HELLO·Android
系统源代码
IT资讯
技术文章
我的收藏
注册
登录
-
我收藏的文章
创建代码块
我的代码块
我的账号
Oreo
|
8.0.0_r4
下载
查看原文件
收藏
根目录
external
clang
lib
Sema
SemaChecking.cpp
//===--- SemaChecking.cpp - Extra Semantic Checking -----------------------===// // // The LLVM Compiler Infrastructure // // This file is distributed under the University of Illinois Open Source // License. See LICENSE.TXT for details. // //===----------------------------------------------------------------------===// // // This file implements extra semantic analysis beyond what is enforced // by the C type system. // //===----------------------------------------------------------------------===// #include "clang/Sema/SemaInternal.h" #include "clang/AST/ASTContext.h" #include "clang/AST/CharUnits.h" #include "clang/AST/DeclCXX.h" #include "clang/AST/DeclObjC.h" #include "clang/AST/EvaluatedExprVisitor.h" #include "clang/AST/Expr.h" #include "clang/AST/ExprCXX.h" #include "clang/AST/ExprObjC.h" #include "clang/AST/ExprOpenMP.h" #include "clang/AST/StmtCXX.h" #include "clang/AST/StmtObjC.h" #include "clang/Analysis/Analyses/FormatString.h" #include "clang/Basic/CharInfo.h" #include "clang/Basic/TargetBuiltins.h" #include "clang/Basic/TargetInfo.h" #include "clang/Lex/Lexer.h" // TODO: Extract static functions to fix layering. #include "clang/Sema/Initialization.h" #include "clang/Sema/Lookup.h" #include "clang/Sema/ScopeInfo.h" #include "clang/Sema/Sema.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/SmallBitVector.h" #include "llvm/ADT/SmallString.h" #include "llvm/Support/Format.h" #include "llvm/Support/Locale.h" #include "llvm/Support/ConvertUTF.h" #include "llvm/Support/raw_ostream.h" #include
using namespace clang; using namespace sema; SourceLocation Sema::getLocationOfStringLiteralByte(const StringLiteral *SL, unsigned ByteNo) const { return SL->getLocationOfByte(ByteNo, getSourceManager(), LangOpts, Context.getTargetInfo()); } /// Checks that a call expression's argument count is the desired number. /// This is useful when doing custom type-checking. Returns true on error. static bool checkArgCount(Sema &S, CallExpr *call, unsigned desiredArgCount) { unsigned argCount = call->getNumArgs(); if (argCount == desiredArgCount) return false; if (argCount < desiredArgCount) return S.Diag(call->getLocEnd(), diag::err_typecheck_call_too_few_args) << 0 /*function call*/ << desiredArgCount << argCount << call->getSourceRange(); // Highlight all the excess arguments. SourceRange range(call->getArg(desiredArgCount)->getLocStart(), call->getArg(argCount - 1)->getLocEnd()); return S.Diag(range.getBegin(), diag::err_typecheck_call_too_many_args) << 0 /*function call*/ << desiredArgCount << argCount << call->getArg(1)->getSourceRange(); } /// Check that the first argument to __builtin_annotation is an integer /// and the second argument is a non-wide string literal. static bool SemaBuiltinAnnotation(Sema &S, CallExpr *TheCall) { if (checkArgCount(S, TheCall, 2)) return true; // First argument should be an integer. Expr *ValArg = TheCall->getArg(0); QualType Ty = ValArg->getType(); if (!Ty->isIntegerType()) { S.Diag(ValArg->getLocStart(), diag::err_builtin_annotation_first_arg) << ValArg->getSourceRange(); return true; } // Second argument should be a constant string. Expr *StrArg = TheCall->getArg(1)->IgnoreParenCasts(); StringLiteral *Literal = dyn_cast
(StrArg); if (!Literal || !Literal->isAscii()) { S.Diag(StrArg->getLocStart(), diag::err_builtin_annotation_second_arg) << StrArg->getSourceRange(); return true; } TheCall->setType(Ty); return false; } /// Check that the argument to __builtin_addressof is a glvalue, and set the /// result type to the corresponding pointer type. static bool SemaBuiltinAddressof(Sema &S, CallExpr *TheCall) { if (checkArgCount(S, TheCall, 1)) return true; ExprResult Arg(TheCall->getArg(0)); QualType ResultType = S.CheckAddressOfOperand(Arg, TheCall->getLocStart()); if (ResultType.isNull()) return true; TheCall->setArg(0, Arg.get()); TheCall->setType(ResultType); return false; } static bool SemaBuiltinOverflow(Sema &S, CallExpr *TheCall) { if (checkArgCount(S, TheCall, 3)) return true; // First two arguments should be integers. for (unsigned I = 0; I < 2; ++I) { Expr *Arg = TheCall->getArg(I); QualType Ty = Arg->getType(); if (!Ty->isIntegerType()) { S.Diag(Arg->getLocStart(), diag::err_overflow_builtin_must_be_int) << Ty << Arg->getSourceRange(); return true; } } // Third argument should be a pointer to a non-const integer. // IRGen correctly handles volatile, restrict, and address spaces, and // the other qualifiers aren't possible. { Expr *Arg = TheCall->getArg(2); QualType Ty = Arg->getType(); const auto *PtrTy = Ty->getAs
(); if (!(PtrTy && PtrTy->getPointeeType()->isIntegerType() && !PtrTy->getPointeeType().isConstQualified())) { S.Diag(Arg->getLocStart(), diag::err_overflow_builtin_must_be_ptr_int) << Ty << Arg->getSourceRange(); return true; } } return false; } static void SemaBuiltinMemChkCall(Sema &S, FunctionDecl *FDecl, CallExpr *TheCall, unsigned SizeIdx, unsigned DstSizeIdx) { if (TheCall->getNumArgs() <= SizeIdx || TheCall->getNumArgs() <= DstSizeIdx) return; const Expr *SizeArg = TheCall->getArg(SizeIdx); const Expr *DstSizeArg = TheCall->getArg(DstSizeIdx); llvm::APSInt Size, DstSize; // find out if both sizes are known at compile time if (!SizeArg->EvaluateAsInt(Size, S.Context) || !DstSizeArg->EvaluateAsInt(DstSize, S.Context)) return; if (Size.ule(DstSize)) return; // confirmed overflow so generate the diagnostic. IdentifierInfo *FnName = FDecl->getIdentifier(); SourceLocation SL = TheCall->getLocStart(); SourceRange SR = TheCall->getSourceRange(); S.Diag(SL, diag::warn_memcpy_chk_overflow) << SR << FnName; } static bool SemaBuiltinCallWithStaticChain(Sema &S, CallExpr *BuiltinCall) { if (checkArgCount(S, BuiltinCall, 2)) return true; SourceLocation BuiltinLoc = BuiltinCall->getLocStart(); Expr *Builtin = BuiltinCall->getCallee()->IgnoreImpCasts(); Expr *Call = BuiltinCall->getArg(0); Expr *Chain = BuiltinCall->getArg(1); if (Call->getStmtClass() != Stmt::CallExprClass) { S.Diag(BuiltinLoc, diag::err_first_argument_to_cwsc_not_call) << Call->getSourceRange(); return true; } auto CE = cast
(Call); if (CE->getCallee()->getType()->isBlockPointerType()) { S.Diag(BuiltinLoc, diag::err_first_argument_to_cwsc_block_call) << Call->getSourceRange(); return true; } const Decl *TargetDecl = CE->getCalleeDecl(); if (const FunctionDecl *FD = dyn_cast_or_null
(TargetDecl)) if (FD->getBuiltinID()) { S.Diag(BuiltinLoc, diag::err_first_argument_to_cwsc_builtin_call) << Call->getSourceRange(); return true; } if (isa
(CE->getCallee()->IgnoreParens())) { S.Diag(BuiltinLoc, diag::err_first_argument_to_cwsc_pdtor_call) << Call->getSourceRange(); return true; } ExprResult ChainResult = S.UsualUnaryConversions(Chain); if (ChainResult.isInvalid()) return true; if (!ChainResult.get()->getType()->isPointerType()) { S.Diag(BuiltinLoc, diag::err_second_argument_to_cwsc_not_pointer) << Chain->getSourceRange(); return true; } QualType ReturnTy = CE->getCallReturnType(S.Context); QualType ArgTys[2] = { ReturnTy, ChainResult.get()->getType() }; QualType BuiltinTy = S.Context.getFunctionType( ReturnTy, ArgTys, FunctionProtoType::ExtProtoInfo()); QualType BuiltinPtrTy = S.Context.getPointerType(BuiltinTy); Builtin = S.ImpCastExprToType(Builtin, BuiltinPtrTy, CK_BuiltinFnToFnPtr).get(); BuiltinCall->setType(CE->getType()); BuiltinCall->setValueKind(CE->getValueKind()); BuiltinCall->setObjectKind(CE->getObjectKind()); BuiltinCall->setCallee(Builtin); BuiltinCall->setArg(1, ChainResult.get()); return false; } static bool SemaBuiltinSEHScopeCheck(Sema &SemaRef, CallExpr *TheCall, Scope::ScopeFlags NeededScopeFlags, unsigned DiagID) { // Scopes aren't available during instantiation. Fortunately, builtin // functions cannot be template args so they cannot be formed through template // instantiation. Therefore checking once during the parse is sufficient. if (!SemaRef.ActiveTemplateInstantiations.empty()) return false; Scope *S = SemaRef.getCurScope(); while (S && !S->isSEHExceptScope()) S = S->getParent(); if (!S || !(S->getFlags() & NeededScopeFlags)) { auto *DRE = cast
(TheCall->getCallee()->IgnoreParenCasts()); SemaRef.Diag(TheCall->getExprLoc(), DiagID) << DRE->getDecl()->getIdentifier(); return true; } return false; } static inline bool isBlockPointer(Expr *Arg) { return Arg->getType()->isBlockPointerType(); } /// OpenCL C v2.0, s6.13.17.2 - Checks that the block parameters are all local /// void*, which is a requirement of device side enqueue. static bool checkOpenCLBlockArgs(Sema &S, Expr *BlockArg) { const BlockPointerType *BPT = cast
(BlockArg->getType().getCanonicalType()); ArrayRef
Params = BPT->getPointeeType()->getAs
()->getParamTypes(); unsigned ArgCounter = 0; bool IllegalParams = false; // Iterate through the block parameters until either one is found that is not // a local void*, or the block is valid. for (ArrayRef
::iterator I = Params.begin(), E = Params.end(); I != E; ++I, ++ArgCounter) { if (!(*I)->isPointerType() || !(*I)->getPointeeType()->isVoidType() || (*I)->getPointeeType().getQualifiers().getAddressSpace() != LangAS::opencl_local) { // Get the location of the error. If a block literal has been passed // (BlockExpr) then we can point straight to the offending argument, // else we just point to the variable reference. SourceLocation ErrorLoc; if (isa
(BlockArg)) { BlockDecl *BD = cast
(BlockArg)->getBlockDecl(); ErrorLoc = BD->getParamDecl(ArgCounter)->getLocStart(); } else if (isa
(BlockArg)) { ErrorLoc = cast
(BlockArg)->getLocStart(); } S.Diag(ErrorLoc, diag::err_opencl_enqueue_kernel_blocks_non_local_void_args); IllegalParams = true; } } return IllegalParams; } /// OpenCL C v2.0, s6.13.17.6 - Check the argument to the /// get_kernel_work_group_size /// and get_kernel_preferred_work_group_size_multiple builtin functions. static bool SemaOpenCLBuiltinKernelWorkGroupSize(Sema &S, CallExpr *TheCall) { if (checkArgCount(S, TheCall, 1)) return true; Expr *BlockArg = TheCall->getArg(0); if (!isBlockPointer(BlockArg)) { S.Diag(BlockArg->getLocStart(), diag::err_opencl_enqueue_kernel_expected_type) << "block"; return true; } return checkOpenCLBlockArgs(S, BlockArg); } static bool checkOpenCLEnqueueLocalSizeArgs(Sema &S, CallExpr *TheCall, unsigned Start, unsigned End); /// OpenCL v2.0, s6.13.17.1 - Check that sizes are provided for all /// 'local void*' parameter of passed block. static bool checkOpenCLEnqueueVariadicArgs(Sema &S, CallExpr *TheCall, Expr *BlockArg, unsigned NumNonVarArgs) { const BlockPointerType *BPT = cast
(BlockArg->getType().getCanonicalType()); unsigned NumBlockParams = BPT->getPointeeType()->getAs
()->getNumParams(); unsigned TotalNumArgs = TheCall->getNumArgs(); // For each argument passed to the block, a corresponding uint needs to // be passed to describe the size of the local memory. if (TotalNumArgs != NumBlockParams + NumNonVarArgs) { S.Diag(TheCall->getLocStart(), diag::err_opencl_enqueue_kernel_local_size_args); return true; } // Check that the sizes of the local memory are specified by integers. return checkOpenCLEnqueueLocalSizeArgs(S, TheCall, NumNonVarArgs, TotalNumArgs - 1); } /// OpenCL C v2.0, s6.13.17 - Enqueue kernel function contains four different /// overload formats specified in Table 6.13.17.1. /// int enqueue_kernel(queue_t queue, /// kernel_enqueue_flags_t flags, /// const ndrange_t ndrange, /// void (^block)(void)) /// int enqueue_kernel(queue_t queue, /// kernel_enqueue_flags_t flags, /// const ndrange_t ndrange, /// uint num_events_in_wait_list, /// clk_event_t *event_wait_list, /// clk_event_t *event_ret, /// void (^block)(void)) /// int enqueue_kernel(queue_t queue, /// kernel_enqueue_flags_t flags, /// const ndrange_t ndrange, /// void (^block)(local void*, ...), /// uint size0, ...) /// int enqueue_kernel(queue_t queue, /// kernel_enqueue_flags_t flags, /// const ndrange_t ndrange, /// uint num_events_in_wait_list, /// clk_event_t *event_wait_list, /// clk_event_t *event_ret, /// void (^block)(local void*, ...), /// uint size0, ...) static bool SemaOpenCLBuiltinEnqueueKernel(Sema &S, CallExpr *TheCall) { unsigned NumArgs = TheCall->getNumArgs(); if (NumArgs < 4) { S.Diag(TheCall->getLocStart(), diag::err_typecheck_call_too_few_args); return true; } Expr *Arg0 = TheCall->getArg(0); Expr *Arg1 = TheCall->getArg(1); Expr *Arg2 = TheCall->getArg(2); Expr *Arg3 = TheCall->getArg(3); // First argument always needs to be a queue_t type. if (!Arg0->getType()->isQueueT()) { S.Diag(TheCall->getArg(0)->getLocStart(), diag::err_opencl_enqueue_kernel_expected_type) << S.Context.OCLQueueTy; return true; } // Second argument always needs to be a kernel_enqueue_flags_t enum value. if (!Arg1->getType()->isIntegerType()) { S.Diag(TheCall->getArg(1)->getLocStart(), diag::err_opencl_enqueue_kernel_expected_type) << "'kernel_enqueue_flags_t' (i.e. uint)"; return true; } // Third argument is always an ndrange_t type. if (!Arg2->getType()->isNDRangeT()) { S.Diag(TheCall->getArg(2)->getLocStart(), diag::err_opencl_enqueue_kernel_expected_type) << S.Context.OCLNDRangeTy; return true; } // With four arguments, there is only one form that the function could be // called in: no events and no variable arguments. if (NumArgs == 4) { // check that the last argument is the right block type. if (!isBlockPointer(Arg3)) { S.Diag(Arg3->getLocStart(), diag::err_opencl_enqueue_kernel_expected_type) << "block"; return true; } // we have a block type, check the prototype const BlockPointerType *BPT = cast
(Arg3->getType().getCanonicalType()); if (BPT->getPointeeType()->getAs
()->getNumParams() > 0) { S.Diag(Arg3->getLocStart(), diag::err_opencl_enqueue_kernel_blocks_no_args); return true; } return false; } // we can have block + varargs. if (isBlockPointer(Arg3)) return (checkOpenCLBlockArgs(S, Arg3) || checkOpenCLEnqueueVariadicArgs(S, TheCall, Arg3, 4)); // last two cases with either exactly 7 args or 7 args and varargs. if (NumArgs >= 7) { // check common block argument. Expr *Arg6 = TheCall->getArg(6); if (!isBlockPointer(Arg6)) { S.Diag(Arg6->getLocStart(), diag::err_opencl_enqueue_kernel_expected_type) << "block"; return true; } if (checkOpenCLBlockArgs(S, Arg6)) return true; // Forth argument has to be any integer type. if (!Arg3->getType()->isIntegerType()) { S.Diag(TheCall->getArg(3)->getLocStart(), diag::err_opencl_enqueue_kernel_expected_type) << "integer"; return true; } // check remaining common arguments. Expr *Arg4 = TheCall->getArg(4); Expr *Arg5 = TheCall->getArg(5); // Fith argument is always passed as pointers to clk_event_t. if (!Arg4->getType()->getPointeeOrArrayElementType()->isClkEventT()) { S.Diag(TheCall->getArg(4)->getLocStart(), diag::err_opencl_enqueue_kernel_expected_type) << S.Context.getPointerType(S.Context.OCLClkEventTy); return true; } // Sixth argument is always passed as pointers to clk_event_t. if (!(Arg5->getType()->isPointerType() && Arg5->getType()->getPointeeType()->isClkEventT())) { S.Diag(TheCall->getArg(5)->getLocStart(), diag::err_opencl_enqueue_kernel_expected_type) << S.Context.getPointerType(S.Context.OCLClkEventTy); return true; } if (NumArgs == 7) return false; return checkOpenCLEnqueueVariadicArgs(S, TheCall, Arg6, 7); } // None of the specific case has been detected, give generic error S.Diag(TheCall->getLocStart(), diag::err_opencl_enqueue_kernel_incorrect_args); return true; } /// Returns OpenCL access qual. static OpenCLAccessAttr *getOpenCLArgAccess(const Decl *D) { return D->getAttr
(); } /// Returns true if pipe element type is different from the pointer. static bool checkOpenCLPipeArg(Sema &S, CallExpr *Call) { const Expr *Arg0 = Call->getArg(0); // First argument type should always be pipe. if (!Arg0->getType()->isPipeType()) { S.Diag(Call->getLocStart(), diag::err_opencl_builtin_pipe_first_arg) << Call->getDirectCallee() << Arg0->getSourceRange(); return true; } OpenCLAccessAttr *AccessQual = getOpenCLArgAccess(cast
(Arg0)->getDecl()); // Validates the access qualifier is compatible with the call. // OpenCL v2.0 s6.13.16 - The access qualifiers for pipe should only be // read_only and write_only, and assumed to be read_only if no qualifier is // specified. switch (Call->getDirectCallee()->getBuiltinID()) { case Builtin::BIread_pipe: case Builtin::BIreserve_read_pipe: case Builtin::BIcommit_read_pipe: case Builtin::BIwork_group_reserve_read_pipe: case Builtin::BIsub_group_reserve_read_pipe: case Builtin::BIwork_group_commit_read_pipe: case Builtin::BIsub_group_commit_read_pipe: if (!(!AccessQual || AccessQual->isReadOnly())) { S.Diag(Arg0->getLocStart(), diag::err_opencl_builtin_pipe_invalid_access_modifier) << "read_only" << Arg0->getSourceRange(); return true; } break; case Builtin::BIwrite_pipe: case Builtin::BIreserve_write_pipe: case Builtin::BIcommit_write_pipe: case Builtin::BIwork_group_reserve_write_pipe: case Builtin::BIsub_group_reserve_write_pipe: case Builtin::BIwork_group_commit_write_pipe: case Builtin::BIsub_group_commit_write_pipe: if (!(AccessQual && AccessQual->isWriteOnly())) { S.Diag(Arg0->getLocStart(), diag::err_opencl_builtin_pipe_invalid_access_modifier) << "write_only" << Arg0->getSourceRange(); return true; } break; default: break; } return false; } /// Returns true if pipe element type is different from the pointer. static bool checkOpenCLPipePacketType(Sema &S, CallExpr *Call, unsigned Idx) { const Expr *Arg0 = Call->getArg(0); const Expr *ArgIdx = Call->getArg(Idx); const PipeType *PipeTy = cast
(Arg0->getType()); const QualType EltTy = PipeTy->getElementType(); const PointerType *ArgTy = ArgIdx->getType()->getAs
(); // The Idx argument should be a pointer and the type of the pointer and // the type of pipe element should also be the same. if (!ArgTy || !S.Context.hasSameType( EltTy, ArgTy->getPointeeType()->getCanonicalTypeInternal())) { S.Diag(Call->getLocStart(), diag::err_opencl_builtin_pipe_invalid_arg) << Call->getDirectCallee() << S.Context.getPointerType(EltTy) << ArgIdx->getType() << ArgIdx->getSourceRange(); return true; } return false; } // \brief Performs semantic analysis for the read/write_pipe call. // \param S Reference to the semantic analyzer. // \param Call A pointer to the builtin call. // \return True if a semantic error has been found, false otherwise. static bool SemaBuiltinRWPipe(Sema &S, CallExpr *Call) { // OpenCL v2.0 s6.13.16.2 - The built-in read/write // functions have two forms. switch (Call->getNumArgs()) { case 2: { if (checkOpenCLPipeArg(S, Call)) return true; // The call with 2 arguments should be // read/write_pipe(pipe T, T*). // Check packet type T. if (checkOpenCLPipePacketType(S, Call, 1)) return true; } break; case 4: { if (checkOpenCLPipeArg(S, Call)) return true; // The call with 4 arguments should be // read/write_pipe(pipe T, reserve_id_t, uint, T*). // Check reserve_id_t. if (!Call->getArg(1)->getType()->isReserveIDT()) { S.Diag(Call->getLocStart(), diag::err_opencl_builtin_pipe_invalid_arg) << Call->getDirectCallee() << S.Context.OCLReserveIDTy << Call->getArg(1)->getType() << Call->getArg(1)->getSourceRange(); return true; } // Check the index. const Expr *Arg2 = Call->getArg(2); if (!Arg2->getType()->isIntegerType() && !Arg2->getType()->isUnsignedIntegerType()) { S.Diag(Call->getLocStart(), diag::err_opencl_builtin_pipe_invalid_arg) << Call->getDirectCallee() << S.Context.UnsignedIntTy << Arg2->getType() << Arg2->getSourceRange(); return true; } // Check packet type T. if (checkOpenCLPipePacketType(S, Call, 3)) return true; } break; default: S.Diag(Call->getLocStart(), diag::err_opencl_builtin_pipe_arg_num) << Call->getDirectCallee() << Call->getSourceRange(); return true; } return false; } // \brief Performs a semantic analysis on the {work_group_/sub_group_ // /_}reserve_{read/write}_pipe // \param S Reference to the semantic analyzer. // \param Call The call to the builtin function to be analyzed. // \return True if a semantic error was found, false otherwise. static bool SemaBuiltinReserveRWPipe(Sema &S, CallExpr *Call) { if (checkArgCount(S, Call, 2)) return true; if (checkOpenCLPipeArg(S, Call)) return true; // Check the reserve size. if (!Call->getArg(1)->getType()->isIntegerType() && !Call->getArg(1)->getType()->isUnsignedIntegerType()) { S.Diag(Call->getLocStart(), diag::err_opencl_builtin_pipe_invalid_arg) << Call->getDirectCallee() << S.Context.UnsignedIntTy << Call->getArg(1)->getType() << Call->getArg(1)->getSourceRange(); return true; } return false; } // \brief Performs a semantic analysis on {work_group_/sub_group_ // /_}commit_{read/write}_pipe // \param S Reference to the semantic analyzer. // \param Call The call to the builtin function to be analyzed. // \return True if a semantic error was found, false otherwise. static bool SemaBuiltinCommitRWPipe(Sema &S, CallExpr *Call) { if (checkArgCount(S, Call, 2)) return true; if (checkOpenCLPipeArg(S, Call)) return true; // Check reserve_id_t. if (!Call->getArg(1)->getType()->isReserveIDT()) { S.Diag(Call->getLocStart(), diag::err_opencl_builtin_pipe_invalid_arg) << Call->getDirectCallee() << S.Context.OCLReserveIDTy << Call->getArg(1)->getType() << Call->getArg(1)->getSourceRange(); return true; } return false; } // \brief Performs a semantic analysis on the call to built-in Pipe // Query Functions. // \param S Reference to the semantic analyzer. // \param Call The call to the builtin function to be analyzed. // \return True if a semantic error was found, false otherwise. static bool SemaBuiltinPipePackets(Sema &S, CallExpr *Call) { if (checkArgCount(S, Call, 1)) return true; if (!Call->getArg(0)->getType()->isPipeType()) { S.Diag(Call->getLocStart(), diag::err_opencl_builtin_pipe_first_arg) << Call->getDirectCallee() << Call->getArg(0)->getSourceRange(); return true; } return false; } // \brief OpenCL v2.0 s6.13.9 - Address space qualifier functions. // \brief Performs semantic analysis for the to_global/local/private call. // \param S Reference to the semantic analyzer. // \param BuiltinID ID of the builtin function. // \param Call A pointer to the builtin call. // \return True if a semantic error has been found, false otherwise. static bool SemaOpenCLBuiltinToAddr(Sema &S, unsigned BuiltinID, CallExpr *Call) { if (Call->getNumArgs() != 1) { S.Diag(Call->getLocStart(), diag::err_opencl_builtin_to_addr_arg_num) << Call->getDirectCallee() << Call->getSourceRange(); return true; } auto RT = Call->getArg(0)->getType(); if (!RT->isPointerType() || RT->getPointeeType() .getAddressSpace() == LangAS::opencl_constant) { S.Diag(Call->getLocStart(), diag::err_opencl_builtin_to_addr_invalid_arg) << Call->getArg(0) << Call->getDirectCallee() << Call->getSourceRange(); return true; } RT = RT->getPointeeType(); auto Qual = RT.getQualifiers(); switch (BuiltinID) { case Builtin::BIto_global: Qual.setAddressSpace(LangAS::opencl_global); break; case Builtin::BIto_local: Qual.setAddressSpace(LangAS::opencl_local); break; default: Qual.removeAddressSpace(); } Call->setType(S.Context.getPointerType(S.Context.getQualifiedType( RT.getUnqualifiedType(), Qual))); return false; } ExprResult Sema::CheckBuiltinFunctionCall(FunctionDecl *FDecl, unsigned BuiltinID, CallExpr *TheCall) { ExprResult TheCallResult(TheCall); // Find out if any arguments are required to be integer constant expressions. unsigned ICEArguments = 0; ASTContext::GetBuiltinTypeError Error; Context.GetBuiltinType(BuiltinID, Error, &ICEArguments); if (Error != ASTContext::GE_None) ICEArguments = 0; // Don't diagnose previously diagnosed errors. // If any arguments are required to be ICE's, check and diagnose. for (unsigned ArgNo = 0; ICEArguments != 0; ++ArgNo) { // Skip arguments not required to be ICE's. if ((ICEArguments & (1 << ArgNo)) == 0) continue; llvm::APSInt Result; if (SemaBuiltinConstantArg(TheCall, ArgNo, Result)) return true; ICEArguments &= ~(1 << ArgNo); } switch (BuiltinID) { case Builtin::BI__builtin___CFStringMakeConstantString: assert(TheCall->getNumArgs() == 1 && "Wrong # arguments to builtin CFStringMakeConstantString"); if (CheckObjCString(TheCall->getArg(0))) return ExprError(); break; case Builtin::BI__builtin_stdarg_start: case Builtin::BI__builtin_va_start: if (SemaBuiltinVAStart(TheCall)) return ExprError(); break; case Builtin::BI__va_start: { switch (Context.getTargetInfo().getTriple().getArch()) { case llvm::Triple::arm: case llvm::Triple::thumb: if (SemaBuiltinVAStartARM(TheCall)) return ExprError(); break; default: if (SemaBuiltinVAStart(TheCall)) return ExprError(); break; } break; } case Builtin::BI__builtin_isgreater: case Builtin::BI__builtin_isgreaterequal: case Builtin::BI__builtin_isless: case Builtin::BI__builtin_islessequal: case Builtin::BI__builtin_islessgreater: case Builtin::BI__builtin_isunordered: if (SemaBuiltinUnorderedCompare(TheCall)) return ExprError(); break; case Builtin::BI__builtin_fpclassify: if (SemaBuiltinFPClassification(TheCall, 6)) return ExprError(); break; case Builtin::BI__builtin_isfinite: case Builtin::BI__builtin_isinf: case Builtin::BI__builtin_isinf_sign: case Builtin::BI__builtin_isnan: case Builtin::BI__builtin_isnormal: if (SemaBuiltinFPClassification(TheCall, 1)) return ExprError(); break; case Builtin::BI__builtin_shufflevector: return SemaBuiltinShuffleVector(TheCall); // TheCall will be freed by the smart pointer here, but that's fine, since // SemaBuiltinShuffleVector guts it, but then doesn't release it. case Builtin::BI__builtin_prefetch: if (SemaBuiltinPrefetch(TheCall)) return ExprError(); break; case Builtin::BI__assume: case Builtin::BI__builtin_assume: if (SemaBuiltinAssume(TheCall)) return ExprError(); break; case Builtin::BI__builtin_assume_aligned: if (SemaBuiltinAssumeAligned(TheCall)) return ExprError(); break; case Builtin::BI__builtin_object_size: if (SemaBuiltinConstantArgRange(TheCall, 1, 0, 3)) return ExprError(); break; case Builtin::BI__builtin_longjmp: if (SemaBuiltinLongjmp(TheCall)) return ExprError(); break; case Builtin::BI__builtin_setjmp: if (SemaBuiltinSetjmp(TheCall)) return ExprError(); break; case Builtin::BI_setjmp: case Builtin::BI_setjmpex: if (checkArgCount(*this, TheCall, 1)) return true; break; case Builtin::BI__builtin_classify_type: if (checkArgCount(*this, TheCall, 1)) return true; TheCall->setType(Context.IntTy); break; case Builtin::BI__builtin_constant_p: if (checkArgCount(*this, TheCall, 1)) return true; TheCall->setType(Context.IntTy); break; case Builtin::BI__sync_fetch_and_add: case Builtin::BI__sync_fetch_and_add_1: case Builtin::BI__sync_fetch_and_add_2: case Builtin::BI__sync_fetch_and_add_4: case Builtin::BI__sync_fetch_and_add_8: case Builtin::BI__sync_fetch_and_add_16: case Builtin::BI__sync_fetch_and_sub: case Builtin::BI__sync_fetch_and_sub_1: case Builtin::BI__sync_fetch_and_sub_2: case Builtin::BI__sync_fetch_and_sub_4: case Builtin::BI__sync_fetch_and_sub_8: case Builtin::BI__sync_fetch_and_sub_16: case Builtin::BI__sync_fetch_and_or: case Builtin::BI__sync_fetch_and_or_1: case Builtin::BI__sync_fetch_and_or_2: case Builtin::BI__sync_fetch_and_or_4: case Builtin::BI__sync_fetch_and_or_8: case Builtin::BI__sync_fetch_and_or_16: case Builtin::BI__sync_fetch_and_and: case Builtin::BI__sync_fetch_and_and_1: case Builtin::BI__sync_fetch_and_and_2: case Builtin::BI__sync_fetch_and_and_4: case Builtin::BI__sync_fetch_and_and_8: case Builtin::BI__sync_fetch_and_and_16: case Builtin::BI__sync_fetch_and_xor: case Builtin::BI__sync_fetch_and_xor_1: case Builtin::BI__sync_fetch_and_xor_2: case Builtin::BI__sync_fetch_and_xor_4: case Builtin::BI__sync_fetch_and_xor_8: case Builtin::BI__sync_fetch_and_xor_16: case Builtin::BI__sync_fetch_and_nand: case Builtin::BI__sync_fetch_and_nand_1: case Builtin::BI__sync_fetch_and_nand_2: case Builtin::BI__sync_fetch_and_nand_4: case Builtin::BI__sync_fetch_and_nand_8: case Builtin::BI__sync_fetch_and_nand_16: case Builtin::BI__sync_add_and_fetch: case Builtin::BI__sync_add_and_fetch_1: case Builtin::BI__sync_add_and_fetch_2: case Builtin::BI__sync_add_and_fetch_4: case Builtin::BI__sync_add_and_fetch_8: case Builtin::BI__sync_add_and_fetch_16: case Builtin::BI__sync_sub_and_fetch: case Builtin::BI__sync_sub_and_fetch_1: case Builtin::BI__sync_sub_and_fetch_2: case Builtin::BI__sync_sub_and_fetch_4: case Builtin::BI__sync_sub_and_fetch_8: case Builtin::BI__sync_sub_and_fetch_16: case Builtin::BI__sync_and_and_fetch: case Builtin::BI__sync_and_and_fetch_1: case Builtin::BI__sync_and_and_fetch_2: case Builtin::BI__sync_and_and_fetch_4: case Builtin::BI__sync_and_and_fetch_8: case Builtin::BI__sync_and_and_fetch_16: case Builtin::BI__sync_or_and_fetch: case Builtin::BI__sync_or_and_fetch_1: case Builtin::BI__sync_or_and_fetch_2: case Builtin::BI__sync_or_and_fetch_4: case Builtin::BI__sync_or_and_fetch_8: case Builtin::BI__sync_or_and_fetch_16: case Builtin::BI__sync_xor_and_fetch: case Builtin::BI__sync_xor_and_fetch_1: case Builtin::BI__sync_xor_and_fetch_2: case Builtin::BI__sync_xor_and_fetch_4: case Builtin::BI__sync_xor_and_fetch_8: case Builtin::BI__sync_xor_and_fetch_16: case Builtin::BI__sync_nand_and_fetch: case Builtin::BI__sync_nand_and_fetch_1: case Builtin::BI__sync_nand_and_fetch_2: case Builtin::BI__sync_nand_and_fetch_4: case Builtin::BI__sync_nand_and_fetch_8: case Builtin::BI__sync_nand_and_fetch_16: case Builtin::BI__sync_val_compare_and_swap: case Builtin::BI__sync_val_compare_and_swap_1: case Builtin::BI__sync_val_compare_and_swap_2: case Builtin::BI__sync_val_compare_and_swap_4: case Builtin::BI__sync_val_compare_and_swap_8: case Builtin::BI__sync_val_compare_and_swap_16: case Builtin::BI__sync_bool_compare_and_swap: case Builtin::BI__sync_bool_compare_and_swap_1: case Builtin::BI__sync_bool_compare_and_swap_2: case Builtin::BI__sync_bool_compare_and_swap_4: case Builtin::BI__sync_bool_compare_and_swap_8: case Builtin::BI__sync_bool_compare_and_swap_16: case Builtin::BI__sync_lock_test_and_set: case Builtin::BI__sync_lock_test_and_set_1: case Builtin::BI__sync_lock_test_and_set_2: case Builtin::BI__sync_lock_test_and_set_4: case Builtin::BI__sync_lock_test_and_set_8: case Builtin::BI__sync_lock_test_and_set_16: case Builtin::BI__sync_lock_release: case Builtin::BI__sync_lock_release_1: case Builtin::BI__sync_lock_release_2: case Builtin::BI__sync_lock_release_4: case Builtin::BI__sync_lock_release_8: case Builtin::BI__sync_lock_release_16: case Builtin::BI__sync_swap: case Builtin::BI__sync_swap_1: case Builtin::BI__sync_swap_2: case Builtin::BI__sync_swap_4: case Builtin::BI__sync_swap_8: case Builtin::BI__sync_swap_16: return SemaBuiltinAtomicOverloaded(TheCallResult); case Builtin::BI__builtin_nontemporal_load: case Builtin::BI__builtin_nontemporal_store: return SemaBuiltinNontemporalOverloaded(TheCallResult); #define BUILTIN(ID, TYPE, ATTRS) #define ATOMIC_BUILTIN(ID, TYPE, ATTRS) \ case Builtin::BI##ID: \ return SemaAtomicOpsOverloaded(TheCallResult, AtomicExpr::AO##ID); #include "clang/Basic/Builtins.def" case Builtin::BI__builtin_annotation: if (SemaBuiltinAnnotation(*this, TheCall)) return ExprError(); break; case Builtin::BI__builtin_addressof: if (SemaBuiltinAddressof(*this, TheCall)) return ExprError(); break; case Builtin::BI__builtin_add_overflow: case Builtin::BI__builtin_sub_overflow: case Builtin::BI__builtin_mul_overflow: if (SemaBuiltinOverflow(*this, TheCall)) return ExprError(); break; case Builtin::BI__builtin_operator_new: case Builtin::BI__builtin_operator_delete: if (!getLangOpts().CPlusPlus) { Diag(TheCall->getExprLoc(), diag::err_builtin_requires_language) << (BuiltinID == Builtin::BI__builtin_operator_new ? "__builtin_operator_new" : "__builtin_operator_delete") << "C++"; return ExprError(); } // CodeGen assumes it can find the global new and delete to call, // so ensure that they are declared. DeclareGlobalNewDelete(); break; // check secure string manipulation functions where overflows // are detectable at compile time case Builtin::BI__builtin___memcpy_chk: case Builtin::BI__builtin___memmove_chk: case Builtin::BI__builtin___memset_chk: case Builtin::BI__builtin___strlcat_chk: case Builtin::BI__builtin___strlcpy_chk: case Builtin::BI__builtin___strncat_chk: case Builtin::BI__builtin___strncpy_chk: case Builtin::BI__builtin___stpncpy_chk: SemaBuiltinMemChkCall(*this, FDecl, TheCall, 2, 3); break; case Builtin::BI__builtin___memccpy_chk: SemaBuiltinMemChkCall(*this, FDecl, TheCall, 3, 4); break; case Builtin::BI__builtin___snprintf_chk: case Builtin::BI__builtin___vsnprintf_chk: SemaBuiltinMemChkCall(*this, FDecl, TheCall, 1, 3); break; case Builtin::BI__builtin_call_with_static_chain: if (SemaBuiltinCallWithStaticChain(*this, TheCall)) return ExprError(); break; case Builtin::BI__exception_code: case Builtin::BI_exception_code: if (SemaBuiltinSEHScopeCheck(*this, TheCall, Scope::SEHExceptScope, diag::err_seh___except_block)) return ExprError(); break; case Builtin::BI__exception_info: case Builtin::BI_exception_info: if (SemaBuiltinSEHScopeCheck(*this, TheCall, Scope::SEHFilterScope, diag::err_seh___except_filter)) return ExprError(); break; case Builtin::BI__GetExceptionInfo: if (checkArgCount(*this, TheCall, 1)) return ExprError(); if (CheckCXXThrowOperand( TheCall->getLocStart(), Context.getExceptionObjectType(FDecl->getParamDecl(0)->getType()), TheCall)) return ExprError(); TheCall->setType(Context.VoidPtrTy); break; // OpenCL v2.0, s6.13.16 - Pipe functions case Builtin::BIread_pipe: case Builtin::BIwrite_pipe: // Since those two functions are declared with var args, we need a semantic // check for the argument. if (SemaBuiltinRWPipe(*this, TheCall)) return ExprError(); break; case Builtin::BIreserve_read_pipe: case Builtin::BIreserve_write_pipe: case Builtin::BIwork_group_reserve_read_pipe: case Builtin::BIwork_group_reserve_write_pipe: case Builtin::BIsub_group_reserve_read_pipe: case Builtin::BIsub_group_reserve_write_pipe: if (SemaBuiltinReserveRWPipe(*this, TheCall)) return ExprError(); // Since return type of reserve_read/write_pipe built-in function is // reserve_id_t, which is not defined in the builtin def file , we used int // as return type and need to override the return type of these functions. TheCall->setType(Context.OCLReserveIDTy); break; case Builtin::BIcommit_read_pipe: case Builtin::BIcommit_write_pipe: case Builtin::BIwork_group_commit_read_pipe: case Builtin::BIwork_group_commit_write_pipe: case Builtin::BIsub_group_commit_read_pipe: case Builtin::BIsub_group_commit_write_pipe: if (SemaBuiltinCommitRWPipe(*this, TheCall)) return ExprError(); break; case Builtin::BIget_pipe_num_packets: case Builtin::BIget_pipe_max_packets: if (SemaBuiltinPipePackets(*this, TheCall)) return ExprError(); break; case Builtin::BIto_global: case Builtin::BIto_local: case Builtin::BIto_private: if (SemaOpenCLBuiltinToAddr(*this, BuiltinID, TheCall)) return ExprError(); break; // OpenCL v2.0, s6.13.17 - Enqueue kernel functions. case Builtin::BIenqueue_kernel: if (SemaOpenCLBuiltinEnqueueKernel(*this, TheCall)) return ExprError(); break; case Builtin::BIget_kernel_work_group_size: case Builtin::BIget_kernel_preferred_work_group_size_multiple: if (SemaOpenCLBuiltinKernelWorkGroupSize(*this, TheCall)) return ExprError(); } // Since the target specific builtins for each arch overlap, only check those // of the arch we are compiling for. if (Context.BuiltinInfo.isTSBuiltin(BuiltinID)) { switch (Context.getTargetInfo().getTriple().getArch()) { case llvm::Triple::arm: case llvm::Triple::armeb: case llvm::Triple::thumb: case llvm::Triple::thumbeb: if (CheckARMBuiltinFunctionCall(BuiltinID, TheCall)) return ExprError(); break; case llvm::Triple::aarch64: case llvm::Triple::aarch64_be: if (CheckAArch64BuiltinFunctionCall(BuiltinID, TheCall)) return ExprError(); break; case llvm::Triple::mips: case llvm::Triple::mipsel: case llvm::Triple::mips64: case llvm::Triple::mips64el: if (CheckMipsBuiltinFunctionCall(BuiltinID, TheCall)) return ExprError(); break; case llvm::Triple::systemz: if (CheckSystemZBuiltinFunctionCall(BuiltinID, TheCall)) return ExprError(); break; case llvm::Triple::x86: case llvm::Triple::x86_64: if (CheckX86BuiltinFunctionCall(BuiltinID, TheCall)) return ExprError(); break; case llvm::Triple::ppc: case llvm::Triple::ppc64: case llvm::Triple::ppc64le: if (CheckPPCBuiltinFunctionCall(BuiltinID, TheCall)) return ExprError(); break; default: break; } } return TheCallResult; } // Get the valid immediate range for the specified NEON type code. static unsigned RFT(unsigned t, bool shift = false, bool ForceQuad = false) { NeonTypeFlags Type(t); int IsQuad = ForceQuad ? true : Type.isQuad(); switch (Type.getEltType()) { case NeonTypeFlags::Int8: case NeonTypeFlags::Poly8: return shift ? 