Merge "merge main into amd-staging" into amd-staging

Author: ronlieb

clang/docs/ReleaseNotes.rst

@@ -624,6 +624,7 @@ Bug Fixes to C++ Support
 - Fixed an assertion failure in range calculations for conditional throw expressions. (#GH111854)
 - Clang now correctly ignores previous partial specializations of member templates explicitly specialized for
   an implicitly instantiated class template specialization. (#GH51051)
+- Fixed an assertion failure caused by invalid enum forward declarations. (#GH112208)
 
 Bug Fixes to AST Handling
 ^^^^^^^^^^^^^^^^^^^^^^^^^
@@ -710,6 +711,7 @@ X86 Support
 - Supported intrinsics for ``MOVRS AND AVX10.2``.
   * Supported intrinsics of ``_mm(256|512)_(mask(z))_loadrs_epi(8|16|32|64)``.
 - Support ISA of ``AMX-FP8``.
+- Support ISA of ``AMX-TRANSPOSE``.
 
 Arm and AArch64 Support
 ^^^^^^^^^^^^^^^^^^^^^^^

clang/docs/analyzer/checkers.rst

@@ -3584,7 +3584,7 @@ These are examples of cases that we consider safe:
       RefCountable* uncounted = this; // ok
     }
 
-Here are some examples of situations that we warn about as they *might* be potentially unsafe. The logic is that either we're able to guarantee that an argument is safe or it's considered if not a bug then bug-prone.
+Here are some examples of situations that we warn about as they *might* be potentially unsafe. The logic is that either we're able to guarantee that a local variable is safe or it's considered unsafe.
 
   .. code-block:: cpp
 
@@ -3603,11 +3603,48 @@ Here are some examples of situations that we warn about as they *might* be poten
       RefCountable* uncounted = counted.get(); // warn
     }
 
-We don't warn about these cases - we don't consider them necessarily safe but since they are very common and usually safe we'd introduce a lot of false positives otherwise:
-- variable defined in condition part of an ```if``` statement
-- variable defined in init statement condition of a ```for``` statement
+alpha.webkit.UncheckedLocalVarsChecker
+""""""""""""""""""""""""""""""""""""""
+The goal of this rule is to make sure that any unchecked local variable is backed by a CheckedPtr or CheckedRef with lifetime that is strictly larger than the scope of the unchecked local variable. To be on the safe side we require the scope of an unchecked variable to be embedded in the scope of CheckedPtr/CheckRef object that backs it.
+
+These are examples of cases that we consider safe:
+
+  .. code-block:: cpp
 
-For the time being we also don't warn about uninitialized uncounted local variables.
+    void foo1() {
+      CheckedPtr<RefCountable> counted;
+      // The scope of uncounted is EMBEDDED in the scope of counted.
+      {
+        RefCountable* uncounted = counted.get(); // ok
+      }
+    }
+
+    void foo2(CheckedPtr<RefCountable> counted_param) {
+      RefCountable* uncounted = counted_param.get(); // ok
+    }
+
+    void FooClass::foo_method() {
+      RefCountable* uncounted = this; // ok
+    }
+
+Here are some examples of situations that we warn about as they *might* be potentially unsafe. The logic is that either we're able to guarantee that a local variable is safe or it's considered unsafe.
+
+  .. code-block:: cpp
+
+    void foo1() {
+      RefCountable* uncounted = new RefCountable; // warn
+    }
+
+    RefCountable* global_uncounted;
+    void foo2() {
+      RefCountable* uncounted = global_uncounted; // warn
+    }
+
+    void foo3() {
+      RefPtr<RefCountable> counted;
+      // The scope of uncounted is not EMBEDDED in the scope of counted.
+      RefCountable* uncounted = counted.get(); // warn
+    }
 
