[analyzer] NFC: Move getStmt() and createEndOfPath() out of PathDiagnostic.

These static functions deal with ExplodedNodes which is something we don't want
the PathDiagnostic interface to know anything about, as it's planned to be
moved out of libStaticAnalyzerCore.

Differential Revision: https://reviews.llvm.org/D67382

llvm-svn: 371659

5 files changed, 151 insertions, 154 deletions

diff --git a/clang/lib/StaticAnalyzer/Core/BugReporter.cpp b/clang/lib/StaticAnalyzer/Core/BugReporter.cpp
index c3b76b42ba1..57f639afbab 100644
--- a/clang/lib/StaticAnalyzer/Core/BugReporter.cpp
+++ b/clang/lib/StaticAnalyzer/Core/BugReporter.cpp
@@ -335,26 +335,6 @@ std::string StackHintGeneratorForSymbol::getMessageForArg(const Expr *ArgE,
 }
 
 //===----------------------------------------------------------------------===//
-// Helper routines for walking the ExplodedGraph and fetching statements.
-//===----------------------------------------------------------------------===//
-
-static const Stmt *GetPreviousStmt(const ExplodedNode *N) {
-  for (N = N->getFirstPred(); N; N = N->getFirstPred())
-    if (const Stmt *S = PathDiagnosticLocation::getStmt(N))
-      return S;
-
-  return nullptr;
-}
-
-static inline const Stmt*
-GetCurrentOrPreviousStmt(const ExplodedNode *N) {
-  if (const Stmt *S = PathDiagnosticLocation::getStmt(N))
-    return S;
-
-  return GetPreviousStmt(N);
-}
-
-//===----------------------------------------------------------------------===//
 // Diagnostic cleanup.
 //===----------------------------------------------------------------------===//
 
@@ -593,7 +573,7 @@ static void removePiecesWithInvalidLocations(PathPieces &Pieces) {
 
 PathDiagnosticLocation PathDiagnosticBuilder::ExecutionContinues(
     const PathDiagnosticConstruct &C) const {
-  if (const Stmt *S = PathDiagnosticLocation::getNextStmt(C.getCurrentNode()))
+  if (const Stmt *S = C.getCurrentNode()->getNextStmtForDiagnostics())
     return PathDiagnosticLocation(S, getSourceManager(),
                                   C.getCurrLocationContext());
 
@@ -888,7 +868,7 @@ void PathDiagnosticBuilder::generateMinimalDiagForBlockEdge(
 
   case Stmt::GotoStmtClass:
   case Stmt::IndirectGotoStmtClass: {
-    if (const Stmt *S = PathDiagnosticLocation::getNextStmt(C.getCurrentNode()))
+    if (const Stmt *S = C.getCurrentNode()->getNextStmtForDiagnostics())
       C.getActivePath().push_front(generateDiagForGotoOP(C, S, Start));
     break;
   }
@@ -2177,8 +2157,11 @@ void PathSensitiveBugReport::Profile(llvm::FoldingSetNodeID &hash) const {
   if (UL.isValid()) {
     UL.Profile(hash);
   } else {
-    assert(ErrorNode);
-    hash.AddPointer(GetCurrentOrPreviousStmt(ErrorNode));
+    // TODO: The statement may be null if the report was emitted before any
+    // statements were executed. In particular, some checkers by design
+    // occasionally emit their reports in empty functions (that have no
+    // statements in their body). Do we profile correctly in this case?
+    hash.AddPointer(ErrorNode->getCurrentOrPreviousStmtForDiagnostics());
   }
 
   for (SourceRange range : Ranges) {
@@ -2333,10 +2316,10 @@ const Stmt *PathSensitiveBugReport::getStmt() const {
   if (Optional<BlockEntrance> BE = ProgP.getAs<BlockEntrance>()) {
     CFGBlock &Exit = ProgP.getLocationContext()->getCFG()->getExit();
     if (BE->getBlock() == &Exit)
-      S = GetPreviousStmt(ErrorNode);
+      S = ErrorNode->getPreviousStmtForDiagnostics();
   }
   if (!S)
-    S = PathDiagnosticLocation::getStmt(ErrorNode);
+    S = ErrorNode->getStmtForDiagnostics();
 
