ArrayBoundCheckerV2.cpp revision 360784 
1//== ArrayBoundCheckerV2.cpp ------------------------------------*- C++ -*--==//
2//
3// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4// See https://llvm.org/LICENSE.txt for license information.
5// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6//
7//===----------------------------------------------------------------------===//
8//
9// This file defines ArrayBoundCheckerV2, which is a path-sensitive check
10// which looks for an out-of-bound array element access.
11//
12//===----------------------------------------------------------------------===//
13
14#include "Taint.h"
15#include "clang/StaticAnalyzer/Checkers/BuiltinCheckerRegistration.h"
16#include "clang/AST/CharUnits.h"
17#include "clang/StaticAnalyzer/Core/BugReporter/BugType.h"
18#include "clang/StaticAnalyzer/Core/Checker.h"
19#include "clang/StaticAnalyzer/Core/CheckerManager.h"
20#include "clang/StaticAnalyzer/Core/PathSensitive/APSIntType.h"
21#include "clang/StaticAnalyzer/Core/PathSensitive/CheckerContext.h"
22#include "clang/StaticAnalyzer/Core/PathSensitive/ExprEngine.h"
23#include "llvm/ADT/SmallString.h"
24#include "llvm/Support/raw_ostream.h"
25
26using namespace clang;
27using namespace ento;
28using namespace taint;
29
30namespace {
31class ArrayBoundCheckerV2 :
32    public Checker<check::Location> {
33  mutable std::unique_ptr<BuiltinBug> BT;
34
35  enum OOB_Kind { OOB_Precedes, OOB_Excedes, OOB_Tainted };
36
37  void reportOOB(CheckerContext &C, ProgramStateRef errorState, OOB_Kind kind,
38                 std::unique_ptr<BugReporterVisitor> Visitor = nullptr) const;
39
40public:
41  void checkLocation(SVal l, bool isLoad, const Stmt*S,
42                     CheckerContext &C) const;
43};
44
45// FIXME: Eventually replace RegionRawOffset with this class.
46class RegionRawOffsetV2 {
47private:
48  const SubRegion *baseRegion;
49  SVal byteOffset;
50
51  RegionRawOffsetV2()
52    : baseRegion(nullptr), byteOffset(UnknownVal()) {}
53
54public:
55  RegionRawOffsetV2(const SubRegion* base, SVal offset)
56    : baseRegion(base), byteOffset(offset) {}
57
58  NonLoc getByteOffset() const { return byteOffset.castAs<NonLoc>(); }
59  const SubRegion *getRegion() const { return baseRegion; }
60
61  static RegionRawOffsetV2 computeOffset(ProgramStateRef state,
62                                         SValBuilder &svalBuilder,
63                                         SVal location);
64
65  void dump() const;
66  void dumpToStream(raw_ostream &os) const;
67};
68}
69
70static SVal computeExtentBegin(SValBuilder &svalBuilder,
71                               const MemRegion *region) {
72  const MemSpaceRegion *SR = region->getMemorySpace();
73  if (SR->getKind() == MemRegion::UnknownSpaceRegionKind)
74    return UnknownVal();
75  else
76    return svalBuilder.makeZeroArrayIndex();
77}
78
79// TODO: once the constraint manager is smart enough to handle non simplified
80// symbolic expressions remove this function. Note that this can not be used in
81// the constraint manager as is, since this does not handle overflows. It is
82// safe to assume, however, that memory offsets will not overflow.
83static std::pair<NonLoc, nonloc::ConcreteInt>
84getSimplifiedOffsets(NonLoc offset, nonloc::ConcreteInt extent,
85                     SValBuilder &svalBuilder) {
86  Optional<nonloc::SymbolVal> SymVal = offset.getAs<nonloc::SymbolVal>();
87  if (SymVal && SymVal->isExpression()) {
88    if (const SymIntExpr *SIE = dyn_cast<SymIntExpr>(SymVal->getSymbol())) {
89      llvm::APSInt constant =
90          APSIntType(extent.getValue()).convert(SIE->getRHS());
91      switch (SIE->getOpcode()) {
92      case BO_Mul:
93        // The constant should never be 0 here, since it the result of scaling
94        // based on the size of a type which is never 0.
95        if ((extent.getValue() % constant) != 0)
96          return std::pair<NonLoc, nonloc::ConcreteInt>(offset, extent);
97        else
98          return getSimplifiedOffsets(
99              nonloc::SymbolVal(SIE->getLHS()),
100              svalBuilder.makeIntVal(extent.getValue() / constant),
101              svalBuilder);
102      case BO_Add:
103        return getSimplifiedOffsets(
104            nonloc::SymbolVal(SIE->getLHS()),
105            svalBuilder.makeIntVal(extent.getValue() - constant), svalBuilder);
106      default:
107        break;
108      }
109    }
110  }
111
112  return std::pair<NonLoc, nonloc::ConcreteInt>(offset, extent);
113}
114
115void ArrayBoundCheckerV2::checkLocation(SVal location, bool isLoad,
116                                        const Stmt* LoadS,
117                                        CheckerContext &checkerContext) const {
118
119  // NOTE: Instead of using ProgramState::assumeInBound(), we are prototyping
120  // some new logic here that reasons directly about memory region extents.
