Merge branch 'main' into fix-dxil-container-unbounded-array

Author: hekota

.github/workflows/containers/github-action-ci/Dockerfile

@@ -2,7 +2,7 @@ FROM docker.io/library/ubuntu:24.04 AS base
 ENV LLVM_SYSROOT=/opt/llvm
 
 FROM base AS stage1-toolchain
-ENV LLVM_VERSION=21.1.1
+ENV LLVM_VERSION=21.1.3
 
 RUN apt-get update && \
     apt-get install -y \

clang/docs/OpenMPSupport.rst

@@ -505,7 +505,7 @@ implementation.
 +-------------------------------------------------------------+---------------------------+---------------------------+--------------------------------------------------------------------------+
 | pure directives in DO CONCURRENT                            |                           | :none:`unclaimed`         |                                                                          |
 +-------------------------------------------------------------+---------------------------+---------------------------+--------------------------------------------------------------------------+
-| Optional argument for all clauses                           | :none:`unclaimed`         | :none:`unclaimed`         |                                                                          |
+| Optional argument for all clauses                           | :none:`partial`           | :none:`In Progress`       | Parse/Sema (nowait): https://github.com/llvm/llvm-project/pull/159628    |
 +-------------------------------------------------------------+---------------------------+---------------------------+--------------------------------------------------------------------------+
 | Function references for locator list items                  | :none:`unclaimed`         | :none:`unclaimed`         |                                                                          |
 +-------------------------------------------------------------+---------------------------+---------------------------+--------------------------------------------------------------------------+

clang/docs/ReleaseNotes.rst

@@ -641,6 +641,8 @@ OpenMP Support
 - Added support for ``defaultmap`` directive implicit-behavior ``private``.
 - Added parsing and semantic analysis support for ``groupprivate`` directive.
 - Added support for 'omp fuse' directive.
+- Updated parsing and semantic analysis support for ``nowait`` clause to accept
+  optional argument in OpenMP >= 60.
 
 Improvements
 ^^^^^^^^^^^^

clang/docs/UsersManual.rst

@@ -2780,6 +2780,25 @@ usual build cycle when using sample profilers for optimization:
        /clang:-fdebug-info-for-profiling /clang:-funique-internal-linkage-names ^
        code.cc /Fe:code -fuse-ld=lld /link /debug:dwarf
 
+   [OPTIONAL] Pseudo instrumentation can be used as the anchor for accurate
+   profile mapping with the ``-fpseudo-probe-for-profiling`` option.
+
+   On Linux:
+
+   .. code-block:: console
+
+     $ clang++ -O2 -gline-tables-only \
+       -fpseudo-probe-for-profiling -funique-internal-linkage-names \
+       code.cc -o code
+
+   On Windows:
+
+   .. code-block:: winbatch
+
+     > clang-cl /O2 -gdwarf -gline-tables-only ^
+       -fpseudo-probe-for-profiling /clang:-funique-internal-linkage-names ^
+       code.cc /Fe:code -fuse-ld=lld /link /debug:dwarf
+
 .. note::
 
    :ref:`-funique-internal-linkage-names <funique_internal_linkage_names>`
@@ -2881,6 +2900,25 @@ usual build cycle when using sample profilers for optimization:
        /clang:-fdebug-info-for-profiling /clang:-funique-internal-linkage-names ^
        -fprofile-sample-use=code.prof code.cc /Fe:code
 
+   [OPTIONAL] Pseudo instrumentation can be used as the anchor for accurate
+   profile mapping with the ``-fpseudo-probe-for-profiling`` option.
+
+   On Linux:
+
+   .. code-block:: console
+
+     $ clang++ -O2 \
+        -fpseudo-probe-for-profiling -funique-internal-linkage-names \
+       -fprofile-sample-use=code.prof code.cc -o code
+
+   On Windows:
+
+   .. code-block:: winbatch
+
+     > clang-cl /O2 ^
+       -fpseudo-probe-for-profiling /clang:-funique-internal-linkage-names ^
+       -fprofile-sample-use=code.prof code.cc /Fe:code
+
    [OPTIONAL] Sampling-based profiles can have inaccuracies or missing block/
    edge counters. The profile inference algorithm (profi) can be used to infer
    missing blocks and edge counts, and improve the quality of profile data.