 Debug Checkers
 ---------------

clang/include/clang/Basic/BuiltinsX86_64.def

@@ -128,6 +128,11 @@ TARGET_BUILTIN(__builtin_ia32_tdpbf16ps_internal, "V256iUsUsUsV256iV256iV256i",
 TARGET_BUILTIN(__builtin_ia32_tdpfp16ps_internal, "V256iUsUsUsV256iV256iV256i", "n", "amx-fp16")
 TARGET_BUILTIN(__builtin_ia32_tcmmimfp16ps_internal, "V256iUsUsUsV256iV256iV256i", "n", "amx-complex")
 TARGET_BUILTIN(__builtin_ia32_tcmmrlfp16ps_internal, "V256iUsUsUsV256iV256iV256i", "n", "amx-complex")
+TARGET_BUILTIN(__builtin_ia32_t2rpntlvwz0_internal, "vUsUsUsV256i*V256i*vC*z", "n", "amx-transpose")
+TARGET_BUILTIN(__builtin_ia32_t2rpntlvwz0t1_internal, "vUsUsUsV256i*V256i*vC*z", "n", "amx-transpose")
+TARGET_BUILTIN(__builtin_ia32_t2rpntlvwz1_internal, "vUsUsUsV256i*V256i*vC*z", "n", "amx-transpose")
+TARGET_BUILTIN(__builtin_ia32_t2rpntlvwz1t1_internal, "vUsUsUsV256i*V256i*vC*z", "n", "amx-transpose")
+TARGET_BUILTIN(__builtin_ia32_ttransposed_internal, "V256iUsUsV256i", "n", "amx-transpose")
 // AMX
 TARGET_BUILTIN(__builtin_ia32_tile_loadconfig, "vvC*", "n", "amx-tile")
 TARGET_BUILTIN(__builtin_ia32_tile_storeconfig, "vvC*", "n", "amx-tile")
@@ -148,9 +153,15 @@ TARGET_BUILTIN(__builtin_ia32_ptwrite64, "vUOi", "n", "ptwrite")
 TARGET_BUILTIN(__builtin_ia32_tcmmimfp16ps, "vIUcIUcIUc", "n", "amx-complex")
 TARGET_BUILTIN(__builtin_ia32_tcmmrlfp16ps, "vIUcIUcIUc", "n", "amx-complex")
 
+TARGET_BUILTIN(__builtin_ia32_t2rpntlvwz0, "vIUcvC*z", "n", "amx-transpose")
+TARGET_BUILTIN(__builtin_ia32_t2rpntlvwz0t1, "vIUcvC*z", "n","amx-transpose")
+TARGET_BUILTIN(__builtin_ia32_t2rpntlvwz1, "vIUcvC*z", "n", "amx-transpose")
+TARGET_BUILTIN(__builtin_ia32_t2rpntlvwz1t1, "vIUcvC*z", "n","amx-transpose")
+TARGET_BUILTIN(__builtin_ia32_ttransposed, "vIUcIUc", "n", "amx-transpose")
+
 TARGET_BUILTIN(__builtin_ia32_prefetchi, "vvC*Ui", "nc", "prefetchi")
 TARGET_BUILTIN(__builtin_ia32_cmpccxadd32, "Siv*SiSiIi", "n", "cmpccxadd")
-TARGET_BUILTIN(__builtin_ia32_cmpccxadd64, "SLLiv*SLLiSLLiIi", "n", "cmpccxadd")
+TARGET_BUILTIN(__builtin_ia32_cmpccxadd64, "SLLiSLLi*SLLiSLLiIi", "n", "cmpccxadd")
 
