From de3f60f0d6eca53160ffe68e6c8fc45d21a5cdae Mon Sep 17 00:00:00 2001 From: Devin Coughlin Date: Fri, 28 Aug 2015 22:26:05 +0000 Subject: [PATCH] [analyzer] When memcpy'ing into a fixed-size array, do not invalidate entire region. Change the analyzer's modeling of memcpy to be more precise when copying into fixed-size array fields. With this change, instead of invalidating the entire containing region the analyzer now invalidates only offsets for the array itself when it can show that the memcpy stays within the bounds of the array. This addresses false positive memory leak warnings of the kind reported by krzysztof in https://llvm.org/bugs/show_bug.cgi?id=22954 A patch by Pierre Gousseau! Differential Revision: http://reviews.llvm.org/D11832 git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@246345 91177308-0d34-0410-b5e6-96231b3b80d8 --- .../Core/PathSensitive/MemRegion.h | 4 +- .../Checkers/CStringChecker.cpp | 92 ++- lib/StaticAnalyzer/Core/RegionStore.cpp | 83 ++- test/Analysis/pr22954.c | 697 ++++++++++++++++++ 4 files changed, 866 insertions(+), 10 deletions(-) create mode 100644 test/Analysis/pr22954.c diff --git a/include/clang/StaticAnalyzer/Core/PathSensitive/MemRegion.h b/include/clang/StaticAnalyzer/Core/PathSensitive/MemRegion.h index 4f0712968b..d11efc44cc 100644 --- a/include/clang/StaticAnalyzer/Core/PathSensitive/MemRegion.h +++ b/include/clang/StaticAnalyzer/Core/PathSensitive/MemRegion.h @@ -1333,7 +1333,9 @@ public: /// Tells that a region's contents is not changed. TK_PreserveContents = 0x1, /// Suppress pointer-escaping of a region. - TK_SuppressEscape = 0x2 + TK_SuppressEscape = 0x2, + // Do not invalidate super region. + TK_DoNotInvalidateSuperRegion = 0x4 // Do not forget to extend StorageTypeForKinds if number of traits exceed // the number of bits StorageTypeForKinds can store. diff --git a/lib/StaticAnalyzer/Checkers/CStringChecker.cpp b/lib/StaticAnalyzer/Checkers/CStringChecker.cpp index 54b12410aa..da060414cd 100644 --- a/lib/StaticAnalyzer/Checkers/CStringChecker.cpp +++ b/lib/StaticAnalyzer/Checkers/CStringChecker.cpp @@ -145,7 +145,8 @@ public: static ProgramStateRef InvalidateBuffer(CheckerContext &C, ProgramStateRef state, const Expr *Ex, SVal V, - bool IsSourceBuffer); + bool IsSourceBuffer, + const Expr *Size); static bool SummarizeRegion(raw_ostream &os, ASTContext &Ctx, const MemRegion *MR); @@ -193,6 +194,14 @@ public: ProgramStateRef state, NonLoc left, NonLoc right) const; + + // Return true if destination buffer of copy function is in bound. + // Expects SVal of Size to be positive and unsigned. + // Expects SVal of FirstBuf to be a FieldRegion. + static bool IsFirstBufInBound(CheckerContext &C, + ProgramStateRef state, + const Expr *FirstBuf, + const Expr *Size); }; } //end anonymous namespace @@ -814,10 +823,68 @@ const StringLiteral *CStringChecker::getCStringLiteral(CheckerContext &C, return strRegion->getStringLiteral(); } +bool CStringChecker::IsFirstBufInBound(CheckerContext &C, + ProgramStateRef state, + const Expr *FirstBuf, + const Expr *Size) { + + // Originally copied from CheckBufferAccess and CheckLocation. + SValBuilder &svalBuilder = C.getSValBuilder(); + ASTContext &Ctx = svalBuilder.getContext(); + const LocationContext *LCtx = C.getLocationContext(); + + QualType sizeTy = Size->getType(); + QualType PtrTy = Ctx.getPointerType(Ctx.CharTy); + SVal BufVal = state->getSVal(FirstBuf, LCtx); + + SVal LengthVal = state->getSVal(Size, LCtx); + // Cast is safe as the size argument to copy functions are of integral type. + NonLoc Length = LengthVal.castAs(); + + // Compute the offset of the last element to be accessed: size-1. + NonLoc One = svalBuilder.makeIntVal(1, sizeTy).castAs(); + NonLoc LastOffset = + svalBuilder.evalBinOpNN(state, BO_Sub, Length, One, sizeTy) + .castAs(); + + // Check that the first buffer is sufficiently long. + SVal BufStart = svalBuilder.evalCast(BufVal, PtrTy, FirstBuf->getType()); + // Cast is safe as caller checks BufVal is a MemRegionVal. + Loc BufLoc = BufStart.castAs(); + + SVal BufEnd = + svalBuilder.evalBinOpLN(state, BO_Add, BufLoc, LastOffset, PtrTy); + + // Check for out of bound array element access. + const MemRegion *R = BufEnd.getAsRegion(); + // BufStart is a MemRegionVal so BufEnd should be one too. + assert(R && "BufEnd should be a MemRegion"); + + // Cast is safe as BufVal's region is a FieldRegion. + const ElementRegion *ER = cast(R); + + assert(ER->getValueType() == C.getASTContext().CharTy && + "IsFirstBufInBound should only be called with char* ElementRegions"); + + // Get the size of the array. + const SubRegion *superReg = cast(ER->getSuperRegion()); + SVal Extent = + svalBuilder.convertToArrayIndex(superReg->getExtent(svalBuilder)); + DefinedOrUnknownSVal ExtentSize = Extent.castAs(); + + // Get the index of the accessed element. + DefinedOrUnknownSVal Idx = ER->getIndex().castAs(); + + ProgramStateRef StInBound = state->assumeInBound(Idx, ExtentSize, true); + + return static_cast(StInBound); +} + ProgramStateRef CStringChecker::InvalidateBuffer(CheckerContext &C, ProgramStateRef state, const Expr *E, SVal V, - bool IsSourceBuffer) { + bool IsSourceBuffer, + const Expr *Size) { Optional L = V.getAs(); if (!L) return state; @@ -847,6 +914,16 @@ ProgramStateRef CStringChecker::InvalidateBuffer(CheckerContext &C, RegionAndSymbolInvalidationTraits::TK_PreserveContents); ITraits.setTrait(R, RegionAndSymbolInvalidationTraits::TK_SuppressEscape); CausesPointerEscape = true; + } else { + const MemRegion::Kind& K = R->getKind(); + if (K == MemRegion::FieldRegionKind) + if (Size && IsFirstBufInBound(C, state, E, Size)) { + // If destination buffer is a field region and access is in bound, + // do not invalidate its super region. + ITraits.setTrait( + R, + RegionAndSymbolInvalidationTraits::TK_DoNotInvalidateSuperRegion); + } } return state->invalidateRegions(R, E, C.blockCount(), LCtx, @@ -1000,12 +1077,12 @@ void CStringChecker::evalCopyCommon(CheckerContext &C, // This would probably remove any existing bindings past the end of the // copied region, but that's still an improvement over blank invalidation. state = InvalidateBuffer(C, state, Dest, C.getSVal(Dest), - /*IsSourceBuffer*/false); + /*IsSourceBuffer*/false, Size); // Invalidate the source (const-invalidation without const-pointer-escaping // the address of the top-level region). state = InvalidateBuffer(C, state, Source, C.getSVal(Source), - /*IsSourceBuffer*/true); + /*IsSourceBuffer*/true, nullptr); C.addTransition(state); } @@ -1620,11 +1697,12 @@ void CStringChecker::evalStrcpyCommon(CheckerContext &C, const CallExpr *CE, // This would probably remove any existing bindings past the end of the // string, but that's still an improvement over blank invalidation. state = InvalidateBuffer(C, state, Dst, *dstRegVal, - /*IsSourceBuffer*/false); + /*IsSourceBuffer*/false, nullptr); // Invalidate the source (const-invalidation without const-pointer-escaping // the address of the top-level region). - state = InvalidateBuffer(C, state, srcExpr, srcVal, /*IsSourceBuffer*/true); + state = InvalidateBuffer(C, state, srcExpr, srcVal, /*IsSourceBuffer*/true, + nullptr); // Set the C string length of the destination, if we know it. if (isBounded && !isAppending) { @@ -1848,7 +1926,7 @@ void CStringChecker::evalStrsep(CheckerContext &C, const CallExpr *CE) const { // Invalidate the search string, representing the change of one delimiter // character to NUL. State = InvalidateBuffer(C, State, SearchStrPtr, Result, - /*IsSourceBuffer*/false); + /*IsSourceBuffer*/false, nullptr); // Overwrite the search string pointer. The new value is either an address // further along in the same string, or NULL if there are no more tokens. diff --git a/lib/StaticAnalyzer/Core/RegionStore.cpp b/lib/StaticAnalyzer/Core/RegionStore.cpp index 4178324543..0d403bc190 100644 --- a/lib/StaticAnalyzer/Core/RegionStore.cpp +++ b/lib/StaticAnalyzer/Core/RegionStore.cpp @@ -710,8 +710,7 @@ public: } bool AddToWorkList(const MemRegion *R) { - const MemRegion *BaseR = R->getBaseRegion(); - return AddToWorkList(WorkListElement(BaseR), getCluster(BaseR)); + return static_cast(this)->AddToWorkList(R); } void RunWorkList() { @@ -956,9 +955,20 @@ public: void VisitCluster(const MemRegion *baseR, const ClusterBindings *C); void VisitBinding(SVal V); + + using ClusterAnalysis::AddToWorkList; + + bool AddToWorkList(const MemRegion *R); }; } +bool invalidateRegionsWorker::AddToWorkList(const MemRegion *R) { + bool doNotInvalidateSuperRegion = ITraits.hasTrait( + R, RegionAndSymbolInvalidationTraits::TK_DoNotInvalidateSuperRegion); + const MemRegion *BaseR = doNotInvalidateSuperRegion ? R : R->getBaseRegion(); + return AddToWorkList(WorkListElement(BaseR), getCluster(BaseR)); +} + void invalidateRegionsWorker::VisitBinding(SVal V) { // A symbol? Mark it touched by the invalidation. if (SymbolRef Sym = V.getAsSymbol()) @@ -1071,6 +1081,66 @@ void invalidateRegionsWorker::VisitCluster(const MemRegion *baseR, } if (const ArrayType *AT = Ctx.getAsArrayType(T)) { + bool doNotInvalidateSuperRegion = ITraits.hasTrait( + baseR, + RegionAndSymbolInvalidationTraits::TK_DoNotInvalidateSuperRegion); + + if (doNotInvalidateSuperRegion) { + // We are not doing blank invalidation of the