7 : (8 << IsQuad) - 1; case NeonTypeFlags::Int16: case NeonTypeFlags::Poly16: return shift ? 15 : (4 << IsQuad) - 1; case NeonTypeFlags::Int32: return shift ? 31 : (2 << IsQuad) - 1; case NeonTypeFlags::Int64: case NeonTypeFlags::Poly64: return shift ? 63 : (1 << IsQuad) - 1; case NeonTypeFlags::Poly128: return shift ? 127 : (1 << IsQuad) - 1; case NeonTypeFlags::Float16: assert(!shift && "cannot shift float types!"); return (4 << IsQuad) - 1; case NeonTypeFlags::Float32: assert(!shift && "cannot shift float types!"); return (2 << IsQuad) - 1; case NeonTypeFlags::Float64: assert(!shift && "cannot shift float types!"); return (1 << IsQuad) - 1; } llvm_unreachable("Invalid NeonTypeFlag!"); } /// getNeonEltType - Return the QualType corresponding to the elements of /// the vector type specified by the NeonTypeFlags. This is used to check /// the pointer arguments for Neon load/store intrinsics. static QualType getNeonEltType(NeonTypeFlags Flags, ASTContext &Context, bool IsPolyUnsigned, bool IsInt64Long) { switch (Flags.getEltType()) { case NeonTypeFlags::Int8: return Flags.isUnsigned() ? Context.UnsignedCharTy : Context.SignedCharTy; case NeonTypeFlags::Int16: return Flags.isUnsigned() ? Context.UnsignedShortTy : Context.ShortTy; case NeonTypeFlags::Int32: return Flags.isUnsigned() ? Context.UnsignedIntTy : Context.IntTy; case NeonTypeFlags::Int64: if (IsInt64Long) return Flags.isUnsigned() ? Context.UnsignedLongTy : Context.LongTy; else return Flags.isUnsigned() ? Context.UnsignedLongLongTy : Context.LongLongTy; case NeonTypeFlags::Poly8: return IsPolyUnsigned ? Context.UnsignedCharTy : Context.SignedCharTy; case NeonTypeFlags::Poly16: return IsPolyUnsigned ? Context.UnsignedShortTy : Context.ShortTy; case NeonTypeFlags::Poly64: if (IsInt64Long) return Context.UnsignedLongTy; else return Context.UnsignedLongLongTy; case NeonTypeFlags::Poly128: break; case NeonTypeFlags::Float16: return Context.HalfTy; case NeonTypeFlags::Float32: return Context.FloatTy; case NeonTypeFlags::Float64: return Context.DoubleTy; } llvm_unreachable("Invalid NeonTypeFlag!"); } bool Sema::CheckNeonBuiltinFunctionCall(unsigned BuiltinID, CallExpr *TheCall) { llvm::APSInt Result; uint64_t mask = 0; unsigned TV = 0; int PtrArgNum = -1; bool HasConstPtr = false; switch (BuiltinID) { #define GET_NEON_OVERLOAD_CHECK #include "clang/Basic/arm_neon.inc" #undef GET_NEON_OVERLOAD_CHECK } // For NEON intrinsics which are overloaded on vector element type, validate // the immediate which specifies which variant to emit. unsigned ImmArg = TheCall->getNumArgs()-1; if (mask) { if (SemaBuiltinConstantArg(TheCall, ImmArg, Result)) return true; TV = Result.getLimitedValue(64); if ((TV > 63) || (mask & (1ULL << TV)) == 0) return Diag(TheCall->getLocStart(), diag::err_invalid_neon_type_code) << TheCall->getArg(ImmArg)->getSourceRange(); } if (PtrArgNum >= 0) { // Check that pointer arguments have the specified type. Expr *Arg = TheCall->getArg(PtrArgNum); if (ImplicitCastExpr *ICE = dyn_cast
(Arg)) Arg = ICE->getSubExpr(); ExprResult RHS = DefaultFunctionArrayLvalueConversion(Arg); QualType RHSTy = RHS.get()->getType(); llvm::Triple::ArchType Arch = Context.getTargetInfo().getTriple().getArch(); bool IsPolyUnsigned = Arch == llvm::Triple::aarch64; bool IsInt64Long = Context.getTargetInfo().getInt64Type() == TargetInfo::SignedLong; QualType EltTy = getNeonEltType(NeonTypeFlags(TV), Context, IsPolyUnsigned, IsInt64Long); if (HasConstPtr) EltTy = EltTy.withConst(); QualType LHSTy = Context.getPointerType(EltTy); AssignConvertType ConvTy; ConvTy = CheckSingleAssignmentConstraints(LHSTy, RHS); if (RHS.isInvalid()) return true; if (DiagnoseAssignmentResult(ConvTy, Arg->getLocStart(), LHSTy, RHSTy, RHS.get(), AA_Assigning)) return true; } // For NEON intrinsics which take an immediate value as part of the // instruction, range check them here. unsigned i = 0, l = 0, u = 0; switch (BuiltinID) { default: return false; #define GET_NEON_IMMEDIATE_CHECK #include "clang/Basic/arm_neon.inc" #undef GET_NEON_IMMEDIATE_CHECK } return SemaBuiltinConstantArgRange(TheCall, i, l, u + l); } bool Sema::CheckARMBuiltinExclusiveCall(unsigned BuiltinID, CallExpr *TheCall, unsigned MaxWidth) { assert((BuiltinID == ARM::BI__builtin_arm_ldrex || BuiltinID == ARM::BI__builtin_arm_ldaex || BuiltinID == ARM::BI__builtin_arm_strex || BuiltinID == ARM::BI__builtin_arm_stlex || BuiltinID == AArch64::BI__builtin_arm_ldrex || BuiltinID == AArch64::BI__builtin_arm_ldaex || BuiltinID == AArch64::BI__builtin_arm_strex || BuiltinID == AArch64::BI__builtin_arm_stlex) && "unexpected ARM builtin"); bool IsLdrex = BuiltinID == ARM::BI__builtin_arm_ldrex || BuiltinID == ARM::BI__builtin_arm_ldaex || BuiltinID == AArch64::BI__builtin_arm_ldrex || BuiltinID == AArch64::BI__builtin_arm_ldaex; DeclRefExpr *DRE =cast
(TheCall->getCallee()->IgnoreParenCasts()); // Ensure that we have the proper number of arguments. if (checkArgCount(*this, TheCall, IsLdrex ? 1 : 2)) return true; // Inspect the pointer argument of the atomic builtin. This should always be // a pointer type, whose element is an integral scalar or pointer type. // Because it is a pointer type, we don't have to worry about any implicit // casts here. Expr *PointerArg = TheCall->getArg(IsLdrex ? 0 : 1); ExprResult PointerArgRes = DefaultFunctionArrayLvalueConversion(PointerArg); if (PointerArgRes.isInvalid()) return true; PointerArg = PointerArgRes.get(); const PointerType *pointerType = PointerArg->getType()->getAs
(); if (!pointerType) { Diag(DRE->getLocStart(), diag::err_atomic_builtin_must_be_pointer) << PointerArg->getType() << PointerArg->getSourceRange(); return true; } // ldrex takes a "const volatile T*" and strex takes a "volatile T*". Our next // task is to insert the appropriate casts into the AST. First work out just // what the appropriate type is. QualType ValType = pointerType->getPointeeType(); QualType AddrType = ValType.getUnqualifiedType().withVolatile(); if (IsLdrex) AddrType.addConst(); // Issue a warning if the cast is dodgy. CastKind CastNeeded = CK_NoOp; if (!AddrType.isAtLeastAsQualifiedAs(ValType)) { CastNeeded = CK_BitCast; Diag(DRE->getLocStart(), diag::ext_typecheck_convert_discards_qualifiers) << PointerArg->getType() << Context.getPointerType(AddrType) << AA_Passing << PointerArg->getSourceRange(); } // Finally, do the cast and replace the argument with the corrected version. AddrType = Context.getPointerType(AddrType); PointerArgRes = ImpCastExprToType(PointerArg, AddrType, CastNeeded); if (PointerArgRes.isInvalid()) return true; PointerArg = PointerArgRes.get(); TheCall->setArg(IsLdrex ? 0 : 1, PointerArg); // In general, we allow ints, floats and pointers to be loaded and stored. if (!ValType->isIntegerType() && !ValType->isAnyPointerType() && !ValType->isBlockPointerType() && !ValType->isFloatingType()) { Diag(DRE->getLocStart(), diag::err_atomic_builtin_must_be_pointer_intfltptr) << PointerArg->getType() << PointerArg->getSourceRange(); return true; } // But ARM doesn't have instructions to deal with 128-bit versions. if (Context.getTypeSize(ValType) > MaxWidth) { assert(MaxWidth == 64 && "Diagnostic unexpectedly inaccurate"); Diag(DRE->getLocStart(), diag::err_atomic_exclusive_builtin_pointer_size) << PointerArg->getType() << PointerArg->getSourceRange(); return true; } switch (ValType.getObjCLifetime()) { case Qualifiers::OCL_None: case Qualifiers::OCL_ExplicitNone: // okay break; case Qualifiers::OCL_Weak: case Qualifiers::OCL_Strong: case Qualifiers::OCL_Autoreleasing: Diag(DRE->getLocStart(), diag::err_arc_atomic_ownership) << ValType << PointerArg->getSourceRange(); return true; } if (IsLdrex) { TheCall->setType(ValType); return false; } // Initialize the argument to be stored. ExprResult ValArg = TheCall->getArg(0); InitializedEntity Entity = InitializedEntity::InitializeParameter( Context, ValType, /*consume*/ false); ValArg = PerformCopyInitialization(Entity, SourceLocation(), ValArg); if (ValArg.isInvalid()) return true; TheCall->setArg(0, ValArg.get()); // __builtin_arm_strex always returns an int. It's marked as such in the .def, // but the custom checker bypasses all default analysis. TheCall->setType(Context.IntTy); return false; } bool Sema::CheckARMBuiltinFunctionCall(unsigned BuiltinID, CallExpr *TheCall) { llvm::APSInt Result; if (BuiltinID == ARM::BI__builtin_arm_ldrex || BuiltinID == ARM::BI__builtin_arm_ldaex || BuiltinID == ARM::BI__builtin_arm_strex || BuiltinID == ARM::BI__builtin_arm_stlex) { return CheckARMBuiltinExclusiveCall(BuiltinID, TheCall, 64); } if (BuiltinID == ARM::BI__builtin_arm_prefetch) { return SemaBuiltinConstantArgRange(TheCall, 1, 0, 1) || SemaBuiltinConstantArgRange(TheCall, 2, 0, 1); } if (BuiltinID == ARM::BI__builtin_arm_rsr64 || BuiltinID == ARM::BI__builtin_arm_wsr64) return SemaBuiltinARMSpecialReg(BuiltinID, TheCall, 0, 3, false); if (BuiltinID == ARM::BI__builtin_arm_rsr || BuiltinID == ARM::BI__builtin_arm_rsrp || BuiltinID == ARM::BI__builtin_arm_wsr || BuiltinID == ARM::BI__builtin_arm_wsrp) return SemaBuiltinARMSpecialReg(BuiltinID, TheCall, 0, 5, true); if (CheckNeonBuiltinFunctionCall(BuiltinID, TheCall)) return true; // For intrinsics which take an immediate value as part of the instruction, // range check them here. unsigned i = 0, l = 0, u = 0; switch (BuiltinID) { default: return false; case ARM::BI__builtin_arm_ssat: i = 1; l = 1; u = 31; break; case ARM::BI__builtin_arm_usat: i = 1; u = 31; break; case ARM::BI__builtin_arm_vcvtr_f: case ARM::BI__builtin_arm_vcvtr_d: i = 1; u = 1; break; case ARM::BI__builtin_arm_dmb: case ARM::BI__builtin_arm_dsb: case ARM::BI__builtin_arm_isb: case ARM::BI__builtin_arm_dbg: l = 0; u = 15; break; } // FIXME: VFP Intrinsics should error if VFP not present. return SemaBuiltinConstantArgRange(TheCall, i, l, u + l); } bool Sema::CheckAArch64BuiltinFunctionCall(unsigned BuiltinID, CallExpr *TheCall) { llvm::APSInt Result; if (BuiltinID == AArch64::BI__builtin_arm_ldrex || BuiltinID == AArch64::BI__builtin_arm_ldaex || BuiltinID == AArch64::BI__builtin_arm_strex || BuiltinID == AArch64::BI__builtin_arm_stlex) { return CheckARMBuiltinExclusiveCall(BuiltinID, TheCall, 128); } if (BuiltinID == AArch64::BI__builtin_arm_prefetch) { return SemaBuiltinConstantArgRange(TheCall, 1, 0, 1) || SemaBuiltinConstantArgRange(TheCall, 2, 0, 2) || SemaBuiltinConstantArgRange(TheCall, 3, 0, 1) || SemaBuiltinConstantArgRange(TheCall, 4, 0, 1); } if (BuiltinID == AArch64::BI__builtin_arm_rsr64 || BuiltinID == AArch64::BI__builtin_arm_wsr64) return SemaBuiltinARMSpecialReg(BuiltinID, TheCall, 0, 5, true); if (BuiltinID == AArch64::BI__builtin_arm_rsr || BuiltinID == AArch64::BI__builtin_arm_rsrp || BuiltinID == AArch64::BI__builtin_arm_wsr || BuiltinID == AArch64::BI__builtin_arm_wsrp) return SemaBuiltinARMSpecialReg(BuiltinID, TheCall, 0, 5, true); if (CheckNeonBuiltinFunctionCall(BuiltinID, TheCall)) return true; // For intrinsics which take an immediate value as part of the instruction, // range check them here. unsigned i = 0, l = 0, u = 0; switch (BuiltinID) { default: return false; case AArch64::BI__builtin_arm_dmb: case AArch64::BI__builtin_arm_dsb: case AArch64::BI__builtin_arm_isb: l = 0; u = 15; break; } return SemaBuiltinConstantArgRange(TheCall, i, l, u + l); } bool Sema::CheckMipsBuiltinFunctionCall(unsigned BuiltinID, CallExpr *TheCall) { unsigned i = 0, l = 0, u = 0; switch (BuiltinID) { default: return false; case Mips::BI__builtin_mips_wrdsp: i = 1; l = 0; u = 63; break; case Mips::BI__builtin_mips_rddsp: i = 0; l = 0; u = 63; break; case Mips::BI__builtin_mips_append: i = 2; l = 0; u = 31; break; case Mips::BI__builtin_mips_balign: i = 2; l = 0; u = 3; break; case Mips::BI__builtin_mips_precr_sra_ph_w: i = 2; l = 0; u = 31; break; case Mips::BI__builtin_mips_precr_sra_r_ph_w: i = 2; l = 0; u = 31; break; case Mips::BI__builtin_mips_prepend: i = 2; l = 0; u = 31; break; } return SemaBuiltinConstantArgRange(TheCall, i, l, u); } bool Sema::CheckPPCBuiltinFunctionCall(unsigned BuiltinID, CallExpr *TheCall) { unsigned i = 0, l = 0, u = 0; bool Is64BitBltin = BuiltinID == PPC::BI__builtin_divde || BuiltinID == PPC::BI__builtin_divdeu || BuiltinID == PPC::BI__builtin_bpermd; bool IsTarget64Bit = Context.getTargetInfo() .getTypeWidth(Context .getTargetInfo() .getIntPtrType()) == 64; bool IsBltinExtDiv = BuiltinID == PPC::BI__builtin_divwe || BuiltinID == PPC::BI__builtin_divweu || BuiltinID == PPC::BI__builtin_divde || BuiltinID == PPC::BI__builtin_divdeu; if (Is64BitBltin && !IsTarget64Bit) return Diag(TheCall->getLocStart(), diag::err_64_bit_builtin_32_bit_tgt) << TheCall->getSourceRange(); if ((IsBltinExtDiv && !Context.getTargetInfo().hasFeature("extdiv")) || (BuiltinID == PPC::BI__builtin_bpermd && !Context.getTargetInfo().hasFeature("bpermd"))) return Diag(TheCall->getLocStart(), diag::err_ppc_builtin_only_on_pwr7) << TheCall->getSourceRange(); switch (BuiltinID) { default: return false; case PPC::BI__builtin_altivec_crypto_vshasigmaw: case PPC::BI__builtin_altivec_crypto_vshasigmad: return SemaBuiltinConstantArgRange(TheCall, 1, 0, 1) || SemaBuiltinConstantArgRange(TheCall, 2, 0, 15); case PPC::BI__builtin_tbegin: case PPC::BI__builtin_tend: i = 0; l = 0; u = 1; break; case PPC::BI__builtin_tsr: i = 0; l = 0; u = 7; break; case PPC::BI__builtin_tabortwc: case PPC::BI__builtin_tabortdc: i = 0; l = 0; u = 31; break; case PPC::BI__builtin_tabortwci: case PPC::BI__builtin_tabortdci: return SemaBuiltinConstantArgRange(TheCall, 0, 0, 31) || SemaBuiltinConstantArgRange(TheCall, 2, 0, 31); } return SemaBuiltinConstantArgRange(TheCall, i, l, u); } bool Sema::CheckSystemZBuiltinFunctionCall(unsigned BuiltinID, CallExpr *TheCall) { if (BuiltinID == SystemZ::BI__builtin_tabort) { Expr *Arg = TheCall->getArg(0); llvm::APSInt AbortCode(32); if (Arg->isIntegerConstantExpr(AbortCode, Context) && AbortCode.getSExtValue() >= 0 && AbortCode.getSExtValue() < 256) return Diag(Arg->getLocStart(), diag::err_systemz_invalid_tabort_code) << Arg->getSourceRange(); } // For intrinsics which take an immediate value as part of the instruction, // range check them here. unsigned i = 0, l = 0, u = 0; switch (BuiltinID) { default: return false; case SystemZ::BI__builtin_s390_lcbb: i = 1; l = 0; u = 15; break; case SystemZ::BI__builtin_s390_verimb: case SystemZ::BI__builtin_s390_verimh: case SystemZ::BI__builtin_s390_verimf: case SystemZ::BI__builtin_s390_verimg: i = 3; l = 0; u = 255; break; case SystemZ::BI__builtin_s390_vfaeb: case SystemZ::BI__builtin_s390_vfaeh: case SystemZ::BI__builtin_s390_vfaef: case SystemZ::BI__builtin_s390_vfaebs: case SystemZ::BI__builtin_s390_vfaehs: case SystemZ::BI__builtin_s390_vfaefs: case SystemZ::BI__builtin_s390_vfaezb: case SystemZ::BI__builtin_s390_vfaezh: case SystemZ::BI__builtin_s390_vfaezf: case SystemZ::BI__builtin_s390_vfaezbs: case SystemZ::BI__builtin_s390_vfaezhs: case SystemZ::BI__builtin_s390_vfaezfs: i = 2; l = 0; u = 15; break; case SystemZ::BI__builtin_s390_vfidb: return SemaBuiltinConstantArgRange(TheCall, 1, 0, 15) || SemaBuiltinConstantArgRange(TheCall, 2, 0, 15); case SystemZ::BI__builtin_s390_vftcidb: i = 1; l = 0; u = 4095; break; case SystemZ::BI__builtin_s390_vlbb: i = 1; l = 0; u = 15; break; case SystemZ::BI__builtin_s390_vpdi: i = 2; l = 0; u = 15; break; case SystemZ::BI__builtin_s390_vsldb: i = 2; l = 0; u = 15; break; case SystemZ::BI__builtin_s390_vstrcb: case SystemZ::BI__builtin_s390_vstrch: case SystemZ::BI__builtin_s390_vstrcf: case SystemZ::BI__builtin_s390_vstrczb: case SystemZ::BI__builtin_s390_vstrczh: case SystemZ::BI__builtin_s390_vstrczf: case SystemZ::BI__builtin_s390_vstrcbs: case SystemZ::BI__builtin_s390_vstrchs: case SystemZ::BI__builtin_s390_vstrcfs: case SystemZ::BI__builtin_s390_vstrczbs: case SystemZ::BI__builtin_s390_vstrczhs: case SystemZ::BI__builtin_s390_vstrczfs: i = 3; l = 0; u = 15; break; } return SemaBuiltinConstantArgRange(TheCall, i, l, u); } /// SemaBuiltinCpuSupports - Handle __builtin_cpu_supports(char *). /// This checks that the target supports __builtin_cpu_supports and /// that the string argument is constant and valid. static bool SemaBuiltinCpuSupports(Sema &S, CallExpr *TheCall) { Expr *Arg = TheCall->getArg(0); // Check if the argument is a string literal. if (!isa
(Arg->IgnoreParenImpCasts())) return S.Diag(TheCall->getLocStart(), diag::err_expr_not_string_literal) << Arg->getSourceRange(); // Check the contents of the string. StringRef Feature = cast
(Arg->IgnoreParenImpCasts())->getString(); if (!S.Context.getTargetInfo().validateCpuSupports(Feature)) return S.Diag(TheCall->getLocStart(), diag::err_invalid_cpu_supports) << Arg->getSourceRange(); return false; } bool Sema::CheckX86BuiltinFunctionCall(unsigned BuiltinID, CallExpr *TheCall) { int i = 0, l = 0, u = 0; switch (BuiltinID) { default: return false; case X86::BI__builtin_cpu_supports: return SemaBuiltinCpuSupports(*this, TheCall); case X86::BI__builtin_ms_va_start: return SemaBuiltinMSVAStart(TheCall); case X86::BI__builtin_ia32_extractf64x4_mask: case X86::BI__builtin_ia32_extracti64x4_mask: case X86::BI__builtin_ia32_extractf32x8_mask: case X86::BI__builtin_ia32_extracti32x8_mask: case X86::BI__builtin_ia32_extractf64x2_256_mask: case X86::BI__builtin_ia32_extracti64x2_256_mask: case X86::BI__builtin_ia32_extractf32x4_256_mask: case X86::BI__builtin_ia32_extracti32x4_256_mask: i = 1; l = 0; u = 1; break; case X86::BI_mm_prefetch: case X86::BI__builtin_ia32_extractf32x4_mask: case X86::BI__builtin_ia32_extracti32x4_mask: case X86::BI__builtin_ia32_extractf64x2_512_mask: case X86::BI__builtin_ia32_extracti64x2_512_mask: i = 1; l = 0; u = 3; break; case X86::BI__builtin_ia32_insertf32x8_mask: case X86::BI__builtin_ia32_inserti32x8_mask: case X86::BI__builtin_ia32_insertf64x4_mask: case X86::BI__builtin_ia32_inserti64x4_mask: case X86::BI__builtin_ia32_insertf64x2_256_mask: case X86::BI__builtin_ia32_inserti64x2_256_mask: case X86::BI__builtin_ia32_insertf32x4_256_mask: case X86::BI__builtin_ia32_inserti32x4_256_mask: i = 2; l = 0; u = 1; break; case X86::BI__builtin_ia32_sha1rnds4: case X86::BI__builtin_ia32_shuf_f32x4_256_mask: case X86::BI__builtin_ia32_shuf_f64x2_256_mask: case X86::BI__builtin_ia32_shuf_i32x4_256_mask: case X86::BI__builtin_ia32_shuf_i64x2_256_mask: case X86::BI__builtin_ia32_insertf64x2_512_mask: case X86::BI__builtin_ia32_inserti64x2_512_mask: case X86::BI__builtin_ia32_insertf32x4_mask: case X86::BI__builtin_ia32_inserti32x4_mask: i = 2; l = 0; u = 3; break; case X86::BI__builtin_ia32_vpermil2pd: case X86::BI__builtin_ia32_vpermil2pd256: case X86::BI__builtin_ia32_vpermil2ps: case X86::BI__builtin_ia32_vpermil2ps256: i = 3; l = 0; u = 3; break; case X86::BI__builtin_ia32_cmpb128_mask: case X86::BI__builtin_ia32_cmpw128_mask: case X86::BI__builtin_ia32_cmpd128_mask: case X86::BI__builtin_ia32_cmpq128_mask: case X86::BI__builtin_ia32_cmpb256_mask: case X86::BI__builtin_ia32_cmpw256_mask: case X86::BI__builtin_ia32_cmpd256_mask: case X86::BI__builtin_ia32_cmpq256_mask: case X86::BI__builtin_ia32_cmpb512_mask: case X86::BI__builtin_ia32_cmpw512_mask: case X86::BI__builtin_ia32_cmpd512_mask: case X86::BI__builtin_ia32_cmpq512_mask: case X86::BI__builtin_ia32_ucmpb128_mask: case X86::BI__builtin_ia32_ucmpw128_mask: case X86::BI__builtin_ia32_ucmpd128_mask: case X86::BI__builtin_ia32_ucmpq128_mask: case X86::BI__builtin_ia32_ucmpb256_mask: case X86::BI__builtin_ia32_ucmpw256_mask: case X86::BI__builtin_ia32_ucmpd256_mask: case X86::BI__builtin_ia32_ucmpq256_mask: case X86::BI__builtin_ia32_ucmpb512_mask: case X86::BI__builtin_ia32_ucmpw512_mask: case X86::BI__builtin_ia32_ucmpd512_mask: case X86::BI__builtin_ia32_ucmpq512_mask: case X86::BI__builtin_ia32_vpcomub: case X86::BI__builtin_ia32_vpcomuw: case X86::BI__builtin_ia32_vpcomud: case X86::BI__builtin_ia32_vpcomuq: case X86::BI__builtin_ia32_vpcomb: case X86::BI__builtin_ia32_vpcomw: case X86::BI__builtin_ia32_vpcomd: case X86::BI__builtin_ia32_vpcomq: i = 2; l = 0; u = 7; break; case X86::BI__builtin_ia32_roundps: case X86::BI__builtin_ia32_roundpd: case X86::BI__builtin_ia32_roundps256: case X86::BI__builtin_ia32_roundpd256: i = 1; l = 0; u = 15; break; case X86::BI__builtin_ia32_roundss: case X86::BI__builtin_ia32_roundsd: case X86::BI__builtin_ia32_rangepd128_mask: case X86::BI__builtin_ia32_rangepd256_mask: case X86::BI__builtin_ia32_rangepd512_mask: case X86::BI__builtin_ia32_rangeps128_mask: case X86::BI__builtin_ia32_rangeps256_mask: case X86::BI__builtin_ia32_rangeps512_mask: case X86::BI__builtin_ia32_getmantsd_round_mask: case X86::BI__builtin_ia32_getmantss_round_mask: i = 2; l = 0; u = 15; break; case X86::BI__builtin_ia32_cmpps: case X86::BI__builtin_ia32_cmpss: case X86::BI__builtin_ia32_cmppd: case X86::BI__builtin_ia32_cmpsd: case X86::BI__builtin_ia32_cmpps256: case X86::BI__builtin_ia32_cmppd256: case X86::BI__builtin_ia32_cmpps128_mask: case X86::BI__builtin_ia32_cmppd128_mask: case X86::BI__builtin_ia32_cmpps256_mask: case X86::BI__builtin_ia32_cmppd256_mask: case X86::BI__builtin_ia32_cmpps512_mask: case X86::BI__builtin_ia32_cmppd512_mask: case X86::BI__builtin_ia32_cmpsd_mask: case X86::BI__builtin_ia32_cmpss_mask: i = 2; l = 0; u = 31; break; case X86::BI__builtin_ia32_xabort: i = 0; l = -128; u = 255; break; case X86::BI__builtin_ia32_pshufw: case X86::BI__builtin_ia32_aeskeygenassist128: i = 1; l = -128; u = 255; break; case X86::BI__builtin_ia32_vcvtps2ph: case X86::BI__builtin_ia32_vcvtps2ph256: case X86::BI__builtin_ia32_rndscaleps_128_mask: case X86::BI__builtin_ia32_rndscalepd_128_mask: case X86::BI__builtin_ia32_rndscaleps_256_mask: case X86::BI__builtin_ia32_rndscalepd_256_mask: case X86::BI__builtin_ia32_rndscaleps_mask: case X86::BI__builtin_ia32_rndscalepd_mask: case X86::BI__builtin_ia32_reducepd128_mask: case X86::BI__builtin_ia32_reducepd256_mask: case X86::BI__builtin_ia32_reducepd512_mask: case X86::BI__builtin_ia32_reduceps128_mask: case X86::BI__builtin_ia32_reduceps256_mask: case X86::BI__builtin_ia32_reduceps512_mask: case X86::BI__builtin_ia32_prold512_mask: case X86::BI__builtin_ia32_prolq512_mask: case X86::BI__builtin_ia32_prold128_mask: case X86::BI__builtin_ia32_prold256_mask: case X86::BI__builtin_ia32_prolq128_mask: case X86::BI__builtin_ia32_prolq256_mask: case X86::BI__builtin_ia32_prord128_mask: case X86::BI__builtin_ia32_prord256_mask: case X86::BI__builtin_ia32_prorq128_mask: case X86::BI__builtin_ia32_prorq256_mask: case X86::BI__builtin_ia32_psllwi512_mask: case X86::BI__builtin_ia32_psllwi128_mask: case X86::BI__builtin_ia32_psllwi256_mask: case X86::BI__builtin_ia32_psrldi128_mask: case X86::BI__builtin_ia32_psrldi256_mask: case X86::BI__builtin_ia32_psrldi512_mask: case X86::BI__builtin_ia32_psrlqi128_mask: case X86::BI__builtin_ia32_psrlqi256_mask: case X86::BI__builtin_ia32_psrlqi512_mask: case X86::BI__builtin_ia32_psrawi512_mask: case X86::BI__builtin_ia32_psrawi128_mask: case X86::BI__builtin_ia32_psrawi256_mask: case X86::BI__builtin_ia32_psrlwi512_mask: case X86::BI__builtin_ia32_psrlwi128_mask: case X86::BI__builtin_ia32_psrlwi256_mask: case X86::BI__builtin_ia32_psradi128_mask: case X86::BI__builtin_ia32_psradi256_mask: case X86::BI__builtin_ia32_psradi512_mask: case X86::BI__builtin_ia32_psraqi128_mask: case X86::BI__builtin_ia32_psraqi256_mask: case X86::BI__builtin_ia32_psraqi512_mask: case X86::BI__builtin_ia32_pslldi128_mask: case X86::BI__builtin_ia32_pslldi256_mask: case X86::BI__builtin_ia32_pslldi512_mask: case X86::BI__builtin_ia32_psllqi128_mask: case X86::BI__builtin_ia32_psllqi256_mask: case X86::BI__builtin_ia32_psllqi512_mask: case X86::BI__builtin_ia32_fpclasspd128_mask: case X86::BI__builtin_ia32_fpclasspd256_mask: case X86::BI__builtin_ia32_fpclassps128_mask: case X86::BI__builtin_ia32_fpclassps256_mask: case X86::BI__builtin_ia32_fpclassps512_mask: case X86::BI__builtin_ia32_fpclasspd512_mask: case X86::BI__builtin_ia32_fpclasssd_mask: case X86::BI__builtin_ia32_fpclassss_mask: i = 1; l = 0; u = 255; break; case X86::BI__builtin_ia32_palignr: case X86::BI__builtin_ia32_insertps128: case X86::BI__builtin_ia32_dpps: case X86::BI__builtin_ia32_dppd: case X86::BI__builtin_ia32_dpps256: case X86::BI__builtin_ia32_mpsadbw128: case X86::BI__builtin_ia32_mpsadbw256: case X86::BI__builtin_ia32_pcmpistrm128: case X86::BI__builtin_ia32_pcmpistri128: case X86::BI__builtin_ia32_pcmpistria128: case X86::BI__builtin_ia32_pcmpistric128: case X86::BI__builtin_ia32_pcmpistrio128: case X86::BI__builtin_ia32_pcmpistris128: case X86::BI__builtin_ia32_pcmpistriz128: case X86::BI__builtin_ia32_pclmulqdq128: case X86::BI__builtin_ia32_vperm2f128_pd256: case X86::BI__builtin_ia32_vperm2f128_ps256: case X86::BI__builtin_ia32_vperm2f128_si256: case X86::BI__builtin_ia32_permti256: i = 2; l = -128; u = 255; break; case X86::BI__builtin_ia32_palignr128: case X86::BI__builtin_ia32_palignr256: case X86::BI__builtin_ia32_palignr128_mask: case X86::BI__builtin_ia32_palignr256_mask: case X86::BI__builtin_ia32_palignr512_mask: case X86::BI__builtin_ia32_alignq512_mask: case X86::BI__builtin_ia32_alignd512_mask: case X86::BI__builtin_ia32_alignd128_mask: case X86::BI__builtin_ia32_alignd256_mask: case X86::BI__builtin_ia32_alignq128_mask: case X86::BI__builtin_ia32_alignq256_mask: case X86::BI__builtin_ia32_vcomisd: case X86::BI__builtin_ia32_vcomiss: case X86::BI__builtin_ia32_shuf_f32x4_mask: case X86::BI__builtin_ia32_shuf_f64x2_mask: case X86::BI__builtin_ia32_shuf_i32x4_mask: case X86::BI__builtin_ia32_shuf_i64x2_mask: case X86::BI__builtin_ia32_dbpsadbw128_mask: case X86::BI__builtin_ia32_dbpsadbw256_mask: case X86::BI__builtin_ia32_dbpsadbw512_mask: i = 2; l = 0; u = 255; break; case X86::BI__builtin_ia32_fixupimmpd512_mask: case X86::BI__builtin_ia32_fixupimmpd512_maskz: case X86::BI__builtin_ia32_fixupimmps512_mask: case X86::BI__builtin_ia32_fixupimmps512_maskz: case X86::BI__builtin_ia32_fixupimmsd_mask: case X86::BI__builtin_ia32_fixupimmsd_maskz: case X86::BI__builtin_ia32_fixupimmss_mask: case X86::BI__builtin_ia32_fixupimmss_maskz: case X86::BI__builtin_ia32_fixupimmpd128_mask: case X86::BI__builtin_ia32_fixupimmpd128_maskz: case X86::BI__builtin_ia32_fixupimmpd256_mask: case X86::BI__builtin_ia32_fixupimmpd256_maskz: case X86::BI__builtin_ia32_fixupimmps128_mask: case X86::BI__builtin_ia32_fixupimmps128_maskz: case X86::BI__builtin_ia32_fixupimmps256_mask: case X86::BI__builtin_ia32_fixupimmps256_maskz: case X86::BI__builtin_ia32_pternlogd512_mask: case X86::BI__builtin_ia32_pternlogd512_maskz: case X86::BI__builtin_ia32_pternlogq512_mask: case X86::BI__builtin_ia32_pternlogq512_maskz: case X86::BI__builtin_ia32_pternlogd128_mask: case X86::BI__builtin_ia32_pternlogd128_maskz: case X86::BI__builtin_ia32_pternlogd256_mask: case X86::BI__builtin_ia32_pternlogd256_maskz: case X86::BI__builtin_ia32_pternlogq128_mask: case X86::BI__builtin_ia32_pternlogq128_maskz: case X86::BI__builtin_ia32_pternlogq256_mask: case X86::BI__builtin_ia32_pternlogq256_maskz: i = 3; l = 0; u = 255; break; case X86::BI__builtin_ia32_pcmpestrm128: case X86::BI__builtin_ia32_pcmpestri128: case X86::BI__builtin_ia32_pcmpestria128: case X86::BI__builtin_ia32_pcmpestric128: case X86::BI__builtin_ia32_pcmpestrio128: case X86::BI__builtin_ia32_pcmpestris128: case X86::BI__builtin_ia32_pcmpestriz128: i = 4; l = -128; u = 255; break; case X86::BI__builtin_ia32_rndscalesd_round_mask: case X86::BI__builtin_ia32_rndscaless_round_mask: i = 4; l = 0; u = 255; break; } return SemaBuiltinConstantArgRange(TheCall, i, l, u); } /// Given a FunctionDecl's FormatAttr, attempts to populate the FomatStringInfo /// parameter with the FormatAttr's correct format_idx and firstDataArg. /// Returns true when the format fits the function and the FormatStringInfo has /// been populated. bool Sema::getFormatStringInfo(const FormatAttr *Format, bool IsCXXMember, FormatStringInfo *FSI) { FSI->HasVAListArg = Format->getFirstArg() == 0; FSI->FormatIdx = Format->getFormatIdx() - 1; FSI->FirstDataArg = FSI->HasVAListArg ? 0 : Format->getFirstArg() - 1; // The way the format attribute works in GCC, the implicit this argument // of member functions is counted. However, it doesn't appear in our own // lists, so decrement format_idx in that case. if (IsCXXMember) { if(FSI->FormatIdx == 0) return false; --FSI->FormatIdx; if (FSI->FirstDataArg != 0) --FSI->FirstDataArg; } return true; } /// Checks if a the given expression evaluates to null. /// /// \brief Returns true if the value evaluates to null. static bool CheckNonNullExpr(Sema &S, const Expr *Expr) { // If the expression has non-null type, it doesn't evaluate to null. if (auto nullability = Expr->IgnoreImplicit()->getType()->getNullability(S.Context)) { if (*nullability == NullabilityKind::NonNull) return false; } // As a special case, transparent unions initialized with zero are // considered null for the purposes of the nonnull attribute. if (const RecordType *UT = Expr->getType()->getAsUnionType()) { if (UT->getDecl()->hasAttr
