merge main into amd-staging

Change-Id: I7fa3fdb040e429c77e71bb919c73e08ff002e3ac

clang-tools-extra/clang-tidy/modernize/UseStdPrintCheck.cpp

@@ -23,7 +23,7 @@ AST_MATCHER(StringLiteral, isOrdinary) { return Node.isOrdinary(); }
 } // namespace
 
 UseStdPrintCheck::UseStdPrintCheck(StringRef Name, ClangTidyContext *Context)
-    : ClangTidyCheck(Name, Context),
+    : ClangTidyCheck(Name, Context), PP(nullptr),
       StrictMode(Options.getLocalOrGlobal("StrictMode", false)),
       PrintfLikeFunctions(utils::options::parseStringList(
           Options.get("PrintfLikeFunctions", ""))),
@@ -131,6 +131,7 @@ void UseStdPrintCheck::check(const MatchFinder::MatchResult &Result) {
   utils::FormatStringConverter::Configuration ConverterConfig;
   ConverterConfig.StrictMode = StrictMode;
   ConverterConfig.AllowTrailingNewlineRemoval = true;
+  assert(PP && "Preprocessor should be set by registerPPCallbacks");
   utils::FormatStringConverter Converter(
       Result.Context, Printf, FormatArgOffset, ConverterConfig, getLangOpts(),
       *Result.SourceManager, *PP);

clang/cmake/caches/Fuchsia-stage2.cmake

@@ -332,7 +332,7 @@ foreach(target armv6m-none-eabi;armv7m-none-eabi;armv8m.main-none-eabi;armv8.1m.
   foreach(lang C;CXX;ASM)
     # TODO: The preprocessor defines workaround various issues in libc and libc++ integration.
     # These should be addressed and removed over time.
-    set(RUNTIMES_${target}_CMAKE_${lang}_local_flags "--target=${target} -mthumb -Wno-atomic-alignment \"-Dvfprintf(stream, format, vlist)=vprintf(format, vlist)\" \"-Dfprintf(stream, format, ...)=printf(format)\" \"-Dtimeval=struct timeval{int tv_sec; int tv_usec;}\" \"-Dgettimeofday(tv, tz)\" -D_LIBCPP_PRINT=1")
+    set(RUNTIMES_${target}_CMAKE_${lang}_local_flags "--target=${target} -mthumb -Wno-atomic-alignment \"-Dvfprintf(stream, format, vlist)=vprintf(format, vlist)\" \"-Dfprintf(stream, format, ...)=printf(format)\" \"-Dgettimeofday(tv, tz)\" -D_LIBCPP_PRINT=1")
     if(${target} STREQUAL "armv8m.main-none-eabi")
       set(RUNTIMES_${target}_CMAKE_${lang}_local_flags "${RUNTIMES_${target}_CMAKE_${lang}_local_flags} -mfloat-abi=softfp -march=armv8m.main+fp+dsp -mcpu=cortex-m33" CACHE STRING "")
     endif()
@@ -346,7 +346,6 @@ foreach(target armv6m-none-eabi;armv7m-none-eabi;armv8m.main-none-eabi;armv8.1m.
   endforeach()
   set(RUNTIMES_${target}_LLVM_LIBC_FULL_BUILD ON CACHE BOOL "")
   set(RUNTIMES_${target}_LIBC_ENABLE_USE_BY_CLANG ON CACHE BOOL "")
-  set(RUNTIMES_${target}_LIBC_USE_NEW_HEADER_GEN OFF CACHE BOOL "")
   set(RUNTIMES_${target}_LIBCXX_ABI_VERSION 2 CACHE STRING "")
   set(RUNTIMES_${target}_LIBCXX_CXX_ABI none CACHE STRING "")
   set(RUNTIMES_${target}_LIBCXX_ENABLE_SHARED OFF CACHE BOOL "")
@@ -391,14 +390,13 @@ foreach(target riscv32-unknown-elf)
   foreach(lang C;CXX;ASM)
     # TODO: The preprocessor defines workaround various issues in libc and libc++ integration.
     # These should be addressed and removed over time.
-    set(RUNTIMES_${target}_CMAKE_${lang}_FLAGS "--target=${target} -march=rv32imafc -mabi=ilp32f -Wno-atomic-alignment \"-Dvfprintf(stream, format, vlist)=vprintf(format, vlist)\" \"-Dfprintf(stream, format, ...)=printf(format)\" \"-Dtimeval=struct timeval{int tv_sec; int tv_usec;}\" \"-Dgettimeofday(tv, tz)\" -D_LIBCPP_PRINT=1" CACHE STRING "")
+    set(RUNTIMES_${target}_CMAKE_${lang}_FLAGS "--target=${target} -march=rv32imafc -mabi=ilp32f -Wno-atomic-alignment \"-Dvfprintf(stream, format, vlist)=vprintf(format, vlist)\" \"-Dfprintf(stream, format, ...)=printf(format)\" \"-Dgettimeofday(tv, tz)\" -D_LIBCPP_PRINT=1" CACHE STRING "")
   endforeach()
   foreach(type SHARED;MODULE;EXE)
     set(RUNTIMES_${target}_CMAKE_${type}_LINKER_FLAGS "-fuse-ld=lld" CACHE STRING "")
   endforeach()
   set(RUNTIMES_${target}_LLVM_LIBC_FULL_BUILD ON CACHE BOOL "")
   set(RUNTIMES_${target}_LIBC_ENABLE_USE_BY_CLANG ON CACHE BOOL "")
-  set(RUNTIMES_${target}_LIBC_USE_NEW_HEADER_GEN OFF CACHE BOOL "")
   set(RUNTIMES_${target}_LIBCXX_ABI_VERSION 2 CACHE STRING "")
   set(RUNTIMES_${target}_LIBCXX_CXX_ABI none CACHE STRING "")
   set(RUNTIMES_${target}_LIBCXX_ENABLE_SHARED OFF CACHE BOOL "")