whole array region so we + // have to manually invalidate each elements. + Optional NumElements; + + // Compute lower and upper offsets for region within array. + if (const ConstantArrayType *CAT = dyn_cast(AT)) + NumElements = CAT->getSize().getZExtValue(); + if (!NumElements) // We are not dealing with a constant size array + goto conjure_default; + QualType ElementTy = AT->getElementType(); + uint64_t ElemSize = Ctx.getTypeSize(ElementTy); + const RegionOffset &RO = baseR->getAsOffset(); + const MemRegion *SuperR = baseR->getBaseRegion(); + if (RO.hasSymbolicOffset()) { + // If base region has a symbolic offset, + // we revert to invalidating the super region. + if (SuperR) + AddToWorkList(SuperR); + goto conjure_default; + } + assert(RO.getOffset() >= 0 && "Offset should not be negative"); + uint64_t LowerOffset = RO.getOffset(); + uint64_t UpperOffset = LowerOffset + *NumElements * ElemSize; + + // Invalidate regions which are within array boundaries, + // or have a symbolic offset. + if (!SuperR) + goto conjure_default; + + const ClusterBindings *C = B.lookup(SuperR); + if (!C) + goto conjure_default; + + for (ClusterBindings::iterator I = C->begin(), E = C->end(); I != E; + ++I) { + const BindingKey &BK = I.getKey(); + Optional ROffset = + BK.hasSymbolicOffset() ? Optional() : BK.getOffset(); + // Check offset is not symbolic and within array's boundaries. + // Handles arrays of 0 elements and of 0-sized elements as well. + if (!ROffset || + (ROffset && + ((*ROffset >= LowerOffset && *ROffset < UpperOffset) || + (LowerOffset == UpperOffset && *ROffset == LowerOffset)))) { + B = B.removeBinding(I.getKey()); + // Bound symbolic regions need to be invalidated for dead symbol + // detection. + SVal V = I.getData(); + const MemRegion *R = V.getAsRegion(); + if (R && isa(R)) + VisitBinding(V); + } + } + } + conjure_default: // Set the default value of the array to conjured symbol. DefinedOrUnknownSVal V = svalBuilder.conjureSymbolVal(baseR, Ex, LCtx, @@ -2187,11 +2257,20 @@ public: void VisitCluster(const MemRegion *baseR, const ClusterBindings *C); using ClusterAnalysis::VisitCluster; + using ClusterAnalysis::AddToWorkList; + + bool AddToWorkList(const MemRegion *R); + bool UpdatePostponed(); void VisitBinding(SVal V); }; } +bool removeDeadBindingsWorker::AddToWorkList(const MemRegion *R) { + const MemRegion *BaseR = R->getBaseRegion(); + return AddToWorkList(WorkListElement(BaseR), getCluster(BaseR)); +} + void removeDeadBindingsWorker::VisitAddedToCluster(const MemRegion *baseR, const ClusterBindings &C) { diff --git a/test/Analysis/pr22954.c b/test/Analysis/pr22954.c new file mode 100644 index 0000000000..ae385bb862 --- /dev/null +++ b/test/Analysis/pr22954.c @@ -0,0 +1,697 @@ +// Given code 'struct aa { char s1[4]; char * s2;} a; memcpy(a.s1, ...);', +// this test checks that the CStringChecker only invalidates the destination buffer array a.s1 (instead of a.s1 and a.s2). +// At the moment the whole of the destination array content is invalidated. +// If a.s1 region has a symbolic offset, the whole region of 'a' is invalidated. +// Specific triple set to test structures of size 0. +// RUN: %clang_cc1 -triple x86_64-pc-linux-gnu -analyze -analyzer-checker=core,unix.Malloc,debug.ExprInspection -analyzer-store=region -verify %s + +typedef __typeof(sizeof(int)) size_t; + +char *strdup(const char *s); +void free(void *); +void *memcpy(void *dst, const void *src, size_t n); // expected-note{{passing argument to parameter 'dst' here}} +void *malloc(size_t n); + +void clang_analyzer_eval(int); + +struct aa { + char s1[4]; + char *s2; +}; + +// Test different types of structure initialisation. +int f0() { + struct aa a0 = {{1, 2, 3, 4}, 0}; + a0.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(a0.s1, input, 4); + clang_analyzer_eval(a0.s1[0] == 'a'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a0.s1[1] == 'b'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a0.s1[2] == 'c'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a0.s1[3] == 'd'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a0.s2 == 0); // expected-warning{{UNKNOWN}} + free(a0.s2); // no warning + return 0; +} + +int f1() { + struct aa a1; + a1.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(a1.s1, input, 4); + clang_analyzer_eval(a1.s1[0] == 'a'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a1.s1[1] == 'b'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a1.s1[2] == 'c'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a1.s1[3] == 'd'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a1.s2 == 0); // expected-warning{{UNKNOWN}} + free(a1.s2); // no warning + return 0; +} + +int f2() { + struct aa a2 = {{1, 2}}; + a2.