[OpenMP] Add OpenMP v6.0 API Routines omp_target_memset() and omp_target_memset_sync()

[mjklemm]

This PR adds the newly landed OpenMP API routines omp_target_memset() and omp_target_memset_sync() to fill memory in the device memory. This first implementation is based on a slow path that initializes memory on the host and then issues a H2D transfer to fill the memory on the target device. A better solution is being investigate at the moment.

[github-actions]

⚠️ C/C++ code formatter, clang-format found issues in your code. ⚠️

You can test this locally with the following command:

git-clang-format --diff 816c42d880d9d146c5d7cd6cd48a99324a8e2edf 012952ada089b623a7d361a00f74f787481ee9b0 -- clang/test/Layout/ms-no-unique-address.cpp clang/test/SemaCXX/cxx2a-ms-no-unique-address.cpp libc/src/math/expm1.h libc/src/math/generic/expm1.cpp libc/test/src/math/expm1_test.cpp libc/test/src/math/smoke/expm1_test.cpp llvm/include/llvm/ExecutionEngine/Orc/Debugging/DebugInfoSupport.h llvm/lib/ExecutionEngine/Orc/Debugging/DebugInfoSupport.cpp openmp/libomptarget/test/api/omp_target_memset.c clang/include/clang/Basic/ParsedAttrInfo.h clang/include/clang/Basic/SourceManager.h clang/include/clang/Lex/HeaderSearch.h clang/include/clang/Lex/ModuleMap.h clang/include/clang/Sema/Sema.h clang/lib/AST/ASTContext.cpp clang/lib/AST/Decl.cpp clang/lib/AST/Interp/Boolean.h clang/lib/AST/Interp/ByteCodeExprGen.cpp clang/lib/AST/Interp/Floating.h clang/lib/AST/Interp/Integral.h clang/lib/AST/Interp/Interp.h clang/lib/AST/Interp/InterpBuiltin.cpp clang/lib/AST/Interp/Pointer.h clang/lib/AST/RecordLayoutBuilder.cpp clang/lib/Analysis/ThreadSafety.cpp clang/lib/Basic/SourceManager.cpp clang/lib/Basic/Targets/NVPTX.h clang/lib/CodeGen/Targets/NVPTX.cpp clang/lib/Driver/ToolChains/CommonArgs.cpp clang/lib/Driver/ToolChains/VEToolchain.cpp clang/lib/Format/WhitespaceManager.cpp clang/lib/Frontend/FrontendAction.cpp clang/lib/Interpreter/IncrementalExecutor.cpp clang/lib/Lex/HeaderSearch.cpp clang/lib/Lex/ModuleMap.cpp clang/lib/Lex/PPDirectives.cpp clang/lib/Parse/ParseDeclCXX.cpp clang/lib/Parse/ParseOpenMP.cpp clang/lib/Sema/ParsedAttr.cpp clang/lib/Sema/SemaDeclAttr.cpp clang/lib/Sema/SemaTemplateInstantiateDecl.cpp clang/lib/Serialization/ASTReader.cpp clang/lib/Serialization/ASTWriter.cpp clang/test/AST/Interp/arrays.cpp clang/test/AST/Interp/cxx20.cpp clang/test/Driver/ve-toolchain.c clang/test/Driver/ve-toolchain.cpp clang/test/Preprocessor/has_attribute.cpp clang/test/SemaCXX/sugar-common-types.cpp clang/unittests/Format/FormatTest.cpp clang/utils/TableGen/ClangAttrEmitter.cpp