lib/Checker/Store.cpp

//== Store.cpp - Interface for maps from Locations to Values ----*- C++ -*--==//
//
//                     The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
//  This file defined the types Store and StoreManager.
//
//===----------------------------------------------------------------------===//

#include "clang/Checker/PathSensitive/Store.h"
#include "clang/Checker/PathSensitive/GRState.h"
#include "clang/AST/CharUnits.h"

using namespace clang;

StoreManager::StoreManager(GRStateManager &stateMgr)
  : ValMgr(stateMgr.getValueManager()), StateMgr(stateMgr),
    MRMgr(ValMgr.getRegionManager()), Ctx(stateMgr.getContext()) {}

const MemRegion *StoreManager::MakeElementRegion(const MemRegion *Base,
                                              QualType EleTy, uint64_t index) {
  SVal idx = ValMgr.makeArrayIndex(index);
  return MRMgr.getElementRegion(EleTy, idx, Base, ValMgr.getContext());
}

// FIXME: Merge with the implementation of the same method in MemRegion.cpp
static bool IsCompleteType(ASTContext &Ctx, QualType Ty) {
  if (const RecordType *RT = Ty->getAs<RecordType>()) {
    const RecordDecl *D = RT->getDecl();
    if (!D->getDefinition(Ctx))
      return false;
  }

  return true;
}

const MemRegion *StoreManager::CastRegion(const MemRegion *R, QualType CastToTy) {

  ASTContext& Ctx = StateMgr.getContext();

  // Handle casts to Objective-C objects.
  if (CastToTy->isObjCObjectPointerType())
    return R->StripCasts();

  if (CastToTy->isBlockPointerType()) {
    // FIXME: We may need different solutions, depending on the symbol
    // involved.  Blocks can be casted to/from 'id', as they can be treated
    // as Objective-C objects.  This could possibly be handled by enhancing
    // our reasoning of downcasts of symbolic objects.
    if (isa<CodeTextRegion>(R) || isa<SymbolicRegion>(R))
      return R;

    // We don't know what to make of it.  Return a NULL region, which
    // will be interpretted as UnknownVal.
    return NULL;
  }

  // Now assume we are casting from pointer to pointer. Other cases should
  // already be handled.
  QualType PointeeTy = CastToTy->getAs<PointerType>()->getPointeeType();
  QualType CanonPointeeTy = Ctx.getCanonicalType(PointeeTy);

  // Handle casts to void*.  We just pass the region through.
  if (CanonPointeeTy.getLocalUnqualifiedType() == Ctx.VoidTy)
    return R;

  // Handle casts from compatible types.
  if (R->isBoundable())
    if (const TypedRegion *TR = dyn_cast<TypedRegion>(R)) {
      QualType ObjTy = Ctx.getCanonicalType(TR->getValueType(Ctx));
      if (CanonPointeeTy == ObjTy)
        return R;
    }

  // Process region cast according to the kind of the region being cast.
  switch (R->getKind()) {
    case MemRegion::CXXThisRegionKind:
    case MemRegion::GenericMemSpaceRegionKind:
    case MemRegion::StackLocalsSpaceRegionKind:
    case MemRegion::StackArgumentsSpaceRegionKind:
    case MemRegion::HeapSpaceRegionKind:
    case MemRegion::UnknownSpaceRegionKind:
    case MemRegion::GlobalsSpaceRegionKind: {
      assert(0 && "Invalid region cast");
      break;
    }
    
    case MemRegion::FunctionTextRegionKind:
    case MemRegion::BlockTextRegionKind:
    case MemRegion::BlockDataRegionKind: {
      // CodeTextRegion should be cast to only a function or block pointer type,
      // although they can in practice be casted to anything, e.g, void*, char*,
      // etc.  
      // Just return the region.
      return R;
    }

    case MemRegion::StringRegionKind:
      // FIXME: Need to handle arbitrary downcasts.
    case MemRegion::SymbolicRegionKind:
    case MemRegion::AllocaRegionKind:
    case MemRegion::CompoundLiteralRegionKind:
    case MemRegion::FieldRegionKind:
    case MemRegion::ObjCIvarRegionKind:
    case MemRegion::VarRegionKind:
    case MemRegion::CXXObjectRegionKind:
      return MakeElementRegion(R, PointeeTy);