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(a2.s1, input, 4); + clang_analyzer_eval(a2.s1[0] == 'a'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a2.s1[1] == 'b'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a2.s1[2] == 'c'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a2.s1[3] == 'd'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a2.s2 == 0); // expected-warning{{UNKNOWN}} + free(a2.s2); // no warning + return 0; +} + +int f3() { + struct aa a3 = {{1, 2, 3, 4}, 0}; + a3.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + int * dest = (int*)a3.s1; + memcpy(dest, input, 4); + clang_analyzer_eval(a3.s1[0] == 'a'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(dest[0] == 'a'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a3.s1[1] == 'b'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(dest[1] == 'b'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a3.s1[2] == 'c'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(dest[2] == 'c'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a3.s1[3] == 'd'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(dest[3] == 'd'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a3.s2 == 0); // expected-warning{{UNKNOWN}} + free(a3.s2); // no warning + return 0; +} + +struct bb { + struct aa a; + char * s2; +}; + +int f4() { + struct bb b0 = {{1, 2, 3, 4}, 0}; + b0.s2 = strdup("hello"); + b0.a.s2 = strdup("hola"); + char input[] = {'a', 'b', 'c', 'd'}; + char * dest = (char*)(b0.a.s1); + memcpy(dest, input, 4); + clang_analyzer_eval(b0.a.s1[0] == 'a'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(dest[0] == 'a'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(b0.a.s1[1] == 'b'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(dest[1] == 'b'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(b0.a.s1[2] == 'c'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(dest[2] == 'c'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(b0.a.s1[3] == 'd'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(dest[3] == 'd'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(b0.s2 == 0); // expected-warning{{UNKNOWN}} + free(b0.a.s2); // no warning + free(b0.s2); // no warning + return 0; +} + +// Test that memory leaks are caught. +int f5() { + struct aa a0 = {{1, 2, 3, 4}, 0}; + a0.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(a0.s1, input, 4); + return 0; // expected-warning{{Potential leak of memory pointed to by 'a0.s2'}} +} + +int f6() { + struct aa a1; + a1.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(a1.s1, input, 4); + return 0; // expected-warning{{Potential leak of memory pointed to by 'a1.s2'}} +} + +int f7() { + struct aa a2 = {{1, 2}}; + a2.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(a2.s1, input, 4); + return 0; // expected-warning{{Potential leak of memory pointed to by 'a2.s2'}} +} + +int f8() { + struct aa a3 = {{1, 2, 3, 4}, 0}; + a3.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + int * dest = (int*)a3.s1; + memcpy(dest, input, 4); + return 0; // expected-warning{{Potential leak of memory pointed to by 'a3.s2'}} +} + +int f9() { + struct bb b0 = {{1, 2, 3, 4}, 0}; + b0.s2 = strdup("hello"); + b0.a.s2 = strdup("hola"); + char input[] = {'a', 'b', 'c', 'd'}; + char * dest = (char*)(b0.a.s1); + memcpy(dest, input, 4); + free(b0.a.s2); // expected-warning{{Potential leak of memory pointed to by 'b0.s2'}} + return 0; +} + +int f10() { + struct bb b0 = {{1, 2, 3, 4}, 0}; + b0.s2 = strdup("hello"); + b0.a.s2 = strdup("hola"); + char input[] = {'a', 'b', 'c', 'd'}; + char * dest = (char*)(b0.a.s1); + memcpy(dest, input, 4); + free(b0.s2); // expected-warning{{Potential leak of memory pointed to by 'b0.a.s2'}} + return 0; +} + +// Test invalidating fields being addresses of array. +struct cc { + char * s1; + char * s2; +}; + +int f11() { + char x[4] = {1, 2}; + x[0] = 1; + x[1] = 2; + struct cc c0; + c0.s2 = strdup("hello"); + c0.s1 = &x[0]; + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(c0.s1, input, 4); + clang_analyzer_eval(x[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(x[1] == 2); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(c0.s1[0] == 'a'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(c0.s1[1] == 'b'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(c0.s1[2] == 'c'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(c0.s1[3] == 'd'); // expected-warning{{UNKNOWN}} + free(c0.s2); // no-warning + return 0; +} + +// Test inverting field position between s1 and s2. +struct dd { + char *s2; + char s1[4]; +}; + +int f12() { + struct dd d0 = {0, {1, 2, 3, 4}}; + d0.