compiler-rt/lib/scudo/standalone/chunk.h compiler-rt/lib/scudo/standalone/combined.h compiler-rt/lib/scudo/standalone/report.cpp compiler-rt/lib/scudo/standalone/report.h compiler-rt/lib/scudo/standalone/tests/chunk_test.cpp compiler-rt/lib/scudo/standalone/tests/report_test.cpp flang/include/flang/Runtime/descriptor.h flang/lib/Lower/ConvertVariable.cpp flang/lib/Lower/OpenACC.cpp flang/lib/Optimizer/CodeGen/CodeGen.cpp flang/runtime/descriptor.cpp libc/src/__support/FPUtil/FPBits.h libc/src/__support/FPUtil/except_value_utils.h libc/src/math/generic/exp.cpp libc/src/math/generic/expm1f.cpp libc/src/math/generic/log1pf.cpp llvm/lib/Analysis/InstructionSimplify.cpp llvm/lib/Analysis/ScalarEvolution.cpp llvm/lib/CodeGen/AsmPrinter/DwarfDebug.h llvm/lib/CodeGen/MachineSink.cpp llvm/lib/Target/AArch64/AArch64ISelLowering.cpp llvm/lib/Target/AMDGPU/AMDGPUAsmPrinter.cpp llvm/lib/Target/AMDGPU/AMDGPUAsmPrinter.h llvm/lib/Target/AMDGPU/SIInstrInfo.h llvm/lib/Target/AMDGPU/SILowerSGPRSpills.cpp llvm/lib/Target/NVPTX/NVPTXAsmPrinter.cpp llvm/lib/Target/NVPTX/NVPTXUtilities.cpp llvm/lib/Target/NVPTX/NVPTXUtilities.h llvm/lib/Transforms/IPO/AttributorAttributes.cpp llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp llvm/lib/Transforms/InstCombine/InstCombineShifts.cpp llvm/lib/Transforms/Utils/InlineFunction.cpp llvm/lib/Transforms/Utils/ScalarEvolutionExpander.cpp llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp llvm/tools/lli/lli.cpp llvm/tools/llvm-jitlink/llvm-jitlink.cpp mlir/include/mlir/Dialect/Affine/IR/AffineOps.h mlir/lib/Conversion/AffineToStandard/AffineToStandard.cpp mlir/lib/Conversion/SCFToGPU/SCFToGPU.cpp mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVM.cpp mlir/lib/Dialect/Affine/Analysis/AffineStructures.cpp mlir/lib/Dialect/Affine/Analysis/LoopAnalysis.cpp mlir/lib/Dialect/Affine/IR/AffineOps.cpp mlir/lib/Dialect/Affine/Transforms/PipelineDataTransfer.cpp mlir/lib/Dialect/Affine/Transforms/SuperVectorize.cpp mlir/lib/Dialect/Affine/Utils/LoopUtils.cpp mlir/lib/Dialect/Affine/Utils/Utils.cpp mlir/lib/Target/LLVMIR/DebugImporter.cpp openmp/libomptarget/include/omptarget.h openmp/libomptarget/src/api.cpp openmp/libomptarget/src/private.h openmp/runtime/src/kmp_ftn_os.h llvm/include/llvm/ExecutionEngine/Orc/Debugging/DebuggerSupport.h llvm/include/llvm/ExecutionEngine/Orc/Debugging/DebuggerSupportPlugin.h llvm/include/llvm/ExecutionEngine/Orc/Debugging/PerfSupportPlugin.h llvm/lib/ExecutionEngine/Orc/Debugging/DebuggerSupport.cpp llvm/lib/ExecutionEngine/Orc/Debugging/DebuggerSupportPlugin.cpp llvm/lib/ExecutionEngine/Orc/Debugging/PerfSupportPlugin.cpp

