misc bug fixes + enable LTO support

build.sh

@@ -77,7 +77,8 @@ build_llvm()
             -DCMAKE_BUILD_TYPE=Release \
             -DCMAKE_INSTALL_PREFIX=install \
             -DBUILD_SHARED_LIBS=ON \
-            -DLLVM_TARGETS_TO_BUILD="X86"
+            -DLLVM_TARGETS_TO_BUILD="X86" \
+            -DLLVM_BINUTILS_INCDIR=/usr/include
     fi
     make -j `nproc` install install-clang
     rm -rf "../$RELEASE_NAME"

llvm-4.0.0.src/include/llvm/ExecutionEngine/Orc/OrcRemoteTargetClient.h

@@ -687,8 +687,8 @@ class OrcRemoteTargetClient : public OrcRemoteTargetRPCAPI {
 
   uint32_t getTrampolineSize() const { return RemoteTrampolineSize; }
 
-  Expected<std::vector<char>> readMem(char *Dst, JITTargetAddress Src,
-                                      uint64_t Size) {
+  Expected<std::vector<uint8_t>> readMem(char *Dst, JITTargetAddress Src,
+                                         uint64_t Size) {
     // Check for an 'out-of-band' error, e.g. from an MM destructor.
     if (ExistingError)
       return std::move(ExistingError);

llvm-4.0.0.src/lib/Transforms/Instrumentation/LowFat.cpp

@@ -540,7 +540,12 @@ static Bounds getPtrBounds(const TargetLibraryInfo *TLI, const DataLayout *DL,
     Bounds bounds = Bounds::nonFat();
     if (GetElementPtrInst *GEP = dyn_cast<GetElementPtrInst>(Ptr))
     {
-        bounds = getPtrBounds(TLI, DL, GEP->getPointerOperand(), boundsInfo);
+        // JumpThreading or other optimizations may produce incorrect IR that
+        // has not yet been removed; e.g.
+        // %231 = getelementptr inbounds i8, i8* %231, i64 1
+        bounds = GEP != GEP->getPointerOperand() ?
+                 getPtrBounds(TLI, DL, GEP->getPointerOperand(), boundsInfo) :
+                 Bounds::nonFat();
         if (!bounds.isUnknown() && !bounds.isNonFat())
         {
             APInt offset(64, 0);

llvm-4.0.0.src/projects/compiler-rt/lib/lowfat/CMakeLists.txt.legacy

@@ -5,7 +5,7 @@ set(LOWFAT_SOURCES
 
 include_directories(..)
 
-set(LOWFAT_CFLAGS -std=gnu99 -m64 -mno-bmi -mno-bmi2 -mno-lzcnt -I. -O2 -mcmodel=large -DLOWFAT_LINUX)
+set(LOWFAT_CFLAGS ${SANITIZER_COMMON_CFLAGS} -std=gnu99 -m64 -mno-bmi -mno-bmi2 -mno-lzcnt -I. -O2 -mcmodel=large -DLOWFAT_LINUX)
 
 add_compiler_rt_runtime(clang_rt.lowfat
         STATIC

llvm-4.0.0.src/projects/compiler-rt/lib/lowfat/CMakeLists.txt.modern

@@ -5,7 +5,7 @@ set(LOWFAT_SOURCES
 
 include_directories(..)
 
-set(LOWFAT_CFLAGS -std=gnu99 -m64 -I. -O2 -mbmi -mbmi2 -mlzcnt -mcmodel=large -DLOWFAT_LINUX)
+set(LOWFAT_CFLAGS ${SANITIZER_COMMON_CFLAGS} -std=gnu99 -m64 -I. -O2 -mbmi -mbmi2 -mlzcnt -mcmodel=large -DLOWFAT_LINUX)
 
 add_compiler_rt_runtime(clang_rt.lowfat
         STATIC

llvm-4.0.0.src/projects/compiler-rt/lib/lowfat/lowfat.c

@@ -27,7 +27,7 @@
 #define LOWFAT_PAGE_SIZE        4096
 #define LOWFAT_MAX_ADDRESS      0x1000000000000ull
 
