif

diff --git a/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp b/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp
index 5935c6a49f09694e5a0e8586b2ac69e18b72be35..fc07225eebed47a9ce7dc28b2937fe7d5635dcff 100644 (file)
--- a/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp
+++ b/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.cpp
@@ -100,10 +100,17 @@ void AMDGPUTargetAsmStreamer::EmitRuntimeMetadata(const FeatureBitset &Features,
   OS << "\n\t.end_amdgpu_runtime_metadata\n";
 }
 
-void AMDGPUTargetAsmStreamer::EmitRuntimeMetadata(StringRef Metadata) {
+bool AMDGPUTargetAsmStreamer::EmitRuntimeMetadata(const FeatureBitset &Features,
+                                                  StringRef Metadata) {
+  auto VerifiedMetadata = getRuntimeMDYAMLString(Features, Metadata);
+  if (!VerifiedMetadata)
+    return true;
+
   OS << "\t.amdgpu_runtime_metadata";
-  OS << Metadata;
+  OS << VerifiedMetadata.get();
   OS << "\t.end_amdgpu_runtime_metadata\n";
+
+  return false;
 }
 
 //===----------------------------------------------------------------------===//
@@ -216,7 +223,12 @@ void AMDGPUTargetELFStreamer::EmitAMDGPUHsaProgramScopeGlobal(
   Symbol->setBinding(ELF::STB_GLOBAL);
 }
 
-void AMDGPUTargetELFStreamer::EmitRuntimeMetadata(StringRef Metadata) {
+bool AMDGPUTargetELFStreamer::EmitRuntimeMetadata(const FeatureBitset &Features,
+                                                  StringRef Metadata) {
+  auto VerifiedMetadata = getRuntimeMDYAMLString(Features, Metadata);
+  if (!VerifiedMetadata)
+    return true;
+
   // Create two labels to mark the beginning and end of the desc field
   // and a MCExpr to calculate the size of the desc field.
   auto &Context = getContext();
@@ -231,13 +243,15 @@ void AMDGPUTargetELFStreamer::EmitRuntimeMetadata(StringRef Metadata) {
     PT_NOTE::NT_AMDGPU_HSA_RUNTIME_METADATA,
     [&](MCELFStreamer &OS) {
       OS.EmitLabel(DescBegin);
-      OS.EmitBytes(Metadata);
+      OS.EmitBytes(VerifiedMetadata.get());
       OS.EmitLabel(DescEnd);
     }
   );
+
+  return false;
 }
 
 void AMDGPUTargetELFStreamer::EmitRuntimeMetadata(const FeatureBitset &Features,
                                                   const Module &M) {
-  EmitRuntimeMetadata(getRuntimeMDYAMLString(Features, M));
+  EmitRuntimeMetadata(Features, getRuntimeMDYAMLString(Features, M));
 }

diff --git a/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.h b/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.h
index 8504d1a18cb46b48cd066c37e37e8ca52e7b1c0c..20d7bc113cf51facda641ceb2406df36a3e86c9f 100644 (file)
--- a/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.h
+++ b/lib/Target/AMDGPU/MCTargetDesc/AMDGPUTargetStreamer.h
@@ -50,7 +50,9 @@ public:
   virtual void EmitRuntimeMetadata(const FeatureBitset &Features,
                                    const Module &M) = 0;
 
-  virtual void EmitRuntimeMetadata(StringRef Metadata) = 0;
+  /// \returns False on success, true on failure.
+  virtual bool EmitRuntimeMetadata(const FeatureBitset &Features,
+                                   StringRef Metadata) = 0;
 };
 
 class AMDGPUTargetAsmStreamer : public AMDGPUTargetStreamer {
@@ -75,7 +77,9 @@ public:
   void EmitRuntimeMetadata(const FeatureBitset &Features,
                            const Module &M) override;
 
-  void EmitRuntimeMetadata(StringRef Metadata) override;
+  /// \returns False on success, true on failure.
+  bool EmitRuntimeMetadata(const FeatureBitset &Features,
+                           StringRef Metadata) override;
 };
 
 class AMDGPUTargetELFStreamer : public AMDGPUTargetStreamer {
@@ -107,7 +111,9 @@ public:
   void EmitRuntimeMetadata(const FeatureBitset &Features,
                            const Module &M) override;
 
