Revert "[mlir] Create a gpu.module operation for the GPU Dialect."

This reverts commit 4624a1e. Causing
problems downstream.

(cherry picked from commit 0133cc6)

mlir/include/mlir/Conversion/GPUToCUDA/GPUToCUDAPass.h

@@ -19,17 +19,12 @@ namespace mlir {
 class Location;
 class ModuleOp;
 
-template <typename T>
-class OpPassBase;
-
-namespace gpu {
-class GPUModuleOp;
-} // namespace gpu
-
 namespace LLVM {
 class LLVMDialect;
 } // namespace LLVM
 
+template <typename T> class OpPassBase;
+
 using OwnedCubin = std::unique_ptr<std::vector<char>>;
 using CubinGenerator =
     std::function<OwnedCubin(const std::string &, Location, StringRef)>;
@@ -43,7 +38,7 @@ using CubinGenerator =
 /// attached as a string attribute named 'nvvm.cubin' to the kernel function.
 /// After the transformation, the body of the kernel function is removed (i.e.,
 /// it is turned into a declaration).
-std::unique_ptr<OpPassBase<gpu::GPUModuleOp>>
+std::unique_ptr<OpPassBase<ModuleOp>>
 createConvertGPUKernelToCubinPass(CubinGenerator cubinGenerator);
 
 /// Creates a pass to convert a gpu.launch_func operation into a sequence of

mlir/include/mlir/Conversion/GPUToNVVM/GPUToNVVMPass.h

@@ -14,19 +14,15 @@ namespace mlir {
 class LLVMTypeConverter;
 class OwningRewritePatternList;
 
-template <typename OpT>
-class OpPassBase;
-
-namespace gpu {
-class GPUModuleOp;
-}
+class ModuleOp;
+template <typename OpT> class OpPassBase;
 
 /// Collect a set of patterns to convert from the GPU dialect to NVVM.
 void populateGpuToNVVMConversionPatterns(LLVMTypeConverter &converter,
                                          OwningRewritePatternList &patterns);
 
 /// Creates a pass that lowers GPU dialect operations to NVVM counterparts.
-std::unique_ptr<OpPassBase<gpu::GPUModuleOp>> createLowerGpuOpsToNVVMOpsPass();
+std::unique_ptr<OpPassBase<ModuleOp>> createLowerGpuOpsToNVVMOpsPass();
 
 } // namespace mlir
 

mlir/include/mlir/Dialect/GPU/GPUOps.td

@@ -588,56 +588,4 @@ def GPU_BarrierOp : GPU_Op<"barrier"> {
   let printer = [{ p << getOperationName(); }];
 }
 
-def GPU_GPUModuleOp : GPU_Op<"module", [
-  IsolatedFromAbove, SymbolTable, Symbol,
-  SingleBlockImplicitTerminator<"ModuleEndOp">
-]> {
-  let summary = "A top level compilation unit containing code to be run on a GPU.";
-  let description = [{
-    GPU module contains code that is intended to be run on a GPU. A host device
-    can launch this code through a gpu.launc_func that creates a fully
-    qualified symbol through the gpu.module's symbol and a gpu.func symbol
-    contained in the gpu.module.
-
-    The module's top-level scope is modeled by a single region with a single
-    block. GPU modules are required to have a name that is used for symbol
-    resolution by the gpu.launch_func operation.
-
-    Using an op with a region to define a GPU module enables "embedding" GPU
-    modules with SIMT execution models in other dialects in a clean manner and
-    allows filtering of code regions to execute passes on only code intended to
-    or not intended to be run on the separate device.
-
-    ```
-      gpu.module @symbol_name {
-      gpu.func {}
-        ...
-      gpu.module_end
-    }
-
-    ```
-  }];
-  let builders = [OpBuilder<"Builder *builder, OperationState &result, "
-                            "StringRef name">];
-  let parser = [{ return ::parseGPUModuleOp(parser, result); }];
-  let printer = [{ return ::print(p, *this); }];
-  let regions = (region SizedRegion<1>:$body);
-
-  // We need to ensure the block inside the region is properly terminated;
-  // the auto-generated builders do not guarantee that.
-  let skipDefaultBuilders = 1;
-}
-
-def GPU_ModuleEndOp : GPU_Op<"module_end", [
-  Terminator, HasParent<"GPUModuleOp">
-]> {
-  let summary = "A pseudo op that marks the end of a gpu.module.";
-  let description = [{
-    This op terminates the only block inside the only region of a `gpu.module`.
-  }];
-
-  let parser = [{ return success(); }];
-  let printer = [{ p << getOperationName(); }];
-}
-
 #endif // GPU_OPS