   return S;
 }
@@ -2353,7 +2336,45 @@ PathSensitiveBugReport::getRanges() const {
 
 PathDiagnosticLocation
 PathSensitiveBugReport::getLocation() const {
-  return PathDiagnosticLocation::createEndOfPath(ErrorNode);
+  assert(ErrorNode && "Cannot create a location with a null node.");
+  const Stmt *S = ErrorNode->getStmtForDiagnostics();
+    ProgramPoint P = ErrorNode->getLocation();
+  const LocationContext *LC = P.getLocationContext();
+  SourceManager &SM =
+      ErrorNode->getState()->getStateManager().getContext().getSourceManager();
+
+  if (!S) {
+    // If this is an implicit call, return the implicit call point location.
+    if (Optional<PreImplicitCall> PIE = P.getAs<PreImplicitCall>())
+      return PathDiagnosticLocation(PIE->getLocation(), SM);
+    if (auto FE = P.getAs<FunctionExitPoint>()) {
+      if (const ReturnStmt *RS = FE->getStmt())
+        return PathDiagnosticLocation::createBegin(RS, SM, LC);
+    }
+    S = ErrorNode->getNextStmtForDiagnostics();
+  }
+
+  if (S) {
+    // For member expressions, return the location of the '.' or '->'.
+    if (const auto *ME = dyn_cast<MemberExpr>(S))
+      return PathDiagnosticLocation::createMemberLoc(ME, SM);
+
+    // For binary operators, return the location of the operator.
+    if (const auto *B = dyn_cast<BinaryOperator>(S))
+      return PathDiagnosticLocation::createOperatorLoc(B, SM);
+
+    if (P.getAs<PostStmtPurgeDeadSymbols>())
+      return PathDiagnosticLocation::createEnd(S, SM, LC);
+
+    if (S->getBeginLoc().isValid())
+      return PathDiagnosticLocation(S, SM, LC);
+
+    return PathDiagnosticLocation(
+        PathDiagnosticLocation::getValidSourceLocation(S, LC), SM);
+  }
+
+  return PathDiagnosticLocation::createDeclEnd(ErrorNode->getLocationContext(),
+                                               SM);
 }
 
 //===----------------------------------------------------------------------===//
@@ -3070,7 +3091,7 @@ findExecutedLines(const SourceManager &SM, const ExplodedNode *N) {
       // Inlined function: show signature.
       const Decl* D = CE->getCalleeContext()->getDecl();
       populateExecutedLinesWithFunctionSignature(D, SM, *ExecutedLines);
-    } else if (const Stmt *S = PathDiagnosticLocation::getStmt(N)) {
+    } else if (const Stmt *S = N->getStmtForDiagnostics()) {
       populateExecutedLinesWithStmt(S, SM, *ExecutedLines);
 
       // Show extra context for some parent kinds.
diff --git a/clang/lib/StaticAnalyzer/Core/BugReporterVisitors.cpp b/clang/lib/StaticAnalyzer/Core/BugReporterVisitors.cpp
index 7b26b8fabb9..bf923daeec5 100644
--- a/clang/lib/StaticAnalyzer/Core/BugReporterVisitors.cpp
+++ b/clang/lib/StaticAnalyzer/Core/BugReporterVisitors.cpp
@@ -308,9 +308,7 @@ PathDiagnosticPieceRef
 BugReporterVisitor::getDefaultEndPath(const BugReporterContext &BRC,
                                       const ExplodedNode *EndPathNode,
                                       const PathSensitiveBugReport &BR) {
-  PathDiagnosticLocation L =
-      PathDiagnosticLocation::createEndOfPath(EndPathNode);
-
+  PathDiagnosticLocation L = BR.getLocation();
   const auto &Ranges = BR.getRanges();
 
   // Only add the statement itself as a range if we didn't specify any
@@ -852,7 +850,7 @@ private:
   /// \return Source location of right hand side of an assignment
   /// into \c RegionOfInterest, empty optional if none found.
   Optional<SourceLocation> matchAssignment(const ExplodedNode *N) {
-    const Stmt *S = PathDiagnosticLocation::getStmt(N);
+    const Stmt *S = N->getStmtForDiagnostics();
     ProgramStateRef State = N->getState();
     auto *LCtx = N->getLocationContext();
     if (!S)
@@ -1919,7 +1917,7 @@ static const Expr *peelOffOuterExpr(const Expr *Ex,
 static const ExplodedNode* findNodeForExpression(const ExplodedNode *N,
                                                  const Expr *Inner) {
   while (N) {
-    if (PathDiagnosticLocation::getStmt(N) == Inner)
+    if (N->getStmtForDiagnostics() == Inner)
       return N;
     N = N->getFirstPred();
   }
diff --git a/clang/lib/StaticAnalyzer/Core/ExplodedGraph.cpp b/clang/lib/StaticAnalyzer/Core/ExplodedGraph.cpp
index 03e813e1e67..a8fa7ad2d98 100644
--- a/clang/lib/StaticAnalyzer/Core/ExplodedGraph.cpp
+++ b/clang/lib/StaticAnalyzer/Core/ExplodedGraph.cpp
@@ -299,7 +299,9 @@ const CFGBlock *ExplodedNode::getCFGBlock() const {
     return BEP->getBlock();
 
