Home | History | Annotate | Download | only in Checkers 
      1 //==- UnreachableCodeChecker.cpp - Generalized dead code checker -*- C++ -*-==//
      2 //
      3 //                     The LLVM Compiler Infrastructure
      4 //
      5 // This file is distributed under the University of Illinois Open Source
      6 // License. See LICENSE.TXT for details.
      7 //
      8 //===----------------------------------------------------------------------===//
      9 // This file implements a generalized unreachable code checker using a
     10 // path-sensitive analysis. We mark any path visited, and then walk the CFG as a
     11 // post-analysis to determine what was never visited.
     12 //
     13 // A similar flow-sensitive only check exists in Analysis/ReachableCode.cpp
     14 //===----------------------------------------------------------------------===//
     15 
     16 #include "ClangSACheckers.h"
     17 #include "clang/StaticAnalyzer/Core/Checker.h"
     18 #include "clang/StaticAnalyzer/Core/CheckerManager.h"
     19 #include "clang/StaticAnalyzer/Core/PathSensitive/CheckerContext.h"
     20 #include "clang/StaticAnalyzer/Core/PathSensitive/ExplodedGraph.h"
     21 #include "clang/StaticAnalyzer/Core/PathSensitive/SVals.h"
     22 #include "clang/StaticAnalyzer/Core/PathSensitive/CheckerHelpers.h"
     23 #include "clang/StaticAnalyzer/Core/BugReporter/BugReporter.h"
     24 #include "clang/AST/ParentMap.h"
     25 #include "clang/Basic/Builtins.h"
     26 #include "clang/Basic/SourceManager.h"
     27 #include "llvm/ADT/SmallSet.h"
     28 
     29 // The number of CFGBlock pointers we want to reserve memory for. This is used
     30 // once for each function we analyze.
     31 #define DEFAULT_CFGBLOCKS 256
     32 
     33 using namespace clang;
     34 using namespace ento;
     35 
     36 namespace {
     37 class UnreachableCodeChecker : public Checker<check::EndAnalysis> {
     38 public:
     39   void checkEndAnalysis(ExplodedGraph &G, BugReporter &B,
     40                         ExprEngine &Eng) const;
     41 private:
     42   typedef llvm::SmallSet<unsigned, DEFAULT_CFGBLOCKS> CFGBlocksSet;
     43 
     44   static inline const Stmt *getUnreachableStmt(const CFGBlock *CB);
     45   static void FindUnreachableEntryPoints(const CFGBlock *CB,
     46                                          CFGBlocksSet &reachable,
     47                                          CFGBlocksSet &visited);
     48   static bool isInvalidPath(const CFGBlock *CB, const ParentMap &PM);
     49   static inline bool isEmptyCFGBlock(const CFGBlock *CB);
     50 };
     51 }
     52 
     53 void UnreachableCodeChecker::checkEndAnalysis(ExplodedGraph &G,
     54                                               BugReporter &B,
     55                                               ExprEngine &Eng) const {
     56   CFGBlocksSet reachable, visited;
     57 
     58   if (Eng.hasWorkRemaining())
     59     return;
     60 
     61   const Decl *D = 0;
     62   CFG *C = 0;
     63   ParentMap *PM = 0;
     64   const LocationContext *LC = 0;
     65   // Iterate over ExplodedGraph
     66   for (ExplodedGraph::node_iterator I = G.nodes_begin(), E = G.nodes_end();
     67       I != E; ++I) {
     68     const ProgramPoint &P = I->getLocation();
     69     LC = P.getLocationContext();
     70 
     71     if (!D)
     72       D = LC->getAnalysisDeclContext()->getDecl();
     73     // Save the CFG if we don't have it already
     74     if (!C)
     75       C = LC->getAnalysisDeclContext()->getUnoptimizedCFG();
     76     if (!PM)
     77       PM = &LC->getParentMap();
     78 
     79     if (const BlockEntrance *BE = dyn_cast<BlockEntrance>(&P)) {
     80       const CFGBlock *CB = BE->getBlock();
     81       reachable.insert(CB->getBlockID());
     82     }
     83   }
     84 
     85   // Bail out if we didn't get the CFG or the ParentMap.
     86   if (!D || !C || !PM)
     87     return;
     88 
     89   // Don't do anything for template instantiations.  Proving that code
     90   // in a template instantiation is unreachable means proving that it is
     91   // unreachable in all instantiations.
     92   if (const FunctionDecl *FD = dyn_cast<FunctionDecl>(D))
     93     if (FD->isTemplateInstantiation())
     94       return;
     95 
     96   // Find CFGBlocks that were not covered by any node
     97   for (CFG::const_iterator I = C->begin(), E = C->end(); I != E; ++I) {
     98     const CFGBlock *CB = *I;
     99     // Check if the block is unreachable
    100     if (reachable.count(CB->getBlockID()))
    101       continue;
    102 
    103     // Check if the block is empty (an artificial block)
    104     if (isEmptyCFGBlock(CB))
    105       continue;
    106 
    107     // Find the entry points for this block
    108     if (!visited.count(CB->getBlockID()))
    109       FindUnreachableEntryPoints(CB, reachable, visited);
    110 
    111     // This block may have been pruned; check if we still want to report it
    112     if (reachable.count(CB->getBlockID()))
    113       continue;
    114 
    115     // Check for false positives
    116     if (CB->size() > 0 && isInvalidPath(CB, *PM))
    117       continue;
    118 
    119     // It is good practice to always have a "default" label in a "switch", even
    120     // if we should never get there. It can be used to detect errors, for
    121     // instance. Unreachable code directly under a "default" label is therefore
    122     // likely to be a false positive.