-#define LOWFAT_CONSTRUCTOR      __attribute__((__constructor__(10102)))
+#define LOWFAT_CONSTRUCTOR      __attribute__((__constructor__(0)))
 #define LOWFAT_DESTRUCTOR       __attribute__((__destructor__(10102)))
 #define LOWFAT_NOINLINE         __attribute__((__noinline__))
 #define LOWFAT_NORETURN         __attribute__((__noreturn__))
@@ -55,9 +55,6 @@ static LOWFAT_NOINLINE void lowfat_warning(const char *format, ...);
 static LOWFAT_DATA uint8_t *lowfat_seed = NULL;
 static LOWFAT_DATA size_t lowfat_seed_pos = LOWFAT_PAGE_SIZE;
 static LOWFAT_DATA bool lowfat_malloc_inited = false;
-#ifndef LOWFAT_WINDOWS
-static LOWFAT_DATA char **lowfat_envp = NULL;
-#endif
 
 #ifdef LOWFAT_WINDOWS
 #include "lowfat_windows.c"
@@ -527,10 +524,7 @@ extern LOWFAT_NOINLINE void *lowfat_stack_pivot_2(void *stack_top)
     // that the environment is stored at the base of the stack, which is
     // currently true for Linux.  It is also necessary to copy the argv/env
     // since we patch the stack.
-    if (lowfat_envp == NULL)
-        lowfat_error("failed to get the environment pointer");
-    char **envp = lowfat_envp;
-    lowfat_envp = NULL;
+    char **envp = __environ;
     uint8_t *stack_bottom = (void *)envp;
     while (*envp != NULL)
     {
@@ -579,26 +573,12 @@ __asm__ (
     "\t.type lowfat_stack_pivot,@function\n"
     "lowfat_stack_pivot:\n"
     "\tmovq %rsp, %rdi\n"
+    "\tsub $8, %rsp\n"
     "\tmovabsq $lowfat_stack_pivot_2, %rax\n"
     "\tcallq *%rax\n"
     "\tmovq %rax, %rsp\n"
     "\tretq\n"
 );
 
-/*
- * This bit of magic ensures lowfat_init() is called very early in process
- * startup.  Using the "constructor" attribute is not good enough since shared
- * object constructors/initializers may be called before lowfat_init().  We
- * also save the environment pointer so we can later derive the stack base.
- */
-void lowfat_preinit(int argc, char **argv, char **envp)
-{
-	lowfat_envp = envp;
-	lowfat_init();
-}
-__attribute__((section(".preinit_array"), used))
-void (*__local_effective_preinit)(int argc, char **argv, char **envp) =
-	lowfat_preinit;
-
 #endif
 

llvm-4.0.0.src/projects/compiler-rt/lib/lowfat/lowfat_linux.c

@@ -61,6 +61,7 @@ static void lowfat_random_page(void *buf)
 			path);
 }
 
+#ifdef __GLIBC__
 #include <execinfo.h>
 static LOWFAT_NOINLINE void lowfat_backtrace(void)
 {
@@ -73,6 +74,9 @@ static LOWFAT_NOINLINE void lowfat_backtrace(void)
 	if (len == 0 || len == sizeof(trace) / sizeof(void *))
 	    fprintf(stderr, "...\n");
 }
+#else
+static LOWFAT_NOINLINE void lowfat_backtrace(void) { }
+#endif /* __GLIBC__ */
 
 /*
  * Open a unique+anonymous shared memory object.

llvm-4.0.0.src/projects/compiler-rt/lib/sanitizer_common/sanitizer_platform_limits_posix.cc