   // Find the node's current statement in the CFG.
-  if (const Stmt *S = PathDiagnosticLocation::getStmt(this))
+  // FIXME: getStmtForDiagnostics() does nasty things in order to provide
+  // a valid statement for body farms, do we need this behavior here?
+  if (const Stmt *S = getStmtForDiagnostics())
     return getLocationContext()
         ->getAnalysisDeclContext()
         ->getCFGStmtMap()
@@ -308,6 +310,92 @@ const CFGBlock *ExplodedNode::getCFGBlock() const {
   return nullptr;
 }
 
+static const LocationContext *
+findTopAutosynthesizedParentContext(const LocationContext *LC) {
+  assert(LC->getAnalysisDeclContext()->isBodyAutosynthesized());
+  const LocationContext *ParentLC = LC->getParent();
+  assert(ParentLC && "We don't start analysis from autosynthesized code");
+  while (ParentLC->getAnalysisDeclContext()->isBodyAutosynthesized()) {
+    LC = ParentLC;
+    ParentLC = LC->getParent();
+    assert(ParentLC && "We don't start analysis from autosynthesized code");
+  }
+  return LC;
+}
+
+const Stmt *ExplodedNode::getStmtForDiagnostics() const {
+  // We cannot place diagnostics on autosynthesized code.
+  // Put them onto the call site through which we jumped into autosynthesized
+  // code for the first time.
+  const LocationContext *LC = getLocationContext();
+  if (LC->getAnalysisDeclContext()->isBodyAutosynthesized()) {
+    // It must be a stack frame because we only autosynthesize functions.
+    return cast<StackFrameContext>(findTopAutosynthesizedParentContext(LC))
+        ->getCallSite();
+  }
+  // Otherwise, see if the node's program point directly points to a statement.
+  // FIXME: Refactor into a ProgramPoint method?
+  ProgramPoint P = getLocation();
+  if (auto SP = P.getAs<StmtPoint>())
+    return SP->getStmt();
+  if (auto BE = P.getAs<BlockEdge>())
+    return BE->getSrc()->getTerminatorStmt();
+  if (auto CE = P.getAs<CallEnter>())
+    return CE->getCallExpr();
+  if (auto CEE = P.getAs<CallExitEnd>())
+    return CEE->getCalleeContext()->getCallSite();
+  if (auto PIPP = P.getAs<PostInitializer>())
+    return PIPP->getInitializer()->getInit();
+  if (auto CEB = P.getAs<CallExitBegin>())
+    return CEB->getReturnStmt();
+  if (auto FEP = P.getAs<FunctionExitPoint>())
+    return FEP->getStmt();
+
+  return nullptr;
+}
+
+const Stmt *ExplodedNode::getNextStmtForDiagnostics() const {
+  for (const ExplodedNode *N = getFirstSucc(); N; N = N->getFirstSucc()) {
+    if (const Stmt *S = N->getStmtForDiagnostics()) {
+      // Check if the statement is '?' or '&&'/'||'.  These are "merges",
+      // not actual statement points.
+      switch (S->getStmtClass()) {
+        case Stmt::ChooseExprClass:
+        case Stmt::BinaryConditionalOperatorClass:
+        case Stmt::ConditionalOperatorClass:
+          continue;
+        case Stmt::BinaryOperatorClass: {
+          BinaryOperatorKind Op = cast<BinaryOperator>(S)->getOpcode();
+          if (Op == BO_LAnd || Op == BO_LOr)
+            continue;
+          break;
+        }
+        default:
+          break;
+      }
+      // We found the statement, so return it.
+      return S;
+    }
+  }
+
+  return nullptr;
+}
+
+const Stmt *ExplodedNode::getPreviousStmtForDiagnostics() const {
+  for (const ExplodedNode *N = getFirstPred(); N; N = N->getFirstPred())
+    if (const Stmt *S = N->getStmtForDiagnostics())
+      return S;
+
+  return nullptr;
+}
+
+const Stmt *ExplodedNode::getCurrentOrPreviousStmtForDiagnostics() const {
+  if (const Stmt *S = getStmtForDiagnostics())
+    return S;
+
+  return getPreviousStmtForDiagnostics();
+}
+
 ExplodedNode *ExplodedGraph::getNode(const ProgramPoint &L,
                                      ProgramStateRef State,
                                      bool IsSink,
diff --git a/clang/lib/StaticAnalyzer/Core/LoopUnrolling.cpp b/clang/lib/StaticAnalyzer/Core/LoopUnrolling.cpp
index 9838249ae82..1a09a521f11 100644
--- a/clang/lib/StaticAnalyzer/Core/LoopUnrolling.cpp
+++ b/clang/lib/StaticAnalyzer/Core/LoopUnrolling.cpp
@@ -165,7 +165,9 @@ static bool isPossiblyEscaped(const VarDecl *VD, ExplodedNode *N) {
     return true;
 