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(d0.s1, input, 4); + clang_analyzer_eval(d0.s1[0] == 'a'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(d0.s1[1] == 'b'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(d0.s1[2] == 'c'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(d0.s1[3] == 'd'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(d0.s2 == 0); // expected-warning{{UNKNOWN}} + free(d0.s2); // no warning + return 0; +} + +// Test arrays of structs. +struct ee { + int a; + char b; +}; + +struct EE { + struct ee s1[2]; + char * s2; +}; + +int f13() { + struct EE E0 = {{{1, 2}, {3, 4}}, 0}; + E0.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(E0.s1, input, 4); + clang_analyzer_eval(E0.s1[0].a == 'a'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(E0.s1[0].b == 'b'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(E0.s1[1].a == 'c'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(E0.s1[1].b == 'd'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(E0.s2 == 0); // expected-warning{{UNKNOWN}} + free(E0.s2); // no warning + return 0; +} + +// Test global parameters. +struct aa a15 = {{1, 2, 3, 4}, 0}; + +int f15() { + a15.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(a15.s1, input, 4); + clang_analyzer_eval(a15.s1[0] == 'a'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a15.s1[1] == 'b'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a15.s1[2] == 'c'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a15.s1[3] == 'd'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a15.s2 == 0); // expected-warning{{UNKNOWN}} + free(a15.s2); // no warning + return 0; +} + +// Test array of 0 sized elements. +struct empty {}; +struct gg { + struct empty s1[4]; + char * s2; +}; + +int f16() { + struct gg g0 = {{}, 0}; + g0.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(g0.s1, input, 4); + clang_analyzer_eval(*(int*)(&g0.s1[0]) == 'a'); // expected-warning{{UNKNOWN}}\ + expected-warning{{Potential leak of memory pointed to by 'g0.s2'}} + clang_analyzer_eval(*(int*)(&g0.s1[1]) == 'b'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(*(int*)(&g0.s1[2]) == 'c'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(*(int*)(&g0.s1[3]) == 'd'); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(g0.s2 == 0); // expected-warning{{UNKNOWN}} + free(g0.s2); // no warning + return 0; +} + +// Test array of 0 elements. +struct hh { + char s1[0]; + char * s2; +}; + +int f17() { + struct hh h0; + h0.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(h0.s1, input, 4); + clang_analyzer_eval(h0.s1[0] == 'a'); // expected-warning{{UNKNOWN}}\ + expected-warning{{Potential leak of memory pointed to by 'h0.s2'}} + clang_analyzer_eval(h0.s2 == 0); // expected-warning{{UNKNOWN}} + free(h0.s2); // no warning + return 0; +} + +// Test writing past the array. +struct ii { + char s1[4]; + int i; + int j; + char * s2; +}; + +int f18() { + struct ii i18 = {{1, 2, 3, 4}, 5, 6}; + i18.i = 10; + i18.j = 11; + i18.s2 = strdup("hello"); + char input[100] = {3}; + memcpy(i18.s1, input, 100); + clang_analyzer_eval(i18.s1[0] == 1); // expected-warning{{UNKNOWN}}\ + expected-warning{{Potential leak of memory pointed to by 'i18.s2'}} + clang_analyzer_eval(i18.s1[1] == 2); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(i18.s1[2] == 3); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(i18.s1[3] == 4); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(i18.i == 10); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(i18.j == 11); // expected-warning{{UNKNOWN}} + return 0; +} + +int f181() { + struct ii i181 = {{1, 2, 3, 4}, 5, 6}; + i181.i = 10; + i181.j = 11; + i181.s2 = strdup("hello"); + char input[100] = {3}; + memcpy(i181.s1, input, 5); // invalidate the whole region of i181 + clang_analyzer_eval(i181.s1[0] == 1); // expected-warning{{UNKNOWN}}\ + expected-warning{{Potential leak of memory pointed to by 'i181.s2'}} + clang_analyzer_eval(i181.s1[1] == 2); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(i181.s1[2] == 3); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(i181.s1[3] == 4); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(i181.i == 10); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(i181.j == 11); // expected-warning{{UNKNOWN}} + return 0; +} + +// Test array with a symbolic offset. +struct jj { + char s1[2]; + char * s2; +}; + +struct JJ { + struct jj s1[3]; + char * s2; +}; + +int f19(int i) { + struct JJ J0 = {{{1, 2, 0}, {3, 4, 0}, {5, 6, 0}}, 0}; + J0.s2 = strdup("hello"); + J0.s1[0].s2 = strdup("hello"); + J0.s1[1].s2 = strdup("hi"); + J0.s1[2].s2 = strdup("world"); + char input[2] = {'a', 'b'}; + memcpy(J0.s1[i].s1, input, 2); + clang_analyzer_eval(J0.s1[0].s1[0] == 1); // expected-warning{{UNKNOWN}}\ + expected-warning{{Potential leak of memory pointed to by field 's2'}}\ + expected-warning{{Potential leak of memory pointed to by 'J0.s2'}} + clang_analyzer_eval(J0.s1[0].s1[1] == 2); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(J0.s1[1].s1[0] == 3); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(J0.s1[1].s1[1] == 4); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(J0.s1[2].s1[0] == 5); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(J0.s1[2].s1[1] == 6); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(J0.s1[i].s1[0] == 5); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(J0.s1[i].s1[1] == 6); // expected-warning{{UNKNOWN}} + // FIXME: memory leak warning for J0.s2 should be emitted here instead of after memcpy call. + return 0; // no warning +} + +// Test array with its super region having symbolic offseted regions. +int f20(int i) { + struct aa * a20 = malloc(sizeof(struct aa) * 2); + a20[0].s1[0] = 1; + a20[0].s1[1] = 2; + a20[0].s1[2] = 3; + a20[0].s1[3] = 4; + a20[0].s2 = strdup("hello"); + a20[1].s1[0] = 5; + a20[1].s1[1] = 6; + a20[1].s1[2] = 7; + a20[1].s1[3] = 8; + a20[1].s2 = strdup("world"); + a20[i].s2 = strdup("hola"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(a20[0].s1, input, 4); + clang_analyzer_eval(a20[0].s1[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a20[0].s1[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a20[0].s1[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a20[0].s1[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a20[0].s2 == 0); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a20[1].s1[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a20[1].s1[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a20[1].s1[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a20[1].s1[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a20[1].s2 == 0); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a20[i].s1[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a20[i].s1[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a20[i].s1[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a20[i].s1[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a20[i].s2 == 0); // expected-warning{{UNKNOWN}}\ + expected-warning{{Potential leak of memory pointed to by 'a20'}} + + return 0; +} + +// Test array's region and super region both having symbolic offsets. +int f21(int i) { + struct aa * a21 = malloc(sizeof(struct aa) * 2); + a21[0].s1[0] = 1; + a21[0].s1[1] = 2; + a21[0].s1[2] = 3; + a21[0].s1[3] = 4; + a21[0].s2 = 0; + a21[1].s1[0] = 5; + a21[1].s1[1] = 6; + a21[1].s1[2] = 7; + a21[1].s1[3] = 8; + a21[1].s2 = 0; + a21[i].s2 = strdup("hello"); + a21[i].s1[0] = 1; + a21[i].s1[1] = 2; + a21[i].s1[2] = 3; + a21[i].s1[3] = 4; + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(a21[i].s1, input, 4); + clang_analyzer_eval(a21[0].s1[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a21[0].s1[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a21[0].s1[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a21[0].s1[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a21[0].s2 == 0); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a21[1].s1[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a21[1].s1[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a21[1].s1[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a21[1].s1[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a21[1].s2 == 0); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a21[i].s1[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a21[i].s1[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a21[i].s1[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a21[i].s1[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a21[i].s2 == 0); // expected-warning{{UNKNOWN}}\ + expected-warning{{Potential leak of memory pointed to by 'a21'}} + + return 0; +} + +// Test regions aliasing other regions. +struct ll { + char s1[4]; + char * s2; +}; + +struct mm { + char s3[4]; + char * s4; +}; + +int f24() { + struct ll l24 = {{1, 2, 3, 4}, 0}; + struct mm * m24 = (struct mm *)&l24; + m24->s4 = strdup("hello"); + char input[] = {1, 2, 3, 4}; + memcpy(m24->s3, input, 4); + clang_analyzer_eval(m24->s3[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m24->s3[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m24->s3[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m24->s3[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(l24.s1[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(l24.s1[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(l24.s1[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(l24.s1[3] == 