-  void EmitRuntimeMetadata(StringRef Metadata) override;
+  /// \returns False on success, true on failure.
+  bool EmitRuntimeMetadata(const FeatureBitset &Features,
+                           StringRef Metadata) override;
 };
 
 }

diff --git a/test/MC/AMDGPU/hsa.s b/test/MC/AMDGPU/hsa.s
index 21083f3915db2dff9fec468a2359400d1fb54ec2..a8d850966a757215e983a532f93a74fb0657c384 100644 (file)
--- a/test/MC/AMDGPU/hsa.s
+++ b/test/MC/AMDGPU/hsa.s
@@ -48,13 +48,13 @@
 .end_amdgpu_runtime_metadata
 
 // ASM: .amdgpu_runtime_metadata
-// ASM:     {
-// ASM:         amd.MDVersion: [ 2, 0 ]
-// ASM:         amd.Kernels: [
-// ASM:             { amd.KernelName: amd_kernel_code_t_test_all },
-// ASM:             { amd.KernelName: amd_kernel_code_t_minimal }
-// ASM:         ]
-// ASM:     }
+// ASM:  {
+// ASM:    amd.MDVersion: [ 2, 0 ]
+// ASM:    amd.IsaInfo: { amd.IsaInfoWavefrontSize: 64, amd.IsaInfoLocalMemorySize: 65536, amd.IsaInfoEUsPerCU: 4, amd.IsaInfoMaxWavesPerEU: 10, amd.IsaInfoMaxFlatWorkGroupSize: 2048, amd.IsaInfoSGPRAllocGranule: 8, amd.IsaInfoTotalNumSGPRs: 512, amd.IsaInfoAddressableNumSGPRs: 104, amd.IsaInfoVGPRAllocGranule: 4, amd.IsaInfoTotalNumVGPRs: 256, amd.IsaInfoAddressableNumVGPRs: 256 },
+// ASM:    amd.Kernels:
+// ASM:      - { amd.KernelName: amd_kernel_code_t_test_all }
+// ASM:      - { amd.KernelName: amd_kernel_code_t_minimal }
+// ASM:  }
 // ASM: .end_amdgpu_runtime_metadata
 
 .amdgpu_hsa_kernel amd_kernel_code_t_test_all

diff --git a/test/MC/AMDGPU/metadata.s b/test/MC/AMDGPU/metadata.s
deleted file mode 100644 (file)
index 3c009ff..0000000
--- a/test/MC/AMDGPU/metadata.s
+++ /dev/null
@@ -1,35 +0,0 @@
-// RUN: llvm-mc -triple amdgcn--amdhsa -mcpu=kaveri -show-encoding %s | FileCheck %s --check-prefix=ASM
-
-.amdgpu_runtime_metadata
-    { amd.MDVersion: [ 2, 0 ], amd.PrintfInfo: [ '1:1:4:%d\n', '2:1:8:%g\n' ], amd.Kernels:
-
-    - { amd.KernelName: test_char, amd.Language: OpenCL C, amd.LanguageVersion: [ 2, 0 ], amd.Args:
-        - { amd.ArgSize: 1, amd.ArgAlign: 1, amd.ArgKind: 0, amd.ArgValueType: 1, amd.ArgTypeName: char, amd.ArgAccQual: 0 }
-        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 7, amd.ArgValueType: 9 }
-        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 8, amd.ArgValueType: 9 }
-        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 9, amd.ArgValueType: 9 }
-        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 11, amd.ArgValueType: 1, amd.ArgAddrQual: 1 } }
-
-   - { amd.KernelName: test_ushort2, amd.Language: OpenCL C, amd.LanguageVersion: [ 2, 0 ], amd.Args:
-       - { amd.ArgSize: 4, amd.ArgAlign: 4, amd.ArgKind: 0, amd.ArgValueType: 4, amd.ArgTypeName: ushort2, amd.ArgAccQual: 0 }
-       - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 7, amd.ArgValueType: 9 }
-       - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 8, amd.ArgValueType: 9 }
-       - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 9, amd.ArgValueType: 9 }
-       - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 11, amd.ArgValueType: 1, amd.ArgAddrQual: 1 } }
-   }
-.end_amdgpu_runtime_metadata
-
-// ASM: { amd.MDVersion: [ 2, 0 ], amd.PrintfInfo: [ '1:1:4:%d\n', '2:1:8:%g\n' ], amd.Kernels:
-// ASM: - { amd.KernelName: test_char, amd.Language: OpenCL C, amd.LanguageVersion: [ 2, 0 ], amd.Args:
-// ASM:     - { amd.ArgSize: 1, amd.ArgAlign: 1, amd.ArgKind: 0, amd.ArgValueType: 1, amd.ArgTypeName: char, amd.ArgAccQual: 0 }
-// ASM:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 7, amd.ArgValueType: 9 }
-// ASM:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 8, amd.ArgValueType: 9 }
-// ASM:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 9, amd.ArgValueType: 9 }
-// ASM:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 11, amd.ArgValueType: 1, amd.ArgAddrQual: 1 } }
-// ASM: - { amd.KernelName: test_ushort2, amd.Language: OpenCL C, amd.LanguageVersion: [ 2, 0 ], amd.Args:
-// ASM:    - { amd.ArgSize: 4, amd.ArgAlign: 4, amd.ArgKind: 0, amd.ArgValueType: 4, amd.ArgTypeName: ushort2, amd.ArgAccQual: 0 }
-// ASM:    - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 7, amd.ArgValueType: 9 }
-// ASM:    - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 8, amd.ArgValueType: 9 }
-// ASM:    - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 9, amd.ArgValueType: 9 }
-// ASM:    - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 11, amd.ArgValueType: 1, amd.ArgAddrQual: 1 } }
-// ASM: }