 // AMX_FP16 FP16
 TARGET_BUILTIN(__builtin_ia32_tdpfp16ps, "vIUcIUcIUc", "n", "amx-fp16")

clang/include/clang/Driver/Options.td

@@ -1791,6 +1791,12 @@ defm debug_info_for_profiling : BoolFOption<"debug-info-for-profiling",
   PosFlag<SetTrue, [], [ClangOption, CC1Option],
           "Emit extra debug info to make sample profile more accurate">,
   NegFlag<SetFalse>>;
+def fprofile_generate_cold_function_coverage : Flag<["-"], "fprofile-generate-cold-function-coverage">, 
+    Group<f_Group>, Visibility<[ClangOption, CLOption]>,
+    HelpText<"Generate instrumented code to collect coverage info for cold functions into default.profraw file (overridden by '=' form of option or LLVM_PROFILE_FILE env var)">;
+def fprofile_generate_cold_function_coverage_EQ : Joined<["-"], "fprofile-generate-cold-function-coverage=">, 
+    Group<f_Group>, Visibility<[ClangOption, CLOption]>, MetaVarName<"<directory>">,
+    HelpText<"Generate instrumented code to collect coverage info for cold functions into <directory>/default.profraw (overridden by LLVM_PROFILE_FILE env var)">; 
 def fprofile_instr_generate : Flag<["-"], "fprofile-instr-generate">,
     Group<f_Group>, Visibility<[ClangOption, CLOption]>,
     HelpText<"Generate instrumented code to collect execution counts into default.profraw file (overridden by '=' form of option or LLVM_PROFILE_FILE env var)">;
@@ -6414,6 +6420,8 @@ def mamx_fp8 : Flag<["-"], "mamx-fp8">, Group<m_x86_Features_Group>;
 def mno_amx_fp8 : Flag<["-"], "mno-amx-fp8">, Group<m_x86_Features_Group>;
 def mamx_tile : Flag<["-"], "mamx-tile">, Group<m_x86_Features_Group>;
 def mno_amx_tile : Flag<["-"], "mno-amx-tile">, Group<m_x86_Features_Group>;
+def mamx_transpose : Flag<["-"], "mamx-transpose">, Group<m_x86_Features_Group>;
+def mno_amx_transpose : Flag<["-"], "mno-amx-transpose">, Group<m_x86_Features_Group>;
 def mcmpccxadd : Flag<["-"], "mcmpccxadd">, Group<m_x86_Features_Group>;
 def mno_cmpccxadd : Flag<["-"], "mno-cmpccxadd">, Group<m_x86_Features_Group>;
 def msse : Flag<["-"], "msse">, Group<m_x86_Features_Group>;

clang/include/clang/StaticAnalyzer/Checkers/Checkers.td

@@ -1764,4 +1764,8 @@ def UncountedLocalVarsChecker : Checker<"UncountedLocalVarsChecker">,
   HelpText<"Check uncounted local variables.">,
   Documentation<HasDocumentation>;
 
+def UncheckedLocalVarsChecker : Checker<"UncheckedLocalVarsChecker">,
+  HelpText<"Check unchecked local variables.">,
+  Documentation<HasDocumentation>;
+
 } // end alpha.webkit

clang/lib/AST/ByteCode/IntegralAP.h

@@ -171,10 +171,14 @@ template <bool Signed> class IntegralAP final {
     return IntegralAP<false>(Copy);
   }
 
-  void bitcastToMemory(std::byte *Dest) const { assert(false); }
+  void bitcastToMemory(std::byte *Dest) const {
+    llvm::StoreIntToMemory(V, (uint8_t *)Dest, bitWidth() / 8);
+  }
 
   static IntegralAP bitcastFromMemory(const std::byte *Src, unsigned BitWidth) {
-    return IntegralAP();
+    APInt V(BitWidth, static_cast<uint64_t>(0), Signed);
+    llvm::LoadIntFromMemory(V, (const uint8_t *)Src, BitWidth / 8);
+    return IntegralAP(V);
   }
 
   ComparisonCategoryResult compare(const IntegralAP &RHS) const {

clang/lib/Basic/DiagnosticIDs.cpp

@@ -575,6 +575,12 @@ DiagnosticIDs::getDiagnosticSeverity(unsigned DiagID, SourceLocation Loc,
       DiagID != diag::fatal_too_many_errors && Diag.FatalsAsError)
     Result = diag::Severity::Error;
 