()) if (const CompoundLiteralExpr *CLE = dyn_cast
(Expr)) if (const InitListExpr *ILE = dyn_cast
(CLE->getInitializer())) Expr = ILE->getInit(0); } bool Result; return (!Expr->isValueDependent() && Expr->EvaluateAsBooleanCondition(Result, S.Context) && !Result); } static void CheckNonNullArgument(Sema &S, const Expr *ArgExpr, SourceLocation CallSiteLoc) { if (CheckNonNullExpr(S, ArgExpr)) S.DiagRuntimeBehavior(CallSiteLoc, ArgExpr, S.PDiag(diag::warn_null_arg) << ArgExpr->getSourceRange()); } bool Sema::GetFormatNSStringIdx(const FormatAttr *Format, unsigned &Idx) { FormatStringInfo FSI; if ((GetFormatStringType(Format) == FST_NSString) && getFormatStringInfo(Format, false, &FSI)) { Idx = FSI.FormatIdx; return true; } return false; } /// \brief Diagnose use of %s directive in an NSString which is being passed /// as formatting string to formatting method. static void DiagnoseCStringFormatDirectiveInCFAPI(Sema &S, const NamedDecl *FDecl, Expr **Args, unsigned NumArgs) { unsigned Idx = 0; bool Format = false; ObjCStringFormatFamily SFFamily = FDecl->getObjCFStringFormattingFamily(); if (SFFamily == ObjCStringFormatFamily::SFF_CFString) { Idx = 2; Format = true; } else for (const auto *I : FDecl->specific_attrs
()) { if (S.GetFormatNSStringIdx(I, Idx)) { Format = true; break; } } if (!Format || NumArgs <= Idx) return; const Expr *FormatExpr = Args[Idx]; if (const CStyleCastExpr *CSCE = dyn_cast
(FormatExpr)) FormatExpr = CSCE->getSubExpr(); const StringLiteral *FormatString; if (const ObjCStringLiteral *OSL = dyn_cast
(FormatExpr->IgnoreParenImpCasts())) FormatString = OSL->getString(); else FormatString = dyn_cast
(FormatExpr->IgnoreParenImpCasts()); if (!FormatString) return; if (S.FormatStringHasSArg(FormatString)) { S.Diag(FormatExpr->getExprLoc(), diag::warn_objc_cdirective_format_string) << "%s" << 1 << 1; S.Diag(FDecl->getLocation(), diag::note_entity_declared_at) << FDecl->getDeclName(); } } /// Determine whether the given type has a non-null nullability annotation. static bool isNonNullType(ASTContext &ctx, QualType type) { if (auto nullability = type->getNullability(ctx)) return *nullability == NullabilityKind::NonNull; return false; } static void CheckNonNullArguments(Sema &S, const NamedDecl *FDecl, const FunctionProtoType *Proto, ArrayRef
Args, SourceLocation CallSiteLoc) { assert((FDecl || Proto) && "Need a function declaration or prototype"); // Check the attributes attached to the method/function itself. llvm::SmallBitVector NonNullArgs; if (FDecl) { // Handle the nonnull attribute on the function/method declaration itself. for (const auto *NonNull : FDecl->specific_attrs
()) { if (!NonNull->args_size()) { // Easy case: all pointer arguments are nonnull. for (const auto *Arg : Args) if (S.isValidPointerAttrType(Arg->getType())) CheckNonNullArgument(S, Arg, CallSiteLoc); return; } for (unsigned Val : NonNull->args()) { if (Val >= Args.size()) continue; if (NonNullArgs.empty()) NonNullArgs.resize(Args.size()); NonNullArgs.set(Val); } } } if (FDecl && (isa
(FDecl) || isa
(FDecl))) { // Handle the nonnull attribute on the parameters of the // function/method. ArrayRef
parms; if (const FunctionDecl *FD = dyn_cast
(FDecl)) parms = FD->parameters(); else parms = cast
(FDecl)->parameters(); unsigned ParamIndex = 0; for (ArrayRef
::iterator I = parms.begin(), E = parms.end(); I != E; ++I, ++ParamIndex) { const ParmVarDecl *PVD = *I; if (PVD->hasAttr
() || isNonNullType(S.Context, PVD->getType())) { if (NonNullArgs.empty()) NonNullArgs.resize(Args.size()); NonNullArgs.set(ParamIndex); } } } else { // If we have a non-function, non-method declaration but no // function prototype, try to dig out the function prototype. if (!Proto) { if (const ValueDecl *VD = dyn_cast
(FDecl)) { QualType type = VD->getType().getNonReferenceType(); if (auto pointerType = type->getAs
()) type = pointerType->getPointeeType(); else if (auto blockType = type->getAs
()) type = blockType->getPointeeType(); // FIXME: data member pointers? // Dig out the function prototype, if there is one. Proto = type->getAs
(); } } // Fill in non-null argument information from the nullability // information on the parameter types (if we have them). if (Proto) { unsigned Index = 0; for (auto paramType : Proto->getParamTypes()) { if (isNonNullType(S.Context, paramType)) { if (NonNullArgs.empty()) NonNullArgs.resize(Args.size()); NonNullArgs.set(Index); } ++Index; } } } // Check for non-null arguments. for (unsigned ArgIndex = 0, ArgIndexEnd = NonNullArgs.size(); ArgIndex != ArgIndexEnd; ++ArgIndex) { if (NonNullArgs[ArgIndex]) CheckNonNullArgument(S, Args[ArgIndex], CallSiteLoc); } } /// Handles the checks for format strings, non-POD arguments to vararg /// functions, and NULL arguments passed to non-NULL parameters. void Sema::checkCall(NamedDecl *FDecl, const FunctionProtoType *Proto, ArrayRef
Args, bool IsMemberFunction, SourceLocation Loc, SourceRange Range, VariadicCallType CallType) { // FIXME: We should check as much as we can in the template definition. if (CurContext->isDependentContext()) return; // Printf and scanf checking. llvm::SmallBitVector CheckedVarArgs; if (FDecl) { for (const auto *I : FDecl->specific_attrs
()) { // Only create vector if there are format attributes. CheckedVarArgs.resize(Args.size()); CheckFormatArguments(I, Args, IsMemberFunction, CallType, Loc, Range, CheckedVarArgs); } } // Refuse POD arguments that weren't caught by the format string // checks above. if (CallType != VariadicDoesNotApply) { unsigned NumParams = Proto ? Proto->getNumParams() : FDecl && isa
(FDecl) ? cast
(FDecl)->getNumParams() : FDecl && isa
(FDecl) ? cast
(FDecl)->param_size() : 0; for (unsigned ArgIdx = NumParams; ArgIdx < Args.size(); ++ArgIdx) { // Args[ArgIdx] can be null in malformed code. if (const Expr *Arg = Args[ArgIdx]) { if (CheckedVarArgs.empty() || !CheckedVarArgs[ArgIdx]) checkVariadicArgument(Arg, CallType); } } } if (FDecl || Proto) { CheckNonNullArguments(*this, FDecl, Proto, Args, Loc); // Type safety checking. if (FDecl) { for (const auto *I : FDecl->specific_attrs
()) CheckArgumentWithTypeTag(I, Args.data()); } } } /// CheckConstructorCall - Check a constructor call for correctness and safety /// properties not enforced by the C type system. void Sema::CheckConstructorCall(FunctionDecl *FDecl, ArrayRef
Args, const FunctionProtoType *Proto, SourceLocation Loc) { VariadicCallType CallType = Proto->isVariadic() ? VariadicConstructor : VariadicDoesNotApply; checkCall(FDecl, Proto, Args, /*IsMemberFunction=*/true, Loc, SourceRange(), CallType); } /// CheckFunctionCall - Check a direct function call for various correctness /// and safety properties not strictly enforced by the C type system. bool Sema::CheckFunctionCall(FunctionDecl *FDecl, CallExpr *TheCall, const FunctionProtoType *Proto) { bool IsMemberOperatorCall = isa
(TheCall) && isa
(FDecl); bool IsMemberFunction = isa
(TheCall) || IsMemberOperatorCall; VariadicCallType CallType = getVariadicCallType(FDecl, Proto, TheCall->getCallee()); Expr** Args = TheCall->getArgs(); unsigned NumArgs = TheCall->getNumArgs(); if (IsMemberOperatorCall) { // If this is a call to a member operator, hide the first argument // from checkCall. // FIXME: Our choice of AST representation here is less than ideal. ++Args; --NumArgs; } checkCall(FDecl, Proto, llvm::makeArrayRef(Args, NumArgs), IsMemberFunction, TheCall->getRParenLoc(), TheCall->getCallee()->getSourceRange(), CallType); IdentifierInfo *FnInfo = FDecl->getIdentifier(); // None of the checks below are needed for functions that don't have // simple names (e.g., C++ conversion functions). if (!FnInfo) return false; CheckAbsoluteValueFunction(TheCall, FDecl, FnInfo); if (getLangOpts().ObjC1) DiagnoseCStringFormatDirectiveInCFAPI(*this, FDecl, Args, NumArgs); unsigned CMId = FDecl->getMemoryFunctionKind(); if (CMId == 0) return false; // Handle memory setting and copying functions. if (CMId == Builtin::BIstrlcpy || CMId == Builtin::BIstrlcat) CheckStrlcpycatArguments(TheCall, FnInfo); else if (CMId == Builtin::BIstrncat) CheckStrncatArguments(TheCall, FnInfo); else CheckMemaccessArguments(TheCall, CMId, FnInfo); return false; } bool Sema::CheckObjCMethodCall(ObjCMethodDecl *Method, SourceLocation lbrac, ArrayRef
Args) { VariadicCallType CallType = Method->isVariadic() ? VariadicMethod : VariadicDoesNotApply; checkCall(Method, nullptr, Args, /*IsMemberFunction=*/false, lbrac, Method->getSourceRange(), CallType); return false; } bool Sema::CheckPointerCall(NamedDecl *NDecl, CallExpr *TheCall, const FunctionProtoType *Proto) { QualType Ty; if (const auto *V = dyn_cast
(NDecl)) Ty = V->getType().getNonReferenceType(); else if (const auto *F = dyn_cast
(NDecl)) Ty = F->getType().getNonReferenceType(); else return false; if (!Ty->isBlockPointerType() && !Ty->isFunctionPointerType() && !Ty->isFunctionProtoType()) return false; VariadicCallType CallType; if (!Proto || !Proto->isVariadic()) { CallType = VariadicDoesNotApply; } else if (Ty->isBlockPointerType()) { CallType = VariadicBlock; } else { // Ty->isFunctionPointerType() CallType = VariadicFunction; } checkCall(NDecl, Proto, llvm::makeArrayRef(TheCall->getArgs(), TheCall->getNumArgs()), /*IsMemberFunction=*/false, TheCall->getRParenLoc(), TheCall->getCallee()->getSourceRange(), CallType); return false; } /// Checks function calls when a FunctionDecl or a NamedDecl is not available, /// such as function pointers returned from functions. bool Sema::CheckOtherCall(CallExpr *TheCall, const FunctionProtoType *Proto) { VariadicCallType CallType = getVariadicCallType(/*FDecl=*/nullptr, Proto, TheCall->getCallee()); checkCall(/*FDecl=*/nullptr, Proto, llvm::makeArrayRef(TheCall->getArgs(), TheCall->getNumArgs()), /*IsMemberFunction=*/false, TheCall->getRParenLoc(), TheCall->getCallee()->getSourceRange(), CallType); return false; } static bool isValidOrderingForOp(int64_t Ordering, AtomicExpr::AtomicOp Op) { if (!llvm::isValidAtomicOrderingCABI(Ordering)) return false; auto OrderingCABI = (llvm::AtomicOrderingCABI)Ordering; switch (Op) { case AtomicExpr::AO__c11_atomic_init: llvm_unreachable("There is no ordering argument for an init"); case AtomicExpr::AO__c11_atomic_load: case AtomicExpr::AO__atomic_load_n: case AtomicExpr::AO__atomic_load: return OrderingCABI != llvm::AtomicOrderingCABI::release && OrderingCABI != llvm::AtomicOrderingCABI::acq_rel; case AtomicExpr::AO__c11_atomic_store: case AtomicExpr::AO__atomic_store: case AtomicExpr::AO__atomic_store_n: return OrderingCABI != llvm::AtomicOrderingCABI::consume && OrderingCABI != llvm::AtomicOrderingCABI::acquire && OrderingCABI != llvm::AtomicOrderingCABI::acq_rel; default: return true; } } ExprResult Sema::SemaAtomicOpsOverloaded(ExprResult TheCallResult, AtomicExpr::AtomicOp Op) { CallExpr *TheCall = cast
(TheCallResult.get()); DeclRefExpr *DRE =cast
(TheCall->getCallee()->IgnoreParenCasts()); // All these operations take one of the following forms: enum { // C __c11_atomic_init(A *, C) Init, // C __c11_atomic_load(A *, int) Load, // void __atomic_load(A *, CP, int) LoadCopy, // void __atomic_store(A *, CP, int) Copy, // C __c11_atomic_add(A *, M, int) Arithmetic, // C __atomic_exchange_n(A *, CP, int) Xchg, // void __atomic_exchange(A *, C *, CP, int) GNUXchg, // bool __c11_atomic_compare_exchange_strong(A *, C *, CP, int, int) C11CmpXchg, // bool __atomic_compare_exchange(A *, C *, CP, bool, int, int) GNUCmpXchg } Form = Init; const unsigned NumArgs[] = { 2, 2, 3, 3, 3, 3, 4, 5, 6 }; const unsigned NumVals[] = { 1, 0, 1, 1, 1, 1, 2, 2, 3 }; // where: // C is an appropriate type, // A is volatile _Atomic(C) for __c11 builtins and is C for GNU builtins, // CP is C for __c11 builtins and GNU _n builtins and is C * otherwise, // M is C if C is an integer, and ptrdiff_t if C is a pointer, and // the int parameters are for orderings. static_assert(AtomicExpr::AO__c11_atomic_init == 0 && AtomicExpr::AO__c11_atomic_fetch_xor + 1 == AtomicExpr::AO__atomic_load, "need to update code for modified C11 atomics"); bool IsC11 = Op >= AtomicExpr::AO__c11_atomic_init && Op <= AtomicExpr::AO__c11_atomic_fetch_xor; bool IsN = Op == AtomicExpr::AO__atomic_load_n || Op == AtomicExpr::AO__atomic_store_n || Op == AtomicExpr::AO__atomic_exchange_n || Op == AtomicExpr::AO__atomic_compare_exchange_n; bool IsAddSub = false; switch (Op) { case AtomicExpr::AO__c11_atomic_init: Form = Init; break; case AtomicExpr::AO__c11_atomic_load: case AtomicExpr::AO__atomic_load_n: Form = Load; break; case AtomicExpr::AO__atomic_load: Form = LoadCopy; break; case AtomicExpr::AO__c11_atomic_store: case AtomicExpr::AO__atomic_store: case AtomicExpr::AO__atomic_store_n: Form = Copy; break; case AtomicExpr::AO__c11_atomic_fetch_add: case AtomicExpr::AO__c11_atomic_fetch_sub: case AtomicExpr::AO__atomic_fetch_add: case AtomicExpr::AO__atomic_fetch_sub: case AtomicExpr::AO__atomic_add_fetch: case AtomicExpr::AO__atomic_sub_fetch: IsAddSub = true; // Fall through. case AtomicExpr::AO__c11_atomic_fetch_and: case AtomicExpr::AO__c11_atomic_fetch_or: case AtomicExpr::AO__c11_atomic_fetch_xor: case AtomicExpr::AO__atomic_fetch_and: case AtomicExpr::AO__atomic_fetch_or: case AtomicExpr::AO__atomic_fetch_xor: case AtomicExpr::AO__atomic_fetch_nand: case AtomicExpr::AO__atomic_and_fetch: case AtomicExpr::AO__atomic_or_fetch: case AtomicExpr::AO__atomic_xor_fetch: case AtomicExpr::AO__atomic_nand_fetch: Form = Arithmetic; break; case AtomicExpr::AO__c11_atomic_exchange: case AtomicExpr::AO__atomic_exchange_n: Form = Xchg; break; case AtomicExpr::AO__atomic_exchange: Form = GNUXchg; break; case AtomicExpr::AO__c11_atomic_compare_exchange_strong: case AtomicExpr::AO__c11_atomic_compare_exchange_weak: Form = C11CmpXchg; break; case AtomicExpr::AO__atomic_compare_exchange: case AtomicExpr::AO__atomic_compare_exchange_n: Form = GNUCmpXchg; break; } // Check we have the right number of arguments. if (TheCall->getNumArgs() < NumArgs[Form]) { Diag(TheCall->getLocEnd(), diag::err_typecheck_call_too_few_args) << 0 << NumArgs[Form] << TheCall->getNumArgs() << TheCall->getCallee()->getSourceRange(); return ExprError(); } else if (TheCall->getNumArgs() > NumArgs[Form]) { Diag(TheCall->getArg(NumArgs[Form])->getLocStart(), diag::err_typecheck_call_too_many_args) << 0 << NumArgs[Form] << TheCall->getNumArgs() << TheCall->getCallee()->getSourceRange(); return ExprError(); } // Inspect the first argument of the atomic operation. Expr *Ptr = TheCall->getArg(0); Ptr = DefaultFunctionArrayLvalueConversion(Ptr).get(); const PointerType *pointerType = Ptr->getType()->getAs
(); if (!pointerType) { Diag(DRE->getLocStart(), diag::err_atomic_builtin_must_be_pointer) << Ptr->getType() << Ptr->getSourceRange(); return ExprError(); } // For a __c11 builtin, this should be a pointer to an _Atomic type. QualType AtomTy = pointerType->getPointeeType(); // 'A' QualType ValType = AtomTy; // 'C' if (IsC11) { if (!AtomTy->isAtomicType()) { Diag(DRE->getLocStart(), diag::err_atomic_op_needs_atomic) << Ptr->getType() << Ptr->getSourceRange(); return ExprError(); } if (AtomTy.isConstQualified()) { Diag(DRE->getLocStart(), diag::err_atomic_op_needs_non_const_atomic) << Ptr->getType() << Ptr->getSourceRange(); return ExprError(); } ValType = AtomTy->getAs
()->getValueType(); } else if (Form != Load && Form != LoadCopy) { if (ValType.isConstQualified()) { Diag(DRE->getLocStart(), diag::err_atomic_op_needs_non_const_pointer) << Ptr->getType() << Ptr->getSourceRange(); return ExprError(); } } // For an arithmetic operation, the implied arithmetic must be well-formed. if (Form == Arithmetic) { // gcc does not enforce these rules for GNU atomics, but we do so for sanity. if (IsAddSub && !ValType->isIntegerType() && !ValType->isPointerType()) { Diag(DRE->getLocStart(), diag::err_atomic_op_needs_atomic_int_or_ptr) << IsC11 << Ptr->getType() << Ptr->getSourceRange(); return ExprError(); } if (!IsAddSub && !ValType->isIntegerType()) { Diag(DRE->getLocStart(), diag::err_atomic_op_bitwise_needs_atomic_int) << IsC11 << Ptr->getType() << Ptr->getSourceRange(); return ExprError(); } if (IsC11 && ValType->isPointerType() && RequireCompleteType(Ptr->getLocStart(), ValType->getPointeeType(), diag::err_incomplete_type)) { return ExprError(); } } else if (IsN && !ValType->isIntegerType() && !ValType->isPointerType()) { // For __atomic_*_n operations, the value type must be a scalar integral or // pointer type which is 1, 2, 4, 8 or 16 bytes in length. Diag(DRE->getLocStart(), diag::err_atomic_op_needs_atomic_int_or_ptr) << IsC11 << Ptr->getType() << Ptr->getSourceRange(); return ExprError(); } if (!IsC11 && !AtomTy.isTriviallyCopyableType(Context) && !AtomTy->isScalarType()) { // For GNU atomics, require a trivially-copyable type. This is not part of // the GNU atomics specification, but we enforce it for sanity. Diag(DRE->getLocStart(), diag::err_atomic_op_needs_trivial_copy) << Ptr->getType() << Ptr->getSourceRange(); return ExprError(); } switch (ValType.getObjCLifetime()) { case Qualifiers::OCL_None: case Qualifiers::OCL_ExplicitNone: // okay break; case Qualifiers::OCL_Weak: case Qualifiers::OCL_Strong: case Qualifiers::OCL_Autoreleasing: // FIXME: Can this happen? By this point, ValType should be known // to be trivially copyable. Diag(DRE->getLocStart(), diag::err_arc_atomic_ownership) << ValType << Ptr->getSourceRange(); return ExprError(); } // atomic_fetch_or takes a pointer to a volatile 'A'. We shouldn't let the // volatile-ness of the pointee-type inject itself into the result or the // other operands. Similarly atomic_load can take a pointer to a const 'A'. ValType.removeLocalVolatile(); ValType.removeLocalConst(); QualType ResultType = ValType; if (Form == Copy || Form == LoadCopy || Form == GNUXchg || Form == Init) ResultType = Context.VoidTy; else if (Form == C11CmpXchg || Form == GNUCmpXchg) ResultType = Context.BoolTy; // The type of a parameter passed 'by value'. In the GNU atomics, such // arguments are actually passed as pointers. QualType ByValType = ValType; // 'CP' if (!IsC11 && !IsN) ByValType = Ptr->getType(); // The first argument --- the pointer --- has a fixed type; we // deduce the types of the rest of the arguments accordingly. Walk // the remaining arguments, converting them to the deduced value type. for (unsigned i = 1; i != NumArgs[Form]; ++i) { QualType Ty; if (i < NumVals[Form] + 1) { switch (i) { case 1: // The second argument is the non-atomic operand. For arithmetic, this // is always passed by value, and for a compare_exchange it is always // passed by address. For the rest, GNU uses by-address and C11 uses // by-value. assert(Form != Load); if (Form == Init || (Form == Arithmetic && ValType->isIntegerType())) Ty = ValType; else if (Form == Copy || Form == Xchg) Ty = ByValType; else if (Form == Arithmetic) Ty = Context.getPointerDiffType(); else { Expr *ValArg = TheCall->getArg(i); unsigned AS = 0; // Keep address space of non-atomic pointer type. if (const PointerType *PtrTy = ValArg->getType()->getAs
()) { AS = PtrTy->getPointeeType().getAddressSpace(); } Ty = Context.getPointerType( Context.getAddrSpaceQualType(ValType.getUnqualifiedType(), AS)); } break; case 2: // The third argument to compare_exchange / GNU exchange is a // (pointer to a) desired value. Ty = ByValType; break; case 3: // The fourth argument to GNU compare_exchange is a 'weak' flag. Ty = Context.BoolTy; break; } } else { // The order(s) are always converted to int. Ty = Context.IntTy; } InitializedEntity Entity = InitializedEntity::InitializeParameter(Context, Ty, false); ExprResult Arg = TheCall->getArg(i); Arg = PerformCopyInitialization(Entity, SourceLocation(), Arg); if (Arg.isInvalid()) return true; TheCall->setArg(i, Arg.get()); } // Permute the arguments into a 'consistent' order. SmallVector
SubExprs; SubExprs.push_back(Ptr); switch (Form) { case Init: // Note, AtomicExpr::getVal1() has a special case for this atomic. SubExprs.push_back(TheCall->getArg(1)); // Val1 break; case Load: SubExprs.push_back(TheCall->getArg(1)); // Order break; case LoadCopy: case Copy: case Arithmetic: case Xchg: SubExprs.push_back(TheCall->getArg(2)); // Order SubExprs.push_back(TheCall->getArg(1)); // Val1 break; case GNUXchg: // Note, AtomicExpr::getVal2() has a special case for this atomic. SubExprs.push_back(TheCall->getArg(3)); // Order SubExprs.push_back(TheCall->getArg(1)); // Val1 SubExprs.push_back(TheCall->getArg(2)); // Val2 break; case C11CmpXchg: SubExprs.push_back(TheCall->getArg(3)); // Order SubExprs.push_back(TheCall->getArg(1)); // Val1 SubExprs.push_back(TheCall->getArg(4)); // OrderFail SubExprs.push_back(TheCall->getArg(2)); // Val2 break; case GNUCmpXchg: SubExprs.push_back(TheCall->getArg(4)); // Order SubExprs.push_back(TheCall->getArg(1)); // Val1 SubExprs.push_back(TheCall->getArg(5)); // OrderFail SubExprs.push_back(TheCall->getArg(2)); // Val2 SubExprs.push_back(TheCall->getArg(3)); // Weak break; } if (SubExprs.size() >= 2 && Form != Init) { llvm::APSInt Result(32); if (SubExprs[1]->isIntegerConstantExpr(Result, Context) && !isValidOrderingForOp(Result.getSExtValue(), Op)) Diag(SubExprs[1]->getLocStart(), diag::warn_atomic_op_has_invalid_memory_order) << SubExprs[1]->getSourceRange(); } AtomicExpr *AE = new (Context) AtomicExpr(TheCall->getCallee()->getLocStart(), SubExprs, ResultType, Op, TheCall->getRParenLoc()); if ((Op == AtomicExpr::AO__c11_atomic_load || (Op == AtomicExpr::AO__c11_atomic_store)) && Context.AtomicUsesUnsupportedLibcall(AE)) Diag(AE->getLocStart(), diag::err_atomic_load_store_uses_lib) << ((Op == AtomicExpr::AO__c11_atomic_load) ? 0 : 1); return AE; } /// checkBuiltinArgument - Given a call to a builtin function, perform /// normal type-checking on the given argument, updating the call in /// place. This is useful when a builtin function requires custom /// type-checking for some of its arguments but not necessarily all of /// them. /// /// Returns true on error. static bool checkBuiltinArgument(Sema &S, CallExpr *E, unsigned ArgIndex) { FunctionDecl *Fn = E->getDirectCallee(); assert(Fn && "builtin call without direct callee!"); ParmVarDecl *Param = Fn->getParamDecl(ArgIndex); InitializedEntity Entity = InitializedEntity::InitializeParameter(S.Context, Param); ExprResult Arg = E->getArg(0); Arg = S.PerformCopyInitialization(Entity, SourceLocation(), Arg); if (Arg.isInvalid()) return true; E->setArg(ArgIndex, Arg.get()); return false; } /// SemaBuiltinAtomicOverloaded - We have a call to a function like /// __sync_fetch_and_add, which is an overloaded function based on the pointer /// type of its first argument. The main ActOnCallExpr routines have already /// promoted the types of arguments because all of these calls are prototyped as /// void(...). /// /// This function goes through and does final semantic checking for these /// builtins, ExprResult Sema::SemaBuiltinAtomicOverloaded(ExprResult TheCallResult) { CallExpr *TheCall = (CallExpr *)TheCallResult.get(); DeclRefExpr *DRE =cast
(TheCall->getCallee()->IgnoreParenCasts()); FunctionDecl *FDecl = cast
(DRE->getDecl()); // Ensure that we have at least one argument to do type inference from. if (TheCall->getNumArgs() < 1) { Diag(TheCall->getLocEnd(), diag::err_typecheck_call_too_few_args_at_least) << 0 << 1 << TheCall->getNumArgs() << TheCall->getCallee()->getSourceRange(); return ExprError(); } // Inspect the first argument of the atomic builtin. This should always be // a pointer type, whose element is an integral scalar or pointer type. // Because it is a pointer type, we don't have to worry about any implicit // casts here. // FIXME: We don't allow floating point scalars as input. Expr *FirstArg = TheCall->getArg(0); ExprResult FirstArgResult = DefaultFunctionArrayLvalueConversion(FirstArg); if (FirstArgResult.isInvalid()) return ExprError(); FirstArg = FirstArgResult.get(); TheCall->setArg(0, FirstArg); const PointerType *pointerType = FirstArg->getType()->getAs
(); if (!pointerType) { Diag(DRE->getLocStart(), diag::err_atomic_builtin_must_be_pointer) << FirstArg->getType() << FirstArg->getSourceRange(); return ExprError(); } QualType ValType = pointerType->getPointeeType(); if (!ValType->isIntegerType() && !ValType->isAnyPointerType() && !ValType->isBlockPointerType()) { Diag(DRE->getLocStart(), diag::err_atomic_builtin_must_be_pointer_intptr) << FirstArg->getType() << FirstArg->getSourceRange(); return ExprError(); } switch (ValType.getObjCLifetime()) { case Qualifiers::OCL_None: case Qualifiers::OCL_ExplicitNone: // okay break; case Qualifiers::OCL_Weak: case Qualifiers::OCL_Strong: case Qualifiers::OCL_Autoreleasing: Diag(DRE->getLocStart(), diag::err_arc_atomic_ownership) << ValType << FirstArg->getSourceRange(); return ExprError(); } // Strip any qualifiers off ValType. ValType = ValType.getUnqualifiedType(); // The majority of builtins return a value, but a few have special return // types, so allow them to override appropriately below. QualType ResultType = ValType; // We need to figure out which concrete builtin this maps onto. For example, // __sync_fetch_and_add with a 2 byte object turns into // __sync_fetch_and_add_2. #define BUILTIN_ROW(x) \ { Builtin::BI##x##_1, Builtin::BI##x##_2, Builtin::BI##x##_4, \ Builtin::BI##x##_8, Builtin::BI##x##_16 } static const unsigned BuiltinIndices[][5] = { BUILTIN_ROW(__sync_fetch_and_add), BUILTIN_ROW(__sync_fetch_and_sub), BUILTIN_ROW(__sync_fetch_and_or), BUILTIN_ROW(__sync_fetch_and_and), BUILTIN_ROW(__sync_fetch_and_xor), BUILTIN_ROW(__sync_fetch_and_nand), BUILTIN_ROW(__sync_add_and_fetch), BUILTIN_ROW(__sync_sub_and_fetch), BUILTIN_ROW(__sync_and_and_fetch), BUILTIN_ROW(__sync_or_and_fetch), BUILTIN_ROW(__sync_xor_and_fetch), BUILTIN_ROW(__sync_nand_and_fetch), BUILTIN_ROW(__sync_val_compare_and_swap), BUILTIN_ROW(__sync_bool_compare_and_swap), BUILTIN_ROW(__sync_lock_test_and_set), BUILTIN_ROW(__sync_lock_release), BUILTIN_ROW(__sync_swap) }; #undef BUILTIN_ROW // Determine the index of the size. unsigned SizeIndex; switch (Context.getTypeSizeInChars(ValType).getQuantity()) { case 1: SizeIndex = 0; break; case 2: SizeIndex = 1; break; case 4: SizeIndex = 2; break; case 8: SizeIndex = 3; break; case 16: SizeIndex = 4; break; default: Diag(DRE->getLocStart(), diag::err_atomic_builtin_pointer_size) << FirstArg->getType() << FirstArg->getSourceRange(); return ExprError(); } // Each of these builtins has one pointer argument, followed by some number of // values (0, 1 or 2) followed by a potentially empty varags list of stuff // that we ignore. Find out which row of BuiltinIndices to read from as well // as the number of fixed args. unsigned BuiltinID = FDecl->getBuiltinID(); unsigned BuiltinIndex, NumFixed = 1; bool WarnAboutSemanticsChange = false; switch (BuiltinID) { default: llvm_unreachable("Unknown overloaded atomic builtin!"); case Builtin::BI__sync_fetch_and_add: case Builtin::BI__sync_fetch_and_add_1: case Builtin::BI__sync_fetch_and_add_2: case Builtin::BI__sync_fetch_and_add_4: case Builtin::BI__sync_fetch_and_add_8: case Builtin::BI__sync_fetch_and_add_16: BuiltinIndex = 0; break; case Builtin::BI__sync_fetch_and_sub: case Builtin::BI__sync_fetch_and_sub_1: case Builtin::BI__sync_fetch_and_sub_2: case Builtin::BI__sync_fetch_and_sub_4: case Builtin::BI__sync_fetch_and_sub_8: case Builtin::BI__sync_fetch_and_sub_16: BuiltinIndex = 1; break; case Builtin::BI__sync_fetch_and_or: case Builtin::BI__sync_fetch_and_or_1: case Builtin::BI__sync_fetch_and_or_2: case Builtin::BI__sync_fetch_and_or_4: case Builtin::BI__sync_fetch_and_or_8: case Builtin::BI__sync_fetch_and_or_16: BuiltinIndex = 2; break; case Builtin::BI__sync_fetch_and_and: case Builtin::BI__sync_fetch_and_and_1: case Builtin::BI__sync_fetch_and_and_2: case Builtin::BI__sync_fetch_and_and_4: case Builtin::BI__sync_fetch_and_and_8: case Builtin::BI__sync_fetch_and_and_16: BuiltinIndex = 3; break; case Builtin::BI__sync_fetch_and_xor: case Builtin::BI__sync_fetch_and_xor_1: case Builtin::BI__sync_fetch_and_xor_2: case Builtin::BI__sync_fetch_and_xor_4: case Builtin::BI__sync_fetch_and_xor_8: case Builtin::BI__sync_fetch_and_xor_16: BuiltinIndex = 4; break; case Builtin::BI__sync_fetch_and_nand: case Builtin::BI__sync_fetch_and_nand_1: case Builtin::BI__sync_fetch_and_nand_2: case Builtin::BI__sync_fetch_and_nand_4: case Builtin::BI__sync_fetch_and_nand_8: case Builtin::BI__sync_fetch_and_nand_16: BuiltinIndex = 5; WarnAboutSemanticsChange = true; break; case Builtin::BI__sync_add_and_fetch: case Builtin::BI__sync_add_and_fetch_1: case Builtin::BI__sync_add_and_fetch_2: case Builtin::BI__sync_add_and_fetch_4: case Builtin::BI__sync_add_and_fetch_8: case Builtin::BI__sync_add_and_fetch_16: BuiltinIndex = 6; break; case Builtin::BI__sync_sub_and_fetch: case Builtin::BI__sync_sub_and_fetch_1: case Builtin::BI__sync_sub_and_fetch_2: case Builtin::BI__sync_sub_and_fetch_4: case Builtin::BI__sync_sub_and_fetch_8: case Builtin::BI__sync_sub_and_fetch_16: BuiltinIndex = 7; break; case Builtin::BI__sync_and_and_fetch: case Builtin::BI__sync_and_and_fetch_1: case Builtin::BI__sync_and_and_fetch_2: case Builtin::BI__sync_and_and_fetch_4: case Builtin::BI__sync_and_and_fetch_8: case Builtin::BI__sync_and_and_fetch_16: BuiltinIndex = 8; break; case Builtin::BI__sync_or_and_fetch: case Builtin::BI__sync_or_and_fetch_1: case Builtin::BI__sync_or_and_fetch_2: case Builtin::BI__sync_or_and_fetch_4: case Builtin::BI__sync_or_and_fetch_8: case Builtin::BI__sync_or_and_fetch_16: BuiltinIndex = 9; break; case Builtin::BI__sync_xor_and_fetch: case Builtin::BI__sync_xor_and_fetch_1: case Builtin::BI__sync_xor_and_fetch_2: case Builtin::BI__sync_xor_and_fetch_4: case Builtin::BI__sync_xor_and_fetch_8: case Builtin::BI__sync_xor_and_fetch_16: BuiltinIndex = 10; break; case Builtin::BI__sync_nand_and_fetch: case Builtin::BI__sync_nand_and_fetch_1: case Builtin::BI__sync_nand_and_fetch_2: case Builtin::BI__sync_nand_and_fetch_4: case Builtin::BI__sync_nand_and_fetch_8: case Builtin::BI__sync_nand_and_fetch_16: BuiltinIndex = 11; WarnAboutSemanticsChange = true; break; case Builtin::BI__sync_val_compare_and_swap: case Builtin::BI__sync_val_compare_and_swap_1: case Builtin::BI__sync_val_compare_and_swap_2: case Builtin::BI__sync_val_compare_and_swap_4: case Builtin::BI__sync_val_compare_and_swap_8: case Builtin::BI__sync_val_compare_and_swap_16: BuiltinIndex = 12; NumFixed = 2; break; case Builtin::BI__sync_bool_compare_and_swap: case Builtin::BI__sync_bool_compare_and_swap_1: case Builtin::BI__sync_bool_compare_and_swap_2: case Builtin::BI__sync_bool_compare_and_swap_4: case Builtin::BI__sync_bool_compare_and_swap_8: case Builtin::BI__sync_bool_compare_and_swap_16: BuiltinIndex = 13; NumFixed = 2; ResultType = Context.BoolTy; break; case Builtin::BI__sync_lock_test_and_set: case Builtin::BI__sync_lock_test_and_set_1: case Builtin::BI__sync_lock_test_and_set_2: case Builtin::BI__sync_lock_test_and_set_4: case Builtin::BI__sync_lock_test_and_set_8: case Builtin::BI__sync_lock_test_and_set_16: BuiltinIndex = 14; break; case Builtin::BI__sync_lock_release: case Builtin::BI__sync_lock_release_1: case Builtin::BI__sync_lock_release_2: case Builtin::BI__sync_lock_release_4: case Builtin::BI__sync_lock_release_8: case Builtin::BI__sync_lock_release_16: BuiltinIndex = 15; NumFixed = 0; ResultType = Context.VoidTy; break; case Builtin::BI__sync_swap: case Builtin::BI__sync_swap_1: case Builtin::BI__sync_swap_2: case Builtin::BI__sync_swap_4: case Builtin::BI__sync_swap_8: case Builtin::BI__sync_swap_16: BuiltinIndex = 16; break; } // Now that we know how many fixed arguments we expect, first check that we // have at least that many. if (TheCall->getNumArgs() < 1+NumFixed) { Diag(TheCall->getLocEnd(), diag::err_typecheck_call_too_few_args_at_least) << 0 << 1+NumFixed << TheCall->getNumArgs() << TheCall->getCallee()->getSourceRange(); return ExprError(); } if (WarnAboutSemanticsChange) { Diag(TheCall->getLocEnd(), diag::warn_sync_fetch_and_nand_semantics_change) << TheCall->getCallee()->getSourceRange(); } // Get the decl for the concrete builtin from this, we can tell what the // concrete integer type we should convert to is. unsigned NewBuiltinID = BuiltinIndices[BuiltinIndex][SizeIndex]; const char *NewBuiltinName = Context.BuiltinInfo.getName(NewBuiltinID); FunctionDecl *NewBuiltinDecl; if (NewBuiltinID == BuiltinID) NewBuiltinDecl = FDecl; else { // Perform builtin lookup to avoid redeclaring it. DeclarationName DN(&Context.Idents.get(NewBuiltinName)); LookupResult Res(*this, DN, DRE->getLocStart(), LookupOrdinaryName); LookupName(Res, TUScope, /*AllowBuiltinCreation=*/true); assert(Res.getFoundDecl()); NewBuiltinDecl = dyn_cast