   while (!N->pred_empty()) {
-    const Stmt *S = PathDiagnosticLocation::getStmt(N);
+    // FIXME: getStmtForDiagnostics() does nasty things in order to provide
+    // a valid statement for body farms, do we need this behavior here?
+    const Stmt *S = N->getStmtForDiagnostics();
     if (!S) {
       N = N->getFirstPred();
       continue;
diff --git a/clang/lib/StaticAnalyzer/Core/PathDiagnostic.cpp b/clang/lib/StaticAnalyzer/Core/PathDiagnostic.cpp
index 94d1c095b8b..853029ac7b0 100644
--- a/clang/lib/StaticAnalyzer/Core/PathDiagnostic.cpp
+++ b/clang/lib/StaticAnalyzer/Core/PathDiagnostic.cpp
@@ -30,7 +30,6 @@
 #include "clang/Basic/SourceLocation.h"
 #include "clang/Basic/SourceManager.h"
 #include "clang/StaticAnalyzer/Core/PathSensitive/AnalysisManager.h"
-#include "clang/StaticAnalyzer/Core/PathSensitive/ExplodedGraph.h"
 #include "llvm/ADT/ArrayRef.h"
 #include "llvm/ADT/FoldingSet.h"
 #include "llvm/ADT/None.h"
@@ -524,12 +523,12 @@ PathDiagnosticConsumer::FilesMade::getFiles(const PathDiagnostic &PD) {
 // PathDiagnosticLocation methods.
 //===----------------------------------------------------------------------===//
 
-static SourceLocation getValidSourceLocation(const Stmt* S,
-                                             LocationOrAnalysisDeclContext LAC,
-                                             bool UseEnd = false) {
-  SourceLocation L = UseEnd ? S->getEndLoc() : S->getBeginLoc();
-  assert(!LAC.isNull() && "A valid LocationContext or AnalysisDeclContext should "
-                          "be passed to PathDiagnosticLocation upon creation.");
+SourceLocation PathDiagnosticLocation::getValidSourceLocation(
+    const Stmt *S, LocationOrAnalysisDeclContext LAC, bool UseEndOfStatement) {
+  SourceLocation L = UseEndOfStatement ? S->getEndLoc() : S->getBeginLoc();
+  assert(!LAC.isNull() &&
+         "A valid LocationContext or AnalysisDeclContext should be passed to "
+         "PathDiagnosticLocation upon creation.");
 
   // S might be a temporary statement that does not have a location in the
   // source code, so find an enclosing statement and use its location.
@@ -559,7 +558,7 @@ static SourceLocation getValidSourceLocation(const Stmt* S,
         break;
       }
 
-      L = UseEnd ? Parent->getEndLoc() : Parent->getBeginLoc();
+      L = UseEndOfStatement ? Parent->getEndLoc() : Parent->getBeginLoc();
     } while (!L.isValid());
   }
 
@@ -778,117 +777,6 @@ PathDiagnosticLocation::create(const ProgramPoint& P,
   return PathDiagnosticLocation(S, SMng, P.getLocationContext());
 }
 