clang/include/clang/AST/OpenMPClause.h

@@ -2291,18 +2291,68 @@ class OMPOrderedClause final
 /// This represents 'nowait' clause in the '#pragma omp ...' directive.
 ///
 /// \code
-/// #pragma omp for nowait
+/// #pragma omp for nowait (cond)
 /// \endcode
-/// In this example directive '#pragma omp for' has 'nowait' clause.
-class OMPNowaitClause final : public OMPNoChildClause<llvm::omp::OMPC_nowait> {
+/// In this example directive '#pragma omp for' has simple 'nowait' clause with
+/// condition 'cond'.
+class OMPNowaitClause final : public OMPClause {
+  friend class OMPClauseReader;
+
+  /// Location of '('.
+  SourceLocation LParenLoc;
+
+  /// Condition of the 'nowait' clause.
+  Stmt *Condition = nullptr;
+
+  /// Set condition.
+  void setCondition(Expr *Cond) { Condition = Cond; }
+
 public:
-  /// Build 'nowait' clause.
+  /// Build 'nowait' clause with condition \a Cond.
   ///
+  /// \param Cond Condition of the clause.
   /// \param StartLoc Starting location of the clause.
+  /// \param LParenLoc Location of '('.
   /// \param EndLoc Ending location of the clause.
-  OMPNowaitClause(SourceLocation StartLoc = SourceLocation(),
-                  SourceLocation EndLoc = SourceLocation())
-      : OMPNoChildClause(StartLoc, EndLoc) {}
+  OMPNowaitClause(Expr *Cond, SourceLocation StartLoc, SourceLocation LParenLoc,
+                  SourceLocation EndLoc)
+      : OMPClause(llvm::omp::OMPC_nowait, StartLoc, EndLoc),
+        LParenLoc(LParenLoc), Condition(Cond) {}
+
+  /// Build an empty clause.
+  OMPNowaitClause()
+      : OMPClause(llvm::omp::OMPC_nowait, SourceLocation(), SourceLocation()) {}
+
+  /// Sets the location of '('.
+  void setLParenLoc(SourceLocation Loc) { LParenLoc = Loc; }
+
+  /// Returns the location of '('.
+  SourceLocation getLParenLoc() const { return LParenLoc; }
+
+  /// Returns condition.
+  Expr *getCondition() const { return cast_or_null<Expr>(Condition); }
+
+  child_range children() {
+    if (Condition)
+      return child_range(&Condition, &Condition + 1);
+    return child_range(child_iterator(), child_iterator());
+  }
+
+  const_child_range children() const {
+    if (Condition)
+      return const_child_range(&Condition, &Condition + 1);
+    return const_child_range(const_child_iterator(), const_child_iterator());
+  }
+
+  child_range used_children();
+  const_child_range used_children() const {
+    auto Children = const_cast<OMPNowaitClause *>(this)->used_children();
+    return const_child_range(Children.begin(), Children.end());
+  }
+
+  static bool classof(const OMPClause *T) {
+    return T->getClauseKind() == llvm::omp::OMPC_nowait;
+  }
 };
 
 /// This represents 'untied' clause in the '#pragma omp ...' directive.

clang/include/clang/AST/RecursiveASTVisitor.h

@@ -3594,7 +3594,8 @@ bool RecursiveASTVisitor<Derived>::VisitOMPOrderedClause(OMPOrderedClause *C) {
 }
 
 template <typename Derived>
-bool RecursiveASTVisitor<Derived>::VisitOMPNowaitClause(OMPNowaitClause *) {
+bool RecursiveASTVisitor<Derived>::VisitOMPNowaitClause(OMPNowaitClause *C) {
+  TRY_TO(TraverseStmt(C->getCondition()));
   return true;
 }
 

clang/include/clang/CIR/Dialect/IR/CIRAttrs.td

@@ -814,6 +814,14 @@ def CIR_GlobalCtorAttr : CIR_GlobalCtorDtor<"Ctor", "ctor"> {
   }];
 }
 