    123     if (const Stmt *label = CB->getLabel())
    124       if (label->getStmtClass() == Stmt::DefaultStmtClass)
    125         continue;
    126 
    127     // Special case for __builtin_unreachable.
    128     // FIXME: This should be extended to include other unreachable markers,
    129     // such as llvm_unreachable.
    130     if (!CB->empty()) {
    131       bool foundUnreachable = false;
    132       for (CFGBlock::const_iterator ci = CB->begin(), ce = CB->end();
    133            ci != ce; ++ci) {
    134         if (const CFGStmt *S = (*ci).getAs<CFGStmt>())
    135           if (const CallExpr *CE = dyn_cast<CallExpr>(S->getStmt())) {
    136             if (CE->isBuiltinCall() == Builtin::BI__builtin_unreachable) {
    137               foundUnreachable = true;
    138               break;
    139             }
    140           }
    141       }
    142       if (foundUnreachable)
    143         continue;
    144     }
    145 
    146     // We found a block that wasn't covered - find the statement to report
    147     SourceRange SR;
    148     PathDiagnosticLocation DL;
    149     SourceLocation SL;
    150     if (const Stmt *S = getUnreachableStmt(CB)) {
    151       SR = S->getSourceRange();
    152       DL = PathDiagnosticLocation::createBegin(S, B.getSourceManager(), LC);
    153       SL = DL.asLocation();
    154       if (SR.isInvalid() || !SL.isValid())
    155         continue;
    156     }
    157     else
    158       continue;
    159 
    160     // Check if the SourceLocation is in a system header
    161     const SourceManager &SM = B.getSourceManager();
    162     if (SM.isInSystemHeader(SL) || SM.isInExternCSystemHeader(SL))
    163       continue;
    164 
    165     B.EmitBasicReport(D, "Unreachable code", "Dead code",
    166                       "This statement is never executed", DL, SR);
    167   }
    168 }
    169 
    170 // Recursively finds the entry point(s) for this dead CFGBlock.
    171 void UnreachableCodeChecker::FindUnreachableEntryPoints(const CFGBlock *CB,
    172                                                         CFGBlocksSet &reachable,
    173                                                         CFGBlocksSet &visited) {
    174   visited.insert(CB->getBlockID());
    175 
    176   for (CFGBlock::const_pred_iterator I = CB->pred_begin(), E = CB->pred_end();
    177       I != E; ++I) {
    178     if (!reachable.count((*I)->getBlockID())) {
    179       // If we find an unreachable predecessor, mark this block as reachable so
    180       // we don't report this block
    181       reachable.insert(CB->getBlockID());
    182       if (!visited.count((*I)->getBlockID()))
    183         // If we haven't previously visited the unreachable predecessor, recurse
    184         FindUnreachableEntryPoints(*I, reachable, visited);
    185     }
    186   }
    187 }
    188 
    189 // Find the Stmt* in a CFGBlock for reporting a warning
    190 const Stmt *UnreachableCodeChecker::getUnreachableStmt(const CFGBlock *CB) {
    191   for (CFGBlock::const_iterator I = CB->begin(), E = CB->end(); I != E; ++I) {
    192     if (const CFGStmt *S = I->getAs<CFGStmt>())
    193       return S->getStmt();
    194   }
    195   if (const Stmt *S = CB->getTerminator())
    196     return S;
    197   else
    198     return 0;
    199 }
    200 
    201 // Determines if the path to this CFGBlock contained an element that infers this
    202 // block is a false positive. We assume that FindUnreachableEntryPoints has
    203 // already marked only the entry points to any dead code, so we need only to
    204 // find the condition that led to this block (the predecessor of this block.)
    205 // There will never be more than one predecessor.
    206 bool UnreachableCodeChecker::isInvalidPath(const CFGBlock *CB,
    207                                            const ParentMap &PM) {
    208   // We only expect a predecessor size of 0 or 1. If it is >1, then an external
    209   // condition has broken our assumption (for example, a sink being placed by
    210   // another check). In these cases, we choose not to report.
    211   if (CB->pred_size() > 1)
    212     return true;
    213 
    214   // If there are no predecessors, then this block is trivially unreachable
    215   if (CB->pred_size() == 0)
    216     return false;
    217 
    218   const CFGBlock *pred = *CB->pred_begin();
    219 
    220   // Get the predecessor block's terminator conditon
    221   const Stmt *cond = pred->getTerminatorCondition();
    222 
    223   //assert(cond && "CFGBlock's predecessor has a terminator condition");
    224   // The previous assertion is invalid in some cases (eg do/while). Leaving
    225   // reporting of these situations on at the moment to help triage these cases.
    226   if (!cond)
    227     return false;
    228 
    229   // Run each of the checks on the conditions
    230   if (containsMacro(cond) || containsEnum(cond)
    231       || containsStaticLocal(cond) || containsBuiltinOffsetOf(cond)
    232       || containsStmt<UnaryExprOrTypeTraitExpr>(cond))
    233     return true;
    234 
    235   return false;
    236 }
    237 
    238 // Returns true if the given CFGBlock is empty
    239 bool UnreachableCodeChecker::isEmptyCFGBlock(const CFGBlock *CB) {
    240   return CB->getLabel() == 0       // No labels
    241       && CB->size() == 0           // No statements
    242       && CB->getTerminator() == 0; // No terminator
    243 }
    244 
    245 void ento::registerUnreachableCodeChecker(CheckerManager &mgr) {
    246   mgr.registerChecker<UnreachableCodeChecker>();
    247 }
    248