diff --git a/test/MC/AMDGPU/runtime-metadata-1.s b/test/MC/AMDGPU/runtime-metadata-1.s
new file mode 100644 (file)
index 0000000..6e371a8
--- /dev/null
+++ b/test/MC/AMDGPU/runtime-metadata-1.s
@@ -0,0 +1,39 @@
+// RUN: llvm-mc -triple=amdgcn-amd-amdhsa -mcpu=gfx700 -show-encoding %s | FileCheck %s --check-prefix=GCN --check-prefix=GFX700
+// RUN: llvm-mc -triple=amdgcn-amd-amdhsa -mcpu=gfx800 -show-encoding %s | FileCheck %s --check-prefix=GCN --check-prefix=GFX800
+// RUN: llvm-mc -triple=amdgcn-amd-amdhsa -mcpu=gfx900 -show-encoding %s | FileCheck %s --check-prefix=GCN --check-prefix=GFX900
+
+.amdgpu_runtime_metadata
+  { amd.MDVersion: [ 2, 1 ], amd.PrintfInfo: [ '1:1:4:%d\n', '2:1:8:%g\n' ], amd.Kernels:
+    - { amd.KernelName: test_char, amd.Language: OpenCL C, amd.LanguageVersion: [ 2, 0 ], amd.Args:
+        - { amd.ArgSize: 1, amd.ArgAlign: 1, amd.ArgKind: 0, amd.ArgValueType: 1, amd.ArgTypeName: char, amd.ArgAccQual: 0 }
+        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 7, amd.ArgValueType: 9 }
+        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 8, amd.ArgValueType: 9 }
+        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 9, amd.ArgValueType: 9 }
+        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 11, amd.ArgValueType: 1, amd.ArgAddrQual: 1 } }
+    - { amd.KernelName: test_ushort2, amd.Language: OpenCL C, amd.LanguageVersion: [ 2, 0 ], amd.Args:
+        - { amd.ArgSize: 4, amd.ArgAlign: 4, amd.ArgKind: 0, amd.ArgValueType: 4, amd.ArgTypeName: ushort2, amd.ArgAccQual: 0 }
+        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 7, amd.ArgValueType: 9 }
+        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 8, amd.ArgValueType: 9 }
+        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 9, amd.ArgValueType: 9 }
+        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 11, amd.ArgValueType: 1, amd.ArgAddrQual: 1 } }
+  }
+.end_amdgpu_runtime_metadata
+
+// GFX700: { amd.MDVersion: [ 2, 1 ], amd.IsaInfo: { amd.IsaInfoWavefrontSize: 64, amd.IsaInfoLocalMemorySize: 65536, amd.IsaInfoEUsPerCU: 4, amd.IsaInfoMaxWavesPerEU: 10, amd.IsaInfoMaxFlatWorkGroupSize: 2048, amd.IsaInfoSGPRAllocGranule: 8, amd.IsaInfoTotalNumSGPRs: 512, amd.IsaInfoAddressableNumSGPRs: 104, amd.IsaInfoVGPRAllocGranule: 4, amd.IsaInfoTotalNumVGPRs: 256, amd.IsaInfoAddressableNumVGPRs: 256 }, amd.PrintfInfo: [ '1:1:4:%d\n', '2:1:8:%g\n' ], amd.Kernels:
+
+// GFX800: { amd.MDVersion: [ 2, 1 ], amd.IsaInfo: { amd.IsaInfoWavefrontSize: 64, amd.IsaInfoLocalMemorySize: 65536, amd.IsaInfoEUsPerCU: 4, amd.IsaInfoMaxWavesPerEU: 10, amd.IsaInfoMaxFlatWorkGroupSize: 2048, amd.IsaInfoSGPRAllocGranule: 16, amd.IsaInfoTotalNumSGPRs: 800, amd.IsaInfoAddressableNumSGPRs: 96, amd.IsaInfoVGPRAllocGranule: 4, amd.IsaInfoTotalNumVGPRs: 256, amd.IsaInfoAddressableNumVGPRs: 256 }, amd.PrintfInfo: [ '1:1:4:%d\n', '2:1:8:%g\n' ], amd.Kernels:
+
+// GFX900: { amd.MDVersion: [ 2, 1 ], amd.IsaInfo: { amd.IsaInfoWavefrontSize: 64, amd.IsaInfoLocalMemorySize: 65536, amd.IsaInfoEUsPerCU: 4, amd.IsaInfoMaxWavesPerEU: 10, amd.IsaInfoMaxFlatWorkGroupSize: 2048, amd.IsaInfoSGPRAllocGranule: 16, amd.IsaInfoTotalNumSGPRs: 800, amd.IsaInfoAddressableNumSGPRs: 102, amd.IsaInfoVGPRAllocGranule: 4, amd.IsaInfoTotalNumVGPRs: 256, amd.IsaInfoAddressableNumVGPRs: 256 }, amd.PrintfInfo: [ '1:1:4:%d\n', '2:1:8:%g\n' ], amd.Kernels:
+
+// GCN:      - { amd.KernelName: test_char, amd.Language: OpenCL C, amd.LanguageVersion: [ 2, 0 ], amd.Args:
+// GCN-NEXT:     - { amd.ArgSize: 1, amd.ArgAlign: 1, amd.ArgKind: 0, amd.ArgValueType: 1, amd.ArgTypeName: char, amd.ArgAccQual: 0 }
+// GCN-NEXT:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 7, amd.ArgValueType: 9 }
+// GCN-NEXT:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 8, amd.ArgValueType: 9 }
+// GCN-NEXT:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 9, amd.ArgValueType: 9 }
+// GCN-NEXT:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 11, amd.ArgValueType: 1, amd.ArgAddrQual: 1 } }
+// GCN-NEXT: - { amd.KernelName: test_ushort2, amd.Language: OpenCL C, amd.LanguageVersion: [ 2, 0 ], amd.Args:
+// GCN-NEXT:     - { amd.ArgSize: 4, amd.ArgAlign: 4, amd.ArgKind: 0, amd.ArgValueType: 4, amd.ArgTypeName: ushort2, amd.ArgAccQual: 0 }
+// GCN-NEXT:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 7, amd.ArgValueType: 9 }
+// GCN-NEXT:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 8, amd.ArgValueType: 9 }
+// GCN-NEXT:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 9, amd.ArgValueType: 9 }
+// GCN-NEXT:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 11, amd.ArgValueType: 1, amd.ArgAddrQual: 1 } } }