@@ -160,7 +160,6 @@ typedef struct user_fpregs elf_fpregset_t;
 # include <sys/procfs.h>
 #endif
 #include <sys/user.h>
-#include <sys/ustat.h>
 #include <linux/cyclades.h>
 #include <linux/if_eql.h>
 #include <linux/if_plip.h>
@@ -253,7 +252,19 @@ namespace __sanitizer {
 #endif // SANITIZER_LINUX || SANITIZER_FREEBSD
 
 #if SANITIZER_LINUX && !SANITIZER_ANDROID
-  unsigned struct_ustat_sz = sizeof(struct ustat);
+  // Use pre-computed size of struct ustat to avoid <sys/ustat.h> which
+  // has been removed from glibc 2.28.
+#if defined(__aarch64__) || defined(__s390x__) || defined (__mips64) \
+  || defined(__powerpc64__) || defined(__arch64__) || defined(__sparcv9) \
+  || defined(__x86_64__)
+#define SIZEOF_STRUCT_USTAT 32
+#elif defined(__arm__) || defined(__i386__) || defined(__mips__) \
+  || defined(__powerpc__) || defined(__s390__)
+#define SIZEOF_STRUCT_USTAT 20
+#else
+#error Unknown size of struct ustat
+#endif
+  unsigned struct_ustat_sz = SIZEOF_STRUCT_USTAT;
   unsigned struct_rlimit64_sz = sizeof(struct rlimit64);
   unsigned struct_statvfs64_sz = sizeof(struct statvfs64);
 #endif // SANITIZER_LINUX && !SANITIZER_ANDROID
@@ -1147,8 +1158,9 @@ CHECK_SIZE_AND_OFFSET(ipc_perm, uid);
 CHECK_SIZE_AND_OFFSET(ipc_perm, gid);
 CHECK_SIZE_AND_OFFSET(ipc_perm, cuid);
 CHECK_SIZE_AND_OFFSET(ipc_perm, cgid);
-#if !defined(__aarch64__) || !SANITIZER_LINUX || __GLIBC_PREREQ (2, 21)
-/* On aarch64 glibc 2.20 and earlier provided incorrect mode field.  */
+#if !SANITIZER_LINUX || __GLIBC_PREREQ (2, 31)
+/* glibc 2.30 and earlier provided 16-bit mode field instead of 32-bit
+   on many architectures.  */
 CHECK_SIZE_AND_OFFSET(ipc_perm, mode);
 #endif
 

llvm-4.0.0.src/projects/compiler-rt/lib/sanitizer_common/sanitizer_platform_limits_posix.h

@@ -225,8 +225,7 @@ namespace __sanitizer {
     unsigned long __unused1;
     unsigned long __unused2;
 #else
-    unsigned short mode;
-    unsigned short __pad1;
+    unsigned int mode;
     unsigned short __seq;
     unsigned short __pad2;
 #if defined(__x86_64__) && !defined(_LP64)

llvm-4.0.0.src/tools/clang/lib/AST/CXXInheritance.cpp

@@ -88,7 +88,7 @@ bool CXXRecordDecl::isDerivedFrom(const CXXRecordDecl *Base,
   const CXXRecordDecl *BaseDecl = Base->getCanonicalDecl();
   // FIXME: Capturing 'this' is a workaround for name lookup bugs in GCC 4.7.
   return lookupInBases(
-      [this, BaseDecl](const CXXBaseSpecifier *Specifier, CXXBasePath &Path) {
+      [BaseDecl](const CXXBaseSpecifier *Specifier, CXXBasePath &Path) {
         return FindBaseClass(Specifier, Path, BaseDecl);
       },
       Paths);
@@ -109,7 +109,7 @@ bool CXXRecordDecl::isVirtuallyDerivedFrom(const CXXRecordDecl *Base) const {
   const CXXRecordDecl *BaseDecl = Base->getCanonicalDecl();
   // FIXME: Capturing 'this' is a workaround for name lookup bugs in GCC 4.7.
   return lookupInBases(
-      [this, BaseDecl](const CXXBaseSpecifier *Specifier, CXXBasePath &Path) {
+      [BaseDecl](const CXXBaseSpecifier *Specifier, CXXBasePath &Path) {
         return FindVirtualBaseClass(Specifier, Path, BaseDecl);
       },
       Paths);