View the diff from clang-format here.

diff --git a/clang/include/clang/Basic/SourceManager.h b/clang/include/clang/Basic/SourceManager.h
index 431f97f30525..98d194b1f8b0 100644
--- a/clang/include/clang/Basic/SourceManager.h
+++ b/clang/include/clang/Basic/SourceManager.h
@@ -649,7 +649,7 @@ class SourceManager : public RefCountedBase<SourceManager> {
   /// This map allows us to merge ContentCache entries based
   /// on their FileEntry*.  All ContentCache objects will thus have unique,
   /// non-null, FileEntry pointers.
-  llvm::DenseMap<const FileEntry*, SrcMgr::ContentCache*> FileInfos;
+  llvm::DenseMap<const FileEntry *, SrcMgr::ContentCache *> FileInfos;
 
   /// True if the ContentCache for files that are overridden by other
   /// files, should report the original file name. Defaults to true.
@@ -1680,7 +1680,7 @@ public:
 
   // Iterators over FileInfos.
   using fileinfo_iterator =
-      llvm::DenseMap<const FileEntry*, SrcMgr::ContentCache*>::const_iterator;
+      llvm::DenseMap<const FileEntry *, SrcMgr::ContentCache *>::const_iterator;
 
   fileinfo_iterator fileinfo_begin() const { return FileInfos.begin(); }
   fileinfo_iterator fileinfo_end() const { return FileInfos.end(); }
diff --git a/clang/include/clang/Sema/Sema.h b/clang/include/clang/Sema/Sema.h
index e13524b5f3b3..817c823af3df 100644
--- a/clang/include/clang/Sema/Sema.h
+++ b/clang/include/clang/Sema/Sema.h
@@ -2183,14 +2183,11 @@ public:
       const FunctionProtoType *Old, SourceLocation OldLoc,
       const FunctionProtoType *New, SourceLocation NewLoc);
   bool handlerCanCatch(QualType HandlerType, QualType ExceptionType);
-  bool CheckExceptionSpecSubset(const PartialDiagnostic &DiagID,
-                                const PartialDiagnostic &NestedDiagID,
-                                const PartialDiagnostic &NoteID,
-                                const PartialDiagnostic &NoThrowDiagID,
-                                const FunctionProtoType *Superset,
-                                SourceLocation SuperLoc,
-                                const FunctionProtoType *Subset,
-                                SourceLocation SubLoc);
+  bool CheckExceptionSpecSubset(
+      const PartialDiagnostic &DiagID, const PartialDiagnostic &NestedDiagID,
+      const PartialDiagnostic &NoteID, const PartialDiagnostic &NoThrowDiagID,
+      const FunctionProtoType *Superset, SourceLocation SuperLoc,
+      const FunctionProtoType *Subset, SourceLocation SubLoc);
   bool CheckParamExceptionSpec(const PartialDiagnostic &NestedDiagID,
                                const PartialDiagnostic &NoteID,
                                const FunctionProtoType *Target,
@@ -3855,7 +3852,7 @@ public:
   bool isObjCWritebackConversion(QualType FromType, QualType ToType,
                                  QualType &ConvertedType);
   bool IsBlockPointerConversion(QualType FromType, QualType ToType,
-                                QualType& ConvertedType);
+                                QualType &ConvertedType);
   bool FunctionParamTypesAreEqual(const FunctionProtoType *OldType,
                                   const FunctionProtoType *NewType,
                                   unsigned *ArgPos = nullptr,
@@ -4221,8 +4218,7 @@ public:
       QualType DestTypeForComplaining = QualType(),
       unsigned DiagIDForComplaining = 0);
 
-  Expr *FixOverloadedFunctionReference(Expr *E,
-                                       DeclAccessPair FoundDecl,
+  Expr *FixOverloadedFunctionReference(Expr *E, DeclAccessPair FoundDecl,
                                        FunctionDecl *Fn);
   ExprResult FixOverloadedFunctionReference(ExprResult,
                                             DeclAccessPair FoundDecl,
@@ -9222,8 +9218,7 @@ public:
 
   TemplateDeductionResult
   DeduceTemplateArguments(FunctionTemplateDecl *FunctionTemplate,
-                          QualType ToType,
-                          CXXConversionDecl *&Specialization,
+                          QualType ToType, CXXConversionDecl *&Specialization,
                           sema::TemplateDeductionInfo &Info);
 
   TemplateDeductionResult
diff --git a/clang/lib/Basic/SourceManager.cpp b/clang/lib/Basic/SourceManager.cpp
index a630743e0d73..3ceff54c205c 100644
--- a/clang/lib/Basic/SourceManager.cpp
+++ b/clang/lib/Basic/SourceManager.cpp
@@ -324,8 +324,10 @@ SourceManager::~SourceManager() {
       ContentCacheAlloc.Deallocate(MemBufferInfos[i]);
     }
   }
-  for (llvm::DenseMap<const FileEntry*, SrcMgr::ContentCache*>::iterator
-       I = FileInfos.begin(), E = FileInfos.end(); I != E; ++I) {
+  for (llvm::DenseMap<const FileEntry *, SrcMgr::ContentCache *>::iterator
+           I = FileInfos.begin(),
+           E = FileInfos.end();
+       I != E; ++I) {
     if (I->second) {
       I->second->~ContentCache();
       ContentCacheAlloc.Deallocate(I->second);
@@ -2344,11 +2346,11 @@ SourceManager::MemoryBufferSizes SourceManager::getMemoryBufferSizes() const {
 }
 
 size_t SourceManager::getDataStructureSizes() const {
-  size_t size = llvm::capacity_in_bytes(MemBufferInfos)
-    + llvm::capacity_in_bytes(LocalSLocEntryTable)
-    + llvm::capacity_in_bytes(LoadedSLocEntryTable)
-    + llvm::capacity_in_bytes(SLocEntryLoaded)
-    + llvm::capacity_in_bytes(FileInfos);
+  size_t size = llvm::capacity_in_bytes(MemBufferInfos) +
+                llvm::capacity_in_bytes(LocalSLocEntryTable) +
+                llvm::capacity_in_bytes(LoadedSLocEntryTable) +
+                llvm::capacity_in_bytes(SLocEntryLoaded) +
+                llvm::capacity_in_bytes(FileInfos);
 