(Res.getFoundDecl()); if (!NewBuiltinDecl) return ExprError(); } // The first argument --- the pointer --- has a fixed type; we // deduce the types of the rest of the arguments accordingly. Walk // the remaining arguments, converting them to the deduced value type. for (unsigned i = 0; i != NumFixed; ++i) { ExprResult Arg = TheCall->getArg(i+1); // GCC does an implicit conversion to the pointer or integer ValType. This // can fail in some cases (1i -> int**), check for this error case now. // Initialize the argument. InitializedEntity Entity = InitializedEntity::InitializeParameter(Context, ValType, /*consume*/ false); Arg = PerformCopyInitialization(Entity, SourceLocation(), Arg); if (Arg.isInvalid()) return ExprError(); // Okay, we have something that *can* be converted to the right type. Check // to see if there is a potentially weird extension going on here. This can // happen when you do an atomic operation on something like an char* and // pass in 42. The 42 gets converted to char. This is even more strange // for things like 45.123 -> char, etc. // FIXME: Do this check. TheCall->setArg(i+1, Arg.get()); } ASTContext& Context = this->getASTContext(); // Create a new DeclRefExpr to refer to the new decl. DeclRefExpr* NewDRE = DeclRefExpr::Create( Context, DRE->getQualifierLoc(), SourceLocation(), NewBuiltinDecl, /*enclosing*/ false, DRE->getLocation(), Context.BuiltinFnTy, DRE->getValueKind()); // Set the callee in the CallExpr. // FIXME: This loses syntactic information. QualType CalleePtrTy = Context.getPointerType(NewBuiltinDecl->getType()); ExprResult PromotedCall = ImpCastExprToType(NewDRE, CalleePtrTy, CK_BuiltinFnToFnPtr); TheCall->setCallee(PromotedCall.get()); // Change the result type of the call to match the original value type. This // is arbitrary, but the codegen for these builtins ins design to handle it // gracefully. TheCall->setType(ResultType); return TheCallResult; } /// SemaBuiltinNontemporalOverloaded - We have a call to /// __builtin_nontemporal_store or __builtin_nontemporal_load, which is an /// overloaded function based on the pointer type of its last argument. /// /// This function goes through and does final semantic checking for these /// builtins. ExprResult Sema::SemaBuiltinNontemporalOverloaded(ExprResult TheCallResult) { CallExpr *TheCall = (CallExpr *)TheCallResult.get(); DeclRefExpr *DRE = cast
(TheCall->getCallee()->IgnoreParenCasts()); FunctionDecl *FDecl = cast
(DRE->getDecl()); unsigned BuiltinID = FDecl->getBuiltinID(); assert((BuiltinID == Builtin::BI__builtin_nontemporal_store || BuiltinID == Builtin::BI__builtin_nontemporal_load) && "Unexpected nontemporal load/store builtin!"); bool isStore = BuiltinID == Builtin::BI__builtin_nontemporal_store; unsigned numArgs = isStore ? 2 : 1; // Ensure that we have the proper number of arguments. if (checkArgCount(*this, TheCall, numArgs)) return ExprError(); // Inspect the last argument of the nontemporal builtin. This should always // be a pointer type, from which we imply the type of the memory access. // Because it is a pointer type, we don't have to worry about any implicit // casts here. Expr *PointerArg = TheCall->getArg(numArgs - 1); ExprResult PointerArgResult = DefaultFunctionArrayLvalueConversion(PointerArg); if (PointerArgResult.isInvalid()) return ExprError(); PointerArg = PointerArgResult.get(); TheCall->setArg(numArgs - 1, PointerArg); const PointerType *pointerType = PointerArg->getType()->getAs
(); if (!pointerType) { Diag(DRE->getLocStart(), diag::err_nontemporal_builtin_must_be_pointer) << PointerArg->getType() << PointerArg->getSourceRange(); return ExprError(); } QualType ValType = pointerType->getPointeeType(); // Strip any qualifiers off ValType. ValType = ValType.getUnqualifiedType(); if (!ValType->isIntegerType() && !ValType->isAnyPointerType() && !ValType->isBlockPointerType() && !ValType->isFloatingType() && !ValType->isVectorType()) { Diag(DRE->getLocStart(), diag::err_nontemporal_builtin_must_be_pointer_intfltptr_or_vector) << PointerArg->getType() << PointerArg->getSourceRange(); return ExprError(); } if (!isStore) { TheCall->setType(ValType); return TheCallResult; } ExprResult ValArg = TheCall->getArg(0); InitializedEntity Entity = InitializedEntity::InitializeParameter( Context, ValType, /*consume*/ false); ValArg = PerformCopyInitialization(Entity, SourceLocation(), ValArg); if (ValArg.isInvalid()) return ExprError(); TheCall->setArg(0, ValArg.get()); TheCall->setType(Context.VoidTy); return TheCallResult; } /// CheckObjCString - Checks that the argument to the builtin /// CFString constructor is correct /// Note: It might also make sense to do the UTF-16 conversion here (would /// simplify the backend). bool Sema::CheckObjCString(Expr *Arg) { Arg = Arg->IgnoreParenCasts(); StringLiteral *Literal = dyn_cast
(Arg); if (!Literal || !Literal->isAscii()) { Diag(Arg->getLocStart(), diag::err_cfstring_literal_not_string_constant) << Arg->getSourceRange(); return true; } if (Literal->containsNonAsciiOrNull()) { StringRef String = Literal->getString(); unsigned NumBytes = String.size(); SmallVector
ToBuf(NumBytes); const UTF8 *FromPtr = (const UTF8 *)String.data(); UTF16 *ToPtr = &ToBuf[0]; ConversionResult Result = ConvertUTF8toUTF16(&FromPtr, FromPtr + NumBytes, &ToPtr, ToPtr + NumBytes, strictConversion); // Check for conversion failure. if (Result != conversionOK) Diag(Arg->getLocStart(), diag::warn_cfstring_truncated) << Arg->getSourceRange(); } return false; } /// Check the arguments to '__builtin_va_start' or '__builtin_ms_va_start' /// for validity. Emit an error and return true on failure; return false /// on success. bool Sema::SemaBuiltinVAStartImpl(CallExpr *TheCall) { Expr *Fn = TheCall->getCallee(); if (TheCall->getNumArgs() > 2) { Diag(TheCall->getArg(2)->getLocStart(), diag::err_typecheck_call_too_many_args) << 0 /*function call*/ << 2 << TheCall->getNumArgs() << Fn->getSourceRange() << SourceRange(TheCall->getArg(2)->getLocStart(), (*(TheCall->arg_end()-1))->getLocEnd()); return true; } if (TheCall->getNumArgs() < 2) { return Diag(TheCall->getLocEnd(), diag::err_typecheck_call_too_few_args_at_least) << 0 /*function call*/ << 2 << TheCall->getNumArgs(); } // Type-check the first argument normally. if (checkBuiltinArgument(*this, TheCall, 0)) return true; // Determine whether the current function is variadic or not. BlockScopeInfo *CurBlock = getCurBlock(); bool isVariadic; if (CurBlock) isVariadic = CurBlock->TheDecl->isVariadic(); else if (FunctionDecl *FD = getCurFunctionDecl()) isVariadic = FD->isVariadic(); else isVariadic = getCurMethodDecl()->isVariadic(); if (!isVariadic) { Diag(Fn->getLocStart(), diag::err_va_start_used_in_non_variadic_function); return true; } // Verify that the second argument to the builtin is the last argument of the // current function or method. bool SecondArgIsLastNamedArgument = false; const Expr *Arg = TheCall->getArg(1)->IgnoreParenCasts(); // These are valid if SecondArgIsLastNamedArgument is false after the next // block. QualType Type; SourceLocation ParamLoc; bool IsCRegister = false; if (const DeclRefExpr *DR = dyn_cast
(Arg)) { if (const ParmVarDecl *PV = dyn_cast
(DR->getDecl())) { // FIXME: This isn't correct for methods (results in bogus warning). // Get the last formal in the current function. const ParmVarDecl *LastArg; if (CurBlock) LastArg = CurBlock->TheDecl->parameters().back(); else if (FunctionDecl *FD = getCurFunctionDecl()) LastArg = FD->parameters().back(); else LastArg = getCurMethodDecl()->parameters().back(); SecondArgIsLastNamedArgument = PV == LastArg; Type = PV->getType(); ParamLoc = PV->getLocation(); IsCRegister = PV->getStorageClass() == SC_Register && !getLangOpts().CPlusPlus; } } if (!SecondArgIsLastNamedArgument) Diag(TheCall->getArg(1)->getLocStart(), diag::warn_second_arg_of_va_start_not_last_named_param); else if (IsCRegister || Type->isReferenceType() || Type->isPromotableIntegerType() || Type->isSpecificBuiltinType(BuiltinType::Float)) { unsigned Reason = 0; if (Type->isReferenceType()) Reason = 1; else if (IsCRegister) Reason = 2; Diag(Arg->getLocStart(), diag::warn_va_start_type_is_undefined) << Reason; Diag(ParamLoc, diag::note_parameter_type) << Type; } TheCall->setType(Context.VoidTy); return false; } /// Check the arguments to '__builtin_va_start' for validity, and that /// it was called from a function of the native ABI. /// Emit an error and return true on failure; return false on success. bool Sema::SemaBuiltinVAStart(CallExpr *TheCall) { // On x86-64 Unix, don't allow this in Win64 ABI functions. // On x64 Windows, don't allow this in System V ABI functions. // (Yes, that means there's no corresponding way to support variadic // System V ABI functions on Windows.) if (Context.getTargetInfo().getTriple().getArch() == llvm::Triple::x86_64) { unsigned OS = Context.getTargetInfo().getTriple().getOS(); clang::CallingConv CC = CC_C; if (const FunctionDecl *FD = getCurFunctionDecl()) CC = FD->getType()->getAs
()->getCallConv(); if ((OS == llvm::Triple::Win32 && CC == CC_X86_64SysV) || (OS != llvm::Triple::Win32 && CC == CC_X86_64Win64)) return Diag(TheCall->getCallee()->getLocStart(), diag::err_va_start_used_in_wrong_abi_function) << (OS != llvm::Triple::Win32); } return SemaBuiltinVAStartImpl(TheCall); } /// Check the arguments to '__builtin_ms_va_start' for validity, and that /// it was called from a Win64 ABI function. /// Emit an error and return true on failure; return false on success. bool Sema::SemaBuiltinMSVAStart(CallExpr *TheCall) { // This only makes sense for x86-64. const llvm::Triple &TT = Context.getTargetInfo().getTriple(); Expr *Callee = TheCall->getCallee(); if (TT.getArch() != llvm::Triple::x86_64) return Diag(Callee->getLocStart(), diag::err_x86_builtin_32_bit_tgt); // Don't allow this in System V ABI functions. clang::CallingConv CC = CC_C; if (const FunctionDecl *FD = getCurFunctionDecl()) CC = FD->getType()->getAs
()->getCallConv(); if (CC == CC_X86_64SysV || (TT.getOS() != llvm::Triple::Win32 && CC != CC_X86_64Win64)) return Diag(Callee->getLocStart(), diag::err_ms_va_start_used_in_sysv_function); return SemaBuiltinVAStartImpl(TheCall); } bool Sema::SemaBuiltinVAStartARM(CallExpr *Call) { // void __va_start(va_list *ap, const char *named_addr, size_t slot_size, // const char *named_addr); Expr *Func = Call->getCallee(); if (Call->getNumArgs() < 3) return Diag(Call->getLocEnd(), diag::err_typecheck_call_too_few_args_at_least) << 0 /*function call*/ << 3 << Call->getNumArgs(); // Determine whether the current function is variadic or not. bool IsVariadic; if (BlockScopeInfo *CurBlock = getCurBlock()) IsVariadic = CurBlock->TheDecl->isVariadic(); else if (FunctionDecl *FD = getCurFunctionDecl()) IsVariadic = FD->isVariadic(); else if (ObjCMethodDecl *MD = getCurMethodDecl()) IsVariadic = MD->isVariadic(); else llvm_unreachable("unexpected statement type"); if (!IsVariadic) { Diag(Func->getLocStart(), diag::err_va_start_used_in_non_variadic_function); return true; } // Type-check the first argument normally. if (checkBuiltinArgument(*this, Call, 0)) return true; const struct { unsigned ArgNo; QualType Type; } ArgumentTypes[] = { { 1, Context.getPointerType(Context.CharTy.withConst()) }, { 2, Context.getSizeType() }, }; for (const auto &AT : ArgumentTypes) { const Expr *Arg = Call->getArg(AT.ArgNo)->IgnoreParens(); if (Arg->getType().getCanonicalType() == AT.Type.getCanonicalType()) continue; Diag(Arg->getLocStart(), diag::err_typecheck_convert_incompatible) << Arg->getType() << AT.Type << 1 /* different class */ << 0 /* qualifier difference */ << 3 /* parameter mismatch */ << AT.ArgNo + 1 << Arg->getType() << AT.Type; } return false; } /// SemaBuiltinUnorderedCompare - Handle functions like __builtin_isgreater and /// friends. This is declared to take (...), so we have to check everything. bool Sema::SemaBuiltinUnorderedCompare(CallExpr *TheCall) { if (TheCall->getNumArgs() < 2) return Diag(TheCall->getLocEnd(), diag::err_typecheck_call_too_few_args) << 0 << 2 << TheCall->getNumArgs()/*function call*/; if (TheCall->getNumArgs() > 2) return Diag(TheCall->getArg(2)->getLocStart(), diag::err_typecheck_call_too_many_args) << 0 /*function call*/ << 2 << TheCall->getNumArgs() << SourceRange(TheCall->getArg(2)->getLocStart(), (*(TheCall->arg_end()-1))->getLocEnd()); ExprResult OrigArg0 = TheCall->getArg(0); ExprResult OrigArg1 = TheCall->getArg(1); // Do standard promotions between the two arguments, returning their common // type. QualType Res = UsualArithmeticConversions(OrigArg0, OrigArg1, false); if (OrigArg0.isInvalid() || OrigArg1.isInvalid()) return true; // Make sure any conversions are pushed back into the call; this is // type safe since unordered compare builtins are declared as "_Bool // foo(...)". TheCall->setArg(0, OrigArg0.get()); TheCall->setArg(1, OrigArg1.get()); if (OrigArg0.get()->isTypeDependent() || OrigArg1.get()->isTypeDependent()) return false; // If the common type isn't a real floating type, then the arguments were // invalid for this operation. if (Res.isNull() || !Res->isRealFloatingType()) return Diag(OrigArg0.get()->getLocStart(), diag::err_typecheck_call_invalid_ordered_compare) << OrigArg0.get()->getType() << OrigArg1.get()->getType() << SourceRange(OrigArg0.get()->getLocStart(), OrigArg1.get()->getLocEnd()); return false; } /// SemaBuiltinSemaBuiltinFPClassification - Handle functions like /// __builtin_isnan and friends. This is declared to take (...), so we have /// to check everything. We expect the last argument to be a floating point /// value. bool Sema::SemaBuiltinFPClassification(CallExpr *TheCall, unsigned NumArgs) { if (TheCall->getNumArgs() < NumArgs) return Diag(TheCall->getLocEnd(), diag::err_typecheck_call_too_few_args) << 0 << NumArgs << TheCall->getNumArgs()/*function call*/; if (TheCall->getNumArgs() > NumArgs) return Diag(TheCall->getArg(NumArgs)->getLocStart(), diag::err_typecheck_call_too_many_args) << 0 /*function call*/ << NumArgs << TheCall->getNumArgs() << SourceRange(TheCall->getArg(NumArgs)->getLocStart(), (*(TheCall->arg_end()-1))->getLocEnd()); Expr *OrigArg = TheCall->getArg(NumArgs-1); if (OrigArg->isTypeDependent()) return false; // This operation requires a non-_Complex floating-point number. if (!OrigArg->getType()->isRealFloatingType()) return Diag(OrigArg->getLocStart(), diag::err_typecheck_call_invalid_unary_fp) << OrigArg->getType() << OrigArg->getSourceRange(); // If this is an implicit conversion from float -> double, remove it. if (ImplicitCastExpr *Cast = dyn_cast
(OrigArg)) { Expr *CastArg = Cast->getSubExpr(); if (CastArg->getType()->isSpecificBuiltinType(BuiltinType::Float)) { assert(Cast->getType()->isSpecificBuiltinType(BuiltinType::Double) && "promotion from float to double is the only expected cast here"); Cast->setSubExpr(nullptr); TheCall->setArg(NumArgs-1, CastArg); } } return false; } /// SemaBuiltinShuffleVector - Handle __builtin_shufflevector. // This is declared to take (...), so we have to check everything. ExprResult Sema::SemaBuiltinShuffleVector(CallExpr *TheCall) { if (TheCall->getNumArgs() < 2) return ExprError(Diag(TheCall->getLocEnd(), diag::err_typecheck_call_too_few_args_at_least) << 0 /*function call*/ << 2 << TheCall->getNumArgs() << TheCall->getSourceRange()); // Determine which of the following types of shufflevector we're checking: // 1) unary, vector mask: (lhs, mask) // 2) binary, scalar mask: (lhs, rhs, index, ..., index) QualType resType = TheCall->getArg(0)->getType(); unsigned numElements = 0; if (!TheCall->getArg(0)->isTypeDependent() && !TheCall->getArg(1)->isTypeDependent()) { QualType LHSType = TheCall->getArg(0)->getType(); QualType RHSType = TheCall->getArg(1)->getType(); if (!LHSType->isVectorType() || !RHSType->isVectorType()) return ExprError(Diag(TheCall->getLocStart(), diag::err_shufflevector_non_vector) << SourceRange(TheCall->getArg(0)->getLocStart(), TheCall->getArg(1)->getLocEnd())); numElements = LHSType->getAs
()->getNumElements(); unsigned numResElements = TheCall->getNumArgs() - 2; // Check to see if we have a call with 2 vector arguments, the unary shuffle // with mask. If so, verify that RHS is an integer vector type with the // same number of elts as lhs. if (TheCall->getNumArgs() == 2) { if (!RHSType->hasIntegerRepresentation() || RHSType->getAs
()->getNumElements() != numElements) return ExprError(Diag(TheCall->getLocStart(), diag::err_shufflevector_incompatible_vector) << SourceRange(TheCall->getArg(1)->getLocStart(), TheCall->getArg(1)->getLocEnd())); } else if (!Context.hasSameUnqualifiedType(LHSType, RHSType)) { return ExprError(Diag(TheCall->getLocStart(), diag::err_shufflevector_incompatible_vector) << SourceRange(TheCall->getArg(0)->getLocStart(), TheCall->getArg(1)->getLocEnd())); } else if (numElements != numResElements) { QualType eltType = LHSType->getAs
()->getElementType(); resType = Context.getVectorType(eltType, numResElements, VectorType::GenericVector); } } for (unsigned i = 2; i < TheCall->getNumArgs(); i++) { if (TheCall->getArg(i)->isTypeDependent() || TheCall->getArg(i)->isValueDependent()) continue; llvm::APSInt Result(32); if (!TheCall->getArg(i)->isIntegerConstantExpr(Result, Context)) return ExprError(Diag(TheCall->getLocStart(), diag::err_shufflevector_nonconstant_argument) << TheCall->getArg(i)->getSourceRange()); // Allow -1 which will be translated to undef in the IR. if (Result.isSigned() && Result.isAllOnesValue()) continue; if (Result.getActiveBits() > 64 || Result.getZExtValue() >= numElements*2) return ExprError(Diag(TheCall->getLocStart(), diag::err_shufflevector_argument_too_large) << TheCall->getArg(i)->getSourceRange()); } SmallVector
exprs; for (unsigned i = 0, e = TheCall->getNumArgs(); i != e; i++) { exprs.push_back(TheCall->getArg(i)); TheCall->setArg(i, nullptr); } return new (Context) ShuffleVectorExpr(Context, exprs, resType, TheCall->getCallee()->getLocStart(), TheCall->getRParenLoc()); } /// SemaConvertVectorExpr - Handle __builtin_convertvector ExprResult Sema::SemaConvertVectorExpr(Expr *E, TypeSourceInfo *TInfo, SourceLocation BuiltinLoc, SourceLocation RParenLoc) { ExprValueKind VK = VK_RValue; ExprObjectKind OK = OK_Ordinary; QualType DstTy = TInfo->getType(); QualType SrcTy = E->getType(); if (!SrcTy->isVectorType() && !SrcTy->isDependentType()) return ExprError(Diag(BuiltinLoc, diag::err_convertvector_non_vector) << E->getSourceRange()); if (!DstTy->isVectorType() && !DstTy->isDependentType()) return ExprError(Diag(BuiltinLoc, diag::err_convertvector_non_vector_type)); if (!SrcTy->isDependentType() && !DstTy->isDependentType()) { unsigned SrcElts = SrcTy->getAs
()->getNumElements(); unsigned DstElts = DstTy->getAs
()->getNumElements(); if (SrcElts != DstElts) return ExprError(Diag(BuiltinLoc, diag::err_convertvector_incompatible_vector) << E->getSourceRange()); } return new (Context) ConvertVectorExpr(E, TInfo, DstTy, VK, OK, BuiltinLoc, RParenLoc); } /// SemaBuiltinPrefetch - Handle __builtin_prefetch. // This is declared to take (const void*, ...) and can take two // optional constant int args. bool Sema::SemaBuiltinPrefetch(CallExpr *TheCall) { unsigned NumArgs = TheCall->getNumArgs(); if (NumArgs > 3) return Diag(TheCall->getLocEnd(), diag::err_typecheck_call_too_many_args_at_most) << 0 /*function call*/ << 3 << NumArgs << TheCall->getSourceRange(); // Argument 0 is checked for us and the remaining arguments must be // constant integers. for (unsigned i = 1; i != NumArgs; ++i) if (SemaBuiltinConstantArgRange(TheCall, i, 0, i == 1 ? 1 : 3)) return true; return false; } /// SemaBuiltinAssume - Handle __assume (MS Extension). // __assume does not evaluate its arguments, and should warn if its argument // has side effects. bool Sema::SemaBuiltinAssume(CallExpr *TheCall) { Expr *Arg = TheCall->getArg(0); if (Arg->isInstantiationDependent()) return false; if (Arg->HasSideEffects(Context)) Diag(Arg->getLocStart(), diag::warn_assume_side_effects) << Arg->getSourceRange() << cast
(TheCall->getCalleeDecl())->getIdentifier(); return false; } /// Handle __builtin_assume_aligned. This is declared /// as (const void*, size_t, ...) and can take one optional constant int arg. bool Sema::SemaBuiltinAssumeAligned(CallExpr *TheCall) { unsigned NumArgs = TheCall->getNumArgs(); if (NumArgs > 3) return Diag(TheCall->getLocEnd(), diag::err_typecheck_call_too_many_args_at_most) << 0 /*function call*/ << 3 << NumArgs << TheCall->getSourceRange(); // The alignment must be a constant integer. Expr *Arg = TheCall->getArg(1); // We can't check the value of a dependent argument. if (!Arg->isTypeDependent() && !Arg->isValueDependent()) { llvm::APSInt Result; if (SemaBuiltinConstantArg(TheCall, 1, Result)) return true; if (!Result.isPowerOf2()) return Diag(TheCall->getLocStart(), diag::err_alignment_not_power_of_two) << Arg->getSourceRange(); } if (NumArgs > 2) { ExprResult Arg(TheCall->getArg(2)); InitializedEntity Entity = InitializedEntity::InitializeParameter(Context, Context.getSizeType(), false); Arg = PerformCopyInitialization(Entity, SourceLocation(), Arg); if (Arg.isInvalid()) return true; TheCall->setArg(2, Arg.get()); } return false; } /// SemaBuiltinConstantArg - Handle a check if argument ArgNum of CallExpr /// TheCall is a constant expression. bool Sema::SemaBuiltinConstantArg(CallExpr *TheCall, int ArgNum, llvm::APSInt &Result) { Expr *Arg = TheCall->getArg(ArgNum); DeclRefExpr *DRE =cast
(TheCall->getCallee()->IgnoreParenCasts()); FunctionDecl *FDecl = cast
(DRE->getDecl()); if (Arg->isTypeDependent() || Arg->isValueDependent()) return false; if (!Arg->isIntegerConstantExpr(Result, Context)) return Diag(TheCall->getLocStart(), diag::err_constant_integer_arg_type) << FDecl->getDeclName() << Arg->getSourceRange(); return false; } /// SemaBuiltinConstantArgRange - Handle a check if argument ArgNum of CallExpr /// TheCall is a constant expression in the range [Low, High]. bool Sema::SemaBuiltinConstantArgRange(CallExpr *TheCall, int ArgNum, int Low, int High) { llvm::APSInt Result; // We can't check the value of a dependent argument. Expr *Arg = TheCall->getArg(ArgNum); if (Arg->isTypeDependent() || Arg->isValueDependent()) return false; // Check constant-ness first. if (SemaBuiltinConstantArg(TheCall, ArgNum, Result)) return true; if (Result.getSExtValue() < Low || Result.getSExtValue() > High) return Diag(TheCall->getLocStart(), diag::err_argument_invalid_range) << Low << High << Arg->getSourceRange(); return false; } /// SemaBuiltinARMSpecialReg - Handle a check if argument ArgNum of CallExpr /// TheCall is an ARM/AArch64 special register string literal. bool Sema::SemaBuiltinARMSpecialReg(unsigned BuiltinID, CallExpr *TheCall, int ArgNum, unsigned ExpectedFieldNum, bool AllowName) { bool IsARMBuiltin = BuiltinID == ARM::BI__builtin_arm_rsr64 || BuiltinID == ARM::BI__builtin_arm_wsr64 || BuiltinID == ARM::BI__builtin_arm_rsr || BuiltinID == ARM::BI__builtin_arm_rsrp || BuiltinID == ARM::BI__builtin_arm_wsr || BuiltinID == ARM::BI__builtin_arm_wsrp; bool IsAArch64Builtin = BuiltinID == AArch64::BI__builtin_arm_rsr64 || BuiltinID == AArch64::BI__builtin_arm_wsr64 || BuiltinID == AArch64::BI__builtin_arm_rsr || BuiltinID == AArch64::BI__builtin_arm_rsrp || BuiltinID == AArch64::BI__builtin_arm_wsr || BuiltinID == AArch64::BI__builtin_arm_wsrp; assert((IsARMBuiltin || IsAArch64Builtin) && "Unexpected ARM builtin."); // We can't check the value of a dependent argument. Expr *Arg = TheCall->getArg(ArgNum); if (Arg->isTypeDependent() || Arg->isValueDependent()) return false; // Check if the argument is a string literal. if (!isa
(Arg->IgnoreParenImpCasts())) return Diag(TheCall->getLocStart(), diag::err_expr_not_string_literal) << Arg->getSourceRange(); // Check the type of special register given. StringRef Reg = cast
(Arg->IgnoreParenImpCasts())->getString(); SmallVector
Fields; Reg.split(Fields, ":"); if (Fields.size() != ExpectedFieldNum && !(AllowName && Fields.size() == 1)) return Diag(TheCall->getLocStart(), diag::err_arm_invalid_specialreg) << Arg->getSourceRange(); // If the string is the name of a register then we cannot check that it is // valid here but if the string is of one the forms described in ACLE then we // can check that the supplied fields are integers and within the valid // ranges. if (Fields.size() > 1) { bool FiveFields = Fields.size() == 5; bool ValidString = true; if (IsARMBuiltin) { ValidString &= Fields[0].startswith_lower("cp") || Fields[0].startswith_lower("p"); if (ValidString) Fields[0] = Fields[0].drop_front(Fields[0].startswith_lower("cp") ? 2 : 1); ValidString &= Fields[2].startswith_lower("c"); if (ValidString) Fields[2] = Fields[2].drop_front(1); if (FiveFields) { ValidString &= Fields[3].startswith_lower("c"); if (ValidString) Fields[3] = Fields[3].drop_front(1); } } SmallVector
Ranges; if (FiveFields) Ranges.append({IsAArch64Builtin ? 1 : 15, 7, 7, 15, 15}); else Ranges.append({15, 7, 15}); for (unsigned i=0; i
= 0 && IntField <= Ranges[i]); } if (!ValidString) return Diag(TheCall->getLocStart(), diag::err_arm_invalid_specialreg) << Arg->getSourceRange(); } else if (IsAArch64Builtin && Fields.size() == 1) { // If the register name is one of those that appear in the condition below // and the special register builtin being used is one of the write builtins, // then we require that the argument provided for writing to the register // is an integer constant expression. This is because it will be lowered to // an MSR (immediate) instruction, so we need to know the immediate at // compile time. if (TheCall->getNumArgs() != 2) return false; std::string RegLower = Reg.lower(); if (RegLower != "spsel" && RegLower != "daifset" && RegLower != "daifclr" && RegLower != "pan" && RegLower != "uao") return false; return SemaBuiltinConstantArgRange(TheCall, 1, 0, 15); } return false; } /// SemaBuiltinLongjmp - Handle __builtin_longjmp(void *env[5], int val). /// This checks that the target supports __builtin_longjmp and /// that val is a constant 1. bool Sema::SemaBuiltinLongjmp(CallExpr *TheCall) { if (!Context.getTargetInfo().hasSjLjLowering()) return Diag(TheCall->getLocStart(), diag::err_builtin_longjmp_unsupported) << SourceRange(TheCall->getLocStart(), TheCall->getLocEnd()); Expr *Arg = TheCall->getArg(1); llvm::APSInt Result; // TODO: This is less than ideal. Overload this to take a value. if (SemaBuiltinConstantArg(TheCall, 1, Result)) return true; if (Result != 1) return Diag(TheCall->getLocStart(), diag::err_builtin_longjmp_invalid_val) << SourceRange(Arg->getLocStart(), Arg->getLocEnd()); return false; } /// SemaBuiltinSetjmp - Handle __builtin_setjmp(void *env[5]). /// This checks that the target supports __builtin_setjmp. bool Sema::SemaBuiltinSetjmp(CallExpr *TheCall) { if (!Context.getTargetInfo().hasSjLjLowering()) return Diag(TheCall->getLocStart(), diag::err_builtin_setjmp_unsupported) << SourceRange(TheCall->getLocStart(), TheCall->getLocEnd()); return false; } namespace { class UncoveredArgHandler { enum { Unknown = -1, AllCovered = -2 }; signed FirstUncoveredArg; SmallVector
DiagnosticExprs; public: UncoveredArgHandler() : FirstUncoveredArg(Unknown) { } bool hasUncoveredArg() const { return (FirstUncoveredArg >= 0); } unsigned getUncoveredArg() const { assert(hasUncoveredArg() && "no uncovered argument"); return FirstUncoveredArg; } void setAllCovered() { // A string has been found with all arguments covered, so clear out // the diagnostics. DiagnosticExprs.clear(); FirstUncoveredArg = AllCovered; } void Update(signed NewFirstUncoveredArg, const Expr *StrExpr) { assert(NewFirstUncoveredArg >= 0 && "Outside range"); // Don't update if a previous string covers all arguments. if (FirstUncoveredArg == AllCovered) return; // UncoveredArgHandler tracks the highest uncovered argument index // and with it all the strings that match this index. if (NewFirstUncoveredArg == FirstUncoveredArg) DiagnosticExprs.push_back(StrExpr); else if (NewFirstUncoveredArg > FirstUncoveredArg) { DiagnosticExprs.clear(); DiagnosticExprs.push_back(StrExpr); FirstUncoveredArg = NewFirstUncoveredArg; } } void Diagnose(Sema &S, bool IsFunctionCall, const Expr *ArgExpr); }; enum StringLiteralCheckType { SLCT_NotALiteral, SLCT_UncheckedLiteral, SLCT_CheckedLiteral }; } // end anonymous namespace static void CheckFormatString(Sema &S, const StringLiteral *FExpr, const Expr *OrigFormatExpr, ArrayRef
Args, bool HasVAListArg, unsigned format_idx, unsigned firstDataArg, Sema::FormatStringType Type, bool inFunctionCall, Sema::VariadicCallType CallType, llvm::SmallBitVector &CheckedVarArgs, UncoveredArgHandler &UncoveredArg); // Determine if an expression is a string literal or constant string. // If this function returns false on the arguments to a function expecting a // format string, we will usually need to emit a warning. // True string literals are then checked by CheckFormatString. static StringLiteralCheckType checkFormatStringExpr(Sema &S, const Expr *E, ArrayRef
Args, bool HasVAListArg, unsigned format_idx, unsigned firstDataArg, Sema::FormatStringType Type, Sema::VariadicCallType CallType, bool InFunctionCall, llvm::SmallBitVector &CheckedVarArgs, UncoveredArgHandler &UncoveredArg) { tryAgain: if (E->isTypeDependent() || E->isValueDependent()) return SLCT_NotALiteral; E = E->IgnoreParenCasts(); if (E->isNullPointerConstant(S.Context, Expr::NPC_ValueDependentIsNotNull)) // Technically -Wformat-nonliteral does not warn about this case. // The behavior of printf and friends in this case is implementation // dependent. Ideally if the format string cannot be null then // it should have a 'nonnull' attribute in the function prototype. return SLCT_UncheckedLiteral; switch (E->getStmtClass()) { case Stmt::BinaryConditionalOperatorClass: case Stmt::ConditionalOperatorClass: { // The expression is a literal if both sub-expressions were, and it was // completely checked only if both sub-expressions were checked. const AbstractConditionalOperator *C = cast
(E); // Determine whether it is necessary to check both sub-expressions, for // example, because the condition expression is a constant that can be // evaluated at compile time. bool CheckLeft = true, CheckRight = true; bool Cond; if (C->getCond()->EvaluateAsBooleanCondition(Cond, S.getASTContext())) { if (Cond) CheckRight = false; else CheckLeft = false; } StringLiteralCheckType Left; if (!CheckLeft) Left = SLCT_UncheckedLiteral; else { Left = checkFormatStringExpr(S, C->getTrueExpr(), Args, HasVAListArg, format_idx, firstDataArg, Type, CallType, InFunctionCall, CheckedVarArgs, UncoveredArg); if (Left == SLCT_NotALiteral || !CheckRight) return Left; } StringLiteralCheckType Right = checkFormatStringExpr(S, C->getFalseExpr(), Args, HasVAListArg, format_idx, firstDataArg, Type, CallType, InFunctionCall, CheckedVarArgs, UncoveredArg); return (CheckLeft && Left < Right) ? Left : Right; } case Stmt::ImplicitCastExprClass: { E = cast
(E)->getSubExpr(); goto tryAgain; } case Stmt::OpaqueValueExprClass: if (const Expr *src = cast