mlir/lib/Conversion/GPUToCUDA/ConvertKernelFuncToCubin.cpp

@@ -46,15 +46,18 @@ static constexpr const char *kCubinAnnotation = "nvvm.cubin";
 /// IR and further to PTX. A user provided CubinGenerator compiles the PTX to
 /// GPU binary code, which is then attached as an attribute to the function. The
 /// function body is erased.
-class GpuKernelToCubinPass
-    : public OperationPass<GpuKernelToCubinPass, gpu::GPUModuleOp> {
+class GpuKernelToCubinPass : public ModulePass<GpuKernelToCubinPass> {
 public:
   GpuKernelToCubinPass(
       CubinGenerator cubinGenerator = compilePtxToCubinForTesting)
       : cubinGenerator(cubinGenerator) {}
 
-  void runOnOperation() override {
-    gpu::GPUModuleOp module = getOperation();
+  void runOnModule() override {
+    ModuleOp module = getModule();
+    if (!module.getAttrOfType<UnitAttr>(
+            gpu::GPUDialect::getKernelModuleAttrName()) ||
+        !module.getName())
+      return;
 
     // Make sure the NVPTX target is initialized.
     LLVMInitializeNVPTXTarget();
@@ -68,8 +71,8 @@ class GpuKernelToCubinPass
 
     // Translate the module to CUBIN and attach the result as attribute to the
     // module.
-    if (auto cubinAttr = translateGPUModuleToCubinAnnotation(
-            *llvmModule, module.getLoc(), module.getName()))
+    if (auto cubinAttr = translateGpuModuleToCubinAnnotation(
+            *llvmModule, module.getLoc(), *module.getName()))
       module.setAttr(kCubinAnnotation, cubinAttr);
     else
       signalPassFailure();
@@ -89,7 +92,7 @@ class GpuKernelToCubinPass
                                   StringRef name);
 
   /// Translates llvmModule to cubin and returns the result as attribute.
-  StringAttr translateGPUModuleToCubinAnnotation(llvm::Module &llvmModule,
+  StringAttr translateGpuModuleToCubinAnnotation(llvm::Module &llvmModule,
                                                  Location loc, StringRef name);
 
   CubinGenerator cubinGenerator;
@@ -146,15 +149,15 @@ OwnedCubin GpuKernelToCubinPass::convertModuleToCubin(llvm::Module &llvmModule,
   return cubinGenerator(ptx, loc, name);
 }
 
-StringAttr GpuKernelToCubinPass::translateGPUModuleToCubinAnnotation(
+StringAttr GpuKernelToCubinPass::translateGpuModuleToCubinAnnotation(
     llvm::Module &llvmModule, Location loc, StringRef name) {
   auto cubin = convertModuleToCubin(llvmModule, loc, name);
   if (!cubin)
     return {};
   return StringAttr::get({cubin->data(), cubin->size()}, loc->getContext());
 }
 