121  // Once that logic is more mature, we can bring it back to assumeInBound()
122  // for all clients to use.
123  //
124  // The algorithm we are using here for bounds checking is to see if the
125  // memory access is within the extent of the base region.  Since we
126  // have some flexibility in defining the base region, we can achieve
127  // various levels of conservatism in our buffer overflow checking.
128  ProgramStateRef state = checkerContext.getState();
129
130  SValBuilder &svalBuilder = checkerContext.getSValBuilder();
131  const RegionRawOffsetV2 &rawOffset =
132    RegionRawOffsetV2::computeOffset(state, svalBuilder, location);
133
134  if (!rawOffset.getRegion())
135    return;
136
137  NonLoc rawOffsetVal = rawOffset.getByteOffset();
138
139  // CHECK LOWER BOUND: Is byteOffset < extent begin?
140  //  If so, we are doing a load/store
141  //  before the first valid offset in the memory region.
142
143  SVal extentBegin = computeExtentBegin(svalBuilder, rawOffset.getRegion());
144
145  if (Optional<NonLoc> NV = extentBegin.getAs<NonLoc>()) {
146    if (NV->getAs<nonloc::ConcreteInt>()) {
147      std::pair<NonLoc, nonloc::ConcreteInt> simplifiedOffsets =
148          getSimplifiedOffsets(rawOffset.getByteOffset(),
149                               NV->castAs<nonloc::ConcreteInt>(),
150                               svalBuilder);
151      rawOffsetVal = simplifiedOffsets.first;
152      *NV = simplifiedOffsets.second;
153    }
154
155    SVal lowerBound = svalBuilder.evalBinOpNN(state, BO_LT, rawOffsetVal, *NV,
156                                              svalBuilder.getConditionType());
157
158    Optional<NonLoc> lowerBoundToCheck = lowerBound.getAs<NonLoc>();
159    if (!lowerBoundToCheck)
160      return;
161
162    ProgramStateRef state_precedesLowerBound, state_withinLowerBound;
163    std::tie(state_precedesLowerBound, state_withinLowerBound) =
164      state->assume(*lowerBoundToCheck);
165
166    // Are we constrained enough to definitely precede the lower bound?
167    if (state_precedesLowerBound && !state_withinLowerBound) {
168      reportOOB(checkerContext, state_precedesLowerBound, OOB_Precedes);
169      return;
170    }
171
172    // Otherwise, assume the constraint of the lower bound.
173    assert(state_withinLowerBound);
174    state = state_withinLowerBound;
175  }
176
177  do {
178    // CHECK UPPER BOUND: Is byteOffset >= extent(baseRegion)?  If so,
179    // we are doing a load/store after the last valid offset.
180    DefinedOrUnknownSVal extentVal =
181      rawOffset.getRegion()->getExtent(svalBuilder);
182    if (!extentVal.getAs<NonLoc>())
183      break;
184
185    if (extentVal.getAs<nonloc::ConcreteInt>()) {
186      std::pair<NonLoc, nonloc::ConcreteInt> simplifiedOffsets =
187          getSimplifiedOffsets(rawOffset.getByteOffset(),
188                               extentVal.castAs<nonloc::ConcreteInt>(),
189                               svalBuilder);
190      rawOffsetVal = simplifiedOffsets.first;
191      extentVal = simplifiedOffsets.second;
192    }
193
194    SVal upperbound = svalBuilder.evalBinOpNN(state, BO_GE, rawOffsetVal,
195                                              extentVal.castAs<NonLoc>(),
196                                              svalBuilder.getConditionType());
197
198    Optional<NonLoc> upperboundToCheck = upperbound.getAs<NonLoc>();
199    if (!upperboundToCheck)
200      break;
201
202    ProgramStateRef state_exceedsUpperBound, state_withinUpperBound;
203    std::tie(state_exceedsUpperBound, state_withinUpperBound) =
204      state->assume(*upperboundToCheck);
205
206    // If we are under constrained and the index variables are tainted, report.
207    if (state_exceedsUpperBound && state_withinUpperBound) {
208      SVal ByteOffset = rawOffset.getByteOffset();
209      if (isTainted(state, ByteOffset)) {
210        reportOOB(checkerContext, state_exceedsUpperBound, OOB_Tainted,
211                  std::make_unique<TaintBugVisitor>(ByteOffset));
212        return;
213      }
214    } else if (state_exceedsUpperBound) {
215      // If we are constrained enough to definitely exceed the upper bound,
216      // report.