llvm-4.0.0.src/tools/clang/lib/AST/MicrosoftMangle.cpp

@@ -2997,14 +2997,14 @@ void MicrosoftMangleContextImpl::mangleStringLiteral(const StringLiteral *SL,
   // N.B. The length is in terms of bytes, not characters.
   Mangler.mangleNumber(SL->getByteLength() + SL->getCharByteWidth());
 
-  auto GetLittleEndianByte = [&Mangler, &SL](unsigned Index) {
+  auto GetLittleEndianByte = [&SL](unsigned Index) {
     unsigned CharByteWidth = SL->getCharByteWidth();
     uint32_t CodeUnit = SL->getCodeUnit(Index / CharByteWidth);
     unsigned OffsetInCodeUnit = Index % CharByteWidth;
     return static_cast<char>((CodeUnit >> (8 * OffsetInCodeUnit)) & 0xff);
   };
 
-  auto GetBigEndianByte = [&Mangler, &SL](unsigned Index) {
+  auto GetBigEndianByte = [&SL](unsigned Index) {
     unsigned CharByteWidth = SL->getCharByteWidth();
     uint32_t CodeUnit = SL->getCodeUnit(Index / CharByteWidth);
     unsigned OffsetInCodeUnit = (CharByteWidth - 1) - (Index % CharByteWidth);

llvm-4.0.0.src/tools/clang/lib/CodeGen/CGOpenMPRuntime.cpp

@@ -4006,8 +4006,8 @@ void CGOpenMPRuntime::emitTaskCall(CodeGenFunction &CGF, SourceLocation Loc,
     DepTaskArgs[5] = CGF.Builder.getInt32(0);
     DepTaskArgs[6] = llvm::ConstantPointerNull::get(CGF.VoidPtrTy);
   }
-  auto &&ThenCodeGen = [this, Loc, &Data, TDBase, KmpTaskTQTyRD,
-                        NumDependencies, &TaskArgs,
+  auto &&ThenCodeGen = [this, &Data, TDBase, KmpTaskTQTyRD, NumDependencies,
+                        &TaskArgs,
                         &DepTaskArgs](CodeGenFunction &CGF, PrePostActionTy &) {
     if (!Data.Tied) {
       auto PartIdFI = std::next(KmpTaskTQTyRD->field_begin(), KmpTaskTPartId);
@@ -4562,7 +4562,7 @@ void CGOpenMPRuntime::emitReduction(CodeGenFunction &CGF, SourceLocation Loc,
       }
       if (XExpr) {
         auto *VD = cast<VarDecl>(cast<DeclRefExpr>(*ILHS)->getDecl());
-        auto &&AtomicRedGen = [BO, VD, IPriv,
+        auto &&AtomicRedGen = [BO, VD,
                                Loc](CodeGenFunction &CGF, const Expr *XExpr,
                                     const Expr *EExpr, const Expr *UpExpr) {
           LValue X = CGF.EmitLValue(XExpr);
@@ -4572,7 +4572,7 @@ void CGOpenMPRuntime::emitReduction(CodeGenFunction &CGF, SourceLocation Loc,
           CGF.EmitOMPAtomicSimpleUpdateExpr(
               X, E, BO, /*IsXLHSInRHSPart=*/true,
               llvm::AtomicOrdering::Monotonic, Loc,
-              [&CGF, UpExpr, VD, IPriv, Loc](RValue XRValue) {
+              [&CGF, UpExpr, VD, Loc](RValue XRValue) {
                 CodeGenFunction::OMPPrivateScope PrivateScope(CGF);
                 PrivateScope.addPrivate(
                     VD, [&CGF, VD, XRValue, Loc]() -> Address {
@@ -5984,8 +5984,8 @@ void CGOpenMPRuntime::emitTargetCall(CodeGenFunction &CGF,
                         OffloadError);
 
