Merge branch 'upstream' into x86-fast-amd-shld

Author: RKSimon

clang/include/clang/Driver/Options.td

@@ -1392,6 +1392,8 @@ def fhip_emit_relocatable : Flag<["-"], "fhip-emit-relocatable">,
   HelpText<"Compile HIP source to relocatable">;
 def fno_hip_emit_relocatable : Flag<["-"], "fno-hip-emit-relocatable">,
   HelpText<"Do not override toolchain to compile HIP source to relocatable">;
+def flto_partitions_EQ : Joined<["--"], "flto-partitions=">, Group<hip_Group>,
+  HelpText<"Number of partitions to use for parallel full LTO codegen. Use 1 to disable partitioning.">;
 }
 
 // Clang specific/exclusive options for OpenACC.

clang/lib/AST/ByteCode/InterpBuiltin.cpp

@@ -212,11 +212,13 @@ static bool interp__builtin_strcmp(InterpState &S, CodePtr OpPC,
   const Pointer &A = getParam<Pointer>(Frame, 0);
   const Pointer &B = getParam<Pointer>(Frame, 1);
 
-  if (ID == Builtin::BIstrcmp || ID == Builtin::BIstrncmp)
+  if (ID == Builtin::BIstrcmp || ID == Builtin::BIstrncmp ||
+      ID == Builtin::BIwcscmp || ID == Builtin::BIwcsncmp)
     diagnoseNonConstexprBuiltin(S, OpPC, ID);
 
   uint64_t Limit = ~static_cast<uint64_t>(0);
-  if (ID == Builtin::BIstrncmp || ID == Builtin::BI__builtin_strncmp)
+  if (ID == Builtin::BIstrncmp || ID == Builtin::BI__builtin_strncmp ||
+      ID == Builtin::BIwcsncmp || ID == Builtin::BI__builtin_wcsncmp)
     Limit = peekToAPSInt(S.Stk, *S.getContext().classify(Call->getArg(2)))
                 .getZExtValue();
 
@@ -231,6 +233,9 @@ static bool interp__builtin_strcmp(InterpState &S, CodePtr OpPC,
   if (A.isDummy() || B.isDummy())
     return false;
 
+  bool IsWide = ID == Builtin::BIwcscmp || ID == Builtin::BIwcsncmp ||
+                ID == Builtin::BI__builtin_wcscmp ||
+                ID == Builtin::BI__builtin_wcsncmp;
   assert(A.getFieldDesc()->isPrimitiveArray());
   assert(B.getFieldDesc()->isPrimitiveArray());
 
@@ -248,6 +253,21 @@ static bool interp__builtin_strcmp(InterpState &S, CodePtr OpPC,
         !CheckRange(S, OpPC, PB, AK_Read)) {
       return false;
     }
+
+    if (IsWide)
+      INT_TYPE_SWITCH(
+          *S.getContext().classify(S.getASTContext().getWCharType()), {
+            T A = PA.deref<T>();
+            T B = PB.deref<T>();
+            if (A < B) {
+              pushInteger(S, -1, Call->getType());
+              return true;
+            } else if (A > B) {
+              pushInteger(S, 1, Call->getType());
+              return true;
+            }
+          });
+
     uint8_t CA = PA.deref<uint8_t>();
     uint8_t CB = PB.deref<uint8_t>();
 
@@ -1792,6 +1812,17 @@ static bool interp__builtin_memcpy(InterpState &S, CodePtr OpPC,
     return false;
   }
 
+  // Diagnose integral src/dest pointers specially.
+  if (SrcPtr.isIntegralPointer() || DestPtr.isIntegralPointer()) {
+    std::string DiagVal = "(void *)";
+    DiagVal += SrcPtr.isIntegralPointer()
+                   ? std::to_string(SrcPtr.getIntegerRepresentation())
+                   : std::to_string(DestPtr.getIntegerRepresentation());
+    S.FFDiag(S.Current->getSource(OpPC), diag::note_constexpr_memcpy_null)
+        << Move << false << DestPtr.isIntegralPointer() << DiagVal;
+    return false;
+  }
+
   // Can't read from dummy pointers.
   if (DestPtr.isDummy() || SrcPtr.isDummy())
     return false;
@@ -2109,6 +2140,10 @@ bool InterpretBuiltin(InterpState &S, CodePtr OpPC, const Function *F,
   case Builtin::BIstrcmp:
   case Builtin::BI__builtin_strncmp:
   case Builtin::BIstrncmp:
+  case Builtin::BI__builtin_wcsncmp:
+  case Builtin::BIwcsncmp:
+  case Builtin::BI__builtin_wcscmp:
+  case Builtin::BIwcscmp:
     if (!interp__builtin_strcmp(S, OpPC, Frame, F, Call))
       return false;
     break;