   if (OverriddenFilesInfo)
     size += llvm::capacity_in_bytes(OverriddenFilesInfo->OverriddenFiles);
diff --git a/clang/lib/Lex/HeaderSearch.cpp b/clang/lib/Lex/HeaderSearch.cpp
index e54a19ebfdbb..a61f2ea4dca4 100644
--- a/clang/lib/Lex/HeaderSearch.cpp
+++ b/clang/lib/Lex/HeaderSearch.cpp
@@ -927,8 +927,8 @@ OptionalFileEntryRef HeaderSearch::LookupFile(
       // from a module build. We should treat this as a system header if we're
       // building a [system] module.
       bool IncluderIsSystemHeader =
-          Includer ? getFileInfo(Includer).DirInfo != SrcMgr::C_User :
-          BuildSystemModule;
+          Includer ? getFileInfo(Includer).DirInfo != SrcMgr::C_User
+                   : BuildSystemModule;
       if (OptionalFileEntryRef FE = getFileAndSuggestModule(
               TmpDir, IncludeLoc, IncluderAndDir.second, IncluderIsSystemHeader,
               RequestingModule, SuggestedModule)) {
diff --git a/clang/lib/Sema/SemaDeclAttr.cpp b/clang/lib/Sema/SemaDeclAttr.cpp
index 2d7e69946a39..ddcc290651d3 100644
--- a/clang/lib/Sema/SemaDeclAttr.cpp
+++ b/clang/lib/Sema/SemaDeclAttr.cpp
@@ -1336,9 +1336,9 @@ static void handleReturnTypestateAttr(Sema &S, Decl *D, const ParsedAttr &AL) {
   // FIXME: This check is currently being done in the analysis.  It can be
   //        enabled here only after the parser propagates attributes at
   //        template specialization definition, not declaration.
-  //QualType ReturnType;
+  // QualType ReturnType;
   //
-  //if (const ParmVarDecl *Param = dyn_cast<ParmVarDecl>(D)) {
+  // if (const ParmVarDecl *Param = dyn_cast<ParmVarDecl>(D)) {
   //  ReturnType = Param->getType();
   //
   //} else if (const CXXConstructorDecl *Constructor =
@@ -1350,9 +1350,9 @@ static void handleReturnTypestateAttr(Sema &S, Decl *D, const ParsedAttr &AL) {
   //  ReturnType = cast<FunctionDecl>(D)->getCallResultType();
   //}
   //
-  //const CXXRecordDecl *RD = ReturnType->getAsCXXRecordDecl();
+  // const CXXRecordDecl *RD = ReturnType->getAsCXXRecordDecl();
   //
-  //if (!RD || !RD->hasAttr<ConsumableAttr>()) {
+  // if (!RD || !RD->hasAttr<ConsumableAttr>()) {
   //    S.Diag(Attr.getLoc(), diag::warn_return_state_for_unconsumable_type) <<
   //      ReturnType.getAsString();
   //    return;
diff --git a/llvm/lib/ExecutionEngine/Orc/Debugging/DebuggerSupportPlugin.cpp b/llvm/lib/ExecutionEngine/Orc/Debugging/DebuggerSupportPlugin.cpp
index 236ba5114130..4ac1b8b5a72b 100644
--- a/llvm/lib/ExecutionEngine/Orc/Debugging/DebuggerSupportPlugin.cpp
+++ b/llvm/lib/ExecutionEngine/Orc/Debugging/DebuggerSupportPlugin.cpp
@@ -169,15 +169,13 @@ public:
       // Try to parse line data. Consume error on failure.
       if (auto Err = LineTable.parse(DebugLineData, &Offset, *DWARFCtx, nullptr,
                                      consumeError)) {
-        handleAllErrors(
-          std::move(Err),
-          [&](ErrorInfoBase &EIB) {
-            LLVM_DEBUG({
-              dbgs() << "Cannot parse line table for \"" << G.getName() << "\": ";
-              EIB.log(dbgs());
-              dbgs() << "\n";
-            });
+        handleAllErrors(std::move(Err), [&](ErrorInfoBase &EIB) {
+          LLVM_DEBUG({
+            dbgs() << "Cannot parse line table for \"" << G.getName() << "\": ";
+            EIB.log(dbgs());
+            dbgs() << "\n";
           });
+        });
       } else {
         if (!LineTable.Prologue.FileNames.empty())
           FileName = *dwarf::toString(LineTable.Prologue.FileNames[0].Name);

[gregrodgers]

If this was or will be landed upstream, we should be able to pick it up with our merge_from_main.sh process that I just completed. I did not see these merged. However, dry-run of this patch showed some conflict. Run get_pr_patches to see conflicts.

Let me know if this is expected to be committed upstream soon.

[mjklemm]

I have checked the conflicts and it happens in dllexports, where there's an ASO change that causes the conflict. I can easily create a workaround patch for that at some point.

[mjklemm]

Merged via trunk.