clang/docs/LanguageExtensions.rst

@@ -732,6 +732,10 @@ at the end to the next power of 2.
 
 These reductions support both fixed-sized and scalable vector types.
 
+The integer reduction intrinsics, including ``__builtin_reduce_add``,
+``__builtin_reduce_mul``, ``__builtin_reduce_and``, ``__builtin_reduce_or``,
+and ``__builtin_reduce_xor``, can be called in a ``constexpr`` context.
+
 Example:
 
 .. code-block:: c++

clang/docs/RealtimeSanitizer.rst

@@ -167,7 +167,11 @@ A **partial** list of flags RealtimeSanitizer respects:
    * - ``halt_on_error``
      - ``true``
      - boolean
-     - Exit after first reported error. If false (continue after a detected error), deduplicates error stacks so errors appear only once.
+     - Exit after first reported error.
+   * - ``suppress_equal_stacks``
+     - ``true``
+     - boolean
+     - If true, suppress duplicate reports (i.e. only print each unique error once). Only particularly useful when ``halt_on_error=false``.
    * - ``print_stats_on_exit``
      - ``false``
      - boolean
@@ -203,6 +207,44 @@ Some issues with flags can be debugged using the ``verbosity=$NUM`` flag:
    misspelled_flag
    ...
 
+Additional customization
+------------------------
+
+In addition to ``__rtsan_default_options`` outlined above, you can provide definitions of other functions that affect how RTSan operates.
+
+To be notified on every error reported by RTsan, provide a definition of ``__sanitizer_report_error_summary``.
+
+.. code-block:: c
+
+   extern "C" void __sanitizer_report_error_summary(const char *error_summary) {
+      fprintf(stderr, "%s %s\n", "In custom handler! ", error_summary);
+      /* do other custom things */
+   }
+
+The error summary will be of the form: 
+
+.. code-block:: console
+
+   SUMMARY: RealtimeSanitizer: unsafe-library-call main.cpp:8 in process(std::__1::vector<int, std::__1::allocator<int>>&)
+
+To register a callback which will be invoked before a RTSan kills the process:
+
+.. code-block:: c
+
+  extern "C" void __sanitizer_set_death_callback(void (*callback)(void));
+
+  void custom_on_die_callback() {
+    fprintf(stderr, "In custom handler!")
+    /* do other custom things */
+  }
+
+  int main()
+  {
+    __sanitizer_set_death_callback(custom_on_die_callback);
+    ...
+  }
+
+
 Disabling and suppressing
 -------------------------
 

clang/docs/ReleaseNotes.rst

@@ -148,6 +148,17 @@ C++ Specific Potentially Breaking Changes
     // Now diagnoses with an error.
     void f(int& i [[clang::lifetimebound]]);
 