clang/lib/Driver/ToolChains/AMDGPU.cpp

@@ -630,8 +630,11 @@ void amdgpu::Linker::ConstructJob(Compilation &C, const JobAction &JA,
   getToolChain().AddFilePathLibArgs(Args, CmdArgs);
   AddLinkerInputs(getToolChain(), Inputs, Args, CmdArgs, JA);
   if (C.getDriver().isUsingLTO()) {
-    addLTOOptions(getToolChain(), Args, CmdArgs, Output, Inputs[0],
-                  C.getDriver().getLTOMode() == LTOK_Thin);
+    const bool ThinLTO = (C.getDriver().getLTOMode() == LTOK_Thin);
+    addLTOOptions(getToolChain(), Args, CmdArgs, Output, Inputs[0], ThinLTO);
+
+    if (!ThinLTO)
+      addFullLTOPartitionOption(C.getDriver(), Args, CmdArgs);
   } else if (Args.hasArg(options::OPT_mcpu_EQ)) {
     CmdArgs.push_back(Args.MakeArgString(
         "-plugin-opt=mcpu=" +
@@ -708,6 +711,33 @@ void amdgpu::getAMDGPUTargetFeatures(const Driver &D,
                             options::OPT_m_amdgpu_Features_Group);
 }
 
+static unsigned getFullLTOPartitions(const Driver &D, const ArgList &Args) {
+  const Arg *A = Args.getLastArg(options::OPT_flto_partitions_EQ);
+  // In the absence of an option, use 8 as the default.
+  if (!A)
+    return 8;
+  int Value = 0;
+  if (StringRef(A->getValue()).getAsInteger(10, Value) || (Value < 1)) {
+    D.Diag(diag::err_drv_invalid_int_value)
+        << A->getAsString(Args) << A->getValue();
+    return 1;
+  }
+
+  return Value;
+}
+
+void amdgpu::addFullLTOPartitionOption(const Driver &D,
+                                       const llvm::opt::ArgList &Args,
+                                       llvm::opt::ArgStringList &CmdArgs) {
+  // TODO: Should this be restricted to fgpu-rdc only ? Currently we'll
+  //       also do it for non gpu-rdc LTO
+
+  if (unsigned NumParts = getFullLTOPartitions(D, Args); NumParts > 1) {
+    CmdArgs.push_back(
+        Args.MakeArgString("--lto-partitions=" + Twine(NumParts)));
+  }
+}
+
 /// AMDGPU Toolchain
 AMDGPUToolChain::AMDGPUToolChain(const Driver &D, const llvm::Triple &Triple,
                                  const ArgList &Args)

clang/lib/Driver/ToolChains/AMDGPU.h

@@ -41,6 +41,8 @@ void getAMDGPUTargetFeatures(const Driver &D, const llvm::Triple &Triple,
                              const llvm::opt::ArgList &Args,
                              std::vector<StringRef> &Features);
 
+void addFullLTOPartitionOption(const Driver &D, const llvm::opt::ArgList &Args,
+                               llvm::opt::ArgStringList &CmdArgs);
 } // end namespace amdgpu
 } // end namespace tools
 

clang/lib/Driver/ToolChains/Clang.cpp

@@ -9337,6 +9337,11 @@ void LinkerWrapper::ConstructJob(Compilation &C, const JobAction &JA,
         CmdArgs.push_back(
             Args.MakeArgString("--device-linker=" + TC.getTripleString() + "=" +
                                "-lclang_rt.builtins"));
+      bool HasFlangRT = HasCompilerRT && C.getDriver().IsFlangMode();
+      if (HasFlangRT)
+        CmdArgs.push_back(
+            Args.MakeArgString("--device-linker=" + TC.getTripleString() + "=" +
+                               "-lflang_rt.runtime"));
     });
   }
 

clang/lib/Driver/ToolChains/HIPAMD.cpp

@@ -116,6 +116,8 @@ void AMDGCN::Linker::constructLldCommand(Compilation &C, const JobAction &JA,
 
   addLinkerCompressDebugSectionsOption(TC, Args, LldArgs);
 