diff --git a/test/MC/AMDGPU/runtime-metadata-2.s b/test/MC/AMDGPU/runtime-metadata-2.s
new file mode 100644 (file)
index 0000000..569ddd5
--- /dev/null
+++ b/test/MC/AMDGPU/runtime-metadata-2.s
@@ -0,0 +1,39 @@
+// RUN: llvm-mc -triple=amdgcn-amd-amdhsa -mcpu=gfx700 -show-encoding %s | FileCheck %s --check-prefix=GCN --check-prefix=GFX700
+// RUN: llvm-mc -triple=amdgcn-amd-amdhsa -mcpu=gfx800 -show-encoding %s | FileCheck %s --check-prefix=GCN --check-prefix=GFX800
+// RUN: llvm-mc -triple=amdgcn-amd-amdhsa -mcpu=gfx900 -show-encoding %s | FileCheck %s --check-prefix=GCN --check-prefix=GFX900
+
+.amdgpu_runtime_metadata
+  { amd.MDVersion: [ 2, 1 ], amd.IsaInfo: { amd.IsaInfoWavefrontSize: 64, amd.IsaInfoLocalMemorySize: 65536, amd.IsaInfoEUsPerCU: 4, amd.IsaInfoMaxWavesPerEU: 10, amd.IsaInfoMaxFlatWorkGroupSize: 2048, amd.IsaInfoSGPRAllocGranule: 8, amd.IsaInfoTotalNumSGPRs: 512, amd.IsaInfoAddressableNumSGPRs: 104, amd.IsaInfoVGPRAllocGranule: 4, amd.IsaInfoTotalNumVGPRs: 256, amd.IsaInfoAddressableNumVGPRs: 256 }, amd.PrintfInfo: [ '1:1:4:%d\n', '2:1:8:%g\n' ], amd.Kernels:
+    - { amd.KernelName: test_char, amd.Language: OpenCL C, amd.LanguageVersion: [ 2, 0 ], amd.Args:
+        - { amd.ArgSize: 1, amd.ArgAlign: 1, amd.ArgKind: 0, amd.ArgValueType: 1, amd.ArgTypeName: char, amd.ArgAccQual: 0 }
+        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 7, amd.ArgValueType: 9 }
+        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 8, amd.ArgValueType: 9 }
+        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 9, amd.ArgValueType: 9 }
+        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 11, amd.ArgValueType: 1, amd.ArgAddrQual: 1 } }
+    - { amd.KernelName: test_ushort2, amd.Language: OpenCL C, amd.LanguageVersion: [ 2, 0 ], amd.Args:
+        - { amd.ArgSize: 4, amd.ArgAlign: 4, amd.ArgKind: 0, amd.ArgValueType: 4, amd.ArgTypeName: ushort2, amd.ArgAccQual: 0 }
+        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 7, amd.ArgValueType: 9 }
+        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 8, amd.ArgValueType: 9 }
+        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 9, amd.ArgValueType: 9 }
+        - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 11, amd.ArgValueType: 1, amd.ArgAddrQual: 1 } }
+  }
+.end_amdgpu_runtime_metadata
+