+- Clang now rejects all field accesses on null pointers in constant expressions. The following code
+  used to work but will now be rejected:
+
+  .. code-block:: c++
+
+    struct S { int a; int b; };
+    constexpr const int *p = &((S*)nullptr)->b;
+
+  Previously, this code was erroneously accepted.
+
+
 ABI Changes in This Version
 ---------------------------
 
@@ -366,6 +377,7 @@ Non-comprehensive list of changes in this release
 - ``__builtin_reduce_add`` function can now be used in constant expressions.
 - ``__builtin_reduce_mul`` function can now be used in constant expressions.
 - ``__builtin_reduce_and`` function can now be used in constant expressions.
+- ``__builtin_reduce_or`` and ``__builtin_reduce_xor`` functions can now be used in constant expressions.
 
 New Compiler Flags
 ------------------
@@ -594,6 +606,8 @@ Improvements to Clang's diagnostics
 - Clang now correctly recognises code after a call to a ``[[noreturn]]`` constructor
   as unreachable (#GH63009).
 
+- Clang now omits shadowing warnings for parameter names in explicit object member functions (#GH95707).
+
 Improvements to Clang's time-trace
 ----------------------------------
 

clang/include/clang/Basic/Builtins.td

@@ -1486,13 +1486,13 @@ def ReduceMinimum : Builtin {
 
 def ReduceXor : Builtin {
   let Spellings = ["__builtin_reduce_xor"];
-  let Attributes = [NoThrow, Const, CustomTypeChecking];
+  let Attributes = [NoThrow, Const, CustomTypeChecking, Constexpr];
   let Prototype = "void(...)";
 }
 
 def ReduceOr : Builtin {
   let Spellings = ["__builtin_reduce_or"];
-  let Attributes = [NoThrow, Const, CustomTypeChecking];
+  let Attributes = [NoThrow, Const, CustomTypeChecking, Constexpr];
   let Prototype = "void(...)";
 }
 
@@ -4750,6 +4750,12 @@ def HLSLAny : LangBuiltin<"HLSL_LANG"> {
   let Prototype = "bool(...)";
 }
 
+def HLSLWaveActiveAnyTrue : LangBuiltin<"HLSL_LANG"> {
+  let Spellings = ["__builtin_hlsl_wave_active_any_true"];
+  let Attributes = [NoThrow, Const];
+  let Prototype = "bool(bool)";
+}
+
 def HLSLWaveActiveCountBits : LangBuiltin<"HLSL_LANG"> {
   let Spellings = ["__builtin_hlsl_wave_active_count_bits"];
   let Attributes = [NoThrow, Const];

clang/include/clang/Basic/BuiltinsAMDGPU.def

@@ -434,10 +434,14 @@ TARGET_BUILTIN(__builtin_amdgcn_cvt_sr_fp8_f32, "ifiiIi", "nc", "fp8-conversion-
 //===----------------------------------------------------------------------===//
 // GFX950 only builtins.
 //===----------------------------------------------------------------------===//
+TARGET_BUILTIN(__builtin_amdgcn_mfma_scale_f32_16x16x128_f8f6f4, "V4fV8ZiV8ZiV4fIiIiIiiIii", "nc", "gfx950-insts")
+TARGET_BUILTIN(__builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4, "V16fV8ZiV8ZiV16fIiIiIiiIii", "nc", "gfx950-insts")
+
 TARGET_BUILTIN(__builtin_amdgcn_mfma_f32_16x16x32_f16, "V4fV8hV8hV4fIiIiIi", "nc", "gfx950-insts")
 TARGET_BUILTIN(__builtin_amdgcn_mfma_f32_32x32x16_f16, "V16fV8hV8hV16fIiIiIi", "nc", "gfx950-insts")
-
 TARGET_BUILTIN(__builtin_amdgcn_mfma_f32_32x32x16_bf16, "V16fV8yV8yV16fIiIiIi", "nc", "gfx950-insts")
+TARGET_BUILTIN(__builtin_amdgcn_mfma_i32_16x16x64_i8, "V4iV4iV4iV4iIiIiIi", "nc", "gfx950-insts")
+TARGET_BUILTIN(__builtin_amdgcn_mfma_i32_32x32x32_i8, "V16iV4iV4iV16iIiIiIi", "nc", "gfx950-insts")
 