(E)->getSourceExpr()) { E = src; goto tryAgain; } return SLCT_NotALiteral; case Stmt::PredefinedExprClass: // While __func__, etc., are technically not string literals, they // cannot contain format specifiers and thus are not a security // liability. return SLCT_UncheckedLiteral; case Stmt::DeclRefExprClass: { const DeclRefExpr *DR = cast
(E); // As an exception, do not flag errors for variables binding to // const string literals. if (const VarDecl *VD = dyn_cast
(DR->getDecl())) { bool isConstant = false; QualType T = DR->getType(); if (const ArrayType *AT = S.Context.getAsArrayType(T)) { isConstant = AT->getElementType().isConstant(S.Context); } else if (const PointerType *PT = T->getAs
()) { isConstant = T.isConstant(S.Context) && PT->getPointeeType().isConstant(S.Context); } else if (T->isObjCObjectPointerType()) { // In ObjC, there is usually no "const ObjectPointer" type, // so don't check if the pointee type is constant. isConstant = T.isConstant(S.Context); } if (isConstant) { if (const Expr *Init = VD->getAnyInitializer()) { // Look through initializers like const char c[] = { "foo" } if (const InitListExpr *InitList = dyn_cast
(Init)) { if (InitList->isStringLiteralInit()) Init = InitList->getInit(0)->IgnoreParenImpCasts(); } return checkFormatStringExpr(S, Init, Args, HasVAListArg, format_idx, firstDataArg, Type, CallType, /*InFunctionCall*/false, CheckedVarArgs, UncoveredArg); } } // For vprintf* functions (i.e., HasVAListArg==true), we add a // special check to see if the format string is a function parameter // of the function calling the printf function. If the function // has an attribute indicating it is a printf-like function, then we // should suppress warnings concerning non-literals being used in a call // to a vprintf function. For example: // // void // logmessage(char const *fmt __attribute__ (format (printf, 1, 2)), ...){ // va_list ap; // va_start(ap, fmt); // vprintf(fmt, ap); // Do NOT emit a warning about "fmt". // ... // } if (HasVAListArg) { if (const ParmVarDecl *PV = dyn_cast
(VD)) { if (const NamedDecl *ND = dyn_cast
(PV->getDeclContext())) { int PVIndex = PV->getFunctionScopeIndex() + 1; for (const auto *PVFormat : ND->specific_attrs
()) { // adjust for implicit parameter if (const CXXMethodDecl *MD = dyn_cast
(ND)) if (MD->isInstance()) ++PVIndex; // We also check if the formats are compatible. // We can't pass a 'scanf' string to a 'printf' function. if (PVIndex == PVFormat->getFormatIdx() && Type == S.GetFormatStringType(PVFormat)) return SLCT_UncheckedLiteral; } } } } } return SLCT_NotALiteral; } case Stmt::CallExprClass: case Stmt::CXXMemberCallExprClass: { const CallExpr *CE = cast
(E); if (const NamedDecl *ND = dyn_cast_or_null
(CE->getCalleeDecl())) { if (const FormatArgAttr *FA = ND->getAttr
()) { unsigned ArgIndex = FA->getFormatIdx(); if (const CXXMethodDecl *MD = dyn_cast
(ND)) if (MD->isInstance()) --ArgIndex; const Expr *Arg = CE->getArg(ArgIndex - 1); return checkFormatStringExpr(S, Arg, Args, HasVAListArg, format_idx, firstDataArg, Type, CallType, InFunctionCall, CheckedVarArgs, UncoveredArg); } else if (const FunctionDecl *FD = dyn_cast
(ND)) { unsigned BuiltinID = FD->getBuiltinID(); if (BuiltinID == Builtin::BI__builtin___CFStringMakeConstantString || BuiltinID == Builtin::BI__builtin___NSStringMakeConstantString) { const Expr *Arg = CE->getArg(0); return checkFormatStringExpr(S, Arg, Args, HasVAListArg, format_idx, firstDataArg, Type, CallType, InFunctionCall, CheckedVarArgs, UncoveredArg); } } } return SLCT_NotALiteral; } case Stmt::ObjCStringLiteralClass: case Stmt::StringLiteralClass: { const StringLiteral *StrE = nullptr; if (const ObjCStringLiteral *ObjCFExpr = dyn_cast
(E)) StrE = ObjCFExpr->getString(); else StrE = cast
(E); if (StrE) { CheckFormatString(S, StrE, E, Args, HasVAListArg, format_idx, firstDataArg, Type, InFunctionCall, CallType, CheckedVarArgs, UncoveredArg); return SLCT_CheckedLiteral; } return SLCT_NotALiteral; } default: return SLCT_NotALiteral; } } Sema::FormatStringType Sema::GetFormatStringType(const FormatAttr *Format) { return llvm::StringSwitch
(Format->getType()->getName()) .Case("scanf", FST_Scanf) .Cases("printf", "printf0", FST_Printf) .Cases("NSString", "CFString", FST_NSString) .Case("strftime", FST_Strftime) .Case("strfmon", FST_Strfmon) .Cases("kprintf", "cmn_err", "vcmn_err", "zcmn_err", FST_Kprintf) .Case("freebsd_kprintf", FST_FreeBSDKPrintf) .Case("os_trace", FST_OSTrace) .Default(FST_Unknown); } /// CheckFormatArguments - Check calls to printf and scanf (and similar /// functions) for correct use of format strings. /// Returns true if a format string has been fully checked. bool Sema::CheckFormatArguments(const FormatAttr *Format, ArrayRef
Args, bool IsCXXMember, VariadicCallType CallType, SourceLocation Loc, SourceRange Range, llvm::SmallBitVector &CheckedVarArgs) { FormatStringInfo FSI; if (getFormatStringInfo(Format, IsCXXMember, &FSI)) return CheckFormatArguments(Args, FSI.HasVAListArg, FSI.FormatIdx, FSI.FirstDataArg, GetFormatStringType(Format), CallType, Loc, Range, CheckedVarArgs); return false; } bool Sema::CheckFormatArguments(ArrayRef
Args, bool HasVAListArg, unsigned format_idx, unsigned firstDataArg, FormatStringType Type, VariadicCallType CallType, SourceLocation Loc, SourceRange Range, llvm::SmallBitVector &CheckedVarArgs) { // CHECK: printf/scanf-like function is called with no format string. if (format_idx >= Args.size()) { Diag(Loc, diag::warn_missing_format_string) << Range; return false; } const Expr *OrigFormatExpr = Args[format_idx]->IgnoreParenCasts(); // CHECK: format string is not a string literal. // // Dynamically generated format strings are difficult to // automatically vet at compile time. Requiring that format strings // are string literals: (1) permits the checking of format strings by // the compiler and thereby (2) can practically remove the source of // many format string exploits. // Format string can be either ObjC string (e.g. @"%d") or // C string (e.g. "%d") // ObjC string uses the same format specifiers as C string, so we can use // the same format string checking logic for both ObjC and C strings. UncoveredArgHandler UncoveredArg; StringLiteralCheckType CT = checkFormatStringExpr(*this, OrigFormatExpr, Args, HasVAListArg, format_idx, firstDataArg, Type, CallType, /*IsFunctionCall*/true, CheckedVarArgs, UncoveredArg); // Generate a diagnostic where an uncovered argument is detected. if (UncoveredArg.hasUncoveredArg()) { unsigned ArgIdx = UncoveredArg.getUncoveredArg() + firstDataArg; assert(ArgIdx < Args.size() && "ArgIdx outside bounds"); UncoveredArg.Diagnose(*this, /*IsFunctionCall*/true, Args[ArgIdx]); } if (CT != SLCT_NotALiteral) // Literal format string found, check done! return CT == SLCT_CheckedLiteral; // Strftime is particular as it always uses a single 'time' argument, // so it is safe to pass a non-literal string. if (Type == FST_Strftime) return false; // Do not emit diag when the string param is a macro expansion and the // format is either NSString or CFString. This is a hack to prevent // diag when using the NSLocalizedString and CFCopyLocalizedString macros // which are usually used in place of NS and CF string literals. SourceLocation FormatLoc = Args[format_idx]->getLocStart(); if (Type == FST_NSString && SourceMgr.isInSystemMacro(FormatLoc)) return false; // If there are no arguments specified, warn with -Wformat-security, otherwise // warn only with -Wformat-nonliteral. if (Args.size() == firstDataArg) { Diag(FormatLoc, diag::warn_format_nonliteral_noargs) << OrigFormatExpr->getSourceRange(); switch (Type) { default: break; case FST_Kprintf: case FST_FreeBSDKPrintf: case FST_Printf: Diag(FormatLoc, diag::note_format_security_fixit) << FixItHint::CreateInsertion(FormatLoc, "\"%s\", "); break; case FST_NSString: Diag(FormatLoc, diag::note_format_security_fixit) << FixItHint::CreateInsertion(FormatLoc, "@\"%@\", "); break; } } else { Diag(FormatLoc, diag::warn_format_nonliteral) << OrigFormatExpr->getSourceRange(); } return false; } namespace { class CheckFormatHandler : public analyze_format_string::FormatStringHandler { protected: Sema &S; const StringLiteral *FExpr; const Expr *OrigFormatExpr; const unsigned FirstDataArg; const unsigned NumDataArgs; const char *Beg; // Start of format string. const bool HasVAListArg; ArrayRef
Args; unsigned FormatIdx; llvm::SmallBitVector CoveredArgs; bool usesPositionalArgs; bool atFirstArg; bool inFunctionCall; Sema::VariadicCallType CallType; llvm::SmallBitVector &CheckedVarArgs; UncoveredArgHandler &UncoveredArg; public: CheckFormatHandler(Sema &s, const StringLiteral *fexpr, const Expr *origFormatExpr, unsigned firstDataArg, unsigned numDataArgs, const char *beg, bool hasVAListArg, ArrayRef
Args, unsigned formatIdx, bool inFunctionCall, Sema::VariadicCallType callType, llvm::SmallBitVector &CheckedVarArgs, UncoveredArgHandler &UncoveredArg) : S(s), FExpr(fexpr), OrigFormatExpr(origFormatExpr), FirstDataArg(firstDataArg), NumDataArgs(numDataArgs), Beg(beg), HasVAListArg(hasVAListArg), Args(Args), FormatIdx(formatIdx), usesPositionalArgs(false), atFirstArg(true), inFunctionCall(inFunctionCall), CallType(callType), CheckedVarArgs(CheckedVarArgs), UncoveredArg(UncoveredArg) { CoveredArgs.resize(numDataArgs); CoveredArgs.reset(); } void DoneProcessing(); void HandleIncompleteSpecifier(const char *startSpecifier, unsigned specifierLen) override; void HandleInvalidLengthModifier( const analyze_format_string::FormatSpecifier &FS, const analyze_format_string::ConversionSpecifier &CS, const char *startSpecifier, unsigned specifierLen, unsigned DiagID); void HandleNonStandardLengthModifier( const analyze_format_string::FormatSpecifier &FS, const char *startSpecifier, unsigned specifierLen); void HandleNonStandardConversionSpecifier( const analyze_format_string::ConversionSpecifier &CS, const char *startSpecifier, unsigned specifierLen); void HandlePosition(const char *startPos, unsigned posLen) override; void HandleInvalidPosition(const char *startSpecifier, unsigned specifierLen, analyze_format_string::PositionContext p) override; void HandleZeroPosition(const char *startPos, unsigned posLen) override; void HandleNullChar(const char *nullCharacter) override; template
static void EmitFormatDiagnostic(Sema &S, bool inFunctionCall, const Expr *ArgumentExpr, const PartialDiagnostic &PDiag, SourceLocation StringLoc, bool IsStringLocation, Range StringRange, ArrayRef
Fixit = None); protected: bool HandleInvalidConversionSpecifier(unsigned argIndex, SourceLocation Loc, const char *startSpec, unsigned specifierLen, const char *csStart, unsigned csLen); void HandlePositionalNonpositionalArgs(SourceLocation Loc, const char *startSpec, unsigned specifierLen); SourceRange getFormatStringRange(); CharSourceRange getSpecifierRange(const char *startSpecifier, unsigned specifierLen); SourceLocation getLocationOfByte(const char *x); const Expr *getDataArg(unsigned i) const; bool CheckNumArgs(const analyze_format_string::FormatSpecifier &FS, const analyze_format_string::ConversionSpecifier &CS, const char *startSpecifier, unsigned specifierLen, unsigned argIndex); template
void EmitFormatDiagnostic(PartialDiagnostic PDiag, SourceLocation StringLoc, bool IsStringLocation, Range StringRange, ArrayRef
Fixit = None); }; } // end anonymous namespace SourceRange CheckFormatHandler::getFormatStringRange() { return OrigFormatExpr->getSourceRange(); } CharSourceRange CheckFormatHandler:: getSpecifierRange(const char *startSpecifier, unsigned specifierLen) { SourceLocation Start = getLocationOfByte(startSpecifier); SourceLocation End = getLocationOfByte(startSpecifier + specifierLen - 1); // Advance the end SourceLocation by one due to half-open ranges. End = End.getLocWithOffset(1); return CharSourceRange::getCharRange(Start, End); } SourceLocation CheckFormatHandler::getLocationOfByte(const char *x) { return S.getLocationOfStringLiteralByte(FExpr, x - Beg); } void CheckFormatHandler::HandleIncompleteSpecifier(const char *startSpecifier, unsigned specifierLen){ EmitFormatDiagnostic(S.PDiag(diag::warn_printf_incomplete_specifier), getLocationOfByte(startSpecifier), /*IsStringLocation*/true, getSpecifierRange(startSpecifier, specifierLen)); } void CheckFormatHandler::HandleInvalidLengthModifier( const analyze_format_string::FormatSpecifier &FS, const analyze_format_string::ConversionSpecifier &CS, const char *startSpecifier, unsigned specifierLen, unsigned DiagID) { using namespace analyze_format_string; const LengthModifier &LM = FS.getLengthModifier(); CharSourceRange LMRange = getSpecifierRange(LM.getStart(), LM.getLength()); // See if we know how to fix this length modifier. Optional
FixedLM = FS.getCorrectedLengthModifier(); if (FixedLM) { EmitFormatDiagnostic(S.PDiag(DiagID) << LM.toString() << CS.toString(), getLocationOfByte(LM.getStart()), /*IsStringLocation*/true, getSpecifierRange(startSpecifier, specifierLen)); S.Diag(getLocationOfByte(LM.getStart()), diag::note_format_fix_specifier) << FixedLM->toString() << FixItHint::CreateReplacement(LMRange, FixedLM->toString()); } else { FixItHint Hint; if (DiagID == diag::warn_format_nonsensical_length) Hint = FixItHint::CreateRemoval(LMRange); EmitFormatDiagnostic(S.PDiag(DiagID) << LM.toString() << CS.toString(), getLocationOfByte(LM.getStart()), /*IsStringLocation*/true, getSpecifierRange(startSpecifier, specifierLen), Hint); } } void CheckFormatHandler::HandleNonStandardLengthModifier( const analyze_format_string::FormatSpecifier &FS, const char *startSpecifier, unsigned specifierLen) { using namespace analyze_format_string; const LengthModifier &LM = FS.getLengthModifier(); CharSourceRange LMRange = getSpecifierRange(LM.getStart(), LM.getLength()); // See if we know how to fix this length modifier. Optional
FixedLM = FS.getCorrectedLengthModifier(); if (FixedLM) { EmitFormatDiagnostic(S.PDiag(diag::warn_format_non_standard) << LM.toString() << 0, getLocationOfByte(LM.getStart()), /*IsStringLocation*/true, getSpecifierRange(startSpecifier, specifierLen)); S.Diag(getLocationOfByte(LM.getStart()), diag::note_format_fix_specifier) << FixedLM->toString() << FixItHint::CreateReplacement(LMRange, FixedLM->toString()); } else { EmitFormatDiagnostic(S.PDiag(diag::warn_format_non_standard) << LM.toString() << 0, getLocationOfByte(LM.getStart()), /*IsStringLocation*/true, getSpecifierRange(startSpecifier, specifierLen)); } } void CheckFormatHandler::HandleNonStandardConversionSpecifier( const analyze_format_string::ConversionSpecifier &CS, const char *startSpecifier, unsigned specifierLen) { using namespace analyze_format_string; // See if we know how to fix this conversion specifier. Optional
FixedCS = CS.getStandardSpecifier(); if (FixedCS) { EmitFormatDiagnostic(S.PDiag(diag::warn_format_non_standard) << CS.toString() << /*conversion specifier*/1, getLocationOfByte(CS.getStart()), /*IsStringLocation*/true, getSpecifierRange(startSpecifier, specifierLen)); CharSourceRange CSRange = getSpecifierRange(CS.getStart(), CS.getLength()); S.Diag(getLocationOfByte(CS.getStart()), diag::note_format_fix_specifier) << FixedCS->toString() << FixItHint::CreateReplacement(CSRange, FixedCS->toString()); } else { EmitFormatDiagnostic(S.PDiag(diag::warn_format_non_standard) << CS.toString() << /*conversion specifier*/1, getLocationOfByte(CS.getStart()), /*IsStringLocation*/true, getSpecifierRange(startSpecifier, specifierLen)); } } void CheckFormatHandler::HandlePosition(const char *startPos, unsigned posLen) { EmitFormatDiagnostic(S.PDiag(diag::warn_format_non_standard_positional_arg), getLocationOfByte(startPos), /*IsStringLocation*/true, getSpecifierRange(startPos, posLen)); } void CheckFormatHandler::HandleInvalidPosition(const char *startPos, unsigned posLen, analyze_format_string::PositionContext p) { EmitFormatDiagnostic(S.PDiag(diag::warn_format_invalid_positional_specifier) << (unsigned) p, getLocationOfByte(startPos), /*IsStringLocation*/true, getSpecifierRange(startPos, posLen)); } void CheckFormatHandler::HandleZeroPosition(const char *startPos, unsigned posLen) { EmitFormatDiagnostic(S.PDiag(diag::warn_format_zero_positional_specifier), getLocationOfByte(startPos), /*IsStringLocation*/true, getSpecifierRange(startPos, posLen)); } void CheckFormatHandler::HandleNullChar(const char *nullCharacter) { if (!isa
(OrigFormatExpr)) { // The presence of a null character is likely an error. EmitFormatDiagnostic( S.PDiag(diag::warn_printf_format_string_contains_null_char), getLocationOfByte(nullCharacter), /*IsStringLocation*/true, getFormatStringRange()); } } // Note that this may return NULL if there was an error parsing or building // one of the argument expressions. const Expr *CheckFormatHandler::getDataArg(unsigned i) const { return Args[FirstDataArg + i]; } void CheckFormatHandler::DoneProcessing() { // Does the number of data arguments exceed the number of // format conversions in the format string? if (!HasVAListArg) { // Find any arguments that weren't covered. CoveredArgs.flip(); signed notCoveredArg = CoveredArgs.find_first(); if (notCoveredArg >= 0) { assert((unsigned)notCoveredArg < NumDataArgs); UncoveredArg.Update(notCoveredArg, OrigFormatExpr); } else { UncoveredArg.setAllCovered(); } } } void UncoveredArgHandler::Diagnose(Sema &S, bool IsFunctionCall, const Expr *ArgExpr) { assert(hasUncoveredArg() && DiagnosticExprs.size() > 0 && "Invalid state"); if (!ArgExpr) return; SourceLocation Loc = ArgExpr->getLocStart(); if (S.getSourceManager().isInSystemMacro(Loc)) return; PartialDiagnostic PDiag = S.PDiag(diag::warn_printf_data_arg_not_used); for (auto E : DiagnosticExprs) PDiag << E->getSourceRange(); CheckFormatHandler::EmitFormatDiagnostic( S, IsFunctionCall, DiagnosticExprs[0], PDiag, Loc, /*IsStringLocation*/false, DiagnosticExprs[0]->getSourceRange()); } bool CheckFormatHandler::HandleInvalidConversionSpecifier(unsigned argIndex, SourceLocation Loc, const char *startSpec, unsigned specifierLen, const char *csStart, unsigned csLen) { bool keepGoing = true; if (argIndex < NumDataArgs) { // Consider the argument coverered, even though the specifier doesn't // make sense. CoveredArgs.set(argIndex); } else { // If argIndex exceeds the number of data arguments we // don't issue a warning because that is just a cascade of warnings (and // they may have intended '%%' anyway). We don't want to continue processing // the format string after this point, however, as we will like just get // gibberish when trying to match arguments. keepGoing = false; } StringRef Specifier(csStart, csLen); // If the specifier in non-printable, it could be the first byte of a UTF-8 // sequence. In that case, print the UTF-8 code point. If not, print the byte // hex value. std::string CodePointStr; if (!llvm::sys::locale::isPrint(*csStart)) { UTF32 CodePoint; const UTF8 **B = reinterpret_cast
(&csStart); const UTF8 *E = reinterpret_cast
(csStart + csLen); ConversionResult Result = llvm::convertUTF8Sequence(B, E, &CodePoint, strictConversion); if (Result != conversionOK) { unsigned char FirstChar = *csStart; CodePoint = (UTF32)FirstChar; } llvm::raw_string_ostream OS(CodePointStr); if (CodePoint < 256) OS << "\\x" << llvm::format("%02x", CodePoint); else if (CodePoint <= 0xFFFF) OS << "\\u" << llvm::format("%04x", CodePoint); else OS << "\\U" << llvm::format("%08x", CodePoint); OS.flush(); Specifier = CodePointStr; } EmitFormatDiagnostic( S.PDiag(diag::warn_format_invalid_conversion) << Specifier, Loc, /*IsStringLocation*/ true, getSpecifierRange(startSpec, specifierLen)); return keepGoing; } void CheckFormatHandler::HandlePositionalNonpositionalArgs(SourceLocation Loc, const char *startSpec, unsigned specifierLen) { EmitFormatDiagnostic( S.PDiag(diag::warn_format_mix_positional_nonpositional_args), Loc, /*isStringLoc*/true, getSpecifierRange(startSpec, specifierLen)); } bool CheckFormatHandler::CheckNumArgs( const analyze_format_string::FormatSpecifier &FS, const analyze_format_string::ConversionSpecifier &CS, const char *startSpecifier, unsigned specifierLen, unsigned argIndex) { if (argIndex >= NumDataArgs) { PartialDiagnostic PDiag = FS.usesPositionalArg() ? (S.PDiag(diag::warn_printf_positional_arg_exceeds_data_args) << (argIndex+1) << NumDataArgs) : S.PDiag(diag::warn_printf_insufficient_data_args); EmitFormatDiagnostic( PDiag, getLocationOfByte(CS.getStart()), /*IsStringLocation*/true, getSpecifierRange(startSpecifier, specifierLen)); // Since more arguments than conversion tokens are given, by extension // all arguments are covered, so mark this as so. UncoveredArg.setAllCovered(); return false; } return true; } template
void CheckFormatHandler::EmitFormatDiagnostic(PartialDiagnostic PDiag, SourceLocation Loc, bool IsStringLocation, Range StringRange, ArrayRef
FixIt) { EmitFormatDiagnostic(S, inFunctionCall, Args[FormatIdx], PDiag, Loc, IsStringLocation, StringRange, FixIt); } /// \brief If the format string is not within the funcion call, emit a note /// so that the function call and string are in diagnostic messages. /// /// \param InFunctionCall if true, the format string is within the function /// call and only one diagnostic message will be produced. Otherwise, an /// extra note will be emitted pointing to location of the format string. /// /// \param ArgumentExpr the expression that is passed as the format string /// argument in the function call. Used for getting locations when two /// diagnostics are emitted. /// /// \param PDiag the callee should already have provided any strings for the /// diagnostic message. This function only adds locations and fixits /// to diagnostics. /// /// \param Loc primary location for diagnostic. If two diagnostics are /// required, one will be at Loc and a new SourceLocation will be created for /// the other one. /// /// \param IsStringLocation if true, Loc points to the format string should be /// used for the note. Otherwise, Loc points to the argument list and will /// be used with PDiag. /// /// \param StringRange some or all of the string to highlight. This is /// templated so it can accept either a CharSourceRange or a SourceRange. /// /// \param FixIt optional fix it hint for the format string. template
void CheckFormatHandler::EmitFormatDiagnostic( Sema &S, bool InFunctionCall, const Expr *ArgumentExpr, const PartialDiagnostic &PDiag, SourceLocation Loc, bool IsStringLocation, Range StringRange, ArrayRef
FixIt) { if (InFunctionCall) { const Sema::SemaDiagnosticBuilder &D = S.Diag(Loc, PDiag); D << StringRange; D << FixIt; } else { S.Diag(IsStringLocation ? ArgumentExpr->getExprLoc() : Loc, PDiag) << ArgumentExpr->getSourceRange(); const Sema::SemaDiagnosticBuilder &Note = S.Diag(IsStringLocation ? Loc : StringRange.getBegin(), diag::note_format_string_defined); Note << StringRange; Note << FixIt; } } //===--- CHECK: Printf format string checking ------------------------------===// namespace { class CheckPrintfHandler : public CheckFormatHandler { bool ObjCContext; public: CheckPrintfHandler(Sema &s, const StringLiteral *fexpr, const Expr *origFormatExpr, unsigned firstDataArg, unsigned numDataArgs, bool isObjC, const char *beg, bool hasVAListArg, ArrayRef
Args, unsigned formatIdx, bool inFunctionCall, Sema::VariadicCallType CallType, llvm::SmallBitVector &CheckedVarArgs, UncoveredArgHandler &UncoveredArg) : CheckFormatHandler(s, fexpr, origFormatExpr, firstDataArg, numDataArgs, beg, hasVAListArg, Args, formatIdx, inFunctionCall, CallType, CheckedVarArgs, UncoveredArg), ObjCContext(isObjC) {} bool HandleInvalidPrintfConversionSpecifier( const analyze_printf::PrintfSpecifier &FS, const char *startSpecifier, unsigned specifierLen) override; bool HandlePrintfSpecifier(const analyze_printf::PrintfSpecifier &FS, const char *startSpecifier, unsigned specifierLen) override; bool checkFormatExpr(const analyze_printf::PrintfSpecifier &FS, const char *StartSpecifier, unsigned SpecifierLen, const Expr *E); bool HandleAmount(const analyze_format_string::OptionalAmount &Amt, unsigned k, const char *startSpecifier, unsigned specifierLen); void HandleInvalidAmount(const analyze_printf::PrintfSpecifier &FS, const analyze_printf::OptionalAmount &Amt, unsigned type, const char *startSpecifier, unsigned specifierLen); void HandleFlag(const analyze_printf::PrintfSpecifier &FS, const analyze_printf::OptionalFlag &flag, const char *startSpecifier, unsigned specifierLen); void HandleIgnoredFlag(const analyze_printf::PrintfSpecifier &FS, const analyze_printf::OptionalFlag &ignoredFlag, const analyze_printf::OptionalFlag &flag, const char *startSpecifier, unsigned specifierLen); bool checkForCStrMembers(const analyze_printf::ArgType &AT, const Expr *E); void HandleEmptyObjCModifierFlag(const char *startFlag, unsigned flagLen) override; void HandleInvalidObjCModifierFlag(const char *startFlag, unsigned flagLen) override; void HandleObjCFlagsWithNonObjCConversion(const char *flagsStart, const char *flagsEnd, const char *conversionPosition) override; }; } // end anonymous namespace bool CheckPrintfHandler::HandleInvalidPrintfConversionSpecifier( const analyze_printf::PrintfSpecifier &FS, const char *startSpecifier, unsigned specifierLen) { const analyze_printf::PrintfConversionSpecifier &CS = FS.getConversionSpecifier(); return HandleInvalidConversionSpecifier(FS.getArgIndex(), getLocationOfByte(CS.getStart()), startSpecifier, specifierLen, CS.getStart(), CS.getLength()); } bool CheckPrintfHandler::HandleAmount( const analyze_format_string::OptionalAmount &Amt, unsigned k, const char *startSpecifier, unsigned specifierLen) { if (Amt.hasDataArgument()) { if (!HasVAListArg) { unsigned argIndex = Amt.getArgIndex(); if (argIndex >= NumDataArgs) { EmitFormatDiagnostic(S.PDiag(diag::warn_printf_asterisk_missing_arg) << k, getLocationOfByte(Amt.getStart()), /*IsStringLocation*/true, getSpecifierRange(startSpecifier, specifierLen)); // Don't do any more checking. We will just emit // spurious errors. return false; } // Type check the data argument. It should be an 'int'. // Although not in conformance with C99, we also allow the argument to be // an 'unsigned int' as that is a reasonably safe case. GCC also // doesn't emit a warning for that case. CoveredArgs.set(argIndex); const Expr *Arg = getDataArg(argIndex); if (!Arg) return false; QualType T = Arg->getType(); const analyze_printf::ArgType &AT = Amt.getArgType(S.Context); assert(AT.isValid()); if (!AT.matchesType(S.Context, T)) { EmitFormatDiagnostic(S.PDiag(diag::warn_printf_asterisk_wrong_type) << k << AT.getRepresentativeTypeName(S.Context) << T << Arg->getSourceRange(), getLocationOfByte(Amt.getStart()), /*IsStringLocation*/true, getSpecifierRange(startSpecifier, specifierLen)); // Don't do any more checking. We will just emit // spurious errors. return false; } } } return true; } void CheckPrintfHandler::HandleInvalidAmount( const analyze_printf::PrintfSpecifier &FS, const analyze_printf::OptionalAmount &Amt, unsigned type, const char *startSpecifier, unsigned specifierLen) { const analyze_printf::PrintfConversionSpecifier &CS = FS.getConversionSpecifier(); FixItHint fixit = Amt.getHowSpecified() == analyze_printf::OptionalAmount::Constant ? FixItHint::CreateRemoval(getSpecifierRange(Amt.getStart(), Amt.getConstantLength())) : FixItHint(); EmitFormatDiagnostic(S.PDiag(diag::warn_printf_nonsensical_optional_amount) << type << CS.toString(), getLocationOfByte(Amt.getStart()), /*IsStringLocation*/true, getSpecifierRange(startSpecifier, specifierLen), fixit); } void CheckPrintfHandler::HandleFlag(const analyze_printf::PrintfSpecifier &FS, const analyze_printf::OptionalFlag &flag, const char *startSpecifier, unsigned specifierLen) { // Warn about pointless flag with a fixit removal. const analyze_printf::PrintfConversionSpecifier &CS = FS.getConversionSpecifier(); EmitFormatDiagnostic(S.PDiag(diag::warn_printf_nonsensical_flag) << flag.toString() << CS.toString(), getLocationOfByte(flag.getPosition()), /*IsStringLocation*/true, getSpecifierRange(startSpecifier, specifierLen), FixItHint::CreateRemoval( getSpecifierRange(flag.getPosition(), 1))); } void CheckPrintfHandler::HandleIgnoredFlag( const analyze_printf::PrintfSpecifier &FS, const analyze_printf::OptionalFlag &ignoredFlag, const analyze_printf::OptionalFlag &flag, const char *startSpecifier, unsigned specifierLen) { // Warn about ignored flag with a fixit removal. EmitFormatDiagnostic(S.PDiag(diag::warn_printf_ignored_flag) << ignoredFlag.toString() << flag.toString(), getLocationOfByte(ignoredFlag.getPosition()), /*IsStringLocation*/true, getSpecifierRange(startSpecifier, specifierLen), FixItHint::CreateRemoval( getSpecifierRange(ignoredFlag.getPosition(), 1))); } // void EmitFormatDiagnostic(PartialDiagnostic PDiag, SourceLocation StringLoc, // bool IsStringLocation, Range StringRange, // ArrayRef
Fixit = None); void CheckPrintfHandler::HandleEmptyObjCModifierFlag(const char *startFlag, unsigned flagLen) { // Warn about an empty flag. EmitFormatDiagnostic(S.PDiag(diag::warn_printf_empty_objc_flag), getLocationOfByte(startFlag), /*IsStringLocation*/true, getSpecifierRange(startFlag, flagLen)); } void CheckPrintfHandler::HandleInvalidObjCModifierFlag(const char *startFlag, unsigned flagLen) { // Warn about an invalid flag. auto Range = getSpecifierRange(startFlag, flagLen); StringRef flag(startFlag, flagLen); EmitFormatDiagnostic(S.PDiag(diag::warn_printf_invalid_objc_flag) << flag, getLocationOfByte(startFlag), /*IsStringLocation*/true, Range, FixItHint::CreateRemoval(Range)); } void CheckPrintfHandler::HandleObjCFlagsWithNonObjCConversion( const char *flagsStart, const char *flagsEnd, const char *conversionPosition) { // Warn about using '[...]' without a '@' conversion. auto Range = getSpecifierRange(flagsStart, flagsEnd - flagsStart + 1); auto diag = diag::warn_printf_ObjCflags_without_ObjCConversion; EmitFormatDiagnostic(S.PDiag(diag) << StringRef(conversionPosition, 1), getLocationOfByte(conversionPosition), /*IsStringLocation*/true, Range, FixItHint::CreateRemoval(Range)); } // Determines if the specified is a C++ class or struct containing // a member with the specified name and kind (e.g. a CXXMethodDecl named // "c_str()"). template
static llvm::SmallPtrSet
CXXRecordMembersNamed(StringRef Name, Sema &S, QualType Ty) { const RecordType *RT = Ty->getAs
(); llvm::SmallPtrSet
Results; if (!RT) return Results; const CXXRecordDecl *RD = dyn_cast
(RT->getDecl()); if (!RD || !RD->getDefinition()) return Results; LookupResult R(S, &S.Context.Idents.get(Name), SourceLocation(), Sema::LookupMemberName); R.suppressDiagnostics(); // We just need to include all members of the right kind turned up by the // filter, at this point. if (S.LookupQualifiedName(R, RT->getDecl())) for (LookupResult::iterator I = R.begin(), E = R.end(); I != E; ++I) { NamedDecl *decl = (*I)->getUnderlyingDecl(); if (MemberKind *FK = dyn_cast
(decl)) Results.insert(FK); } return Results; } /// Check if we could call '.c_str()' on an object. /// /// FIXME: This returns the wrong results in some cases (if cv-qualifiers don't /// allow the call, or if it would be ambiguous). bool Sema::hasCStrMethod(const Expr *E) { typedef llvm::SmallPtrSet
MethodSet; MethodSet Results = CXXRecordMembersNamed
("c_str", *this, E->getType()); for (MethodSet::iterator MI = Results.begin(), ME = Results.end(); MI != ME; ++MI) if ((*MI)->getMinRequiredArguments() == 0) return true; return false; } // Check if a (w)string was passed when a (w)char* was needed, and offer a // better diagnostic if so. AT is assumed to be valid. // Returns true when a c_str() conversion method is found. bool CheckPrintfHandler::checkForCStrMembers( const analyze_printf::ArgType &AT, const Expr *E) { typedef llvm::SmallPtrSet
MethodSet; MethodSet Results = CXXRecordMembersNamed