-std::unique_ptr<OpPassBase<gpu::GPUModuleOp>>
+std::unique_ptr<OpPassBase<ModuleOp>>
 mlir::createConvertGPUKernelToCubinPass(CubinGenerator cubinGenerator) {
   return std::make_unique<GpuKernelToCubinPass>(cubinGenerator);
 }

mlir/lib/Conversion/GPUToCUDA/ConvertLaunchFuncToCudaCalls.cpp

@@ -132,9 +132,9 @@ class GpuLaunchFuncToCudaCallsPass
 
     // GPU kernel modules are no longer necessary since we have a global
     // constant with the CUBIN data.
-    for (auto m :
-         llvm::make_early_inc_range(getModule().getOps<gpu::GPUModuleOp>()))
-      m.erase();
+    for (auto m : llvm::make_early_inc_range(getModule().getOps<ModuleOp>()))
+      if (m.getAttrOfType<UnitAttr>(gpu::GPUDialect::getKernelModuleAttrName()))
+        m.erase();
   }
 
 private:
@@ -343,8 +343,8 @@ void GpuLaunchFuncToCudaCallsPass::translateGpuLaunchCalls(
                                                builder.getI32IntegerAttr(0));
   // Create an LLVM global with CUBIN extracted from the kernel annotation and
   // obtain a pointer to the first byte in it.
-  auto kernelModule = getModule().lookupSymbol<gpu::GPUModuleOp>(
-      launchOp.getKernelModuleName());
+  auto kernelModule =
+      getModule().lookupSymbol<ModuleOp>(launchOp.getKernelModuleName());
   assert(kernelModule && "expected a kernel module");
 
   auto cubinAttr = kernelModule.getAttrOfType<StringAttr>(kCubinAnnotation);
@@ -354,7 +354,8 @@ void GpuLaunchFuncToCudaCallsPass::translateGpuLaunchCalls(
     return signalPassFailure();
   }
 
-  SmallString<128> nameBuffer(kernelModule.getName());
+  assert(kernelModule.getName() && "expected a named module");
+  SmallString<128> nameBuffer(*kernelModule.getName());
   nameBuffer.append(kCubinStorageSuffix);
   Value data = LLVM::createGlobalString(
       loc, builder, nameBuffer.str(), cubinAttr.getValue(),

mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp

@@ -200,7 +200,7 @@ struct GPUAllReduceOpLowering : public LLVMOpLowering {
     auto type = operand.getType().cast<LLVM::LLVMType>();
 
     // Create shared memory array to store the warp reduction.
-    auto module = operand.getDefiningOp()->getParentOfType<gpu::GPUModuleOp>();
+    auto module = operand.getDefiningOp()->getParentOfType<ModuleOp>();
     assert(module && "op must belong to a module");
     Value sharedMemPtr =
         createSharedMemoryArray(loc, module, type, kWarpSize, rewriter);
@@ -391,10 +391,10 @@ struct GPUAllReduceOpLowering : public LLVMOpLowering {
   }
 
   /// Creates a global array stored in shared memory.
-  Value createSharedMemoryArray(Location loc, gpu::GPUModuleOp module,
+  Value createSharedMemoryArray(Location loc, ModuleOp module,
                                 LLVM::LLVMType elementType, int numElements,
                                 ConversionPatternRewriter &rewriter) const {
-    OpBuilder builder(module.body());
+    OpBuilder builder(module.getBodyRegion());
 