-static const LocationContext *
-findTopAutosynthesizedParentContext(const LocationContext *LC) {
-  assert(LC->getAnalysisDeclContext()->isBodyAutosynthesized());
-  const LocationContext *ParentLC = LC->getParent();
-  assert(ParentLC && "We don't start analysis from autosynthesized code");
-  while (ParentLC->getAnalysisDeclContext()->isBodyAutosynthesized()) {
-    LC = ParentLC;
-    ParentLC = LC->getParent();
-    assert(ParentLC && "We don't start analysis from autosynthesized code");
-  }
-  return LC;
-}
-
-const Stmt *PathDiagnosticLocation::getStmt(const ExplodedNode *N) {
-  // We cannot place diagnostics on autosynthesized code.
-  // Put them onto the call site through which we jumped into autosynthesized
-  // code for the first time.
-  const LocationContext *LC = N->getLocationContext();
-  if (LC->getAnalysisDeclContext()->isBodyAutosynthesized()) {
-    // It must be a stack frame because we only autosynthesize functions.
-    return cast<StackFrameContext>(findTopAutosynthesizedParentContext(LC))
-        ->getCallSite();
-  }
-  // Otherwise, see if the node's program point directly points to a statement.
-  ProgramPoint P = N->getLocation();
-  if (auto SP = P.getAs<StmtPoint>())
-    return SP->getStmt();
-  if (auto BE = P.getAs<BlockEdge>())
-    return BE->getSrc()->getTerminatorStmt();
-  if (auto CE = P.getAs<CallEnter>())
-    return CE->getCallExpr();
-  if (auto CEE = P.getAs<CallExitEnd>())
-    return CEE->getCalleeContext()->getCallSite();
-  if (auto PIPP = P.getAs<PostInitializer>())
-    return PIPP->getInitializer()->getInit();
-  if (auto CEB = P.getAs<CallExitBegin>())
-    return CEB->getReturnStmt();
-  if (auto FEP = P.getAs<FunctionExitPoint>())
-    return FEP->getStmt();
-
-  return nullptr;
-}
-
-const Stmt *PathDiagnosticLocation::getNextStmt(const ExplodedNode *N) {
-  for (N = N->getFirstSucc(); N; N = N->getFirstSucc()) {
-    if (const Stmt *S = getStmt(N)) {
-      // Check if the statement is '?' or '&&'/'||'.  These are "merges",
-      // not actual statement points.
-      switch (S->getStmtClass()) {
-        case Stmt::ChooseExprClass:
-        case Stmt::BinaryConditionalOperatorClass:
-        case Stmt::ConditionalOperatorClass:
-          continue;
-        case Stmt::BinaryOperatorClass: {
-          BinaryOperatorKind Op = cast<BinaryOperator>(S)->getOpcode();
-          if (Op == BO_LAnd || Op == BO_LOr)
-            continue;
-          break;
-        }
-        default:
-          break;
-      }
-      // We found the statement, so return it.
-      return S;
-    }
-  }
-
-  return nullptr;
-}
-
-PathDiagnosticLocation
-PathDiagnosticLocation::createEndOfPath(const ExplodedNode *N) {
-  assert(N && "Cannot create a location with a null node.");
-  const Stmt *S = getStmt(N);
-  const LocationContext *LC = N->getLocationContext();
-  SourceManager &SM =
-      N->getState()->getStateManager().getContext().getSourceManager();
-
-  if (!S) {
-    // If this is an implicit call, return the implicit call point location.
-    if (Optional<PreImplicitCall> PIE = N->getLocationAs<PreImplicitCall>())
-      return PathDiagnosticLocation(PIE->getLocation(), SM);
-    if (auto FE = N->getLocationAs<FunctionExitPoint>()) {
-      if (const ReturnStmt *RS = FE->getStmt())
-        return PathDiagnosticLocation::createBegin(RS, SM, LC);
-    }
-    S = getNextStmt(N);
-  }
-
-  if (S) {
-    ProgramPoint P = N->getLocation();
-
-    // For member expressions, return the location of the '.' or '->'.
-    if (const auto *ME = dyn_cast<MemberExpr>(S))
-      return PathDiagnosticLocation::createMemberLoc(ME, SM);
-
-    // For binary operators, return the location of the operator.
-    if (const auto *B = dyn_cast<BinaryOperator>(S))
-      return PathDiagnosticLocation::createOperatorLoc(B, SM);
-
-    if (P.getAs<PostStmtPurgeDeadSymbols>())
-      return PathDiagnosticLocation::createEnd(S, SM, LC);
-
-    if (S->getBeginLoc().isValid())
-      return PathDiagnosticLocation(S, SM, LC);
-    return PathDiagnosticLocation(getValidSourceLocation(S, LC), SM);
-  }
-
-  return createDeclEnd(N->getLocationContext(), SM);
-}
-
 PathDiagnosticLocation PathDiagnosticLocation::createSingleLocation(
                                            const PathDiagnosticLocation &PDL) {
   FullSourceLoc L = PDL.asLocation();