+  // Rest of the mappings are only applicable for diagnostics associated with a
+  // SourceLocation, bail out early for others.
+  if (!Diag.hasSourceManager())
+    return Result;
+
+  const auto &SM = Diag.getSourceManager();
   // Custom diagnostics always are emitted in system headers.
   bool ShowInSystemHeader =
       !GetDiagInfo(DiagID) || GetDiagInfo(DiagID)->WarnShowInSystemHeader;
@@ -583,15 +589,14 @@ DiagnosticIDs::getDiagnosticSeverity(unsigned DiagID, SourceLocation Loc,
   // because we also want to ignore extensions and warnings in -Werror and
   // -pedantic-errors modes, which *map* warnings/extensions to errors.
   if (State->SuppressSystemWarnings && !ShowInSystemHeader && Loc.isValid() &&
-      Diag.getSourceManager().isInSystemHeader(
-          Diag.getSourceManager().getExpansionLoc(Loc)))
+      SM.isInSystemHeader(SM.getExpansionLoc(Loc)))
     return diag::Severity::Ignored;
 
   // We also ignore warnings due to system macros
   bool ShowInSystemMacro =
       !GetDiagInfo(DiagID) || GetDiagInfo(DiagID)->WarnShowInSystemMacro;
   if (State->SuppressSystemWarnings && !ShowInSystemMacro && Loc.isValid() &&
-      Diag.getSourceManager().isInSystemMacro(Loc))
+      SM.isInSystemMacro(Loc))
     return diag::Severity::Ignored;
 
   return Result;

clang/lib/Basic/Targets/X86.cpp

@@ -430,6 +430,8 @@ bool X86TargetInfo::handleTargetFeatures(std::vector<std::string> &Features,
       HasAMXCOMPLEX = true;
     } else if (Feature == "+amx-fp8") {
       HasAMXFP8 = true;
+    } else if (Feature == "+amx-transpose") {
+      HasAMXTRANSPOSE = true;
     } else if (Feature == "+cmpccxadd") {
       HasCMPCCXADD = true;
     } else if (Feature == "+raoint") {
@@ -951,6 +953,8 @@ void X86TargetInfo::getTargetDefines(const LangOptions &Opts,
     Builder.defineMacro("__AMX_COMPLEX__");
   if (HasAMXFP8)
     Builder.defineMacro("__AMX_FP8__");
+  if (HasAMXTRANSPOSE)
+    Builder.defineMacro("__AMX_TRANSPOSE__");
   if (HasCMPCCXADD)
     Builder.defineMacro("__CMPCCXADD__");
   if (HasRAOINT)
@@ -1079,9 +1083,10 @@ bool X86TargetInfo::isValidFeatureName(StringRef Name) const {
       .Case("amx-bf16", true)
       .Case("amx-complex", true)
       .Case("amx-fp16", true)
+      .Case("amx-fp8", true)
       .Case("amx-int8", true)
       .Case("amx-tile", true)
-      .Case("amx-fp8", true)
+      .Case("amx-transpose", true)
       .Case("avx", true)
       .Case("avx10.1-256", true)
       .Case("avx10.1-512", true)
@@ -1198,9 +1203,10 @@ bool X86TargetInfo::hasFeature(StringRef Feature) const {
       .Case("amx-bf16", HasAMXBF16)
       .Case("amx-complex", HasAMXCOMPLEX)
       .Case("amx-fp16", HasAMXFP16)
+      .Case("amx-fp8", HasAMXFP8)
       .Case("amx-int8", HasAMXINT8)
       .Case("amx-tile", HasAMXTILE)
-      .Case("amx-fp8", HasAMXFP8)
+      .Case("amx-transpose", HasAMXTRANSPOSE)
       .Case("avx", SSELevel >= AVX)
       .Case("avx10.1-256", HasAVX10_1)
       .Case("avx10.1-512", HasAVX10_1_512)