+  amdgpu::addFullLTOPartitionOption(D, Args, LldArgs);
+
   // Given that host and device linking happen in separate processes, the device
   // linker doesn't always have the visibility as to which device symbols are
   // needed by a program, especially for the device symbol dependencies that are

clang/lib/StaticAnalyzer/Checkers/ContainerModeling.cpp

@@ -157,8 +157,11 @@ void ContainerModeling::checkPostCall(const CallEvent &Call,
   if (Func->isOverloadedOperator()) {
     const auto Op = Func->getOverloadedOperator();
     if (Op == OO_Equal) {
-      // Overloaded 'operator=' must be a non-static member function.
-      const auto *InstCall = cast<CXXInstanceCall>(&Call);
+      // Only handle the assignment operator with implicit this
+      const auto *InstCall = dyn_cast<CXXInstanceCall>(&Call);
+      if (!InstCall)
+        return;
+
       if (cast<CXXMethodDecl>(Func)->isMoveAssignmentOperator()) {
         handleAssignment(C, InstCall->getCXXThisVal(), Call.getOriginExpr(),
                      Call.getArgSVal(0));

clang/test/AST/ByteCode/builtin-functions.cpp

@@ -22,6 +22,8 @@ extern "C" {
   extern char *strchr(const char *s, int c);
   extern wchar_t *wmemchr(const wchar_t *s, wchar_t c, size_t n);
   extern wchar_t *wcschr(const wchar_t *s, wchar_t c);
+  extern int wcscmp(const wchar_t *s1, const wchar_t *s2);
+  extern int wcsncmp(const wchar_t *s1, const wchar_t *s2, size_t n);
 }
 
 namespace strcmp {
@@ -66,6 +68,50 @@ namespace strcmp {
   static_assert(__builtin_strncmp("abab\0banana", "abab\0canada", 100) == 0);
 }
 