("c_str", S, E->getType()); for (MethodSet::iterator MI = Results.begin(), ME = Results.end(); MI != ME; ++MI) { const CXXMethodDecl *Method = *MI; if (Method->getMinRequiredArguments() == 0 && AT.matchesType(S.Context, Method->getReturnType())) { // FIXME: Suggest parens if the expression needs them. SourceLocation EndLoc = S.getLocForEndOfToken(E->getLocEnd()); S.Diag(E->getLocStart(), diag::note_printf_c_str) << "c_str()" << FixItHint::CreateInsertion(EndLoc, ".c_str()"); return true; } } return false; } bool CheckPrintfHandler::HandlePrintfSpecifier(const analyze_printf::PrintfSpecifier &FS, const char *startSpecifier, unsigned specifierLen) { using namespace analyze_format_string; using namespace analyze_printf; const PrintfConversionSpecifier &CS = FS.getConversionSpecifier(); if (FS.consumesDataArgument()) { if (atFirstArg) { atFirstArg = false; usesPositionalArgs = FS.usesPositionalArg(); } else if (usesPositionalArgs != FS.usesPositionalArg()) { HandlePositionalNonpositionalArgs(getLocationOfByte(CS.getStart()), startSpecifier, specifierLen); return false; } } // First check if the field width, precision, and conversion specifier // have matching data arguments. if (!HandleAmount(FS.getFieldWidth(), /* field width */ 0, startSpecifier, specifierLen)) { return false; } if (!HandleAmount(FS.getPrecision(), /* precision */ 1, startSpecifier, specifierLen)) { return false; } if (!CS.consumesDataArgument()) { // FIXME: Technically specifying a precision or field width here // makes no sense. Worth issuing a warning at some point. return true; } // Consume the argument. unsigned argIndex = FS.getArgIndex(); if (argIndex < NumDataArgs) { // The check to see if the argIndex is valid will come later. // We set the bit here because we may exit early from this // function if we encounter some other error. CoveredArgs.set(argIndex); } // FreeBSD kernel extensions. if (CS.getKind() == ConversionSpecifier::FreeBSDbArg || CS.getKind() == ConversionSpecifier::FreeBSDDArg) { // We need at least two arguments. if (!CheckNumArgs(FS, CS, startSpecifier, specifierLen, argIndex + 1)) return false; // Claim the second argument. CoveredArgs.set(argIndex + 1); // Type check the first argument (int for %b, pointer for %D) const Expr *Ex = getDataArg(argIndex); const analyze_printf::ArgType &AT = (CS.getKind() == ConversionSpecifier::FreeBSDbArg) ? ArgType(S.Context.IntTy) : ArgType::CPointerTy; if (AT.isValid() && !AT.matchesType(S.Context, Ex->getType())) EmitFormatDiagnostic( S.PDiag(diag::warn_format_conversion_argument_type_mismatch) << AT.getRepresentativeTypeName(S.Context) << Ex->getType() << false << Ex->getSourceRange(), Ex->getLocStart(), /*IsStringLocation*/false, getSpecifierRange(startSpecifier, specifierLen)); // Type check the second argument (char * for both %b and %D) Ex = getDataArg(argIndex + 1); const analyze_printf::ArgType &AT2 = ArgType::CStrTy; if (AT2.isValid() && !AT2.matchesType(S.Context, Ex->getType())) EmitFormatDiagnostic( S.PDiag(diag::warn_format_conversion_argument_type_mismatch) << AT2.getRepresentativeTypeName(S.Context) << Ex->getType() << false << Ex->getSourceRange(), Ex->getLocStart(), /*IsStringLocation*/false, getSpecifierRange(startSpecifier, specifierLen)); return true; } // Check for using an Objective-C specific conversion specifier // in a non-ObjC literal. if (!ObjCContext && CS.isObjCArg()) { return HandleInvalidPrintfConversionSpecifier(FS, startSpecifier, specifierLen); } // Check for invalid use of field width if (!FS.hasValidFieldWidth()) { HandleInvalidAmount(FS, FS.getFieldWidth(), /* field width */ 0, startSpecifier, specifierLen); } // Check for invalid use of precision if (!FS.hasValidPrecision()) { HandleInvalidAmount(FS, FS.getPrecision(), /* precision */ 1, startSpecifier, specifierLen); } // Check each flag does not conflict with any other component. if (!FS.hasValidThousandsGroupingPrefix()) HandleFlag(FS, FS.hasThousandsGrouping(), startSpecifier, specifierLen); if (!FS.hasValidLeadingZeros()) HandleFlag(FS, FS.hasLeadingZeros(), startSpecifier, specifierLen); if (!FS.hasValidPlusPrefix()) HandleFlag(FS, FS.hasPlusPrefix(), startSpecifier, specifierLen); if (!FS.hasValidSpacePrefix()) HandleFlag(FS, FS.hasSpacePrefix(), startSpecifier, specifierLen); if (!FS.hasValidAlternativeForm()) HandleFlag(FS, FS.hasAlternativeForm(), startSpecifier, specifierLen); if (!FS.hasValidLeftJustified()) HandleFlag(FS, FS.isLeftJustified(), startSpecifier, specifierLen); // Check that flags are not ignored by another flag if (FS.hasSpacePrefix() && FS.hasPlusPrefix()) // ' ' ignored by '+' HandleIgnoredFlag(FS, FS.hasSpacePrefix(), FS.hasPlusPrefix(), startSpecifier, specifierLen); if (FS.hasLeadingZeros() && FS.isLeftJustified()) // '0' ignored by '-' HandleIgnoredFlag(FS, FS.hasLeadingZeros(), FS.isLeftJustified(), startSpecifier, specifierLen); // Check the length modifier is valid with the given conversion specifier. if (!FS.hasValidLengthModifier(S.getASTContext().getTargetInfo())) HandleInvalidLengthModifier(FS, CS, startSpecifier, specifierLen, diag::warn_format_nonsensical_length); else if (!FS.hasStandardLengthModifier()) HandleNonStandardLengthModifier(FS, startSpecifier, specifierLen); else if (!FS.hasStandardLengthConversionCombination()) HandleInvalidLengthModifier(FS, CS, startSpecifier, specifierLen, diag::warn_format_non_standard_conversion_spec); if (!FS.hasStandardConversionSpecifier(S.getLangOpts())) HandleNonStandardConversionSpecifier(CS, startSpecifier, specifierLen); // The remaining checks depend on the data arguments. if (HasVAListArg) return true; if (!CheckNumArgs(FS, CS, startSpecifier, specifierLen, argIndex)) return false; const Expr *Arg = getDataArg(argIndex); if (!Arg) return true; return checkFormatExpr(FS, startSpecifier, specifierLen, Arg); } static bool requiresParensToAddCast(const Expr *E) { // FIXME: We should have a general way to reason about operator // precedence and whether parens are actually needed here. // Take care of a few common cases where they aren't. const Expr *Inside = E->IgnoreImpCasts(); if (const PseudoObjectExpr *POE = dyn_cast
(Inside)) Inside = POE->getSyntacticForm()->IgnoreImpCasts(); switch (Inside->getStmtClass()) { case Stmt::ArraySubscriptExprClass: case Stmt::CallExprClass: case Stmt::CharacterLiteralClass: case Stmt::CXXBoolLiteralExprClass: case Stmt::DeclRefExprClass: case Stmt::FloatingLiteralClass: case Stmt::IntegerLiteralClass: case Stmt::MemberExprClass: case Stmt::ObjCArrayLiteralClass: case Stmt::ObjCBoolLiteralExprClass: case Stmt::ObjCBoxedExprClass: case Stmt::ObjCDictionaryLiteralClass: case Stmt::ObjCEncodeExprClass: case Stmt::ObjCIvarRefExprClass: case Stmt::ObjCMessageExprClass: case Stmt::ObjCPropertyRefExprClass: case Stmt::ObjCStringLiteralClass: case Stmt::ObjCSubscriptRefExprClass: case Stmt::ParenExprClass: case Stmt::StringLiteralClass: case Stmt::UnaryOperatorClass: return false; default: return true; } } static std::pair
shouldNotPrintDirectly(const ASTContext &Context, QualType IntendedTy, const Expr *E) { // Use a 'while' to peel off layers of typedefs. QualType TyTy = IntendedTy; while (const TypedefType *UserTy = TyTy->getAs
()) { StringRef Name = UserTy->getDecl()->getName(); QualType CastTy = llvm::StringSwitch
(Name) .Case("NSInteger", Context.LongTy) .Case("NSUInteger", Context.UnsignedLongTy) .Case("SInt32", Context.IntTy) .Case("UInt32", Context.UnsignedIntTy) .Default(QualType()); if (!CastTy.isNull()) return std::make_pair(CastTy, Name); TyTy = UserTy->desugar(); } // Strip parens if necessary. if (const ParenExpr *PE = dyn_cast
(E)) return shouldNotPrintDirectly(Context, PE->getSubExpr()->getType(), PE->getSubExpr()); // If this is a conditional expression, then its result type is constructed // via usual arithmetic conversions and thus there might be no necessary // typedef sugar there. Recurse to operands to check for NSInteger & // Co. usage condition. if (const ConditionalOperator *CO = dyn_cast
(E)) { QualType TrueTy, FalseTy; StringRef TrueName, FalseName; std::tie(TrueTy, TrueName) = shouldNotPrintDirectly(Context, CO->getTrueExpr()->getType(), CO->getTrueExpr()); std::tie(FalseTy, FalseName) = shouldNotPrintDirectly(Context, CO->getFalseExpr()->getType(), CO->getFalseExpr()); if (TrueTy == FalseTy) return std::make_pair(TrueTy, TrueName); else if (TrueTy.isNull()) return std::make_pair(FalseTy, FalseName); else if (FalseTy.isNull()) return std::make_pair(TrueTy, TrueName); } return std::make_pair(QualType(), StringRef()); } bool CheckPrintfHandler::checkFormatExpr(const analyze_printf::PrintfSpecifier &FS, const char *StartSpecifier, unsigned SpecifierLen, const Expr *E) { using namespace analyze_format_string; using namespace analyze_printf; // Now type check the data expression that matches the // format specifier. const analyze_printf::ArgType &AT = FS.getArgType(S.Context, ObjCContext); if (!AT.isValid()) return true; QualType ExprTy = E->getType(); while (const TypeOfExprType *TET = dyn_cast
(ExprTy)) { ExprTy = TET->getUnderlyingExpr()->getType(); } analyze_printf::ArgType::MatchKind match = AT.matchesType(S.Context, ExprTy); if (match == analyze_printf::ArgType::Match) { return true; } // Look through argument promotions for our error message's reported type. // This includes the integral and floating promotions, but excludes array // and function pointer decay; seeing that an argument intended to be a // string has type 'char [6]' is probably more confusing than 'char *'. if (const ImplicitCastExpr *ICE = dyn_cast
(E)) { if (ICE->getCastKind() == CK_IntegralCast || ICE->getCastKind() == CK_FloatingCast) { E = ICE->getSubExpr(); ExprTy = E->getType(); // Check if we didn't match because of an implicit cast from a 'char' // or 'short' to an 'int'. This is done because printf is a varargs // function. if (ICE->getType() == S.Context.IntTy || ICE->getType() == S.Context.UnsignedIntTy) { // All further checking is done on the subexpression. if (AT.matchesType(S.Context, ExprTy)) return true; } } } else if (const CharacterLiteral *CL = dyn_cast
(E)) { // Special case for 'a', which has type 'int' in C. // Note, however, that we do /not/ want to treat multibyte constants like // 'MooV' as characters! This form is deprecated but still exists. if (ExprTy == S.Context.IntTy) if (llvm::isUIntN(S.Context.getCharWidth(), CL->getValue())) ExprTy = S.Context.CharTy; } // Look through enums to their underlying type. bool IsEnum = false; if (auto EnumTy = ExprTy->getAs
()) { ExprTy = EnumTy->getDecl()->getIntegerType(); IsEnum = true; } // %C in an Objective-C context prints a unichar, not a wchar_t. // If the argument is an integer of some kind, believe the %C and suggest // a cast instead of changing the conversion specifier. QualType IntendedTy = ExprTy; if (ObjCContext && FS.getConversionSpecifier().getKind() == ConversionSpecifier::CArg) { if (ExprTy->isIntegralOrUnscopedEnumerationType() && !ExprTy->isCharType()) { // 'unichar' is defined as a typedef of unsigned short, but we should // prefer using the typedef if it is visible. IntendedTy = S.Context.UnsignedShortTy; // While we are here, check if the value is an IntegerLiteral that happens // to be within the valid range. if (const IntegerLiteral *IL = dyn_cast
(E)) { const llvm::APInt &V = IL->getValue(); if (V.getActiveBits() <= S.Context.getTypeSize(IntendedTy)) return true; } LookupResult Result(S, &S.Context.Idents.get("unichar"), E->getLocStart(), Sema::LookupOrdinaryName); if (S.LookupName(Result, S.getCurScope())) { NamedDecl *ND = Result.getFoundDecl(); if (TypedefNameDecl *TD = dyn_cast
(ND)) if (TD->getUnderlyingType() == IntendedTy) IntendedTy = S.Context.getTypedefType(TD); } } } // Special-case some of Darwin's platform-independence types by suggesting // casts to primitive types that are known to be large enough. bool ShouldNotPrintDirectly = false; StringRef CastTyName; if (S.Context.getTargetInfo().getTriple().isOSDarwin()) { QualType CastTy; std::tie(CastTy, CastTyName) = shouldNotPrintDirectly(S.Context, IntendedTy, E); if (!CastTy.isNull()) { IntendedTy = CastTy; ShouldNotPrintDirectly = true; } } // We may be able to offer a FixItHint if it is a supported type. PrintfSpecifier fixedFS = FS; bool success = fixedFS.fixType(IntendedTy, S.getLangOpts(), S.Context, ObjCContext); if (success) { // Get the fix string from the fixed format specifier SmallString<16> buf; llvm::raw_svector_ostream os(buf); fixedFS.toString(os); CharSourceRange SpecRange = getSpecifierRange(StartSpecifier, SpecifierLen); if (IntendedTy == ExprTy && !ShouldNotPrintDirectly) { unsigned diag = diag::warn_format_conversion_argument_type_mismatch; if (match == analyze_format_string::ArgType::NoMatchPedantic) { diag = diag::warn_format_conversion_argument_type_mismatch_pedantic; } // In this case, the specifier is wrong and should be changed to match // the argument. EmitFormatDiagnostic(S.PDiag(diag) << AT.getRepresentativeTypeName(S.Context) << IntendedTy << IsEnum << E->getSourceRange(), E->getLocStart(), /*IsStringLocation*/ false, SpecRange, FixItHint::CreateReplacement(SpecRange, os.str())); } else { // The canonical type for formatting this value is different from the // actual type of the expression. (This occurs, for example, with Darwin's // NSInteger on 32-bit platforms, where it is typedef'd as 'int', but // should be printed as 'long' for 64-bit compatibility.) // Rather than emitting a normal format/argument mismatch, we want to // add a cast to the recommended type (and correct the format string // if necessary). SmallString<16> CastBuf; llvm::raw_svector_ostream CastFix(CastBuf); CastFix << "("; IntendedTy.print(CastFix, S.Context.getPrintingPolicy()); CastFix << ")"; SmallVector
Hints; if (!AT.matchesType(S.Context, IntendedTy)) Hints.push_back(FixItHint::CreateReplacement(SpecRange, os.str())); if (const CStyleCastExpr *CCast = dyn_cast
(E)) { // If there's already a cast present, just replace it. SourceRange CastRange(CCast->getLParenLoc(), CCast->getRParenLoc()); Hints.push_back(FixItHint::CreateReplacement(CastRange, CastFix.str())); } else if (!requiresParensToAddCast(E)) { // If the expression has high enough precedence, // just write the C-style cast. Hints.push_back(FixItHint::CreateInsertion(E->getLocStart(), CastFix.str())); } else { // Otherwise, add parens around the expression as well as the cast. CastFix << "("; Hints.push_back(FixItHint::CreateInsertion(E->getLocStart(), CastFix.str())); SourceLocation After = S.getLocForEndOfToken(E->getLocEnd()); Hints.push_back(FixItHint::CreateInsertion(After, ")")); } if (ShouldNotPrintDirectly) { // The expression has a type that should not be printed directly. // We extract the name from the typedef because we don't want to show // the underlying type in the diagnostic. StringRef Name; if (const TypedefType *TypedefTy = dyn_cast
(ExprTy)) Name = TypedefTy->getDecl()->getName(); else Name = CastTyName; EmitFormatDiagnostic(S.PDiag(diag::warn_format_argument_needs_cast) << Name << IntendedTy << IsEnum << E->getSourceRange(), E->getLocStart(), /*IsStringLocation=*/false, SpecRange, Hints); } else { // In this case, the expression could be printed using a different // specifier, but we've decided that the specifier is probably correct // and we should cast instead. Just use the normal warning message. EmitFormatDiagnostic( S.PDiag(diag::warn_format_conversion_argument_type_mismatch) << AT.getRepresentativeTypeName(S.Context) << ExprTy << IsEnum << E->getSourceRange(), E->getLocStart(), /*IsStringLocation*/false, SpecRange, Hints); } } } else { const CharSourceRange &CSR = getSpecifierRange(StartSpecifier, SpecifierLen); // Since the warning for passing non-POD types to variadic functions // was deferred until now, we emit a warning for non-POD // arguments here. switch (S.isValidVarArgType(ExprTy)) { case Sema::VAK_Valid: case Sema::VAK_ValidInCXX11: { unsigned diag = diag::warn_format_conversion_argument_type_mismatch; if (match == analyze_printf::ArgType::NoMatchPedantic) { diag = diag::warn_format_conversion_argument_type_mismatch_pedantic; } EmitFormatDiagnostic( S.PDiag(diag) << AT.getRepresentativeTypeName(S.Context) << ExprTy << IsEnum << CSR << E->getSourceRange(), E->getLocStart(), /*IsStringLocation*/ false, CSR); break; } case Sema::VAK_Undefined: case Sema::VAK_MSVCUndefined: EmitFormatDiagnostic( S.PDiag(diag::warn_non_pod_vararg_with_format_string) << S.getLangOpts().CPlusPlus11 << ExprTy << CallType << AT.getRepresentativeTypeName(S.Context) << CSR << E->getSourceRange(), E->getLocStart(), /*IsStringLocation*/false, CSR); checkForCStrMembers(AT, E); break; case Sema::VAK_Invalid: if (ExprTy->isObjCObjectType()) EmitFormatDiagnostic( S.PDiag(diag::err_cannot_pass_objc_interface_to_vararg_format) << S.getLangOpts().CPlusPlus11 << ExprTy << CallType << AT.getRepresentativeTypeName(S.Context) << CSR << E->getSourceRange(), E->getLocStart(), /*IsStringLocation*/false, CSR); else // FIXME: If this is an initializer list, suggest removing the braces // or inserting a cast to the target type. S.Diag(E->getLocStart(), diag::err_cannot_pass_to_vararg_format) << isa
(E) << ExprTy << CallType << AT.getRepresentativeTypeName(S.Context) << E->getSourceRange(); break; } assert(FirstDataArg + FS.getArgIndex() < CheckedVarArgs.size() && "format string specifier index out of range"); CheckedVarArgs[FirstDataArg + FS.getArgIndex()] = true; } return true; } //===--- CHECK: Scanf format string checking ------------------------------===// namespace { class CheckScanfHandler : public CheckFormatHandler { public: CheckScanfHandler(Sema &s, const StringLiteral *fexpr, const Expr *origFormatExpr, unsigned firstDataArg, unsigned numDataArgs, const char *beg, bool hasVAListArg, ArrayRef
Args, unsigned formatIdx, bool inFunctionCall, Sema::VariadicCallType CallType, llvm::SmallBitVector &CheckedVarArgs, UncoveredArgHandler &UncoveredArg) : CheckFormatHandler(s, fexpr, origFormatExpr, firstDataArg, numDataArgs, beg, hasVAListArg, Args, formatIdx, inFunctionCall, CallType, CheckedVarArgs, UncoveredArg) {} bool HandleScanfSpecifier(const analyze_scanf::ScanfSpecifier &FS, const char *startSpecifier, unsigned specifierLen) override; bool HandleInvalidScanfConversionSpecifier( const analyze_scanf::ScanfSpecifier &FS, const char *startSpecifier, unsigned specifierLen) override; void HandleIncompleteScanList(const char *start, const char *end) override; }; } // end anonymous namespace void CheckScanfHandler::HandleIncompleteScanList(const char *start, const char *end) { EmitFormatDiagnostic(S.PDiag(diag::warn_scanf_scanlist_incomplete), getLocationOfByte(end), /*IsStringLocation*/true, getSpecifierRange(start, end - start)); } bool CheckScanfHandler::HandleInvalidScanfConversionSpecifier( const analyze_scanf::ScanfSpecifier &FS, const char *startSpecifier, unsigned specifierLen) { const analyze_scanf::ScanfConversionSpecifier &CS = FS.getConversionSpecifier(); return HandleInvalidConversionSpecifier(FS.getArgIndex(), getLocationOfByte(CS.getStart()), startSpecifier, specifierLen, CS.getStart(), CS.getLength()); } bool CheckScanfHandler::HandleScanfSpecifier( const analyze_scanf::ScanfSpecifier &FS, const char *startSpecifier, unsigned specifierLen) { using namespace analyze_scanf; using namespace analyze_format_string; const ScanfConversionSpecifier &CS = FS.getConversionSpecifier(); // Handle case where '%' and '*' don't consume an argument. These shouldn't // be used to decide if we are using positional arguments consistently. if (FS.consumesDataArgument()) { if (atFirstArg) { atFirstArg = false; usesPositionalArgs = FS.usesPositionalArg(); } else if (usesPositionalArgs != FS.usesPositionalArg()) { HandlePositionalNonpositionalArgs(getLocationOfByte(CS.getStart()), startSpecifier, specifierLen); return false; } } // Check if the field with is non-zero. const OptionalAmount &Amt = FS.getFieldWidth(); if (Amt.getHowSpecified() == OptionalAmount::Constant) { if (Amt.getConstantAmount() == 0) { const CharSourceRange &R = getSpecifierRange(Amt.getStart(), Amt.getConstantLength()); EmitFormatDiagnostic(S.PDiag(diag::warn_scanf_nonzero_width), getLocationOfByte(Amt.getStart()), /*IsStringLocation*/true, R, FixItHint::CreateRemoval(R)); } } if (!FS.consumesDataArgument()) { // FIXME: Technically specifying a precision or field width here // makes no sense. Worth issuing a warning at some point. return true; } // Consume the argument. unsigned argIndex = FS.getArgIndex(); if (argIndex < NumDataArgs) { // The check to see if the argIndex is valid will come later. // We set the bit here because we may exit early from this // function if we encounter some other error. CoveredArgs.set(argIndex); } // Check the length modifier is valid with the given conversion specifier. if (!FS.hasValidLengthModifier(S.getASTContext().getTargetInfo())) HandleInvalidLengthModifier(FS, CS, startSpecifier, specifierLen, diag::warn_format_nonsensical_length); else if (!FS.hasStandardLengthModifier()) HandleNonStandardLengthModifier(FS, startSpecifier, specifierLen); else if (!FS.hasStandardLengthConversionCombination()) HandleInvalidLengthModifier(FS, CS, startSpecifier, specifierLen, diag::warn_format_non_standard_conversion_spec); if (!FS.hasStandardConversionSpecifier(S.getLangOpts())) HandleNonStandardConversionSpecifier(CS, startSpecifier, specifierLen); // The remaining checks depend on the data arguments. if (HasVAListArg) return true; if (!CheckNumArgs(FS, CS, startSpecifier, specifierLen, argIndex)) return false; // Check that the argument type matches the format specifier. const Expr *Ex = getDataArg(argIndex); if (!Ex) return true; const analyze_format_string::ArgType &AT = FS.getArgType(S.Context); if (!AT.isValid()) { return true; } analyze_format_string::ArgType::MatchKind match = AT.matchesType(S.Context, Ex->getType()); if (match == analyze_format_string::ArgType::Match) { return true; } ScanfSpecifier fixedFS = FS; bool success = fixedFS.fixType(Ex->getType(), Ex->IgnoreImpCasts()->getType(), S.getLangOpts(), S.Context); unsigned diag = diag::warn_format_conversion_argument_type_mismatch; if (match == analyze_format_string::ArgType::NoMatchPedantic) { diag = diag::warn_format_conversion_argument_type_mismatch_pedantic; } if (success) { // Get the fix string from the fixed format specifier. SmallString<128> buf; llvm::raw_svector_ostream os(buf); fixedFS.toString(os); EmitFormatDiagnostic( S.PDiag(diag) << AT.getRepresentativeTypeName(S.Context) << Ex->getType() << false << Ex->getSourceRange(), Ex->getLocStart(), /*IsStringLocation*/ false, getSpecifierRange(startSpecifier, specifierLen), FixItHint::CreateReplacement( getSpecifierRange(startSpecifier, specifierLen), os.str())); } else { EmitFormatDiagnostic(S.PDiag(diag) << AT.getRepresentativeTypeName(S.Context) << Ex->getType() << false << Ex->getSourceRange(), Ex->getLocStart(), /*IsStringLocation*/ false, getSpecifierRange(startSpecifier, specifierLen)); } return true; } static void CheckFormatString(Sema &S, const StringLiteral *FExpr, const Expr *OrigFormatExpr, ArrayRef
Args, bool HasVAListArg, unsigned format_idx, unsigned firstDataArg, Sema::FormatStringType Type, bool inFunctionCall, Sema::VariadicCallType CallType, llvm::SmallBitVector &CheckedVarArgs, UncoveredArgHandler &UncoveredArg) { // CHECK: is the format string a wide literal? if (!FExpr->isAscii() && !FExpr->isUTF8()) { CheckFormatHandler::EmitFormatDiagnostic( S, inFunctionCall, Args[format_idx], S.PDiag(diag::warn_format_string_is_wide_literal), FExpr->getLocStart(), /*IsStringLocation*/true, OrigFormatExpr->getSourceRange()); return; } // Str - The format string. NOTE: this is NOT null-terminated! StringRef StrRef = FExpr->getString(); const char *Str = StrRef.data(); // Account for cases where the string literal is truncated in a declaration. const ConstantArrayType *T = S.Context.getAsConstantArrayType(FExpr->getType()); assert(T && "String literal not of constant array type!"); size_t TypeSize = T->getSize().getZExtValue(); size_t StrLen = std::min(std::max(TypeSize, size_t(1)) - 1, StrRef.size()); const unsigned numDataArgs = Args.size() - firstDataArg; // Emit a warning if the string literal is truncated and does not contain an // embedded null character. if (TypeSize <= StrRef.size() && StrRef.substr(0, TypeSize).find('\0') == StringRef::npos) { CheckFormatHandler::EmitFormatDiagnostic( S, inFunctionCall, Args[format_idx], S.PDiag(diag::warn_printf_format_string_not_null_terminated), FExpr->getLocStart(), /*IsStringLocation=*/true, OrigFormatExpr->getSourceRange()); return; } // CHECK: empty format string? if (StrLen == 0 && numDataArgs > 0) { CheckFormatHandler::EmitFormatDiagnostic( S, inFunctionCall, Args[format_idx], S.PDiag(diag::warn_empty_format_string), FExpr->getLocStart(), /*IsStringLocation*/true, OrigFormatExpr->getSourceRange()); return; } if (Type == Sema::FST_Printf || Type == Sema::FST_NSString || Type == Sema::FST_FreeBSDKPrintf || Type == Sema::FST_OSTrace) { CheckPrintfHandler H(S, FExpr, OrigFormatExpr, firstDataArg, numDataArgs, (Type == Sema::FST_NSString || Type == Sema::FST_OSTrace), Str, HasVAListArg, Args, format_idx, inFunctionCall, CallType, CheckedVarArgs, UncoveredArg); if (!analyze_format_string::ParsePrintfString(H, Str, Str + StrLen, S.getLangOpts(), S.Context.getTargetInfo(), Type == Sema::FST_FreeBSDKPrintf)) H.DoneProcessing(); } else if (Type == Sema::FST_Scanf) { CheckScanfHandler H(S, FExpr, OrigFormatExpr, firstDataArg, numDataArgs, Str, HasVAListArg, Args, format_idx, inFunctionCall, CallType, CheckedVarArgs, UncoveredArg); if (!analyze_format_string::ParseScanfString(H, Str, Str + StrLen, S.getLangOpts(), S.Context.getTargetInfo())) H.DoneProcessing(); } // TODO: handle other formats } bool Sema::FormatStringHasSArg(const StringLiteral *FExpr) { // Str - The format string. NOTE: this is NOT null-terminated! StringRef StrRef = FExpr->getString(); const char *Str = StrRef.data(); // Account for cases where the string literal is truncated in a declaration. const ConstantArrayType *T = Context.getAsConstantArrayType(FExpr->getType()); assert(T && "String literal not of constant array type!"); size_t TypeSize = T->getSize().getZExtValue(); size_t StrLen = std::min(std::max(TypeSize, size_t(1)) - 1, StrRef.size()); return analyze_format_string::ParseFormatStringHasSArg(Str, Str + StrLen, getLangOpts(), Context.getTargetInfo()); } //===--- CHECK: Warn on use of wrong absolute value function. -------------===// // Returns the related absolute value function that is larger, of 0 if one // does not exist. static unsigned getLargerAbsoluteValueFunction(unsigned AbsFunction) { switch (AbsFunction) { default: return 0; case Builtin::BI__builtin_abs: return Builtin::BI__builtin_labs; case Builtin::BI__builtin_labs: return Builtin::BI__builtin_llabs; case Builtin::BI__builtin_llabs: return 0; case Builtin::BI__builtin_fabsf: return Builtin::BI__builtin_fabs; case Builtin::BI__builtin_fabs: return Builtin::BI__builtin_fabsl; case Builtin::BI__builtin_fabsl: return 0; case Builtin::BI__builtin_cabsf: return Builtin::BI__builtin_cabs; case Builtin::BI__builtin_cabs: return Builtin::BI__builtin_cabsl; case Builtin::BI__builtin_cabsl: return 0; case Builtin::BIabs: return Builtin::BIlabs; case Builtin::BIlabs: return Builtin::BIllabs; case Builtin::BIllabs: return 0; case Builtin::BIfabsf: return Builtin::BIfabs; case Builtin::BIfabs: return Builtin::BIfabsl; case Builtin::BIfabsl: return 0; case Builtin::BIcabsf: return Builtin::BIcabs; case Builtin::BIcabs: return Builtin::BIcabsl; case Builtin::BIcabsl: return 0; } } // Returns the argument type of the absolute value function. static QualType getAbsoluteValueArgumentType(ASTContext &Context, unsigned AbsType) { if (AbsType == 0) return QualType(); ASTContext::GetBuiltinTypeError Error = ASTContext::GE_None; QualType BuiltinType = Context.GetBuiltinType(AbsType, Error); if (Error != ASTContext::GE_None) return QualType(); const FunctionProtoType *FT = BuiltinType->getAs
(); if (!FT) return QualType(); if (FT->getNumParams() != 1) return QualType(); return FT->getParamType(0); } // Returns the best absolute value function, or zero, based on type and // current absolute value function. static unsigned getBestAbsFunction(ASTContext &Context, QualType ArgType, unsigned AbsFunctionKind) { unsigned BestKind = 0; uint64_t ArgSize = Context.getTypeSize(ArgType); for (unsigned Kind = AbsFunctionKind; Kind != 0; Kind = getLargerAbsoluteValueFunction(Kind)) { QualType ParamType = getAbsoluteValueArgumentType(Context, Kind); if (Context.getTypeSize(ParamType) >= ArgSize) { if (BestKind == 0) BestKind = Kind; else if (Context.hasSameType(ParamType, ArgType)) { BestKind = Kind; break; } } } return BestKind; } enum AbsoluteValueKind { AVK_Integer, AVK_Floating, AVK_Complex }; static AbsoluteValueKind getAbsoluteValueKind(QualType T) { if (T->isIntegralOrEnumerationType()) return AVK_Integer; if (T->isRealFloatingType()) return AVK_Floating; if (T->isAnyComplexType()) return AVK_Complex; llvm_unreachable("Type not integer, floating, or complex"); } // Changes the absolute value function to a different type. Preserves whether // the function is a builtin. static unsigned changeAbsFunction(unsigned AbsKind, AbsoluteValueKind ValueKind) { switch (ValueKind) { case AVK_Integer: switch (AbsKind) { default: return 0; case Builtin::BI__builtin_fabsf: case Builtin::BI__builtin_fabs: case Builtin::BI__builtin_fabsl: case Builtin::BI__builtin_cabsf: case Builtin::BI__builtin_cabs: case Builtin::BI__builtin_cabsl: return Builtin::BI__builtin_abs; case Builtin::BIfabsf: case Builtin::BIfabs: case Builtin::BIfabsl: case Builtin::BIcabsf: case Builtin::BIcabs: case Builtin::BIcabsl: return Builtin::BIabs; } case AVK_Floating: switch (AbsKind) { default: return 0; case Builtin::BI__builtin_abs: case Builtin::BI__builtin_labs: case Builtin::BI__builtin_llabs: case Builtin::BI__builtin_cabsf: case Builtin::BI__builtin_cabs: case Builtin::BI__builtin_cabsl: return Builtin::BI__builtin_fabsf; case Builtin::BIabs: case Builtin::BIlabs: case Builtin::BIllabs: case Builtin::BIcabsf: case Builtin::BIcabs: case Builtin::BIcabsl: return Builtin::BIfabsf; } case AVK_Complex: switch (AbsKind) { default: return 0; case Builtin::BI__builtin_abs: case Builtin::BI__builtin_labs: case Builtin::BI__builtin_llabs: case Builtin::BI__builtin_fabsf: case Builtin::BI__builtin_fabs: case Builtin::BI__builtin_fabsl: return Builtin::BI__builtin_cabsf; case Builtin::BIabs: case Builtin::BIlabs: case Builtin::BIllabs: case Builtin::BIfabsf: case Builtin::BIfabs: case Builtin::BIfabsl: return Builtin::BIcabsf; } } llvm_unreachable("Unable to convert function"); } static unsigned getAbsoluteValueFunctionKind(const FunctionDecl *FDecl) { const IdentifierInfo *FnInfo = FDecl->getIdentifier(); if (!FnInfo) return 0; switch (FDecl->getBuiltinID()) { default: return 0; case Builtin::BI__builtin_abs: case Builtin::BI__builtin_fabs: case Builtin::BI__builtin_fabsf: case Builtin::BI__builtin_fabsl: case Builtin::BI__builtin_labs: case Builtin::BI__builtin_llabs: case Builtin::BI__builtin_cabs: case Builtin::BI__builtin_cabsf: case Builtin::BI__builtin_cabsl: case Builtin::BIabs: case Builtin::BIlabs: case Builtin::BIllabs: case Builtin::BIfabs: case Builtin::BIfabsf: case Builtin::BIfabsl: case Builtin::BIcabs: case Builtin::BIcabsf: case Builtin::BIcabsl: return FDecl->getBuiltinID(); } llvm_unreachable("Unknown Builtin type"); } // If the replacement is valid, emit a note with replacement function. // Additionally, suggest including the proper header if not already included. static void emitReplacement(Sema &S, SourceLocation Loc, SourceRange Range, unsigned AbsKind, QualType ArgType) { bool EmitHeaderHint = true; const char *HeaderName = nullptr; const char *FunctionName = nullptr; if (S.getLangOpts().CPlusPlus && !ArgType->isAnyComplexType()) { FunctionName = "std::abs"; if (ArgType->isIntegralOrEnumerationType()) { HeaderName = "cstdlib"; } else if (ArgType->isRealFloatingType()) { HeaderName = "cmath"; } else { llvm_unreachable("Invalid Type"); } // Lookup all std::abs if (NamespaceDecl *Std = S.getStdNamespace()) { LookupResult R(S, &S.Context.Idents.get("abs"), Loc, Sema::LookupAnyName); R.suppressDiagnostics(); S.LookupQualifiedName(R, Std); for (const auto *I : R) { const FunctionDecl *FDecl = nullptr; if (const UsingShadowDecl *UsingD = dyn_cast
(I)) { FDecl = dyn_cast
(UsingD->getTargetDecl()); } else { FDecl = dyn_cast
(I); } if (!FDecl) continue; // Found std::abs(), check that they are the right ones. if (FDecl->getNumParams() != 1) continue; // Check that the parameter type can handle the argument. QualType ParamType = FDecl->getParamDecl(0)->getType(); if (getAbsoluteValueKind(ArgType) == getAbsoluteValueKind(ParamType) && S.Context.getTypeSize(ArgType) <= S.Context.getTypeSize(ParamType)) { // Found a function, don't need the header hint. EmitHeaderHint = false; break; } } } } else { FunctionName = S.Context.BuiltinInfo.getName(AbsKind); HeaderName = S.Context.BuiltinInfo.getHeaderName(AbsKind); if (HeaderName) { DeclarationName DN(&S.Context.Idents.get(FunctionName)); LookupResult R(S, DN, Loc, Sema::LookupAnyName); R.suppressDiagnostics(); S.LookupName(R, S.getCurScope()); if (R.isSingleResult()) { FunctionDecl *FD = dyn_cast