clang/lib/Basic/Targets/X86.h

@@ -158,6 +158,7 @@ class LLVM_LIBRARY_VISIBILITY X86TargetInfo : public TargetInfo {
   bool HasAMXBF16 = false;
   bool HasAMXCOMPLEX = false;
   bool HasAMXFP8 = false;
+  bool HasAMXTRANSPOSE = false;
   bool HasSERIALIZE = false;
   bool HasTSXLDTRK = false;
   bool HasUSERMSR = false;

clang/lib/CodeGen/CGBuiltin.cpp

@@ -16991,6 +16991,58 @@ Value *CodeGenFunction::EmitX86BuiltinExpr(unsigned BuiltinID,
     // instruction, but it will create a memset that won't be optimized away.
     return Builder.CreateMemSet(Ops[0], Ops[1], Ops[2], Align(1), true);
   }
+  // Corresponding to intrisics which will return 2 tiles (tile0_tile1).
+  case X86::BI__builtin_ia32_t2rpntlvwz0_internal:
+  case X86::BI__builtin_ia32_t2rpntlvwz0t1_internal:
+  case X86::BI__builtin_ia32_t2rpntlvwz1_internal:
+  case X86::BI__builtin_ia32_t2rpntlvwz1t1_internal: {
+    Intrinsic::ID IID;
+    switch (BuiltinID) {
+    default:
+      llvm_unreachable("Unsupported intrinsic!");
+    case X86::BI__builtin_ia32_t2rpntlvwz0_internal:
+      IID = Intrinsic::x86_t2rpntlvwz0_internal;
+      break;
+    case X86::BI__builtin_ia32_t2rpntlvwz0t1_internal:
+      IID = Intrinsic::x86_t2rpntlvwz0t1_internal;
+      break;
+    case X86::BI__builtin_ia32_t2rpntlvwz1_internal:
+      IID = Intrinsic::x86_t2rpntlvwz1_internal;
+      break;
+    case X86::BI__builtin_ia32_t2rpntlvwz1t1_internal:
+      IID = Intrinsic::x86_t2rpntlvwz1t1_internal;
+      break;
+    }
+
+    // Ops = (Row0, Col0, Col1, DstPtr0, DstPtr1, SrcPtr, Stride)
+    Value *Call = Builder.CreateCall(CGM.getIntrinsic(IID),
+                                     {Ops[0], Ops[1], Ops[2], Ops[5], Ops[6]});
+
+    auto *PtrTy = E->getArg(3)->getType()->getAs<PointerType>();
+    assert(PtrTy && "arg3 must be of pointer type");
+    QualType PtreeTy = PtrTy->getPointeeType();
+    llvm::Type *TyPtee = ConvertType(PtreeTy);
+
+    // Bitcast amx type (x86_amx) to vector type (256 x i32)
+    // Then store tile0 into DstPtr0
+    Value *T0 = Builder.CreateExtractValue(Call, 0);
+    Value *VecT0 = Builder.CreateIntrinsic(Intrinsic::x86_cast_tile_to_vector,
+                                           {TyPtee}, {T0});
+    Builder.CreateDefaultAlignedStore(VecT0, Ops[3]);
+
+    // Then store tile1 into DstPtr1
+    Value *T1 = Builder.CreateExtractValue(Call, 1);
+    Value *VecT1 = Builder.CreateIntrinsic(Intrinsic::x86_cast_tile_to_vector,
+                                           {TyPtee}, {T1});
+    Value *Store = Builder.CreateDefaultAlignedStore(VecT1, Ops[4]);
+
+    // Note: Here we escape directly use x86_tilestored64_internal to store
+    // the results due to it can't make sure the Mem written scope. This may
+    // cause shapes reloads after first amx intrinsic, which current amx reg-
+    // ister allocation has no ability to handle it.
+
+    return Store;
+  }
   case X86::BI__ud2:
     // llvm.trap makes a ud2a instruction on x86.
     return EmitTrapCall(Intrinsic::trap);