1); // expected-warning{{UNKNOWN}}\ + expected-warning{{Potential leak of memory pointed to by field 's4'}} + return 0; +} + +// Test region with potential aliasing and symbolic offsets. +// Store assumes no aliasing. +int f25(int i, int j, struct ll * l, struct mm * m) { + m->s4 = strdup("hola"); // m->s4 not tracked + m->s3[0] = 1; + m->s3[1] = 2; + m->s3[2] = 3; + m->s3[3] = 4; + m->s3[j] = 5; // invalidates m->s3 + l->s2 = strdup("hello"); // l->s2 not tracked + l->s1[0] = 6; + l->s1[1] = 7; + l->s1[2] = 8; + l->s1[3] = 9; + l->s1[i] = 10; // invalidates l->s1 + char input[] = {1, 2, 3, 4}; + memcpy(m->s3, input, 4); // does not invalidate l->s1[i] + clang_analyzer_eval(m->s3[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m->s3[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m->s3[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m->s3[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m->s3[i] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m->s3[j] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(l->s1[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(l->s1[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(l->s1[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(l->s1[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(l->s1[i] == 1); // expected-warning{{FALSE}} + clang_analyzer_eval(l->s1[j] == 1); // expected-warning{{UNKNOWN}} + return 0; +} + +// Test size with symbolic size argument. +int f26(int i) { + struct aa a26 = {{1, 2, 3, 4}, 0}; + a26.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(a26.s1, input, i); // i assumed in bound + clang_analyzer_eval(a26.s1[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a26.s1[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a26.s1[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a26.s1[3] == 1); // expected-warning{{UNKNOWN}}\ + expected-warning{{Potential leak of memory pointed to by 'a26.s2'}} + return 0; +} + +// Test sizeof as a size argument. +int f261() { + struct aa a261 = {{1, 2, 3, 4}, 0}; + a261.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(a261.s1, input, sizeof(a261.s1)); + clang_analyzer_eval(a261.s1[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a261.s1[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a261.s1[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a261.s1[3] == 1); // expected-warning{{UNKNOWN}}\ + expected-warning{{Potential leak of memory pointed to by 'a261.s2'}} + return 0; +} + +// Test negative size argument. +int f262() { + struct aa a262 = {{1, 2, 3, 4}, 0}; + a262.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(a262.s1, input, -1); + clang_analyzer_eval(a262.s1[0] == 1); // expected-warning{{UNKNOWN}}\ + expected-warning{{Potential leak of memory pointed to by 'a262.s2'}} + clang_analyzer_eval(a262.s1[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a262.s1[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(a262.s1[3] == 1); // expected-warning{{UNKNOWN}} + return 0; +} + +// Test casting regions with symbolic offseted sub regions. +int f27(int i) { + struct mm m27 = {{1, 2, 3, 4}, 0}; + m27.s4 = strdup("hello"); + m27.s3[i] = 5; + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(((struct ll*)(&m27))->s1, input, 4); + clang_analyzer_eval(m27.s3[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m27.s3[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m27.s3[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m27.s3[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m27.s3[i] == 1); // expected-warning{{UNKNOWN}}\ + expected-warning{{Potential leak of memory pointed to by 'm27.s4'}} + return 0; +} + +int f28(int i, int j, int k, int l) { + struct mm m28[2]; + m28[i].s4 = strdup("hello"); + m28[j].s3[k] = 1; + struct ll * l28 = (struct ll*)(&m28[1]); + l28->s1[l] = 2; + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(l28->s1, input, 4); + clang_analyzer_eval(m28[0].s3[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m28[0].s3[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m28[0].s3[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m28[0].s3[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m28[1].s3[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m28[1].s3[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m28[1].s3[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m28[1].s3[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m28[i].s3[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m28[i].s3[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m28[i].s3[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m28[i].s3[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m28[j].s3[k] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(l28->s1[l] == 