+// GFX700: { amd.MDVersion: [ 2, 1 ], amd.IsaInfo: { amd.IsaInfoWavefrontSize: 64, amd.IsaInfoLocalMemorySize: 65536, amd.IsaInfoEUsPerCU: 4, amd.IsaInfoMaxWavesPerEU: 10, amd.IsaInfoMaxFlatWorkGroupSize: 2048, amd.IsaInfoSGPRAllocGranule: 8, amd.IsaInfoTotalNumSGPRs: 512, amd.IsaInfoAddressableNumSGPRs: 104, amd.IsaInfoVGPRAllocGranule: 4, amd.IsaInfoTotalNumVGPRs: 256, amd.IsaInfoAddressableNumVGPRs: 256 }, amd.PrintfInfo: [ '1:1:4:%d\n', '2:1:8:%g\n' ], amd.Kernels:
+
+// GFX800: { amd.MDVersion: [ 2, 1 ], amd.IsaInfo: { amd.IsaInfoWavefrontSize: 64, amd.IsaInfoLocalMemorySize: 65536, amd.IsaInfoEUsPerCU: 4, amd.IsaInfoMaxWavesPerEU: 10, amd.IsaInfoMaxFlatWorkGroupSize: 2048, amd.IsaInfoSGPRAllocGranule: 16, amd.IsaInfoTotalNumSGPRs: 800, amd.IsaInfoAddressableNumSGPRs: 96, amd.IsaInfoVGPRAllocGranule: 4, amd.IsaInfoTotalNumVGPRs: 256, amd.IsaInfoAddressableNumVGPRs: 256 }, amd.PrintfInfo: [ '1:1:4:%d\n', '2:1:8:%g\n' ], amd.Kernels:
+
+// GFX900: { amd.MDVersion: [ 2, 1 ], amd.IsaInfo: { amd.IsaInfoWavefrontSize: 64, amd.IsaInfoLocalMemorySize: 65536, amd.IsaInfoEUsPerCU: 4, amd.IsaInfoMaxWavesPerEU: 10, amd.IsaInfoMaxFlatWorkGroupSize: 2048, amd.IsaInfoSGPRAllocGranule: 16, amd.IsaInfoTotalNumSGPRs: 800, amd.IsaInfoAddressableNumSGPRs: 102, amd.IsaInfoVGPRAllocGranule: 4, amd.IsaInfoTotalNumVGPRs: 256, amd.IsaInfoAddressableNumVGPRs: 256 }, amd.PrintfInfo: [ '1:1:4:%d\n', '2:1:8:%g\n' ], amd.Kernels:
+
+// GCN:      - { amd.KernelName: test_char, amd.Language: OpenCL C, amd.LanguageVersion: [ 2, 0 ], amd.Args:
+// GCN-NEXT:     - { amd.ArgSize: 1, amd.ArgAlign: 1, amd.ArgKind: 0, amd.ArgValueType: 1, amd.ArgTypeName: char, amd.ArgAccQual: 0 }
+// GCN-NEXT:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 7, amd.ArgValueType: 9 }
+// GCN-NEXT:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 8, amd.ArgValueType: 9 }
+// GCN-NEXT:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 9, amd.ArgValueType: 9 }
+// GCN-NEXT:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 11, amd.ArgValueType: 1, amd.ArgAddrQual: 1 } }
+// GCN-NEXT: - { amd.KernelName: test_ushort2, amd.Language: OpenCL C, amd.LanguageVersion: [ 2, 0 ], amd.Args:
+// GCN-NEXT:     - { amd.ArgSize: 4, amd.ArgAlign: 4, amd.ArgKind: 0, amd.ArgValueType: 4, amd.ArgTypeName: ushort2, amd.ArgAccQual: 0 }
+// GCN-NEXT:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 7, amd.ArgValueType: 9 }
+// GCN-NEXT:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 8, amd.ArgValueType: 9 }
+// GCN-NEXT:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 9, amd.ArgValueType: 9 }
+// GCN-NEXT:     - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 11, amd.ArgValueType: 1, amd.ArgAddrQual: 1 } } }