(R.getFoundDecl()); if (FD && FD->getBuiltinID() == AbsKind) { EmitHeaderHint = false; } else { return; } } else if (!R.empty()) { return; } } } S.Diag(Loc, diag::note_replace_abs_function) << FunctionName << FixItHint::CreateReplacement(Range, FunctionName); if (!HeaderName) return; if (!EmitHeaderHint) return; S.Diag(Loc, diag::note_include_header_or_declare) << HeaderName << FunctionName; } static bool IsFunctionStdAbs(const FunctionDecl *FDecl) { if (!FDecl) return false; if (!FDecl->getIdentifier() || !FDecl->getIdentifier()->isStr("abs")) return false; const NamespaceDecl *ND = dyn_cast
(FDecl->getDeclContext()); while (ND && ND->isInlineNamespace()) { ND = dyn_cast
(ND->getDeclContext()); } if (!ND || !ND->getIdentifier() || !ND->getIdentifier()->isStr("std")) return false; if (!isa
(ND->getDeclContext())) return false; return true; } // Warn when using the wrong abs() function. void Sema::CheckAbsoluteValueFunction(const CallExpr *Call, const FunctionDecl *FDecl, IdentifierInfo *FnInfo) { if (Call->getNumArgs() != 1) return; unsigned AbsKind = getAbsoluteValueFunctionKind(FDecl); bool IsStdAbs = IsFunctionStdAbs(FDecl); if (AbsKind == 0 && !IsStdAbs) return; QualType ArgType = Call->getArg(0)->IgnoreParenImpCasts()->getType(); QualType ParamType = Call->getArg(0)->getType(); // Unsigned types cannot be negative. Suggest removing the absolute value // function call. if (ArgType->isUnsignedIntegerType()) { const char *FunctionName = IsStdAbs ? "std::abs" : Context.BuiltinInfo.getName(AbsKind); Diag(Call->getExprLoc(), diag::warn_unsigned_abs) << ArgType << ParamType; Diag(Call->getExprLoc(), diag::note_remove_abs) << FunctionName << FixItHint::CreateRemoval(Call->getCallee()->getSourceRange()); return; } // Taking the absolute value of a pointer is very suspicious, they probably // wanted to index into an array, dereference a pointer, call a function, etc. if (ArgType->isPointerType() || ArgType->canDecayToPointerType()) { unsigned DiagType = 0; if (ArgType->isFunctionType()) DiagType = 1; else if (ArgType->isArrayType()) DiagType = 2; Diag(Call->getExprLoc(), diag::warn_pointer_abs) << DiagType << ArgType; return; } // std::abs has overloads which prevent most of the absolute value problems // from occurring. if (IsStdAbs) return; AbsoluteValueKind ArgValueKind = getAbsoluteValueKind(ArgType); AbsoluteValueKind ParamValueKind = getAbsoluteValueKind(ParamType); // The argument and parameter are the same kind. Check if they are the right // size. if (ArgValueKind == ParamValueKind) { if (Context.getTypeSize(ArgType) <= Context.getTypeSize(ParamType)) return; unsigned NewAbsKind = getBestAbsFunction(Context, ArgType, AbsKind); Diag(Call->getExprLoc(), diag::warn_abs_too_small) << FDecl << ArgType << ParamType; if (NewAbsKind == 0) return; emitReplacement(*this, Call->getExprLoc(), Call->getCallee()->getSourceRange(), NewAbsKind, ArgType); return; } // ArgValueKind != ParamValueKind // The wrong type of absolute value function was used. Attempt to find the // proper one. unsigned NewAbsKind = changeAbsFunction(AbsKind, ArgValueKind); NewAbsKind = getBestAbsFunction(Context, ArgType, NewAbsKind); if (NewAbsKind == 0) return; Diag(Call->getExprLoc(), diag::warn_wrong_absolute_value_type) << FDecl << ParamValueKind << ArgValueKind; emitReplacement(*this, Call->getExprLoc(), Call->getCallee()->getSourceRange(), NewAbsKind, ArgType); } //===--- CHECK: Standard memory functions ---------------------------------===// /// \brief Takes the expression passed to the size_t parameter of functions /// such as memcmp, strncat, etc and warns if it's a comparison. /// /// This is to catch typos like `if (memcmp(&a, &b, sizeof(a) > 0))`. static bool CheckMemorySizeofForComparison(Sema &S, const Expr *E, IdentifierInfo *FnName, SourceLocation FnLoc, SourceLocation RParenLoc) { const BinaryOperator *Size = dyn_cast
(E); if (!Size) return false; // if E is binop and op is >, <, >=, <=, ==, &&, ||: if (!Size->isComparisonOp() && !Size->isEqualityOp() && !Size->isLogicalOp()) return false; SourceRange SizeRange = Size->getSourceRange(); S.Diag(Size->getOperatorLoc(), diag::warn_memsize_comparison) << SizeRange << FnName; S.Diag(FnLoc, diag::note_memsize_comparison_paren) << FnName << FixItHint::CreateInsertion( S.getLocForEndOfToken(Size->getLHS()->getLocEnd()), ")") << FixItHint::CreateRemoval(RParenLoc); S.Diag(SizeRange.getBegin(), diag::note_memsize_comparison_cast_silence) << FixItHint::CreateInsertion(SizeRange.getBegin(), "(size_t)(") << FixItHint::CreateInsertion(S.getLocForEndOfToken(SizeRange.getEnd()), ")"); return true; } /// \brief Determine whether the given type is or contains a dynamic class type /// (e.g., whether it has a vtable). static const CXXRecordDecl *getContainedDynamicClass(QualType T, bool &IsContained) { // Look through array types while ignoring qualifiers. const Type *Ty = T->getBaseElementTypeUnsafe(); IsContained = false; const CXXRecordDecl *RD = Ty->getAsCXXRecordDecl(); RD = RD ? RD->getDefinition() : nullptr; if (!RD || RD->isInvalidDecl()) return nullptr; if (RD->isDynamicClass()) return RD; // Check all the fields. If any bases were dynamic, the class is dynamic. // It's impossible for a class to transitively contain itself by value, so // infinite recursion is impossible. for (auto *FD : RD->fields()) { bool SubContained; if (const CXXRecordDecl *ContainedRD = getContainedDynamicClass(FD->getType(), SubContained)) { IsContained = true; return ContainedRD; } } return nullptr; } /// \brief If E is a sizeof expression, returns its argument expression, /// otherwise returns NULL. static const Expr *getSizeOfExprArg(const Expr *E) { if (const UnaryExprOrTypeTraitExpr *SizeOf = dyn_cast
(E)) if (SizeOf->getKind() == clang::UETT_SizeOf && !SizeOf->isArgumentType()) return SizeOf->getArgumentExpr()->IgnoreParenImpCasts(); return nullptr; } /// \brief If E is a sizeof expression, returns its argument type. static QualType getSizeOfArgType(const Expr *E) { if (const UnaryExprOrTypeTraitExpr *SizeOf = dyn_cast
(E)) if (SizeOf->getKind() == clang::UETT_SizeOf) return SizeOf->getTypeOfArgument(); return QualType(); } /// \brief Check for dangerous or invalid arguments to memset(). /// /// This issues warnings on known problematic, dangerous or unspecified /// arguments to the standard 'memset', 'memcpy', 'memmove', and 'memcmp' /// function calls. /// /// \param Call The call expression to diagnose. void Sema::CheckMemaccessArguments(const CallExpr *Call, unsigned BId, IdentifierInfo *FnName) { assert(BId != 0); // It is possible to have a non-standard definition of memset. Validate // we have enough arguments, and if not, abort further checking. unsigned ExpectedNumArgs = (BId == Builtin::BIstrndup ? 2 : 3); if (Call->getNumArgs() < ExpectedNumArgs) return; unsigned LastArg = (BId == Builtin::BImemset || BId == Builtin::BIstrndup ? 1 : 2); unsigned LenArg = (BId == Builtin::BIstrndup ? 1 : 2); const Expr *LenExpr = Call->getArg(LenArg)->IgnoreParenImpCasts(); if (CheckMemorySizeofForComparison(*this, LenExpr, FnName, Call->getLocStart(), Call->getRParenLoc())) return; // We have special checking when the length is a sizeof expression. QualType SizeOfArgTy = getSizeOfArgType(LenExpr); const Expr *SizeOfArg = getSizeOfExprArg(LenExpr); llvm::FoldingSetNodeID SizeOfArgID; for (unsigned ArgIdx = 0; ArgIdx != LastArg; ++ArgIdx) { const Expr *Dest = Call->getArg(ArgIdx)->IgnoreParenImpCasts(); SourceRange ArgRange = Call->getArg(ArgIdx)->getSourceRange(); QualType DestTy = Dest->getType(); QualType PointeeTy; if (const PointerType *DestPtrTy = DestTy->getAs
()) { PointeeTy = DestPtrTy->getPointeeType(); // Never warn about void type pointers. This can be used to suppress // false positives. if (PointeeTy->isVoidType()) continue; // Catch "memset(p, 0, sizeof(p))" -- needs to be sizeof(*p). Do this by // actually comparing the expressions for equality. Because computing the // expression IDs can be expensive, we only do this if the diagnostic is // enabled. if (SizeOfArg && !Diags.isIgnored(diag::warn_sizeof_pointer_expr_memaccess, SizeOfArg->getExprLoc())) { // We only compute IDs for expressions if the warning is enabled, and // cache the sizeof arg's ID. if (SizeOfArgID == llvm::FoldingSetNodeID()) SizeOfArg->Profile(SizeOfArgID, Context, true); llvm::FoldingSetNodeID DestID; Dest->Profile(DestID, Context, true); if (DestID == SizeOfArgID) { // TODO: For strncpy() and friends, this could suggest sizeof(dst) // over sizeof(src) as well. unsigned ActionIdx = 0; // Default is to suggest dereferencing. StringRef ReadableName = FnName->getName(); if (const UnaryOperator *UnaryOp = dyn_cast
(Dest)) if (UnaryOp->getOpcode() == UO_AddrOf) ActionIdx = 1; // If its an address-of operator, just remove it. if (!PointeeTy->isIncompleteType() && (Context.getTypeSize(PointeeTy) == Context.getCharWidth())) ActionIdx = 2; // If the pointee's size is sizeof(char), // suggest an explicit length. // If the function is defined as a builtin macro, do not show macro // expansion. SourceLocation SL = SizeOfArg->getExprLoc(); SourceRange DSR = Dest->getSourceRange(); SourceRange SSR = SizeOfArg->getSourceRange(); SourceManager &SM = getSourceManager(); if (SM.isMacroArgExpansion(SL)) { ReadableName = Lexer::getImmediateMacroName(SL, SM, LangOpts); SL = SM.getSpellingLoc(SL); DSR = SourceRange(SM.getSpellingLoc(DSR.getBegin()), SM.getSpellingLoc(DSR.getEnd())); SSR = SourceRange(SM.getSpellingLoc(SSR.getBegin()), SM.getSpellingLoc(SSR.getEnd())); } DiagRuntimeBehavior(SL, SizeOfArg, PDiag(diag::warn_sizeof_pointer_expr_memaccess) << ReadableName << PointeeTy << DestTy << DSR << SSR); DiagRuntimeBehavior(SL, SizeOfArg, PDiag(diag::warn_sizeof_pointer_expr_memaccess_note) << ActionIdx << SSR); break; } } // Also check for cases where the sizeof argument is the exact same // type as the memory argument, and where it points to a user-defined // record type. if (SizeOfArgTy != QualType()) { if (PointeeTy->isRecordType() && Context.typesAreCompatible(SizeOfArgTy, DestTy)) { DiagRuntimeBehavior(LenExpr->getExprLoc(), Dest, PDiag(diag::warn_sizeof_pointer_type_memaccess) << FnName << SizeOfArgTy << ArgIdx << PointeeTy << Dest->getSourceRange() << LenExpr->getSourceRange()); break; } } } else if (DestTy->isArrayType()) { PointeeTy = DestTy; } if (PointeeTy == QualType()) continue; // Always complain about dynamic classes. bool IsContained; if (const CXXRecordDecl *ContainedRD = getContainedDynamicClass(PointeeTy, IsContained)) { unsigned OperationType = 0; // "overwritten" if we're warning about the destination for any call // but memcmp; otherwise a verb appropriate to the call. if (ArgIdx != 0 || BId == Builtin::BImemcmp) { if (BId == Builtin::BImemcpy) OperationType = 1; else if(BId == Builtin::BImemmove) OperationType = 2; else if (BId == Builtin::BImemcmp) OperationType = 3; } DiagRuntimeBehavior( Dest->getExprLoc(), Dest, PDiag(diag::warn_dyn_class_memaccess) << (BId == Builtin::BImemcmp ? ArgIdx + 2 : ArgIdx) << FnName << IsContained << ContainedRD << OperationType << Call->getCallee()->getSourceRange()); } else if (PointeeTy.hasNonTrivialObjCLifetime() && BId != Builtin::BImemset) DiagRuntimeBehavior( Dest->getExprLoc(), Dest, PDiag(diag::warn_arc_object_memaccess) << ArgIdx << FnName << PointeeTy << Call->getCallee()->getSourceRange()); else continue; DiagRuntimeBehavior( Dest->getExprLoc(), Dest, PDiag(diag::note_bad_memaccess_silence) << FixItHint::CreateInsertion(ArgRange.getBegin(), "(void*)")); break; } } // A little helper routine: ignore addition and subtraction of integer literals. // This intentionally does not ignore all integer constant expressions because // we don't want to remove sizeof(). static const Expr *ignoreLiteralAdditions(const Expr *Ex, ASTContext &Ctx) { Ex = Ex->IgnoreParenCasts(); for (;;) { const BinaryOperator * BO = dyn_cast
(Ex); if (!BO || !BO->isAdditiveOp()) break; const Expr *RHS = BO->getRHS()->IgnoreParenCasts(); const Expr *LHS = BO->getLHS()->IgnoreParenCasts(); if (isa
(RHS)) Ex = LHS; else if (isa
(LHS)) Ex = RHS; else break; } return Ex; } static bool isConstantSizeArrayWithMoreThanOneElement(QualType Ty, ASTContext &Context) { // Only handle constant-sized or VLAs, but not flexible members. if (const ConstantArrayType *CAT = Context.getAsConstantArrayType(Ty)) { // Only issue the FIXIT for arrays of size > 1. if (CAT->getSize().getSExtValue() <= 1) return false; } else if (!Ty->isVariableArrayType()) { return false; } return true; } // Warn if the user has made the 'size' argument to strlcpy or strlcat // be the size of the source, instead of the destination. void Sema::CheckStrlcpycatArguments(const CallExpr *Call, IdentifierInfo *FnName) { // Don't crash if the user has the wrong number of arguments unsigned NumArgs = Call->getNumArgs(); if ((NumArgs != 3) && (NumArgs != 4)) return; const Expr *SrcArg = ignoreLiteralAdditions(Call->getArg(1), Context); const Expr *SizeArg = ignoreLiteralAdditions(Call->getArg(2), Context); const Expr *CompareWithSrc = nullptr; if (CheckMemorySizeofForComparison(*this, SizeArg, FnName, Call->getLocStart(), Call->getRParenLoc())) return; // Look for 'strlcpy(dst, x, sizeof(x))' if (const Expr *Ex = getSizeOfExprArg(SizeArg)) CompareWithSrc = Ex; else { // Look for 'strlcpy(dst, x, strlen(x))' if (const CallExpr *SizeCall = dyn_cast
(SizeArg)) { if (SizeCall->getBuiltinCallee() == Builtin::BIstrlen && SizeCall->getNumArgs() == 1) CompareWithSrc = ignoreLiteralAdditions(SizeCall->getArg(0), Context); } } if (!CompareWithSrc) return; // Determine if the argument to sizeof/strlen is equal to the source // argument. In principle there's all kinds of things you could do // here, for instance creating an == expression and evaluating it with // EvaluateAsBooleanCondition, but this uses a more direct technique: const DeclRefExpr *SrcArgDRE = dyn_cast
(SrcArg); if (!SrcArgDRE) return; const DeclRefExpr *CompareWithSrcDRE = dyn_cast
(CompareWithSrc); if (!CompareWithSrcDRE || SrcArgDRE->getDecl() != CompareWithSrcDRE->getDecl()) return; const Expr *OriginalSizeArg = Call->getArg(2); Diag(CompareWithSrcDRE->getLocStart(), diag::warn_strlcpycat_wrong_size) << OriginalSizeArg->getSourceRange() << FnName; // Output a FIXIT hint if the destination is an array (rather than a // pointer to an array). This could be enhanced to handle some // pointers if we know the actual size, like if DstArg is 'array+2' // we could say 'sizeof(array)-2'. const Expr *DstArg = Call->getArg(0)->IgnoreParenImpCasts(); if (!isConstantSizeArrayWithMoreThanOneElement(DstArg->getType(), Context)) return; SmallString<128> sizeString; llvm::raw_svector_ostream OS(sizeString); OS << "sizeof("; DstArg->printPretty(OS, nullptr, getPrintingPolicy()); OS << ")"; Diag(OriginalSizeArg->getLocStart(), diag::note_strlcpycat_wrong_size) << FixItHint::CreateReplacement(OriginalSizeArg->getSourceRange(), OS.str()); } /// Check if two expressions refer to the same declaration. static bool referToTheSameDecl(const Expr *E1, const Expr *E2) { if (const DeclRefExpr *D1 = dyn_cast_or_null
(E1)) if (const DeclRefExpr *D2 = dyn_cast_or_null
(E2)) return D1->getDecl() == D2->getDecl(); return false; } static const Expr *getStrlenExprArg(const Expr *E) { if (const CallExpr *CE = dyn_cast
(E)) { const FunctionDecl *FD = CE->getDirectCallee(); if (!FD || FD->getMemoryFunctionKind() != Builtin::BIstrlen) return nullptr; return CE->getArg(0)->IgnoreParenCasts(); } return nullptr; } // Warn on anti-patterns as the 'size' argument to strncat. // The correct size argument should look like following: // strncat(dst, src, sizeof(dst) - strlen(dest) - 1); void Sema::CheckStrncatArguments(const CallExpr *CE, IdentifierInfo *FnName) { // Don't crash if the user has the wrong number of arguments. if (CE->getNumArgs() < 3) return; const Expr *DstArg = CE->getArg(0)->IgnoreParenCasts(); const Expr *SrcArg = CE->getArg(1)->IgnoreParenCasts(); const Expr *LenArg = CE->getArg(2)->IgnoreParenCasts(); if (CheckMemorySizeofForComparison(*this, LenArg, FnName, CE->getLocStart(), CE->getRParenLoc())) return; // Identify common expressions, which are wrongly used as the size argument // to strncat and may lead to buffer overflows. unsigned PatternType = 0; if (const Expr *SizeOfArg = getSizeOfExprArg(LenArg)) { // - sizeof(dst) if (referToTheSameDecl(SizeOfArg, DstArg)) PatternType = 1; // - sizeof(src) else if (referToTheSameDecl(SizeOfArg, SrcArg)) PatternType = 2; } else if (const BinaryOperator *BE = dyn_cast
(LenArg)) { if (BE->getOpcode() == BO_Sub) { const Expr *L = BE->getLHS()->IgnoreParenCasts(); const Expr *R = BE->getRHS()->IgnoreParenCasts(); // - sizeof(dst) - strlen(dst) if (referToTheSameDecl(DstArg, getSizeOfExprArg(L)) && referToTheSameDecl(DstArg, getStrlenExprArg(R))) PatternType = 1; // - sizeof(src) - (anything) else if (referToTheSameDecl(SrcArg, getSizeOfExprArg(L))) PatternType = 2; } } if (PatternType == 0) return; // Generate the diagnostic. SourceLocation SL = LenArg->getLocStart(); SourceRange SR = LenArg->getSourceRange(); SourceManager &SM = getSourceManager(); // If the function is defined as a builtin macro, do not show macro expansion. if (SM.isMacroArgExpansion(SL)) { SL = SM.getSpellingLoc(SL); SR = SourceRange(SM.getSpellingLoc(SR.getBegin()), SM.getSpellingLoc(SR.getEnd())); } // Check if the destination is an array (rather than a pointer to an array). QualType DstTy = DstArg->getType(); bool isKnownSizeArray = isConstantSizeArrayWithMoreThanOneElement(DstTy, Context); if (!isKnownSizeArray) { if (PatternType == 1) Diag(SL, diag::warn_strncat_wrong_size) << SR; else Diag(SL, diag::warn_strncat_src_size) << SR; return; } if (PatternType == 1) Diag(SL, diag::warn_strncat_large_size) << SR; else Diag(SL, diag::warn_strncat_src_size) << SR; SmallString<128> sizeString; llvm::raw_svector_ostream OS(sizeString); OS << "sizeof("; DstArg->printPretty(OS, nullptr, getPrintingPolicy()); OS << ") - "; OS << "strlen("; DstArg->printPretty(OS, nullptr, getPrintingPolicy()); OS << ") - 1"; Diag(SL, diag::note_strncat_wrong_size) << FixItHint::CreateReplacement(SR, OS.str()); } //===--- CHECK: Return Address of Stack Variable --------------------------===// static const Expr *EvalVal(const Expr *E, SmallVectorImpl
&refVars, const Decl *ParentDecl); static const Expr *EvalAddr(const Expr *E, SmallVectorImpl
&refVars, const Decl *ParentDecl); /// CheckReturnStackAddr - Check if a return statement returns the address /// of a stack variable. static void CheckReturnStackAddr(Sema &S, Expr *RetValExp, QualType lhsType, SourceLocation ReturnLoc) { const Expr *stackE = nullptr; SmallVector
refVars; // Perform checking for returned stack addresses, local blocks, // label addresses or references to temporaries. if (lhsType->isPointerType() || (!S.getLangOpts().ObjCAutoRefCount && lhsType->isBlockPointerType())) { stackE = EvalAddr(RetValExp, refVars, /*ParentDecl=*/nullptr); } else if (lhsType->isReferenceType()) { stackE = EvalVal(RetValExp, refVars, /*ParentDecl=*/nullptr); } if (!stackE) return; // Nothing suspicious was found. SourceLocation diagLoc; SourceRange diagRange; if (refVars.empty()) { diagLoc = stackE->getLocStart(); diagRange = stackE->getSourceRange(); } else { // We followed through a reference variable. 'stackE' contains the // problematic expression but we will warn at the return statement pointing // at the reference variable. We will later display the "trail" of // reference variables using notes. diagLoc = refVars[0]->getLocStart(); diagRange = refVars[0]->getSourceRange(); } if (const DeclRefExpr *DR = dyn_cast
(stackE)) { // address of local var S.Diag(diagLoc, diag::warn_ret_stack_addr_ref) << lhsType->isReferenceType() << DR->getDecl()->getDeclName() << diagRange; } else if (isa
(stackE)) { // local block. S.Diag(diagLoc, diag::err_ret_local_block) << diagRange; } else if (isa
(stackE)) { // address of label. S.Diag(diagLoc, diag::warn_ret_addr_label) << diagRange; } else { // local temporary. S.Diag(diagLoc, diag::warn_ret_local_temp_addr_ref) << lhsType->isReferenceType() << diagRange; } // Display the "trail" of reference variables that we followed until we // found the problematic expression using notes. for (unsigned i = 0, e = refVars.size(); i != e; ++i) { const VarDecl *VD = cast
(refVars[i]->getDecl()); // If this var binds to another reference var, show the range of the next // var, otherwise the var binds to the problematic expression, in which case // show the range of the expression. SourceRange range = (i < e - 1) ? refVars[i + 1]->getSourceRange() : stackE->getSourceRange(); S.Diag(VD->getLocation(), diag::note_ref_var_local_bind) << VD->getDeclName() << range; } } /// EvalAddr - EvalAddr and EvalVal are mutually recursive functions that /// check if the expression in a return statement evaluates to an address /// to a location on the stack, a local block, an address of a label, or a /// reference to local temporary. The recursion is used to traverse the /// AST of the return expression, with recursion backtracking when we /// encounter a subexpression that (1) clearly does not lead to one of the /// above problematic expressions (2) is something we cannot determine leads to /// a problematic expression based on such local checking. /// /// Both EvalAddr and EvalVal follow through reference variables to evaluate /// the expression that they point to. Such variables are added to the /// 'refVars' vector so that we know what the reference variable "trail" was. /// /// EvalAddr processes expressions that are pointers that are used as /// references (and not L-values). EvalVal handles all other values. /// At the base case of the recursion is a check for the above problematic /// expressions. /// /// This implementation handles: /// /// * pointer-to-pointer casts /// * implicit conversions from array references to pointers /// * taking the address of fields /// * arbitrary interplay between "&" and "*" operators /// * pointer arithmetic from an address of a stack variable /// * taking the address of an array element where the array is on the stack static const Expr *EvalAddr(const Expr *E, SmallVectorImpl
&refVars, const Decl *ParentDecl) { if (E->isTypeDependent()) return nullptr; // We should only be called for evaluating pointer expressions. assert((E->getType()->isAnyPointerType() || E->getType()->isBlockPointerType() || E->getType()->isObjCQualifiedIdType()) && "EvalAddr only works on pointers"); E = E->IgnoreParens(); // Our "symbolic interpreter" is just a dispatch off the currently // viewed AST node. We then recursively traverse the AST by calling // EvalAddr and EvalVal appropriately. switch (E->getStmtClass()) { case Stmt::DeclRefExprClass: { const DeclRefExpr *DR = cast
(E); // If we leave the immediate function, the lifetime isn't about to end. if (DR->refersToEnclosingVariableOrCapture()) return nullptr; if (const VarDecl *V = dyn_cast
(DR->getDecl())) // If this is a reference variable, follow through to the expression that // it points to. if (V->hasLocalStorage() && V->getType()->isReferenceType() && V->hasInit()) { // Add the reference variable to the "trail". refVars.push_back(DR); return EvalAddr(V->getInit(), refVars, ParentDecl); } return nullptr; } case Stmt::UnaryOperatorClass: { // The only unary operator that make sense to handle here // is AddrOf. All others don't make sense as pointers. const UnaryOperator *U = cast
(E); if (U->getOpcode() == UO_AddrOf) return EvalVal(U->getSubExpr(), refVars, ParentDecl); return nullptr; } case Stmt::BinaryOperatorClass: { // Handle pointer arithmetic. All other binary operators are not valid // in this context. const BinaryOperator *B = cast
(E); BinaryOperatorKind op = B->getOpcode(); if (op != BO_Add && op != BO_Sub) return nullptr; const Expr *Base = B->getLHS(); // Determine which argument is the real pointer base. It could be // the RHS argument instead of the LHS. if (!Base->getType()->isPointerType()) Base = B->getRHS(); assert(Base->getType()->isPointerType()); return EvalAddr(Base, refVars, ParentDecl); } // For conditional operators we need to see if either the LHS or RHS are // valid DeclRefExpr*s. If one of them is valid, we return it. case Stmt::ConditionalOperatorClass: { const ConditionalOperator *C = cast
(E); // Handle the GNU extension for missing LHS. // FIXME: That isn't a ConditionalOperator, so doesn't get here. if (const Expr *LHSExpr = C->getLHS()) { // In C++, we can have a throw-expression, which has 'void' type. if (!LHSExpr->getType()->isVoidType()) if (const Expr *LHS = EvalAddr(LHSExpr, refVars, ParentDecl)) return LHS; } // In C++, we can have a throw-expression, which has 'void' type. if (C->getRHS()->getType()->isVoidType()) return nullptr; return EvalAddr(C->getRHS(), refVars, ParentDecl); } case Stmt::BlockExprClass: if (cast
(E)->getBlockDecl()->hasCaptures()) return E; // local block. return nullptr; case Stmt::AddrLabelExprClass: return E; // address of label. case Stmt::ExprWithCleanupsClass: return EvalAddr(cast
(E)->getSubExpr(), refVars, ParentDecl); // For casts, we need to handle conversions from arrays to // pointer values, and pointer-to-pointer conversions. case Stmt::ImplicitCastExprClass: case Stmt::CStyleCastExprClass: case Stmt::CXXFunctionalCastExprClass: case Stmt::ObjCBridgedCastExprClass: case Stmt::CXXStaticCastExprClass: case Stmt::CXXDynamicCastExprClass: case Stmt::CXXConstCastExprClass: case Stmt::CXXReinterpretCastExprClass: { const Expr* SubExpr = cast
(E)->getSubExpr(); switch (cast
(E)->getCastKind()) { case CK_LValueToRValue: case CK_NoOp: case CK_BaseToDerived: case CK_DerivedToBase: case CK_UncheckedDerivedToBase: case CK_Dynamic: case CK_CPointerToObjCPointerCast: case CK_BlockPointerToObjCPointerCast: case CK_AnyPointerToBlockPointerCast: return EvalAddr(SubExpr, refVars, ParentDecl); case CK_ArrayToPointerDecay: return EvalVal(SubExpr, refVars, ParentDecl); case CK_BitCast: if (SubExpr->getType()->isAnyPointerType() || SubExpr->getType()->isBlockPointerType() || SubExpr->getType()->isObjCQualifiedIdType()) return EvalAddr(SubExpr, refVars, ParentDecl); else return nullptr; default: return nullptr; } } case Stmt::MaterializeTemporaryExprClass: if (const Expr *Result = EvalAddr(cast
(E)->GetTemporaryExpr(), refVars, ParentDecl)) return Result; return E; // Everything else: we simply don't reason about them. default: return nullptr; } } /// EvalVal - This function is complements EvalAddr in the mutual recursion. /// See the comments for EvalAddr for more details. static const Expr *EvalVal(const Expr *E, SmallVectorImpl
&refVars, const Decl *ParentDecl) { do { // We should only be called for evaluating non-pointer expressions, or // expressions with a pointer type that are not used as references but // instead // are l-values (e.g., DeclRefExpr with a pointer type). // Our "symbolic interpreter" is just a dispatch off the currently // viewed AST node. We then recursively traverse the AST by calling // EvalAddr and EvalVal appropriately. E = E->IgnoreParens(); switch (E->getStmtClass()) { case Stmt::ImplicitCastExprClass: { const ImplicitCastExpr *IE = cast
(E); if (IE->getValueKind() == VK_LValue) { E = IE->getSubExpr(); continue; } return nullptr; } case Stmt::ExprWithCleanupsClass: return EvalVal(cast
(E)->getSubExpr(), refVars, ParentDecl); case Stmt::DeclRefExprClass: { // When we hit a DeclRefExpr we are looking at code that refers to a // variable's name. If it's not a reference variable we check if it has // local storage within the function, and if so, return the expression. const DeclRefExpr *DR = cast
(E); // If we leave the immediate function, the lifetime isn't about to end. if (DR->refersToEnclosingVariableOrCapture()) return nullptr; if (const VarDecl *V = dyn_cast
(DR->getDecl())) { // Check if it refers to itself, e.g. "int& i = i;". if (V == ParentDecl) return DR; if (V->hasLocalStorage()) { if (!V->getType()->isReferenceType()) return DR; // Reference variable, follow through to the expression that // it points to. if (V->hasInit()) { // Add the reference variable to the "trail". refVars.push_back(DR); return EvalVal(V->getInit(), refVars, V); } } } return nullptr; } case Stmt::UnaryOperatorClass: { // The only unary operator that make sense to handle here // is Deref. All others don't resolve to a "name." This includes // handling all sorts of rvalues passed to a unary operator. const UnaryOperator *U = cast
(E); if (U->getOpcode() == UO_Deref) return EvalAddr(U->getSubExpr(), refVars, ParentDecl); return nullptr; } case Stmt::ArraySubscriptExprClass: { // Array subscripts are potential references to data on the stack. We // retrieve the DeclRefExpr* for the array variable if it indeed // has local storage. const auto *ASE = cast
(E); if (ASE->isTypeDependent()) return nullptr; return EvalAddr(ASE->getBase(), refVars, ParentDecl); } case Stmt::OMPArraySectionExprClass: { return EvalAddr(cast
(E)->getBase(), refVars, ParentDecl); } case Stmt::ConditionalOperatorClass: { // For conditional operators we need to see if either the LHS or RHS are // non-NULL Expr's. If one is non-NULL, we return it. const ConditionalOperator *C = cast
(E); // Handle the GNU extension for missing LHS. if (const Expr *LHSExpr = C->getLHS()) { // In C++, we can have a throw-expression, which has 'void' type. if (!LHSExpr->getType()->isVoidType()) if (const Expr *LHS = EvalVal(LHSExpr, refVars, ParentDecl)) return LHS; } // In C++, we can have a throw-expression, which has 'void' type. if (C->getRHS()->getType()->isVoidType()) return nullptr; return EvalVal(C->getRHS(), refVars, ParentDecl); } // Accesses to members are potential references to data on the stack. case Stmt::MemberExprClass: { const MemberExpr *M = cast
(E); // Check for indirect access. We only want direct field accesses. if (M->isArrow()) return nullptr; // Check whether the member type is itself a reference, in which case // we're not going to refer to the member, but to what the member refers // to. if (M->getMemberDecl()->getType()->isReferenceType()) return nullptr; return EvalVal(M->getBase(), refVars, ParentDecl); } case Stmt::MaterializeTemporaryExprClass: if (const Expr *Result = EvalVal(cast
(E)->GetTemporaryExpr(), refVars, ParentDecl)) return Result; return E; default: // Check that we don't return or take the address of a reference to a // temporary. This is only useful in C++. if (!E->isTypeDependent() && E->isRValue()) return E; // Everything else: we simply don't reason about them. return nullptr; } } while (true); } void Sema::CheckReturnValExpr(Expr *RetValExp, QualType lhsType, SourceLocation ReturnLoc, bool isObjCMethod, const AttrVec *Attrs, const FunctionDecl *FD) { CheckReturnStackAddr(*this, RetValExp, lhsType, ReturnLoc); // Check if the return value is null but should not be. if (((Attrs && hasSpecificAttr
(*Attrs)) || (!isObjCMethod && isNonNullType(Context, lhsType))) && CheckNonNullExpr(*this, RetValExp)) Diag(ReturnLoc, diag::warn_null_ret) << (isObjCMethod ? 1 : 0) << RetValExp->getSourceRange(); // C++11 [basic.stc.dynamic.allocation]p4: // If an allocation function declared with a non-throwing // exception-specification fails to allocate storage, it shall return // a null pointer. Any other allocation function that fails to allocate // storage shall indicate failure only by throwing an exception [...] if (FD) { OverloadedOperatorKind Op = FD->getOverloadedOperator(); if (Op == OO_New || Op == OO_Array_New) { const FunctionProtoType *Proto = FD->getType()->castAs
(); if (!Proto->isNothrow(Context, /*ResultIfDependent*/true) && CheckNonNullExpr(*this, RetValExp)) Diag(ReturnLoc, diag::warn_operator_new_returns_null) << FD << getLangOpts().CPlusPlus11; } } } //===--- CHECK: Floating-Point comparisons (-Wfloat-equal) ---------------===// /// Check for comparisons of floating point operands using != and ==. /// Issue a warning if these are no self-comparisons, as they are not likely /// to do what the programmer intended. void Sema::CheckFloatComparison(SourceLocation Loc, Expr* LHS, Expr *RHS) { Expr* LeftExprSansParen = LHS->IgnoreParenImpCasts(); Expr* RightExprSansParen = RHS->IgnoreParenImpCasts(); // Special case: check for x == x (which is OK). // Do not emit warnings for such cases. if (DeclRefExpr* DRL = dyn_cast