217      assert(!state_withinUpperBound);
218      reportOOB(checkerContext, state_exceedsUpperBound, OOB_Excedes);
219      return;
220    }
221
222    assert(state_withinUpperBound);
223    state = state_withinUpperBound;
224  }
225  while (false);
226
227  checkerContext.addTransition(state);
228}
229
230void ArrayBoundCheckerV2::reportOOB(
231    CheckerContext &checkerContext, ProgramStateRef errorState, OOB_Kind kind,
232    std::unique_ptr<BugReporterVisitor> Visitor) const {
233
234  ExplodedNode *errorNode = checkerContext.generateErrorNode(errorState);
235  if (!errorNode)
236    return;
237
238  if (!BT)
239    BT.reset(new BuiltinBug(this, "Out-of-bound access"));
240
241  // FIXME: This diagnostics are preliminary.  We should get far better
242  // diagnostics for explaining buffer overruns.
243
244  SmallString<256> buf;
245  llvm::raw_svector_ostream os(buf);
246  os << "Out of bound memory access ";
247  switch (kind) {
248  case OOB_Precedes:
249    os << "(accessed memory precedes memory block)";
250    break;
251  case OOB_Excedes:
252    os << "(access exceeds upper limit of memory block)";
253    break;
254  case OOB_Tainted:
255    os << "(index is tainted)";
256    break;
257  }
258
259  auto BR = std::make_unique<PathSensitiveBugReport>(*BT, os.str(), errorNode);
260  BR->addVisitor(std::move(Visitor));
261  checkerContext.emitReport(std::move(BR));
262}
263
264#ifndef NDEBUG
265LLVM_DUMP_METHOD void RegionRawOffsetV2::dump() const {
266  dumpToStream(llvm::errs());
267}
268
269void RegionRawOffsetV2::dumpToStream(raw_ostream &os) const {
270  os << "raw_offset_v2{" << getRegion() << ',' << getByteOffset() << '}';
271}
272#endif
273
274// Lazily computes a value to be used by 'computeOffset'.  If 'val'
275// is unknown or undefined, we lazily substitute '0'.  Otherwise,
276// return 'val'.
277static inline SVal getValue(SVal val, SValBuilder &svalBuilder) {
278  return val.getAs<UndefinedVal>() ? svalBuilder.makeArrayIndex(0) : val;
279}
280
281// Scale a base value by a scaling factor, and return the scaled
282// value as an SVal.  Used by 'computeOffset'.
283static inline SVal scaleValue(ProgramStateRef state,
284                              NonLoc baseVal, CharUnits scaling,
285                              SValBuilder &sb) {
286  return sb.evalBinOpNN(state, BO_Mul, baseVal,
287                        sb.makeArrayIndex(scaling.getQuantity()),
288                        sb.getArrayIndexType());
289}
290
291// Add an SVal to another, treating unknown and undefined values as
292// summing to UnknownVal.  Used by 'computeOffset'.
293static SVal addValue(ProgramStateRef state, SVal x, SVal y,
294                     SValBuilder &svalBuilder) {
295  // We treat UnknownVals and UndefinedVals the same here because we
296  // only care about computing offsets.
297  if (x.isUnknownOrUndef() || y.isUnknownOrUndef())
298    return UnknownVal();
299
300  return svalBuilder.evalBinOpNN(state, BO_Add, x.castAs<NonLoc>(),
301                                 y.castAs<NonLoc>(),
302                                 svalBuilder.getArrayIndexType());
303}
304
305/// Compute a raw byte offset from a base region.  Used for array bounds
306/// checking.
307RegionRawOffsetV2 RegionRawOffsetV2::computeOffset(ProgramStateRef state,
308                                                   SValBuilder &svalBuilder,
309                                                   SVal location)
310{
311  const MemRegion *region = location.getAsRegion();
312  SVal offset = UndefinedVal();
313
314  while (region) {
315    switch (region->getKind()) {
316      default: {
317        if (const SubRegion *subReg = dyn_cast<SubRegion>(region)) {
318          offset = getValue(offset, svalBuilder);
319          if (!offset.isUnknownOrUndef())
320            return RegionRawOffsetV2(subReg, offset);
321        }
322        return RegionRawOffsetV2();
323      }
324      case MemRegion::ElementRegionKind: {
325        const ElementRegion *elemReg = cast<ElementRegion>(region);
326        SVal index = elemReg->getIndex();
327        if (!index.getAs<NonLoc>())
328          return RegionRawOffsetV2();
329        QualType elemType = elemReg->getElementType();
330        // If the element is an incomplete type, go no further.
331        ASTContext &astContext = svalBuilder.getContext();
332        if (elemType->isIncompleteType())
333          return RegionRawOffsetV2();
334
335        // Update the offset.
336        offset = addValue(state,
337                          getValue(offset, svalBuilder),
338                          scaleValue(state,
339                          index.castAs<NonLoc>(),
340                          astContext.getTypeSizeInChars(elemType),
341                          svalBuilder),
342                          svalBuilder);
343
344        if (offset.isUnknownOrUndef())
345          return RegionRawOffsetV2();
346
347        region = elemReg->getSuperRegion();
348        continue;
349      }
350    }
351  }
352  return RegionRawOffsetV2();
353}
354
355void ento::registerArrayBoundCheckerV2(CheckerManager &mgr) {
356  mgr.registerChecker<ArrayBoundCheckerV2>();
357}
358
359bool ento::shouldRegisterArrayBoundCheckerV2(const LangOptions &LO) {
360  return true;
361}
362