    case MemRegion::ElementRegionKind: {
      // If we are casting from an ElementRegion to another type, the
      // algorithm is as follows:
      //
      // (1) Compute the "raw offset" of the ElementRegion from the
      //     base region.  This is done by calling 'getAsRawOffset()'.
      //
      // (2a) If we get a 'RegionRawOffset' after calling
      //      'getAsRawOffset()', determine if the absolute offset
      //      can be exactly divided into chunks of the size of the
      //      casted-pointee type.  If so, create a new ElementRegion with
      //      the pointee-cast type as the new ElementType and the index
      //      being the offset divded by the chunk size.  If not, create
      //      a new ElementRegion at offset 0 off the raw offset region.
      //
      // (2b) If we don't a get a 'RegionRawOffset' after calling
      //      'getAsRawOffset()', it means that we are at offset 0.
      //
      // FIXME: Handle symbolic raw offsets.

      const ElementRegion *elementR = cast<ElementRegion>(R);
      const RegionRawOffset &rawOff = elementR->getAsRawOffset();
      const MemRegion *baseR = rawOff.getRegion();

      // If we cannot compute a raw offset, throw up our hands and return
      // a NULL MemRegion*.
      if (!baseR)
        return NULL;

      CharUnits off = CharUnits::fromQuantity(rawOff.getByteOffset());

      if (off.isZero()) {
        // Edge case: we are at 0 bytes off the beginning of baseR.  We
        // check to see if type we are casting to is the same as the base
        // region.  If so, just return the base region.
        if (const TypedRegion *TR = dyn_cast<TypedRegion>(baseR)) {
          QualType ObjTy = Ctx.getCanonicalType(TR->getValueType(Ctx));
          QualType CanonPointeeTy = Ctx.getCanonicalType(PointeeTy);
          if (CanonPointeeTy == ObjTy)
            return baseR;
        }

        // Otherwise, create a new ElementRegion at offset 0.
        return MakeElementRegion(baseR, PointeeTy);
      }

      // We have a non-zero offset from the base region.  We want to determine
      // if the offset can be evenly divided by sizeof(PointeeTy).  If so,
      // we create an ElementRegion whose index is that value.  Otherwise, we
      // create two ElementRegions, one that reflects a raw offset and the other
      // that reflects the cast.

      // Compute the index for the new ElementRegion.
      int64_t newIndex = 0;
      const MemRegion *newSuperR = 0;

      // We can only compute sizeof(PointeeTy) if it is a complete type.
      if (IsCompleteType(Ctx, PointeeTy)) {
        // Compute the size in **bytes**.
        CharUnits pointeeTySize = Ctx.getTypeSizeInChars(PointeeTy);

        // Is the offset a multiple of the size?  If so, we can layer the
        // ElementRegion (with elementType == PointeeTy) directly on top of
        // the base region.
        if (off % pointeeTySize == 0) {
          newIndex = off / pointeeTySize;
          newSuperR = baseR;
        }
      }

      if (!newSuperR) {
        // Create an intermediate ElementRegion to represent the raw byte.
        // This will be the super region of the final ElementRegion.
        newSuperR = MakeElementRegion(baseR, Ctx.CharTy, off.getQuantity());
      }

      return MakeElementRegion(newSuperR, PointeeTy, newIndex);
    }
  }

  assert(0 && "unreachable");
  return 0;
}


/// CastRetrievedVal - Used by subclasses of StoreManager to implement
///  implicit casts that arise from loads from regions that are reinterpreted
///  as another region.
SVal StoreManager::CastRetrievedVal(SVal V, const TypedRegion *R,
                                    QualType castTy, bool performTestOnly) {
  
  if (castTy.isNull())
    return V;
  