   // Fill up the pointer arrays and transfer execution to the device.
-  auto &&ThenGen = [&Ctx, &BasePointers, &Pointers, &Sizes, &MapTypes, Device,
-                    OutlinedFnID, OffloadError, OffloadErrorQType,
+  auto &&ThenGen = [&BasePointers, &Pointers, &Sizes, &MapTypes, Device,
+                    OutlinedFnID, OffloadError,
                     &D](CodeGenFunction &CGF, PrePostActionTy &) {
     auto &RT = CGF.CGM.getOpenMPRuntime();
     // Emit the offloading arrays.
@@ -6271,8 +6271,8 @@ void CGOpenMPRuntime::emitTargetDataCalls(
   // Generate the code for the opening of the data environment. Capture all the
   // arguments of the runtime call by reference because they are used in the
   // closing of the region.
-  auto &&BeginThenGen = [&D, &CGF, Device, &Info, &CodeGen, &NoPrivAction](
-      CodeGenFunction &CGF, PrePostActionTy &) {
+  auto &&BeginThenGen = [&D, Device, &Info, &CodeGen](CodeGenFunction &CGF,
+                                                      PrePostActionTy &) {
     // Fill up the arrays with all the mapped variables.
     MappableExprsHandler::MapBaseValuesArrayTy BasePointers;
     MappableExprsHandler::MapValuesArrayTy Pointers;
@@ -6318,8 +6318,7 @@ void CGOpenMPRuntime::emitTargetDataCalls(
   };
 
   // Generate code for the closing of the data region.
-  auto &&EndThenGen = [&CGF, Device, &Info](CodeGenFunction &CGF,
-                                            PrePostActionTy &) {
+  auto &&EndThenGen = [Device, &Info](CodeGenFunction &CGF, PrePostActionTy &) {
     assert(Info.isValid() && "Invalid data environment closing arguments.");
 
     llvm::Value *BasePointersArrayArg = nullptr;
@@ -6397,7 +6396,7 @@ void CGOpenMPRuntime::emitTargetDataStandAloneCall(
          "Expecting either target enter, exit data, or update directives.");
 
   // Generate the code for the opening of the data environment.
-  auto &&ThenGen = [&D, &CGF, Device](CodeGenFunction &CGF, PrePostActionTy &) {
+  auto &&ThenGen = [&D, Device](CodeGenFunction &CGF, PrePostActionTy &) {
     // Fill up the arrays with all the mapped variables.
     MappableExprsHandler::MapBaseValuesArrayTy BasePointers;
     MappableExprsHandler::MapValuesArrayTy Pointers;

llvm-4.0.0.src/tools/clang/lib/CodeGen/CGOpenMPRuntimeNVPTX.cpp

@@ -533,8 +533,7 @@ void CGOpenMPRuntimeNVPTX::emitGenericParallelCall(
     ArrayRef<llvm::Value *> CapturedVars, const Expr *IfCond) {
   llvm::Function *Fn = cast<llvm::Function>(OutlinedFn);
 
-  auto &&L0ParallelGen = [this, Fn, &CapturedVars](CodeGenFunction &CGF,
-                                                   PrePostActionTy &) {
+  auto &&L0ParallelGen = [this, Fn](CodeGenFunction &CGF, PrePostActionTy &) {
     CGBuilderTy &Bld = CGF.Builder;
 