diff --git a/test/MC/AMDGPU/runtime-metadata-invalid-1.s b/test/MC/AMDGPU/runtime-metadata-invalid-1.s
new file mode 100644 (file)
index 0000000..66bbbd7
--- /dev/null
+++ b/test/MC/AMDGPU/runtime-metadata-invalid-1.s
@@ -0,0 +1,106 @@
+; RUN: not llvm-mc -triple=amdgcn-amd-amdhsa -mcpu=gfx700 %s 2>&1 | FileCheck %s
+; RUN: not llvm-mc -triple=amdgcn-amd-amdhsa -mcpu=gfx700 -filetype=obj %s 2>&1 | FileCheck %s
+; RUN: not llvm-mc -triple=amdgcn-amd-amdhsa -mcpu=gfx800 %s 2>&1 | FileCheck %s
+; RUN: not llvm-mc -triple=amdgcn-amd-amdhsa -mcpu=gfx800 -filetype=obj %s 2>&1 | FileCheck %s
+; RUN: not llvm-mc -triple=amdgcn-amd-amdhsa -mcpu=gfx900 %s 2>&1 | FileCheck %s
+; RUN: not llvm-mc -triple=amdgcn-amd-amdhsa -mcpu=gfx900 -filetype=obj %s 2>&1 | FileCheck %s
+
+; CHECK: error: unknown key 'amd.RandomUnknownKey'
+
+       .text
+       .hsa_code_object_version 2,1
+       .hsa_code_object_isa 8,0,3,"AMD","AMDGPU"
+       .amdgpu_runtime_metadata
+---
+{ amd.MDVersion: [ 2, 1 ], amd.RandomUnknownKey, amd.IsaInfo: { amd.IsaInfoWavefrontSize: 64, amd.IsaInfoLocalMemorySize: 65536, amd.IsaInfoEUsPerCU: 4, amd.IsaInfoMaxWavesPerEU: 10, amd.IsaInfoMaxFlatWorkGroupSize: 2048, amd.IsaInfoSGPRAllocGranule: 16, amd.IsaInfoTotalNumSGPRs: 800, amd.IsaInfoAddressableNumSGPRs: 102, amd.IsaInfoVGPRAllocGranule: 4, amd.IsaInfoTotalNumVGPRs: 256, amd.IsaInfoAddressableNumVGPRs: 256 }, amd.Kernels:
+  - { amd.KernelName: test, amd.Language: OpenCL C, amd.LanguageVersion: [ 1, 0 ], amd.Args:
+      - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 1, amd.ArgValueType: 6, amd.ArgTypeName: 'int*', amd.ArgAddrQual: 1, amd.ArgAccQual: 0 }
+      - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 7, amd.ArgValueType: 9 }
+      - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 8, amd.ArgValueType: 9 }
+      - { amd.ArgSize: 8, amd.ArgAlign: 8, amd.ArgKind: 9, amd.ArgValueType: 9 } } }
+...
+
+       .end_amdgpu_runtime_metadata
+       .globl  test
+       .p2align        8
+       .type   test,@function
+       .amdgpu_hsa_kernel test
+test:                                   ; @test
+       .amd_kernel_code_t
+               amd_code_version_major = 1
+               amd_code_version_minor = 0
+               amd_machine_kind = 1
+               amd_machine_version_major = 8
+               amd_machine_version_minor = 0
+               amd_machine_version_stepping = 3
+               kernel_code_entry_byte_offset = 256
+               kernel_code_prefetch_byte_size = 0
+               max_scratch_backing_memory_byte_size = 0
+               granulated_workitem_vgpr_count = 0
+               granulated_wavefront_sgpr_count = 0
+               priority = 0
+               float_mode = 192
+               priv = 0
+               enable_dx10_clamp = 1
+               debug_mode = 0
+               enable_ieee_mode = 1
+               enable_sgpr_private_segment_wave_byte_offset = 0
+               user_sgpr_count = 6
+               enable_trap_handler = 1
+               enable_sgpr_workgroup_id_x = 1
+               enable_sgpr_workgroup_id_y = 0
+               enable_sgpr_workgroup_id_z = 0
+               enable_sgpr_workgroup_info = 0
+               enable_vgpr_workitem_id = 0
+               enable_exception_msb = 0
+               granulated_lds_size = 0
+               enable_exception = 0
+               enable_sgpr_private_segment_buffer = 1
+               enable_sgpr_dispatch_ptr = 0
+               enable_sgpr_queue_ptr = 0
+               enable_sgpr_kernarg_segment_ptr = 1
+               enable_sgpr_dispatch_id = 0
+               enable_sgpr_flat_scratch_init = 0
+               enable_sgpr_private_segment_size = 0
+               enable_sgpr_grid_workgroup_count_x = 0
+               enable_sgpr_grid_workgroup_count_y = 0
+               enable_sgpr_grid_workgroup_count_z = 0
+               enable_ordered_append_gds = 0
+               private_element_size = 1
+               is_ptr64 = 1
+               is_dynamic_callstack = 0
+               is_debug_enabled = 0
+               is_xnack_enabled = 0
+               workitem_private_segment_byte_size = 0
+               workgroup_group_segment_byte_size = 0
+               gds_segment_byte_size = 0
+               kernarg_segment_byte_size = 8
+               workgroup_fbarrier_count = 0
+               wavefront_sgpr_count = 6
+               workitem_vgpr_count = 3
+               reserved_vgpr_first = 0
+               reserved_vgpr_count = 0
+               reserved_sgpr_first = 0
+               reserved_sgpr_count = 0
+               debug_wavefront_private_segment_offset_sgpr = 0
+               debug_private_segment_buffer_sgpr = 0
+               kernarg_segment_alignment = 4
+               group_segment_alignment = 4
+               private_segment_alignment = 4
+               wavefront_size = 6
+               call_convention = -1
+               runtime_loader_kernel_symbol = 0
+       .end_amd_kernel_code_t
+; BB#0:                                 ; %entry
+       s_load_dwordx2 s[0:1], s[4:5], 0x0
+       v_mov_b32_e32 v2, 0x309
+       s_waitcnt lgkmcnt(0)
+       v_mov_b32_e32 v0, s0
+       v_mov_b32_e32 v1, s1
+       flat_store_dword v[0:1], v2
+       s_endpgm
+.Lfunc_end0:
+       .size   test, .Lfunc_end0-test
+
+       .ident  ""
+       .section        ".note.GNU-stack"