     auto arrayType = LLVM::LLVMType::getArrayTy(elementType, numElements);
     StringRef name = "reduce_buffer";
@@ -699,11 +699,13 @@ struct GPUReturnOpLowering : public LLVMOpLowering {
 ///
 /// This pass only handles device code and is not meant to be run on GPU host
 /// code.
-class LowerGpuOpsToNVVMOpsPass
-    : public OperationPass<LowerGpuOpsToNVVMOpsPass, gpu::GPUModuleOp> {
+class LowerGpuOpsToNVVMOpsPass : public ModulePass<LowerGpuOpsToNVVMOpsPass> {
 public:
-  void runOnOperation() override {
-    gpu::GPUModuleOp m = getOperation();
+  void runOnModule() override {
+    ModuleOp m = getModule();
+    if (!m.getAttrOfType<UnitAttr>(gpu::GPUDialect::getKernelModuleAttrName()))
+      return;
+
     OwningRewritePatternList patterns;
     NVVMTypeConverter converter(m.getContext());
     populateStdToLLVMConversionPatterns(converter, patterns);
@@ -716,7 +718,7 @@ class LowerGpuOpsToNVVMOpsPass
     target.addLegalDialect<LLVM::LLVMDialect>();
     target.addLegalDialect<NVVM::NVVMDialect>();
     // TODO(csigg): Remove once we support replacing non-root ops.
-    target.addLegalOp<gpu::YieldOp, gpu::GPUModuleOp, gpu::ModuleEndOp>();
+    target.addLegalOp<gpu::YieldOp>();
     if (failed(applyPartialConversion(m, target, patterns, &converter)))
       signalPassFailure();
   }
@@ -748,8 +750,7 @@ void mlir::populateGpuToNVVMConversionPatterns(
                                                "__nv_exp");
 }
 
-std::unique_ptr<OpPassBase<gpu::GPUModuleOp>>
-mlir::createLowerGpuOpsToNVVMOpsPass() {
+std::unique_ptr<OpPassBase<ModuleOp>> mlir::createLowerGpuOpsToNVVMOpsPass() {
   return std::make_unique<LowerGpuOpsToNVVMOpsPass>();
 }
 

mlir/lib/Conversion/GPUToSPIRV/CMakeLists.txt

@@ -1,15 +1,8 @@
-set(LLVM_TARGET_DEFINITIONS GPUToSPIRV.td)
-mlir_tablegen(GPUToSPIRV.cpp.inc -gen-rewriters)
-add_public_tablegen_target(MLIRGPUToSPIRVIncGen)
-
 add_llvm_library(MLIRGPUtoSPIRVTransforms
   ConvertGPUToSPIRV.cpp
   ConvertGPUToSPIRVPass.cpp
   )
 
-add_dependencies(MLIRGPUtoSPIRVTransforms
-  MLIRGPUToSPIRVIncGen)
-
 target_link_libraries(MLIRGPUtoSPIRVTransforms
   MLIRGPU
   MLIRIR

mlir/lib/Conversion/GPUToSPIRV/ConvertGPUToSPIRV.cpp

@@ -63,13 +63,27 @@ class KernelFnConversion final : public SPIRVOpLowering<gpu::GPUFuncOp> {
   SmallVector<int32_t, 3> workGroupSizeAsInt32;
 };
 
-/// Pattern to convert a gpu.module to a spv.module.
-class GPUModuleConversion final : public SPIRVOpLowering<gpu::GPUModuleOp> {
+/// Pattern to convert a module with gpu.kernel_module attribute to a
+/// spv.module.
+class KernelModuleConversion final : public SPIRVOpLowering<ModuleOp> {
 public:
-  using SPIRVOpLowering<gpu::GPUModuleOp>::SPIRVOpLowering;
+  using SPIRVOpLowering<ModuleOp>::SPIRVOpLowering;
 
   PatternMatchResult
-  matchAndRewrite(gpu::GPUModuleOp moduleOp, ArrayRef<Value> operands,
+  matchAndRewrite(ModuleOp moduleOp, ArrayRef<Value> operands,
+                  ConversionPatternRewriter &rewriter) const override;
+};
+
+/// Pattern to convert a module terminator op to a terminator of spv.module op.
+// TODO: Move this into DRR, but that requires ModuleTerminatorOp to be defined
+// in ODS.
+class KernelModuleTerminatorConversion final
+    : public SPIRVOpLowering<ModuleTerminatorOp> {
+public:
+  using SPIRVOpLowering<ModuleTerminatorOp>::SPIRVOpLowering;
+
+  PatternMatchResult
+  matchAndRewrite(ModuleTerminatorOp terminatorOp, ArrayRef<Value> operands,
                   ConversionPatternRewriter &rewriter) const override;
 };
 