+def CIR_GlobalDtorAttr : CIR_GlobalCtorDtor<"Dtor", "dtor"> {
+  let summary = "Marks a function as a global destructor";
+  let description = [{
+      Marks a function as a global destructor in the module dtors list.
+      The function will be executed before the module unloading.
+  }];
+}
+
 //===----------------------------------------------------------------------===//
 // BitfieldInfoAttr
 //===----------------------------------------------------------------------===//

clang/include/clang/CIR/Dialect/IR/CIRDialect.td

@@ -43,6 +43,7 @@ def CIR_Dialect : Dialect {
     static llvm::StringRef getSideEffectAttrName() { return "side_effect"; }
     static llvm::StringRef getModuleLevelAsmAttrName() { return "cir.module_asm"; }
     static llvm::StringRef getGlobalCtorsAttrName() { return "cir.global_ctors"; }
+    static llvm::StringRef getGlobalDtorsAttrName() { return "cir.global_dtors"; }
 
     void registerAttributes();
     void registerTypes();

clang/include/clang/CIR/Dialect/IR/CIROps.td

@@ -88,6 +88,19 @@ class CIR_Op<string mnemonic, list<Trait> traits = []> :
   code extraLLVMLoweringPatternDecl = "";
 }
 
+//===----------------------------------------------------------------------===//
+// CIR Operation Traits
+//===----------------------------------------------------------------------===//
+
+class HasAtMostOneOfAttrsPred<list<string> names> :
+  CPred<!foldl("0", names, acc, name,  acc # " + (" # name # " ? 1 : 0)")
+        # " <= 1">;
+
+class HasAtMostOneOfAttrs<list<string> names> : PredOpTrait<
+  "has only one of the optional attributes: " # !interleave(names, ", "),
+  HasAtMostOneOfAttrsPred<!foreach(name, names, "$" # name)>
+>;
+
 //===----------------------------------------------------------------------===//
 // CastOp
 //===----------------------------------------------------------------------===//
@@ -2422,9 +2435,17 @@ def CIR_GetMemberOp : CIR_Op<"get_member"> {
 // TODO(CIR): FuncOp is still a tiny shell of what it will become.  Many more
 // properties and attributes will be added as upstreaming continues.
 
+def CIR_OptionalPriorityAttr : OptionalAttr<
+  DefaultValuedAttr<
+    ConfinedAttr<I32Attr, [IntMinValue<101>, IntMaxValue<65535>]>,
+    "65535"
+  >
+>;
+
 def CIR_FuncOp : CIR_Op<"func", [
   AutomaticAllocationScope, CallableOpInterface, FunctionOpInterface,
   DeclareOpInterfaceMethods<CIRGlobalValueInterface>,
+  HasAtMostOneOfAttrs<["global_ctor_priority", "global_dtor_priority"]>,
   IsolatedFromAbove
 ]> {
   let summary = "Declare or define a function";
@@ -2449,6 +2470,12 @@ def CIR_FuncOp : CIR_Op<"func", [
     without a prototype and, consequently, may contain calls with invalid
     arguments and undefined behavior.
 
+    The `global_ctor` keyword indicates whether a function should execute before
+    `main()` function, as specified by `__attribute__((constructor))`. An
+    execution priority can also be specified `global_ctor(<priority>)`.
+    Similarly, for global destructors both `global_dtor` and
+    `global_dtor(<priority>)` are available.
+
     Example:
 
     ```mlir
@@ -2487,7 +2514,9 @@ def CIR_FuncOp : CIR_Op<"func", [
                        UnitAttr:$comdat,
                        OptionalAttr<DictArrayAttr>:$arg_attrs,
                        OptionalAttr<DictArrayAttr>:$res_attrs,
-                       OptionalAttr<FlatSymbolRefAttr>:$aliasee);
+                       OptionalAttr<FlatSymbolRefAttr>:$aliasee,
+                       CIR_OptionalPriorityAttr:$global_ctor_priority,
+                       CIR_OptionalPriorityAttr:$global_dtor_priority);
 
   let regions = (region AnyRegion:$body);
 

clang/include/clang/CIR/MissingFeatures.h

@@ -39,7 +39,6 @@ struct MissingFeatures {
   static bool opGlobalUsedOrCompilerUsed() { return false; }
   static bool opGlobalAnnotations() { return false; }
   static bool opGlobalCtorPriority() { return false; }
-  static bool opGlobalDtorList() { return false; }
   static bool setDSOLocal() { return false; }
   static bool setComdat() { return false; }
 
@@ -175,6 +174,10 @@ struct MissingFeatures {
   static bool atomicScope() { return false; }
   static bool atomicSyncScopeID() { return false; }
 
+  // Global ctor handling
+  static bool globalCtorLexOrder() { return false; }
+  static bool globalCtorAssociatedData() { return false; }
+
   // Misc
   static bool abiArgInfo() { return false; }
   static bool addHeapAllocSiteMetadata() { return false; }