+namespace WcsCmp {
+  constexpr wchar_t kFoobar[6] = {L'f',L'o',L'o',L'b',L'a',L'r'};
+  constexpr wchar_t kFoobazfoobar[12] = {L'f',L'o',L'o',L'b',L'a',L'z',L'f',L'o',L'o',L'b',L'a',L'r'};
+
+  static_assert(__builtin_wcscmp(L"abab", L"abab") == 0);
+  static_assert(__builtin_wcscmp(L"abab", L"abba") == -1);
+  static_assert(__builtin_wcscmp(L"abab", L"abaa") == 1);
+  static_assert(__builtin_wcscmp(L"ababa", L"abab") == 1);
+  static_assert(__builtin_wcscmp(L"abab", L"ababa") == -1);
+  static_assert(__builtin_wcscmp(L"abab\0banana", L"abab") == 0);
+  static_assert(__builtin_wcscmp(L"abab", L"abab\0banana") == 0);
+  static_assert(__builtin_wcscmp(L"abab\0banana", L"abab\0canada") == 0);
+#if __WCHAR_WIDTH__ == 32
+  static_assert(__builtin_wcscmp(L"a\x83838383", L"a") == (wchar_t)-1U >> 31);
+#endif
+  static_assert(__builtin_wcscmp(0, L"abab") == 0); // both-error {{not an integral constant}} \
+                                                    // both-note {{dereferenced null}}
+  static_assert(__builtin_wcscmp(L"abab", 0) == 0); // both-error {{not an integral constant}} \
+                                                    // both-note {{dereferenced null}}
+
+  static_assert(__builtin_wcscmp(kFoobar, kFoobazfoobar) == -1);
+  static_assert(__builtin_wcscmp(kFoobar, kFoobazfoobar + 6) == 0); // both-error {{not an integral constant}} \
+                                                                    // both-note {{dereferenced one-past-the-end}}
+
+  static_assert(__builtin_wcsncmp(L"abaa", L"abba", 5) == -1);
+  static_assert(__builtin_wcsncmp(L"abaa", L"abba", 4) == -1);
+  static_assert(__builtin_wcsncmp(L"abaa", L"abba", 3) == -1);
+  static_assert(__builtin_wcsncmp(L"abaa", L"abba", 2) == 0);
+  static_assert(__builtin_wcsncmp(L"abaa", L"abba", 1) == 0);
+  static_assert(__builtin_wcsncmp(L"abaa", L"abba", 0) == 0);
+  static_assert(__builtin_wcsncmp(0, 0, 0) == 0);
+  static_assert(__builtin_wcsncmp(L"abab\0banana", L"abab\0canada", 100) == 0);
+#if __WCHAR_WIDTH__ == 32
+  static_assert(__builtin_wcsncmp(L"a\x83838383", L"aa", 2) ==
+                (wchar_t)-1U >> 31);
+#endif
+
+  static_assert(__builtin_wcsncmp(kFoobar, kFoobazfoobar, 6) == -1);
+  static_assert(__builtin_wcsncmp(kFoobar, kFoobazfoobar, 7) == -1);
+  static_assert(__builtin_wcsncmp(kFoobar, kFoobazfoobar + 6, 6) == 0);
+  static_assert(__builtin_wcsncmp(kFoobar, kFoobazfoobar + 6, 7) == 0); // both-error {{not an integral constant}} \
+                                                                        // both-note {{dereferenced one-past-the-end}}
+}
+
 /// Copied from constant-expression-cxx11.cpp
 namespace strlen {
 constexpr const char *a = "foo\0quux";
@@ -1290,6 +1336,12 @@ namespace BuiltinMemcpy {
     return Result1 && Result2;
   }
   static_assert(memmoveOverlapping());
+
+#define fold(x) (__builtin_constant_p(0) ? (x) : (x))
+  static_assert(__builtin_memcpy(&global, fold((wchar_t*)123), sizeof(wchar_t))); // both-error {{not an integral constant expression}} \
+                                                                                  // both-note {{source of 'memcpy' is (void *)123}}
+  static_assert(__builtin_memcpy(fold(reinterpret_cast<wchar_t*>(123)), &global, sizeof(wchar_t))); // both-error {{not an integral constant expression}} \
+                                                                                                    // both-note {{destination of 'memcpy' is (void *)123}}
 }
 
 namespace Memcmp {

clang/test/Analysis/invalidated-iterator.cpp

@@ -1,5 +1,6 @@
 // RUN: %clang_analyze_cc1 -std=c++11 -analyzer-checker=core,cplusplus,alpha.cplusplus.InvalidatedIterator -analyzer-config aggressive-binary-operation-simplification=true -analyzer-config c++-container-inlining=false %s -verify
 // RUN: %clang_analyze_cc1 -std=c++11 -analyzer-checker=core,cplusplus,alpha.cplusplus.InvalidatedIterator -analyzer-config aggressive-binary-operation-simplification=true -analyzer-config c++-container-inlining=true -DINLINE=1 %s -verify
+// RUN: %clang_analyze_cc1 -std=c++23 -analyzer-checker=core,cplusplus,alpha.cplusplus.InvalidatedIterator -analyzer-config aggressive-binary-operation-simplification=true -analyzer-config c++-container-inlining=true -DINLINE=1 %s -verify
 
 #include "Inputs/system-header-simulator-cxx.h"
 
@@ -204,4 +205,26 @@ void invalidated_subscript_end_ptr_iterator(cont_with_ptr_iterator<int> &C) {
   auto i = C.begin();
   C.erase(i);
   (void) i[1]; // expected-warning{{Invalidated iterator accessed}}
-}
+}
+
+#if __cplusplus >= 202302L
+namespace GH116372 {
+  class ExplicitThis {
+    int f = 0;
+  public:
+    ExplicitThis();
+    ExplicitThis(ExplicitThis& other);
+  
+    ExplicitThis& operator=(this ExplicitThis& self, ExplicitThis const& other) { // no crash
+      self.f = other.f;
+      return self;
+    }
+  
+    ~ExplicitThis();
+  };
+  
+  void func(ExplicitThis& obj1) {
+      obj1 = obj1;
+  }
+}
+#endif