  ASTContext &Ctx = ValMgr.getContext();

  if (performTestOnly) {  
    // Automatically translate references to pointers.
    QualType T = R->getValueType(Ctx);
    if (const ReferenceType *RT = T->getAs<ReferenceType>())
      T = Ctx.getPointerType(RT->getPointeeType());
    
    assert(ValMgr.getContext().hasSameUnqualifiedType(castTy, T));
    return V;
  }
  
  if (const Loc *L = dyn_cast<Loc>(&V))
    return ValMgr.getSValuator().EvalCastL(*L, castTy);
  else if (const NonLoc *NL = dyn_cast<NonLoc>(&V))
    return ValMgr.getSValuator().EvalCastNL(*NL, castTy);
  
  return V;
}

Store StoreManager::InvalidateRegions(Store store,
                                      const MemRegion * const *I,
                                      const MemRegion * const *End,
                                      const Expr *E, unsigned Count,
                                      InvalidatedSymbols *IS) {
  for ( ; I != End ; ++I)
    store = InvalidateRegion(store, *I, E, Count, IS);
  
  return store;
}

SVal StoreManager::getLValueFieldOrIvar(const Decl* D, SVal Base) {
  if (Base.isUnknownOrUndef())
    return Base;

  Loc BaseL = cast<Loc>(Base);
  const MemRegion* BaseR = 0;

  switch (BaseL.getSubKind()) {
  case loc::MemRegionKind:
    BaseR = cast<loc::MemRegionVal>(BaseL).getRegion();
    break;

  case loc::GotoLabelKind:
    // These are anormal cases. Flag an undefined value.
    return UndefinedVal();

  case loc::ConcreteIntKind:
    // While these seem funny, this can happen through casts.
    // FIXME: What we should return is the field offset.  For example,
    //  add the field offset to the integer value.  That way funny things
    //  like this work properly:  &(((struct foo *) 0xa)->f)
    return Base;

  default:
    assert(0 && "Unhandled Base.");
    return Base;
  }

  // NOTE: We must have this check first because ObjCIvarDecl is a subclass
  // of FieldDecl.
  if (const ObjCIvarDecl *ID = dyn_cast<ObjCIvarDecl>(D))
    return loc::MemRegionVal(MRMgr.getObjCIvarRegion(ID, BaseR));

  return loc::MemRegionVal(MRMgr.getFieldRegion(cast<FieldDecl>(D), BaseR));
}

SVal StoreManager::getLValueElement(QualType elementType, SVal Offset, 
                                    SVal Base) {

  // If the base is an unknown or undefined value, just return it back.
  // FIXME: For absolute pointer addresses, we just return that value back as
  //  well, although in reality we should return the offset added to that
  //  value.
  if (Base.isUnknownOrUndef() || isa<loc::ConcreteInt>(Base))
    return Base;

  // Only handle integer offsets... for now.
  if (!isa<nonloc::ConcreteInt>(Offset))
    return UnknownVal();

  const MemRegion* BaseRegion = cast<loc::MemRegionVal>(Base).getRegion();

  // Pointer of any type can be cast and used as array base.
  const ElementRegion *ElemR = dyn_cast<ElementRegion>(BaseRegion);

  // Convert the offset to the appropriate size and signedness.
  Offset = ValMgr.convertToArrayIndex(Offset);

  if (!ElemR) {
    //
    // If the base region is not an ElementRegion, create one.
    // This can happen in the following example:
    //
    //   char *p = __builtin_alloc(10);
    //   p[1] = 8;
    //
    //  Observe that 'p' binds to an AllocaRegion.
    //
    return loc::MemRegionVal(MRMgr.getElementRegion(elementType, Offset,
                                                    BaseRegion, Ctx));
  }

  SVal BaseIdx = ElemR->getIndex();

  if (!isa<nonloc::ConcreteInt>(BaseIdx))
    return UnknownVal();

  const llvm::APSInt& BaseIdxI = cast<nonloc::ConcreteInt>(BaseIdx).getValue();
  const llvm::APSInt& OffI = cast<nonloc::ConcreteInt>(Offset).getValue();
  assert(BaseIdxI.isSigned());

  // Compute the new index.
  SVal NewIdx = nonloc::ConcreteInt(
                      ValMgr.getBasicValueFactory().getValue(BaseIdxI + OffI));

  // Construct the new ElementRegion.
  const MemRegion *ArrayR = ElemR->getSuperRegion();
  return loc::MemRegionVal(MRMgr.getElementRegion(elementType, NewIdx, ArrayR,
                                                  Ctx));
}