2); // expected-warning{{UNKNOWN}} + return 0; +} + +int f29(int i, int j, int k, int l, int m) { + struct mm m29[2]; + m29[i].s4 = strdup("hello"); + m29[j].s3[k] = 1; + struct ll * l29 = (struct ll*)(&m29[l]); + l29->s1[m] = 2; + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(l29->s1, input, 4); + clang_analyzer_eval(m29[0].s3[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m29[0].s3[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m29[0].s3[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m29[0].s3[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m29[1].s3[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m29[1].s3[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m29[1].s3[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m29[1].s3[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m29[i].s3[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m29[i].s3[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m29[i].s3[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m29[i].s3[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(m29[j].s3[k] == 1); // expected-warning{{TRUE}}\ + expected-warning{{Potential leak of memory pointed to by field 's4'}} + clang_analyzer_eval(l29->s1[m] == 2); // expected-warning{{UNKNOWN}} + return 0; +} + +// Test unions' fields. +union uu { + char x; + char s1[4]; +}; + +int f30() { + union uu u30 = { .s1 = {1, 2, 3, 4}}; + char input[] = {1, 2, 3, 4}; + memcpy(u30.s1, input, 4); + clang_analyzer_eval(u30.s1[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(u30.s1[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(u30.s1[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(u30.s1[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(u30.x == 1); // expected-warning{{UNKNOWN}} + return 0; +} + +struct kk { + union uu u; + char * s2; +}; + +int f31() { + struct kk k31; + k31.s2 = strdup("hello"); + k31.u.x = 1; + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(k31.u.s1, input, 4); + clang_analyzer_eval(k31.u.s1[0] == 1); // expected-warning{{UNKNOWN}}\ + expected-warning{{Potential leak of memory pointed to by 'k31.s2'}} + clang_analyzer_eval(k31.u.s1[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(k31.u.s1[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(k31.u.s1[3] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(k31.u.x == 1); // expected-warning{{UNKNOWN}} + // FIXME: memory leak warning for k31.s2 should be emitted here. + return 0; +} + +union vv { + int x; + char * s2; +}; + +int f32() { + union vv v32; + v32.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(v32.s2, input, 4); + clang_analyzer_eval(v32.s2[0] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(v32.s2[1] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(v32.s2[2] == 1); // expected-warning{{UNKNOWN}} + clang_analyzer_eval(v32.s2[3] == 1); // expected-warning{{UNKNOWN}}\ + expected-warning{{Potential leak of memory pointed to by 'v32.s2'}} + return 0; +} + +struct nn { + int s1; + int i; + int j; + int k; + char * s2; +}; + +// Test bad types to dest buffer. +int f33() { + struct nn n33 = {1, 2, 3, 4, 0}; + n33.s2 = strdup("hello"); + char input[] = {'a', 'b', 'c', 'd'}; + memcpy(n33.s1, input, 4); // expected-warning{{incompatible integer to pointer conversion passing 'int' to parameter of type 'void *'}} + clang_analyzer_eval(n33.i == 2); // expected-warning{{TRUE}} + clang_analyzer_eval(n33.j == 3); // expected-warning{{TRUE}} + clang_analyzer_eval(n33.k == 4); // expected-warning{{TRUE}} + clang_analyzer_eval(((char*)(n33.s1))[0] == 1); // expected-warning{{UNKNOWN}}\ + expected-warning{{cast to 'char *' from smaller integer type 'int'}} + clang_analyzer_eval(((char*)(n33.s1))[1] == 1); // expected-warning{{UNKNOWN}}\ + expected-warning{{cast to 'char *' from smaller integer type 'int'}} + clang_analyzer_eval(((char*)(n33.s1))[2] == 1); // expected-warning{{UNKNOWN}}\ + expected-warning{{cast to 'char *' from smaller integer type 'int'}} + clang_analyzer_eval(((char*)(n33.s1))[3] == 1); // expected-warning{{UNKNOWN}}\ + expected-warning{{cast to 'char *' from smaller integer type 'int'}} + clang_analyzer_eval(n33.s2 == 0); //expected-warning{{UNKNOWN}} + return 0; // expected-warning{{Potential leak of memory pointed to by 'n33.s2'}} +} -- 2.40.0