 //===----------------------------------------------------------------------===//
 // GFX12+ only builtins.

clang/lib/AST/ByteCode/Interp.cpp

@@ -1370,9 +1370,15 @@ bool CallBI(InterpState &S, CodePtr OpPC, const Function *Func,
   S.Current = NewFrame.get();
 
   if (InterpretBuiltin(S, OpPC, Func, CE, BuiltinID)) {
-    NewFrame.release();
+    // Release ownership of NewFrame to prevent it from being deleted.
+    NewFrame.release(); // Frame was deleted already.
+    // Ensure that S.Current is correctly reset to the previous frame.
+    assert(S.Current == FrameBefore);
     return true;
   }
+
+  // Interpreting the function failed somehow. Reset to
+  // previous state.
   S.Current = FrameBefore;
   return false;
 }
@@ -1588,7 +1594,7 @@ bool CheckBitCast(InterpState &S, CodePtr OpPC, bool HasIndeterminateBits,
 }
 
 // https://github.com/llvm/llvm-project/issues/102513
-#if defined(_WIN32) && !defined(__clang__) && !defined(NDEBUG)
+#if defined(_MSC_VER) && !defined(__clang__) && !defined(NDEBUG)
 #pragma optimize("", off)
 #endif
 bool Interpret(InterpState &S, APValue &Result) {
@@ -1616,7 +1622,7 @@ bool Interpret(InterpState &S, APValue &Result) {
   }
 }
 // https://github.com/llvm/llvm-project/issues/102513
-#if defined(_WIN32) && !defined(__clang__) && !defined(NDEBUG)
+#if defined(_MSC_VER) && !defined(__clang__) && !defined(NDEBUG)
 #pragma optimize("", on)
 #endif
 

clang/lib/AST/ByteCode/Interp.h

@@ -3084,6 +3084,9 @@ inline bool BitCastPtr(InterpState &S, CodePtr OpPC) {
   const Pointer &FromPtr = S.Stk.pop<Pointer>();
   Pointer &ToPtr = S.Stk.peek<Pointer>();
 
+  if (!CheckLoad(S, OpPC, FromPtr))
+    return false;
+
   if (!DoBitCastPtr(S, OpPC, FromPtr, ToPtr))
     return false;
 

clang/lib/AST/ExprConstant.cpp

@@ -3261,8 +3261,8 @@ static bool HandleLValueDirectBase(EvalInfo &Info, const Expr *E, LValue &Obj,
     RL = &Info.Ctx.getASTRecordLayout(Derived);
   }
 
-  Obj.getLValueOffset() += RL->getBaseClassOffset(Base);
   Obj.addDecl(Info, E, Base, /*Virtual*/ false);
+  Obj.getLValueOffset() += RL->getBaseClassOffset(Base);
   return true;
 }
 
@@ -3286,8 +3286,8 @@ static bool HandleLValueBase(EvalInfo &Info, const Expr *E, LValue &Obj,
   // Find the virtual base class.
   if (DerivedDecl->isInvalidDecl()) return false;
   const ASTRecordLayout &Layout = Info.Ctx.getASTRecordLayout(DerivedDecl);
-  Obj.getLValueOffset() += Layout.getVBaseClassOffset(BaseDecl);
   Obj.addDecl(Info, E, BaseDecl, /*Virtual*/ true);
+  Obj.getLValueOffset() += Layout.getVBaseClassOffset(BaseDecl);
   return true;
 }
 