@@ -270,12 +284,16 @@ KernelFnConversion::matchAndRewrite(gpu::GPUFuncOp funcOp,
 }
 
 //===----------------------------------------------------------------------===//
-// ModuleOp with gpu.module.
+// ModuleOp with gpu.kernel_module.
 //===----------------------------------------------------------------------===//
 
-PatternMatchResult GPUModuleConversion::matchAndRewrite(
-    gpu::GPUModuleOp moduleOp, ArrayRef<Value> operands,
+PatternMatchResult KernelModuleConversion::matchAndRewrite(
+    ModuleOp moduleOp, ArrayRef<Value> operands,
     ConversionPatternRewriter &rewriter) const {
+  if (!moduleOp.getAttrOfType<UnitAttr>(
+          gpu::GPUDialect::getKernelModuleAttrName())) {
+    return matchFailure();
+  }
   // TODO : Generalize this to account for different extensions,
   // capabilities, extended_instruction_sets, other addressing models
   // and memory models.
@@ -284,8 +302,8 @@ PatternMatchResult GPUModuleConversion::matchAndRewrite(
       spirv::MemoryModel::GLSL450, spirv::Capability::Shader,
       spirv::Extension::SPV_KHR_storage_buffer_storage_class);
   // Move the region from the module op into the SPIR-V module.
-  Region &spvModuleRegion = spvModule.body();
-  rewriter.inlineRegionBefore(moduleOp.body(), spvModuleRegion,
+  Region &spvModuleRegion = spvModule.getOperation()->getRegion(0);
+  rewriter.inlineRegionBefore(moduleOp.getBodyRegion(), spvModuleRegion,
                               spvModuleRegion.begin());
   // The spv.module build method adds a block with a terminator. Remove that
   // block. The terminator of the module op in the remaining block will be
@@ -295,6 +313,17 @@ PatternMatchResult GPUModuleConversion::matchAndRewrite(
   return matchSuccess();
 }
 
+//===----------------------------------------------------------------------===//
+// ModuleTerminatorOp for gpu.kernel_module.
+//===----------------------------------------------------------------------===//
+
+PatternMatchResult KernelModuleTerminatorConversion::matchAndRewrite(
+    ModuleTerminatorOp terminatorOp, ArrayRef<Value> operands,
+    ConversionPatternRewriter &rewriter) const {
+  rewriter.replaceOpWithNewOp<spirv::ModuleEndOp>(terminatorOp);
+  return matchSuccess();
+}
+
 //===----------------------------------------------------------------------===//
 // GPU return inside kernel functions to SPIR-V return.
 //===----------------------------------------------------------------------===//
@@ -313,18 +342,14 @@ PatternMatchResult GPUReturnOpConversion::matchAndRewrite(
 // GPU To SPIRV Patterns.
 //===----------------------------------------------------------------------===//
 
-namespace {
-#include "GPUToSPIRV.cpp.inc"
-}
-
 void mlir::populateGPUToSPIRVPatterns(MLIRContext *context,
                                       SPIRVTypeConverter &typeConverter,
                                       OwningRewritePatternList &patterns,
                                       ArrayRef<int64_t> workGroupSize) {
-  populateWithGenerated(context, &patterns);
   patterns.insert<KernelFnConversion>(context, typeConverter, workGroupSize);
   patterns.insert<
-      GPUReturnOpConversion, ForOpConversion, GPUModuleConversion,
+      GPUReturnOpConversion, ForOpConversion, KernelModuleConversion,
+      KernelModuleTerminatorConversion,
       LaunchConfigConversion<gpu::BlockDimOp, spirv::BuiltIn::WorkgroupSize>,
       LaunchConfigConversion<gpu::BlockIdOp, spirv::BuiltIn::WorkgroupId>,
       LaunchConfigConversion<gpu::GridDimOp, spirv::BuiltIn::NumWorkgroups>,