     // Prepare for parallel region. Indicate the outlined function.
@@ -565,8 +564,8 @@ void CGOpenMPRuntimeNVPTX::emitGenericParallelCall(
 
   auto &&SeqGen = [this, Fn, &CapturedVars, &Args](CodeGenFunction &CGF,
                                                    PrePostActionTy &) {
-    auto &&CodeGen = [this, Fn, &CapturedVars, &Args](CodeGenFunction &CGF,
-                                                      PrePostActionTy &Action) {
+    auto &&CodeGen = [this, Fn, &CapturedVars](CodeGenFunction &CGF,
+                                               PrePostActionTy &Action) {
       Action.Enter(CGF);
 
       llvm::SmallVector<llvm::Value *, 16> OutlinedFnArgs;

llvm-4.0.0.src/tools/clang/lib/CodeGen/CGStmtOpenMP.cpp

@@ -986,7 +986,7 @@ void CodeGenFunction::EmitOMPReductionClauseInit(
                         OriginalBaseLValue);
         // Store the address of the original variable associated with the LHS
         // implicit variable.
-        PrivateScope.addPrivate(LHSVD, [this, OASELValueLB]() -> Address {
+        PrivateScope.addPrivate(LHSVD, [OASELValueLB]() -> Address {
           return OASELValueLB.getAddress();
         });
         // Emit reduction copy.
@@ -1040,9 +1040,8 @@ void CodeGenFunction::EmitOMPReductionClauseInit(
             *this, OrigVD->getType(), ASELValue.getType(), OriginalBaseLValue);
         // Store the address of the original variable associated with the LHS
         // implicit variable.
-        PrivateScope.addPrivate(LHSVD, [this, ASELValue]() -> Address {
-          return ASELValue.getAddress();
-        });
+        PrivateScope.addPrivate(
+            LHSVD, [ASELValue]() -> Address { return ASELValue.getAddress(); });
         // Emit reduction copy.
         bool IsRegistered = PrivateScope.addPrivate(
             OrigVD, [this, OrigVD, PrivateVD, BaseLValue, ASELValue,
@@ -2633,7 +2632,7 @@ void CodeGenFunction::EmitOMPTaskBasedDirective(const OMPExecutableDirective &S,
   for (const auto *C : S.getClausesOfKind<OMPDependClause>())
     for (auto *IRef : C->varlists())
       Data.Dependences.push_back(std::make_pair(C->getDependencyKind(), IRef));
-  auto &&CodeGen = [PartId, &S, &Data, CS, &BodyGen, &LastprivateDstsOrigs](
+  auto &&CodeGen = [&Data, CS, &BodyGen, &LastprivateDstsOrigs](
       CodeGenFunction &CGF, PrePostActionTy &Action) {
     // Set proper addresses for generated private copies.
     OMPPrivateScope Scope(CGF);
@@ -3250,7 +3249,7 @@ static void EmitOMPAtomicCaptureExpr(CodeGenFunction &CGF, bool IsSeqCst,
     NewVValType = XRValExpr->getType();
     auto *ERValExpr = IsXLHSInRHSPart ? RHS : LHS;
     auto &&Gen = [&CGF, &NewVVal, UE, ExprRValue, XRValExpr, ERValExpr,
-                  IsSeqCst, IsPostfixUpdate](RValue XRValue) -> RValue {
+                  IsPostfixUpdate](RValue XRValue) -> RValue {
       CodeGenFunction::OpaqueValueMapping MapExpr(CGF, ERValExpr, ExprRValue);
       CodeGenFunction::OpaqueValueMapping MapX(CGF, XRValExpr, XRValue);
       RValue Res = CGF.EmitAnyExpr(UE);
@@ -3277,7 +3276,7 @@ static void EmitOMPAtomicCaptureExpr(CodeGenFunction &CGF, bool IsSeqCst,
     NewVValType = X->getType().getNonReferenceType();
     ExprRValue = convertToType(CGF, ExprRValue, E->getType(),
                                X->getType().getNonReferenceType(), Loc);
-    auto &&Gen = [&CGF, &NewVVal, ExprRValue](RValue XRValue) -> RValue {
+    auto &&Gen = [&NewVVal, ExprRValue](RValue XRValue) -> RValue {
       NewVVal = XRValue;
       return ExprRValue;
     };