@@ -3330,8 +3330,8 @@ static bool HandleLValueMember(EvalInfo &Info, const Expr *E, LValue &LVal,
   }
 
   unsigned I = FD->getFieldIndex();
-  LVal.adjustOffset(Info.Ctx.toCharUnitsFromBits(RL->getFieldOffset(I)));
   LVal.addDecl(Info, E, FD);
+  LVal.adjustOffset(Info.Ctx.toCharUnitsFromBits(RL->getFieldOffset(I)));
   return true;
 }
 
@@ -13529,7 +13529,9 @@ bool IntExprEvaluator::VisitBuiltinCallExpr(const CallExpr *E,
 
   case Builtin::BI__builtin_reduce_add:
   case Builtin::BI__builtin_reduce_mul:
-  case Builtin::BI__builtin_reduce_and: {
+  case Builtin::BI__builtin_reduce_and:
+  case Builtin::BI__builtin_reduce_or:
+  case Builtin::BI__builtin_reduce_xor: {
     APValue Source;
     if (!EvaluateAsRValue(Info, E->getArg(0), Source))
       return false;
@@ -13558,6 +13560,14 @@ bool IntExprEvaluator::VisitBuiltinCallExpr(const CallExpr *E,
         Reduced &= Source.getVectorElt(EltNum).getInt();
         break;
       }
+      case Builtin::BI__builtin_reduce_or: {
+        Reduced |= Source.getVectorElt(EltNum).getInt();
+        break;
+      }
+      case Builtin::BI__builtin_reduce_xor: {
+        Reduced ^= Source.getVectorElt(EltNum).getInt();
+        break;
+      }
       }
     }
 

clang/lib/CodeGen/CGBuiltin.cpp

@@ -19049,8 +19049,8 @@ Value *CodeGenFunction::EmitHLSLBuiltinExpr(unsigned BuiltinID,
            "cross operands must have a float representation");
     // make sure each vector has exactly 3 elements
     assert(
-        E->getArg(0)->getType()->getAs<VectorType>()->getNumElements() == 3 &&
-        E->getArg(1)->getType()->getAs<VectorType>()->getNumElements() == 3 &&
+        E->getArg(0)->getType()->castAs<VectorType>()->getNumElements() == 3 &&
+        E->getArg(1)->getType()->castAs<VectorType>()->getNumElements() == 3 &&
         "input vectors must have 3 elements each");
     return Builder.CreateIntrinsic(
         /*ReturnType=*/Op0->getType(), CGM.getHLSLRuntime().getCrossIntrinsic(),
@@ -19279,6 +19279,15 @@ case Builtin::BI__builtin_hlsl_elementwise_isinf: {
         /*ReturnType=*/Op0->getType(), CGM.getHLSLRuntime().getStepIntrinsic(),
         ArrayRef<Value *>{Op0, Op1}, nullptr, "hlsl.step");
   }
+  case Builtin::BI__builtin_hlsl_wave_active_any_true: {
+    Value *Op = EmitScalarExpr(E->getArg(0));
+    assert(Op->getType()->isIntegerTy(1) &&
+           "Intrinsic WaveActiveAnyTrue operand must be a bool");
+
+    Intrinsic::ID ID = CGM.getHLSLRuntime().getWaveActiveAnyTrueIntrinsic();
+    return EmitRuntimeCall(
+        Intrinsic::getOrInsertDeclaration(&CGM.getModule(), ID), {Op});
+  }
   case Builtin::BI__builtin_hlsl_wave_active_count_bits: {
     Value *OpExpr = EmitScalarExpr(E->getArg(0));
     Intrinsic::ID ID = CGM.getHLSLRuntime().getWaveActiveCountBitsIntrinsic();
@@ -19726,7 +19735,20 @@ Value *CodeGenFunction::EmitAMDGPUBuiltinExpr(unsigned BuiltinID,
                                             (uint64_t)0);
     return Builder.CreateInsertElement(I0, A, 1);
   }
-
+  case AMDGPU::BI__builtin_amdgcn_mfma_scale_f32_16x16x128_f8f6f4:
+  case AMDGPU::BI__builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4: {
+    llvm::FixedVectorType *VT = FixedVectorType::get(Builder.getInt32Ty(), 8);
+    Function *F = CGM.getIntrinsic(
+        BuiltinID == AMDGPU::BI__builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4
+            ? Intrinsic::amdgcn_mfma_scale_f32_32x32x64_f8f6f4
+            : Intrinsic::amdgcn_mfma_scale_f32_16x16x128_f8f6f4,
+        {VT, VT});
+
+    SmallVector<Value *, 9> Args;
+    for (unsigned I = 0, N = E->getNumArgs(); I != N; ++I)
+      Args.push_back(EmitScalarExpr(E->getArg(I)));
+    return Builder.CreateCall(F, Args);
+  }
   case AMDGPU::BI__builtin_amdgcn_wmma_bf16_16x16x16_bf16_w32:
   case AMDGPU::BI__builtin_amdgcn_wmma_bf16_16x16x16_bf16_tied_w32:
   case AMDGPU::BI__builtin_amdgcn_wmma_bf16_16x16x16_bf16_w64:

clang/lib/CodeGen/CGExpr.cpp

@@ -4370,7 +4370,7 @@ LValue CodeGenFunction::EmitArraySubscriptExpr(const ArraySubscriptExpr *E,
           ME &&
           ME->isFlexibleArrayMemberLike(getContext(), StrictFlexArraysLevel) &&
           ME->getMemberDecl()->getType()->isCountAttributedType()) {
-        const FieldDecl *FAMDecl = dyn_cast<FieldDecl>(ME->getMemberDecl());
+        const FieldDecl *FAMDecl = cast<FieldDecl>(ME->getMemberDecl());
         if (const FieldDecl *CountFD = FAMDecl->findCountedByField()) {
           if (std::optional<int64_t> Diff =
                   getOffsetDifferenceInBits(*this, CountFD, FAMDecl)) {

clang/lib/CodeGen/CGHLSLRuntime.h

@@ -91,6 +91,7 @@ class CGHLSLRuntime {
   GENERATE_HLSL_INTRINSIC_FUNCTION(UDot, udot)
   GENERATE_HLSL_INTRINSIC_FUNCTION(Dot4AddI8Packed, dot4add_i8packed)
   GENERATE_HLSL_INTRINSIC_FUNCTION(Dot4AddU8Packed, dot4add_u8packed)
+  GENERATE_HLSL_INTRINSIC_FUNCTION(WaveActiveAnyTrue, wave_any)
   GENERATE_HLSL_INTRINSIC_FUNCTION(WaveActiveCountBits, wave_active_countbits)
   GENERATE_HLSL_INTRINSIC_FUNCTION(WaveIsFirstLane, wave_is_first_lane)
   GENERATE_HLSL_INTRINSIC_FUNCTION(WaveReadLaneAt, wave_readlane)

clang/lib/Headers/hlsl/hlsl_intrinsics.h

@@ -2223,6 +2223,15 @@ float4 trunc(float4);
 // Wave* builtins
 //===----------------------------------------------------------------------===//
 
+/// \brief Returns true if the expression is true in any active lane in the
+/// current wave.
+///
+/// \param Val The boolean expression to evaluate.
+/// \return True if the expression is true in any lane.
+_HLSL_AVAILABILITY(shadermodel, 6.0)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_wave_active_any_true)
+__attribute__((convergent)) bool WaveActiveAnyTrue(bool Val);
+
 /// \brief Counts the number of boolean variables which evaluate to true across
 /// all active lanes in the current wave.
 ///

clang/lib/Sema/AnalysisBasedWarnings.cpp

@@ -2257,7 +2257,7 @@ class UnsafeBufferUsageReporter : public UnsafeBufferUsageHandler {
       } else if (isa<MemberExpr>(Operation)) {
         // note_unsafe_buffer_operation doesn't have this mode yet.
         assert(!IsRelatedToDecl && "Not implemented yet!");
-        auto ME = dyn_cast<MemberExpr>(Operation);
+        auto *ME = cast<MemberExpr>(Operation);
         D = ME->getMemberDecl();
         MsgParam = 5;
       } else if (const auto *ECE = dyn_cast<ExplicitCastExpr>(Operation)) {