(LeftExprSansParen)) if (DeclRefExpr* DRR = dyn_cast
(RightExprSansParen)) if (DRL->getDecl() == DRR->getDecl()) return; // Special case: check for comparisons against literals that can be exactly // represented by APFloat. In such cases, do not emit a warning. This // is a heuristic: often comparison against such literals are used to // detect if a value in a variable has not changed. This clearly can // lead to false negatives. if (FloatingLiteral* FLL = dyn_cast
(LeftExprSansParen)) { if (FLL->isExact()) return; } else if (FloatingLiteral* FLR = dyn_cast
(RightExprSansParen)) if (FLR->isExact()) return; // Check for comparisons with builtin types. if (CallExpr* CL = dyn_cast
(LeftExprSansParen)) if (CL->getBuiltinCallee()) return; if (CallExpr* CR = dyn_cast
(RightExprSansParen)) if (CR->getBuiltinCallee()) return; // Emit the diagnostic. Diag(Loc, diag::warn_floatingpoint_eq) << LHS->getSourceRange() << RHS->getSourceRange(); } //===--- CHECK: Integer mixed-sign comparisons (-Wsign-compare) --------===// //===--- CHECK: Lossy implicit conversions (-Wconversion) --------------===// namespace { /// Structure recording the 'active' range of an integer-valued /// expression. struct IntRange { /// The number of bits active in the int. unsigned Width; /// True if the int is known not to have negative values. bool NonNegative; IntRange(unsigned Width, bool NonNegative) : Width(Width), NonNegative(NonNegative) {} /// Returns the range of the bool type. static IntRange forBoolType() { return IntRange(1, true); } /// Returns the range of an opaque value of the given integral type. static IntRange forValueOfType(ASTContext &C, QualType T) { return forValueOfCanonicalType(C, T->getCanonicalTypeInternal().getTypePtr()); } /// Returns the range of an opaque value of a canonical integral type. static IntRange forValueOfCanonicalType(ASTContext &C, const Type *T) { assert(T->isCanonicalUnqualified()); if (const VectorType *VT = dyn_cast
(T)) T = VT->getElementType().getTypePtr(); if (const ComplexType *CT = dyn_cast
(T)) T = CT->getElementType().getTypePtr(); if (const AtomicType *AT = dyn_cast
(T)) T = AT->getValueType().getTypePtr(); // For enum types, use the known bit width of the enumerators. if (const EnumType *ET = dyn_cast
(T)) { EnumDecl *Enum = ET->getDecl(); if (!Enum->isCompleteDefinition()) return IntRange(C.getIntWidth(QualType(T, 0)), false); unsigned NumPositive = Enum->getNumPositiveBits(); unsigned NumNegative = Enum->getNumNegativeBits(); if (NumNegative == 0) return IntRange(NumPositive, true/*NonNegative*/); else return IntRange(std::max(NumPositive + 1, NumNegative), false/*NonNegative*/); } const BuiltinType *BT = cast
(T); assert(BT->isInteger()); return IntRange(C.getIntWidth(QualType(T, 0)), BT->isUnsignedInteger()); } /// Returns the "target" range of a canonical integral type, i.e. /// the range of values expressible in the type. /// /// This matches forValueOfCanonicalType except that enums have the /// full range of their type, not the range of their enumerators. static IntRange forTargetOfCanonicalType(ASTContext &C, const Type *T) { assert(T->isCanonicalUnqualified()); if (const VectorType *VT = dyn_cast
(T)) T = VT->getElementType().getTypePtr(); if (const ComplexType *CT = dyn_cast
(T)) T = CT->getElementType().getTypePtr(); if (const AtomicType *AT = dyn_cast
(T)) T = AT->getValueType().getTypePtr(); if (const EnumType *ET = dyn_cast
(T)) T = C.getCanonicalType(ET->getDecl()->getIntegerType()).getTypePtr(); const BuiltinType *BT = cast
(T); assert(BT->isInteger()); return IntRange(C.getIntWidth(QualType(T, 0)), BT->isUnsignedInteger()); } /// Returns the supremum of two ranges: i.e. their conservative merge. static IntRange join(IntRange L, IntRange R) { return IntRange(std::max(L.Width, R.Width), L.NonNegative && R.NonNegative); } /// Returns the infinum of two ranges: i.e. their aggressive merge. static IntRange meet(IntRange L, IntRange R) { return IntRange(std::min(L.Width, R.Width), L.NonNegative || R.NonNegative); } }; IntRange GetValueRange(ASTContext &C, llvm::APSInt &value, unsigned MaxWidth) { if (value.isSigned() && value.isNegative()) return IntRange(value.getMinSignedBits(), false); if (value.getBitWidth() > MaxWidth) value = value.trunc(MaxWidth); // isNonNegative() just checks the sign bit without considering // signedness. return IntRange(value.getActiveBits(), true); } IntRange GetValueRange(ASTContext &C, APValue &result, QualType Ty, unsigned MaxWidth) { if (result.isInt()) return GetValueRange(C, result.getInt(), MaxWidth); if (result.isVector()) { IntRange R = GetValueRange(C, result.getVectorElt(0), Ty, MaxWidth); for (unsigned i = 1, e = result.getVectorLength(); i != e; ++i) { IntRange El = GetValueRange(C, result.getVectorElt(i), Ty, MaxWidth); R = IntRange::join(R, El); } return R; } if (result.isComplexInt()) { IntRange R = GetValueRange(C, result.getComplexIntReal(), MaxWidth); IntRange I = GetValueRange(C, result.getComplexIntImag(), MaxWidth); return IntRange::join(R, I); } // This can happen with lossless casts to intptr_t of "based" lvalues. // Assume it might use arbitrary bits. // FIXME: The only reason we need to pass the type in here is to get // the sign right on this one case. It would be nice if APValue // preserved this. assert(result.isLValue() || result.isAddrLabelDiff()); return IntRange(MaxWidth, Ty->isUnsignedIntegerOrEnumerationType()); } QualType GetExprType(const Expr *E) { QualType Ty = E->getType(); if (const AtomicType *AtomicRHS = Ty->getAs
()) Ty = AtomicRHS->getValueType(); return Ty; } /// Pseudo-evaluate the given integer expression, estimating the /// range of values it might take. /// /// \param MaxWidth - the width to which the value will be truncated IntRange GetExprRange(ASTContext &C, const Expr *E, unsigned MaxWidth) { E = E->IgnoreParens(); // Try a full evaluation first. Expr::EvalResult result; if (E->EvaluateAsRValue(result, C)) return GetValueRange(C, result.Val, GetExprType(E), MaxWidth); // I think we only want to look through implicit casts here; if the // user has an explicit widening cast, we should treat the value as // being of the new, wider type. if (const auto *CE = dyn_cast
(E)) { if (CE->getCastKind() == CK_NoOp || CE->getCastKind() == CK_LValueToRValue) return GetExprRange(C, CE->getSubExpr(), MaxWidth); IntRange OutputTypeRange = IntRange::forValueOfType(C, GetExprType(CE)); bool isIntegerCast = CE->getCastKind() == CK_IntegralCast || CE->getCastKind() == CK_BooleanToSignedIntegral; // Assume that non-integer casts can span the full range of the type. if (!isIntegerCast) return OutputTypeRange; IntRange SubRange = GetExprRange(C, CE->getSubExpr(), std::min(MaxWidth, OutputTypeRange.Width)); // Bail out if the subexpr's range is as wide as the cast type. if (SubRange.Width >= OutputTypeRange.Width) return OutputTypeRange; // Otherwise, we take the smaller width, and we're non-negative if // either the output type or the subexpr is. return IntRange(SubRange.Width, SubRange.NonNegative || OutputTypeRange.NonNegative); } if (const auto *CO = dyn_cast
(E)) { // If we can fold the condition, just take that operand. bool CondResult; if (CO->getCond()->EvaluateAsBooleanCondition(CondResult, C)) return GetExprRange(C, CondResult ? CO->getTrueExpr() : CO->getFalseExpr(), MaxWidth); // Otherwise, conservatively merge. IntRange L = GetExprRange(C, CO->getTrueExpr(), MaxWidth); IntRange R = GetExprRange(C, CO->getFalseExpr(), MaxWidth); return IntRange::join(L, R); } if (const auto *BO = dyn_cast
(E)) { switch (BO->getOpcode()) { // Boolean-valued operations are single-bit and positive. case BO_LAnd: case BO_LOr: case BO_LT: case BO_GT: case BO_LE: case BO_GE: case BO_EQ: case BO_NE: return IntRange::forBoolType(); // The type of the assignments is the type of the LHS, so the RHS // is not necessarily the same type. case BO_MulAssign: case BO_DivAssign: case BO_RemAssign: case BO_AddAssign: case BO_SubAssign: case BO_XorAssign: case BO_OrAssign: // TODO: bitfields? return IntRange::forValueOfType(C, GetExprType(E)); // Simple assignments just pass through the RHS, which will have // been coerced to the LHS type. case BO_Assign: // TODO: bitfields? return GetExprRange(C, BO->getRHS(), MaxWidth); // Operations with opaque sources are black-listed. case BO_PtrMemD: case BO_PtrMemI: return IntRange::forValueOfType(C, GetExprType(E)); // Bitwise-and uses the *infinum* of the two source ranges. case BO_And: case BO_AndAssign: return IntRange::meet(GetExprRange(C, BO->getLHS(), MaxWidth), GetExprRange(C, BO->getRHS(), MaxWidth)); // Left shift gets black-listed based on a judgement call. case BO_Shl: // ...except that we want to treat '1 << (blah)' as logically // positive. It's an important idiom. if (IntegerLiteral *I = dyn_cast
(BO->getLHS()->IgnoreParenCasts())) { if (I->getValue() == 1) { IntRange R = IntRange::forValueOfType(C, GetExprType(E)); return IntRange(R.Width, /*NonNegative*/ true); } } // fallthrough case BO_ShlAssign: return IntRange::forValueOfType(C, GetExprType(E)); // Right shift by a constant can narrow its left argument. case BO_Shr: case BO_ShrAssign: { IntRange L = GetExprRange(C, BO->getLHS(), MaxWidth); // If the shift amount is a positive constant, drop the width by // that much. llvm::APSInt shift; if (BO->getRHS()->isIntegerConstantExpr(shift, C) && shift.isNonNegative()) { unsigned zext = shift.getZExtValue(); if (zext >= L.Width) L.Width = (L.NonNegative ? 0 : 1); else L.Width -= zext; } return L; } // Comma acts as its right operand. case BO_Comma: return GetExprRange(C, BO->getRHS(), MaxWidth); // Black-list pointer subtractions. case BO_Sub: if (BO->getLHS()->getType()->isPointerType()) return IntRange::forValueOfType(C, GetExprType(E)); break; // The width of a division result is mostly determined by the size // of the LHS. case BO_Div: { // Don't 'pre-truncate' the operands. unsigned opWidth = C.getIntWidth(GetExprType(E)); IntRange L = GetExprRange(C, BO->getLHS(), opWidth); // If the divisor is constant, use that. llvm::APSInt divisor; if (BO->getRHS()->isIntegerConstantExpr(divisor, C)) { unsigned log2 = divisor.logBase2(); // floor(log_2(divisor)) if (log2 >= L.Width) L.Width = (L.NonNegative ? 0 : 1); else L.Width = std::min(L.Width - log2, MaxWidth); return L; } // Otherwise, just use the LHS's width. IntRange R = GetExprRange(C, BO->getRHS(), opWidth); return IntRange(L.Width, L.NonNegative && R.NonNegative); } // The result of a remainder can't be larger than the result of // either side. case BO_Rem: { // Don't 'pre-truncate' the operands. unsigned opWidth = C.getIntWidth(GetExprType(E)); IntRange L = GetExprRange(C, BO->getLHS(), opWidth); IntRange R = GetExprRange(C, BO->getRHS(), opWidth); IntRange meet = IntRange::meet(L, R); meet.Width = std::min(meet.Width, MaxWidth); return meet; } // The default behavior is okay for these. case BO_Mul: case BO_Add: case BO_Xor: case BO_Or: break; } // The default case is to treat the operation as if it were closed // on the narrowest type that encompasses both operands. IntRange L = GetExprRange(C, BO->getLHS(), MaxWidth); IntRange R = GetExprRange(C, BO->getRHS(), MaxWidth); return IntRange::join(L, R); } if (const auto *UO = dyn_cast
(E)) { switch (UO->getOpcode()) { // Boolean-valued operations are white-listed. case UO_LNot: return IntRange::forBoolType(); // Operations with opaque sources are black-listed. case UO_Deref: case UO_AddrOf: // should be impossible return IntRange::forValueOfType(C, GetExprType(E)); default: return GetExprRange(C, UO->getSubExpr(), MaxWidth); } } if (const auto *OVE = dyn_cast
(E)) return GetExprRange(C, OVE->getSourceExpr(), MaxWidth); if (const auto *BitField = E->getSourceBitField()) return IntRange(BitField->getBitWidthValue(C), BitField->getType()->isUnsignedIntegerOrEnumerationType()); return IntRange::forValueOfType(C, GetExprType(E)); } IntRange GetExprRange(ASTContext &C, const Expr *E) { return GetExprRange(C, E, C.getIntWidth(GetExprType(E))); } /// Checks whether the given value, which currently has the given /// source semantics, has the same value when coerced through the /// target semantics. bool IsSameFloatAfterCast(const llvm::APFloat &value, const llvm::fltSemantics &Src, const llvm::fltSemantics &Tgt) { llvm::APFloat truncated = value; bool ignored; truncated.convert(Src, llvm::APFloat::rmNearestTiesToEven, &ignored); truncated.convert(Tgt, llvm::APFloat::rmNearestTiesToEven, &ignored); return truncated.bitwiseIsEqual(value); } /// Checks whether the given value, which currently has the given /// source semantics, has the same value when coerced through the /// target semantics. /// /// The value might be a vector of floats (or a complex number). bool IsSameFloatAfterCast(const APValue &value, const llvm::fltSemantics &Src, const llvm::fltSemantics &Tgt) { if (value.isFloat()) return IsSameFloatAfterCast(value.getFloat(), Src, Tgt); if (value.isVector()) { for (unsigned i = 0, e = value.getVectorLength(); i != e; ++i) if (!IsSameFloatAfterCast(value.getVectorElt(i), Src, Tgt)) return false; return true; } assert(value.isComplexFloat()); return (IsSameFloatAfterCast(value.getComplexFloatReal(), Src, Tgt) && IsSameFloatAfterCast(value.getComplexFloatImag(), Src, Tgt)); } void AnalyzeImplicitConversions(Sema &S, Expr *E, SourceLocation CC); bool IsZero(Sema &S, Expr *E) { // Suppress cases where we are comparing against an enum constant. if (const DeclRefExpr *DR = dyn_cast
(E->IgnoreParenImpCasts())) if (isa
(DR->getDecl())) return false; // Suppress cases where the '0' value is expanded from a macro. if (E->getLocStart().isMacroID()) return false; llvm::APSInt Value; return E->isIntegerConstantExpr(Value, S.Context) && Value == 0; } bool HasEnumType(Expr *E) { // Strip off implicit integral promotions. while (ImplicitCastExpr *ICE = dyn_cast
(E)) { if (ICE->getCastKind() != CK_IntegralCast && ICE->getCastKind() != CK_NoOp) break; E = ICE->getSubExpr(); } return E->getType()->isEnumeralType(); } void CheckTrivialUnsignedComparison(Sema &S, BinaryOperator *E) { // Disable warning in template instantiations. if (!S.ActiveTemplateInstantiations.empty()) return; BinaryOperatorKind op = E->getOpcode(); if (E->isValueDependent()) return; if (op == BO_LT && IsZero(S, E->getRHS())) { S.Diag(E->getOperatorLoc(), diag::warn_lunsigned_always_true_comparison) << "< 0" << "false" << HasEnumType(E->getLHS()) << E->getLHS()->getSourceRange() << E->getRHS()->getSourceRange(); } else if (op == BO_GE && IsZero(S, E->getRHS())) { S.Diag(E->getOperatorLoc(), diag::warn_lunsigned_always_true_comparison) << ">= 0" << "true" << HasEnumType(E->getLHS()) << E->getLHS()->getSourceRange() << E->getRHS()->getSourceRange(); } else if (op == BO_GT && IsZero(S, E->getLHS())) { S.Diag(E->getOperatorLoc(), diag::warn_runsigned_always_true_comparison) << "0 >" << "false" << HasEnumType(E->getRHS()) << E->getLHS()->getSourceRange() << E->getRHS()->getSourceRange(); } else if (op == BO_LE && IsZero(S, E->getLHS())) { S.Diag(E->getOperatorLoc(), diag::warn_runsigned_always_true_comparison) << "0 <=" << "true" << HasEnumType(E->getRHS()) << E->getLHS()->getSourceRange() << E->getRHS()->getSourceRange(); } } void DiagnoseOutOfRangeComparison(Sema &S, BinaryOperator *E, Expr *Constant, Expr *Other, const llvm::APSInt &Value, bool RhsConstant) { // Disable warning in template instantiations. if (!S.ActiveTemplateInstantiations.empty()) return; // TODO: Investigate using GetExprRange() to get tighter bounds // on the bit ranges. QualType OtherT = Other->getType(); if (const auto *AT = OtherT->getAs
()) OtherT = AT->getValueType(); IntRange OtherRange = IntRange::forValueOfType(S.Context, OtherT); unsigned OtherWidth = OtherRange.Width; bool OtherIsBooleanType = Other->isKnownToHaveBooleanValue(); // 0 values are handled later by CheckTrivialUnsignedComparison(). if ((Value == 0) && (!OtherIsBooleanType)) return; BinaryOperatorKind op = E->getOpcode(); bool IsTrue = true; // Used for diagnostic printout. enum { LiteralConstant = 0, CXXBoolLiteralTrue, CXXBoolLiteralFalse } LiteralOrBoolConstant = LiteralConstant; if (!OtherIsBooleanType) { QualType ConstantT = Constant->getType(); QualType CommonT = E->getLHS()->getType(); if (S.Context.hasSameUnqualifiedType(OtherT, ConstantT)) return; assert((OtherT->isIntegerType() && ConstantT->isIntegerType()) && "comparison with non-integer type"); bool ConstantSigned = ConstantT->isSignedIntegerType(); bool CommonSigned = CommonT->isSignedIntegerType(); bool EqualityOnly = false; if (CommonSigned) { // The common type is signed, therefore no signed to unsigned conversion. if (!OtherRange.NonNegative) { // Check that the constant is representable in type OtherT. if (ConstantSigned) { if (OtherWidth >= Value.getMinSignedBits()) return; } else { // !ConstantSigned if (OtherWidth >= Value.getActiveBits() + 1) return; } } else { // !OtherSigned // Check that the constant is representable in type OtherT. // Negative values are out of range. if (ConstantSigned) { if (Value.isNonNegative() && OtherWidth >= Value.getActiveBits()) return; } else { // !ConstantSigned if (OtherWidth >= Value.getActiveBits()) return; } } } else { // !CommonSigned if (OtherRange.NonNegative) { if (OtherWidth >= Value.getActiveBits()) return; } else { // OtherSigned assert(!ConstantSigned && "Two signed types converted to unsigned types."); // Check to see if the constant is representable in OtherT. if (OtherWidth > Value.getActiveBits()) return; // Check to see if the constant is equivalent to a negative value // cast to CommonT. if (S.Context.getIntWidth(ConstantT) == S.Context.getIntWidth(CommonT) && Value.isNegative() && Value.getMinSignedBits() <= OtherWidth) return; // The constant value rests between values that OtherT can represent // after conversion. Relational comparison still works, but equality // comparisons will be tautological. EqualityOnly = true; } } bool PositiveConstant = !ConstantSigned || Value.isNonNegative(); if (op == BO_EQ || op == BO_NE) { IsTrue = op == BO_NE; } else if (EqualityOnly) { return; } else if (RhsConstant) { if (op == BO_GT || op == BO_GE) IsTrue = !PositiveConstant; else // op == BO_LT || op == BO_LE IsTrue = PositiveConstant; } else { if (op == BO_LT || op == BO_LE) IsTrue = !PositiveConstant; else // op == BO_GT || op == BO_GE IsTrue = PositiveConstant; } } else { // Other isKnownToHaveBooleanValue enum CompareBoolWithConstantResult { AFals, ATrue, Unkwn }; enum ConstantValue { LT_Zero, Zero, One, GT_One, SizeOfConstVal }; enum ConstantSide { Lhs, Rhs, SizeOfConstSides }; static const struct LinkedConditions { CompareBoolWithConstantResult BO_LT_OP[SizeOfConstSides][SizeOfConstVal]; CompareBoolWithConstantResult BO_GT_OP[SizeOfConstSides][SizeOfConstVal]; CompareBoolWithConstantResult BO_LE_OP[SizeOfConstSides][SizeOfConstVal]; CompareBoolWithConstantResult BO_GE_OP[SizeOfConstSides][SizeOfConstVal]; CompareBoolWithConstantResult BO_EQ_OP[SizeOfConstSides][SizeOfConstVal]; CompareBoolWithConstantResult BO_NE_OP[SizeOfConstSides][SizeOfConstVal]; } TruthTable = { // Constant on LHS. | Constant on RHS. | // LT_Zero| Zero | One |GT_One| LT_Zero| Zero | One |GT_One| { { ATrue, Unkwn, AFals, AFals }, { AFals, AFals, Unkwn, ATrue } }, { { AFals, AFals, Unkwn, ATrue }, { ATrue, Unkwn, AFals, AFals } }, { { ATrue, ATrue, Unkwn, AFals }, { AFals, Unkwn, ATrue, ATrue } }, { { AFals, Unkwn, ATrue, ATrue }, { ATrue, ATrue, Unkwn, AFals } }, { { AFals, Unkwn, Unkwn, AFals }, { AFals, Unkwn, Unkwn, AFals } }, { { ATrue, Unkwn, Unkwn, ATrue }, { ATrue, Unkwn, Unkwn, ATrue } } }; bool ConstantIsBoolLiteral = isa
(Constant); enum ConstantValue ConstVal = Zero; if (Value.isUnsigned() || Value.isNonNegative()) { if (Value == 0) { LiteralOrBoolConstant = ConstantIsBoolLiteral ? CXXBoolLiteralFalse : LiteralConstant; ConstVal = Zero; } else if (Value == 1) { LiteralOrBoolConstant = ConstantIsBoolLiteral ? CXXBoolLiteralTrue : LiteralConstant; ConstVal = One; } else { LiteralOrBoolConstant = LiteralConstant; ConstVal = GT_One; } } else { ConstVal = LT_Zero; } CompareBoolWithConstantResult CmpRes; switch (op) { case BO_LT: CmpRes = TruthTable.BO_LT_OP[RhsConstant][ConstVal]; break; case BO_GT: CmpRes = TruthTable.BO_GT_OP[RhsConstant][ConstVal]; break; case BO_LE: CmpRes = TruthTable.BO_LE_OP[RhsConstant][ConstVal]; break; case BO_GE: CmpRes = TruthTable.BO_GE_OP[RhsConstant][ConstVal]; break; case BO_EQ: CmpRes = TruthTable.BO_EQ_OP[RhsConstant][ConstVal]; break; case BO_NE: CmpRes = TruthTable.BO_NE_OP[RhsConstant][ConstVal]; break; default: CmpRes = Unkwn; break; } if (CmpRes == AFals) { IsTrue = false; } else if (CmpRes == ATrue) { IsTrue = true; } else { return; } } // If this is a comparison to an enum constant, include that // constant in the diagnostic. const EnumConstantDecl *ED = nullptr; if (const DeclRefExpr *DR = dyn_cast
(Constant)) ED = dyn_cast
(DR->getDecl()); SmallString<64> PrettySourceValue; llvm::raw_svector_ostream OS(PrettySourceValue); if (ED) OS << '\'' << *ED << "' (" << Value << ")"; else OS << Value; S.DiagRuntimeBehavior( E->getOperatorLoc(), E, S.PDiag(diag::warn_out_of_range_compare) << OS.str() << LiteralOrBoolConstant << OtherT << (OtherIsBooleanType && !OtherT->isBooleanType()) << IsTrue << E->getLHS()->getSourceRange() << E->getRHS()->getSourceRange()); } /// Analyze the operands of the given comparison. Implements the /// fallback case from AnalyzeComparison. void AnalyzeImpConvsInComparison(Sema &S, BinaryOperator *E) { AnalyzeImplicitConversions(S, E->getLHS(), E->getOperatorLoc()); AnalyzeImplicitConversions(S, E->getRHS(), E->getOperatorLoc()); } /// \brief Implements -Wsign-compare. /// /// \param E the binary operator to check for warnings void AnalyzeComparison(Sema &S, BinaryOperator *E) { // The type the comparison is being performed in. QualType T = E->getLHS()->getType(); // Only analyze comparison operators where both sides have been converted to // the same type. if (!S.Context.hasSameUnqualifiedType(T, E->getRHS()->getType())) return AnalyzeImpConvsInComparison(S, E); // Don't analyze value-dependent comparisons directly. if (E->isValueDependent()) return AnalyzeImpConvsInComparison(S, E); Expr *LHS = E->getLHS()->IgnoreParenImpCasts(); Expr *RHS = E->getRHS()->IgnoreParenImpCasts(); bool IsComparisonConstant = false; // Check whether an integer constant comparison results in a value // of 'true' or 'false'. if (T->isIntegralType(S.Context)) { llvm::APSInt RHSValue; bool IsRHSIntegralLiteral = RHS->isIntegerConstantExpr(RHSValue, S.Context); llvm::APSInt LHSValue; bool IsLHSIntegralLiteral = LHS->isIntegerConstantExpr(LHSValue, S.Context); if (IsRHSIntegralLiteral && !IsLHSIntegralLiteral) DiagnoseOutOfRangeComparison(S, E, RHS, LHS, RHSValue, true); else if (!IsRHSIntegralLiteral && IsLHSIntegralLiteral) DiagnoseOutOfRangeComparison(S, E, LHS, RHS, LHSValue, false); else IsComparisonConstant = (IsRHSIntegralLiteral && IsLHSIntegralLiteral); } else if (!T->hasUnsignedIntegerRepresentation()) IsComparisonConstant = E->isIntegerConstantExpr(S.Context); // We don't do anything special if this isn't an unsigned integral // comparison: we're only interested in integral comparisons, and // signed comparisons only happen in cases we don't care to warn about. // // We also don't care about value-dependent expressions or expressions // whose result is a constant. if (!T->hasUnsignedIntegerRepresentation() || IsComparisonConstant) return AnalyzeImpConvsInComparison(S, E); // Check to see if one of the (unmodified) operands is of different // signedness. Expr *signedOperand, *unsignedOperand; if (LHS->getType()->hasSignedIntegerRepresentation()) { assert(!RHS->getType()->hasSignedIntegerRepresentation() && "unsigned comparison between two signed integer expressions?"); signedOperand = LHS; unsignedOperand = RHS; } else if (RHS->getType()->hasSignedIntegerRepresentation()) { signedOperand = RHS; unsignedOperand = LHS; } else { CheckTrivialUnsignedComparison(S, E); return AnalyzeImpConvsInComparison(S, E); } // Otherwise, calculate the effective range of the signed operand. IntRange signedRange = GetExprRange(S.Context, signedOperand); // Go ahead and analyze implicit conversions in the operands. Note // that we skip the implicit conversions on both sides. AnalyzeImplicitConversions(S, LHS, E->getOperatorLoc()); AnalyzeImplicitConversions(S, RHS, E->getOperatorLoc()); // If the signed range is non-negative, -Wsign-compare won't fire, // but we should still check for comparisons which are always true // or false. if (signedRange.NonNegative) return CheckTrivialUnsignedComparison(S, E); // For (in)equality comparisons, if the unsigned operand is a // constant which cannot collide with a overflowed signed operand, // then reinterpreting the signed operand as unsigned will not // change the result of the comparison. if (E->isEqualityOp()) { unsigned comparisonWidth = S.Context.getIntWidth(T); IntRange unsignedRange = GetExprRange(S.Context, unsignedOperand); // We should never be unable to prove that the unsigned operand is // non-negative. assert(unsignedRange.NonNegative && "unsigned range includes negative?"); if (unsignedRange.Width < comparisonWidth) return; } S.DiagRuntimeBehavior(E->getOperatorLoc(), E, S.PDiag(diag::warn_mixed_sign_comparison) << LHS->getType() << RHS->getType() << LHS->getSourceRange() << RHS->getSourceRange()); } /// Analyzes an attempt to assign the given value to a bitfield. /// /// Returns true if there was something fishy about the attempt. bool AnalyzeBitFieldAssignment(Sema &S, FieldDecl *Bitfield, Expr *Init, SourceLocation InitLoc) { assert(Bitfield->isBitField()); if (Bitfield->isInvalidDecl()) return false; // White-list bool bitfields. if (Bitfield->getType()->isBooleanType()) return false; // Ignore value- or type-dependent expressions. if (Bitfield->getBitWidth()->isValueDependent() || Bitfield->getBitWidth()->isTypeDependent() || Init->isValueDependent() || Init->isTypeDependent()) return false; Expr *OriginalInit = Init->IgnoreParenImpCasts(); llvm::APSInt Value; if (!OriginalInit->EvaluateAsInt(Value, S.Context, Expr::SE_AllowSideEffects)) return false; unsigned OriginalWidth = Value.getBitWidth(); unsigned FieldWidth = Bitfield->getBitWidthValue(S.Context); if (OriginalWidth <= FieldWidth) return false; // Compute the value which the bitfield will contain. llvm::APSInt TruncatedValue = Value.trunc(FieldWidth); TruncatedValue.setIsSigned(Bitfield->getType()->isSignedIntegerType()); // Check whether the stored value is equal to the original value. TruncatedValue = TruncatedValue.extend(OriginalWidth); if (llvm::APSInt::isSameValue(Value, TruncatedValue)) return false; // Special-case bitfields of width 1: booleans are naturally 0/1, and // therefore don't strictly fit into a signed bitfield of width 1. if (FieldWidth == 1 && Value == 1) return false; std::string PrettyValue = Value.toString(10); std::string PrettyTrunc = TruncatedValue.toString(10); S.Diag(InitLoc, diag::warn_impcast_bitfield_precision_constant) << PrettyValue << PrettyTrunc << OriginalInit->getType() << Init->getSourceRange(); return true; } /// Analyze the given simple or compound assignment for warning-worthy /// operations. void AnalyzeAssignment(Sema &S, BinaryOperator *E) { // Just recurse on the LHS. AnalyzeImplicitConversions(S, E->getLHS(), E->getOperatorLoc()); // We want to recurse on the RHS as normal unless we're assigning to // a bitfield. if (FieldDecl *Bitfield = E->getLHS()->getSourceBitField()) { if (AnalyzeBitFieldAssignment(S, Bitfield, E->getRHS(), E->getOperatorLoc())) { // Recurse, ignoring any implicit conversions on the RHS. return AnalyzeImplicitConversions(S, E->getRHS()->IgnoreParenImpCasts(), E->getOperatorLoc()); } } AnalyzeImplicitConversions(S, E->getRHS(), E->getOperatorLoc()); } /// Diagnose an implicit cast; purely a helper for CheckImplicitConversion. void DiagnoseImpCast(Sema &S, Expr *E, QualType SourceType, QualType T, SourceLocation CContext, unsigned diag, bool pruneControlFlow = false) { if (pruneControlFlow) { S.DiagRuntimeBehavior(E->getExprLoc(), E, S.PDiag(diag) << SourceType << T << E->getSourceRange() << SourceRange(CContext)); return; } S.Diag(E->getExprLoc(), diag) << SourceType << T << E->getSourceRange() << SourceRange(CContext); } /// Diagnose an implicit cast; purely a helper for CheckImplicitConversion. void DiagnoseImpCast(Sema &S, Expr *E, QualType T, SourceLocation CContext, unsigned diag, bool pruneControlFlow = false) { DiagnoseImpCast(S, E, E->getType(), T, CContext, diag, pruneControlFlow); } /// Diagnose an implicit cast from a floating point value to an integer value. void DiagnoseFloatingImpCast(Sema &S, Expr *E, QualType T, SourceLocation CContext) { const bool IsBool = T->isSpecificBuiltinType(BuiltinType::Bool); const bool PruneWarnings = !S.ActiveTemplateInstantiations.empty(); Expr *InnerE = E->IgnoreParenImpCasts(); // We also want to warn on, e.g., "int i = -1.234" if (UnaryOperator *UOp = dyn_cast
(InnerE)) if (UOp->getOpcode() == UO_Minus || UOp->getOpcode() == UO_Plus) InnerE = UOp->getSubExpr()->IgnoreParenImpCasts(); const bool IsLiteral = isa
(E) || isa
(InnerE); llvm::APFloat Value(0.0); bool IsConstant = E->EvaluateAsFloat(Value, S.Context, Expr::SE_AllowSideEffects); if (!IsConstant) { return DiagnoseImpCast(S, E, T, CContext, diag::warn_impcast_float_integer, PruneWarnings); } bool isExact = false; llvm::APSInt IntegerValue(S.Context.getIntWidth(T), T->hasUnsignedIntegerRepresentation()); if (Value.convertToInteger(IntegerValue, llvm::APFloat::rmTowardZero, &isExact) == llvm::APFloat::opOK && isExact) { if (IsLiteral) return; return DiagnoseImpCast(S, E, T, CContext, diag::warn_impcast_float_integer, PruneWarnings); } unsigned DiagID = 0; if (IsLiteral) { // Warn on floating point literal to integer. DiagID = diag::warn_impcast_literal_float_to_integer; } else if (IntegerValue == 0) { if (Value.isZero()) { // Skip -0.0 to 0 conversion. return DiagnoseImpCast(S, E, T, CContext, diag::warn_impcast_float_integer, PruneWarnings); } // Warn on non-zero to zero conversion. DiagID = diag::warn_impcast_float_to_integer_zero; } else { if (IntegerValue.isUnsigned()) { if (!IntegerValue.isMaxValue()) { return DiagnoseImpCast(S, E, T, CContext, diag::warn_impcast_float_integer, PruneWarnings); } } else { // IntegerValue.isSigned() if (!IntegerValue.isMaxSignedValue() && !IntegerValue.isMinSignedValue()) { return DiagnoseImpCast(S, E, T, CContext, diag::warn_impcast_float_integer, PruneWarnings); } } // Warn on evaluatable floating point expression to integer conversion. DiagID = diag::warn_impcast_float_to_integer; } // FIXME: Force the precision of the source value down so we don't print // digits which are usually useless (we don't really care here if we // truncate a digit by accident in edge cases). Ideally, APFloat::toString // would automatically print the shortest representation, but it's a bit // tricky to implement. SmallString<16> PrettySourceValue; unsigned precision = llvm::APFloat::semanticsPrecision(Value.getSemantics()); precision = (precision * 59 + 195) / 196; Value.toString(PrettySourceValue, precision); SmallString<16> PrettyTargetValue; if (IsBool) PrettyTargetValue = Value.isZero() ? "false" : "true"; else IntegerValue.toString(PrettyTargetValue); if (PruneWarnings) { S.DiagRuntimeBehavior(E->getExprLoc(), E, S.PDiag(DiagID) << E->getType() << T.getUnqualifiedType() << PrettySourceValue << PrettyTargetValue << E->getSourceRange() << SourceRange(CContext)); } else { S.Diag(E->getExprLoc(), DiagID) << E->getType() << T.getUnqualifiedType() << PrettySourceValue << PrettyTargetValue << E->getSourceRange() << SourceRange(CContext); } } std::string PrettyPrintInRange(const llvm::APSInt &Value, IntRange Range) { if (!Range.Width) return "0"; llvm::APSInt ValueInRange = Value; ValueInRange.setIsSigned(!Range.NonNegative); ValueInRange = ValueInRange.trunc(Range.Width); return ValueInRange.toString(10); } bool IsImplicitBoolFloatConversion(Sema &S, Expr *Ex, bool ToBool) { if (!isa