[clangd] Dont run raw-lexer for OOB source locations

[kadircet]

Summary

We can get stale source locations from preamble, make sure we don't
access those locations without checking first.

Fixes https://github.com/clangd/clangd/issues/1636.

clang-tools-extra/clangd/Diagnostics.cpp

bool locationInRange(SourceLocation L, CharSourceRange R,
   if (!R.isValid() || M.getFileID(R.getBegin()) != M.getFileID(R.getEnd()) ||
       M.getFileID(R.getBegin()) != M.getFileID(L))
     return false;
   return L != R.getEnd() && M.isPointWithin(L, R.getBegin(), R.getEnd());
 }
 
 // Clang diags have a location (shown as ^) and 0 or more ranges (~~~~).
 // LSP needs a single range.
-Range diagnosticRange(const clang::Diagnostic &D, const LangOptions &L) {
+std::optional<Range> diagnosticRange(const clang::Diagnostic &D,
+                                     const LangOptions &L) {
   auto &M = D.getSourceManager();
   auto PatchedRange = [&M](CharSourceRange &R) {
     R.setBegin(translatePreamblePatchLocation(R.getBegin(), M));
     R.setEnd(translatePreamblePatchLocation(R.getEnd(), M));
     return R;
   };
   auto Loc = M.getFileLoc(D.getLocation());
   for (const auto &CR : D.getRanges()) {
     auto R = Lexer::makeFileCharRange(CR, M, L);
     if (locationInRange(Loc, R, M))
       return halfOpenToRange(M, PatchedRange(R));
   }
   // The range may be given as a fixit hint instead.
   for (const auto &F : D.getFixItHints()) {
     auto R = Lexer::makeFileCharRange(F.RemoveRange, M, L);
     if (locationInRange(Loc, R, M))
       return halfOpenToRange(M, PatchedRange(R));
   }
+  // Source locations from stale preambles might become OOB.
+  // FIXME: These diagnostics might point to wrong locations even when they're
+  // not OOB.
+  auto [FID, Offset] = M.getDecomposedLoc(Loc);
+  if (Offset > M.getBufferData(FID).size())
+    return std::nullopt;
   // If the token at the location is not a comment, we use the token.
   // If we can't get the token at the location, fall back to using the location
   auto R = CharSourceRange::getCharRange(Loc);
   Token Tok;
-  if (!Lexer::getRawToken(Loc, Tok, M, L, true) && Tok.isNot(tok::comment)) {
+  if (!Lexer::getRawToken(Loc, Tok, M, L, true) && Tok.isNot(tok::comment))
     R = CharSourceRange::getTokenRange(Tok.getLocation(), Tok.getEndLoc());
-  }
   return halfOpenToRange(M, PatchedRange(R));
 }
 
 // Try to find a location in the main-file to report the diagnostic D.
 // Returns a description like "in included file", or nullptr on failure.
 const char *getMainFileRange(const Diag &D, const SourceManager &SM,
                              SourceLocation DiagLoc, Range &R) {
   // Look for a note in the main file indicating template instantiation.
void StoreDiags::HandleDiagnostic(DiagnosticsEngine::Level DiagLevel,
   SourceManager &SM = Info.getSourceManager();
 
   auto FillDiagBase = [&](DiagBase &D) {
     fillNonLocationData(DiagLevel, Info, D);
 
     SourceLocation PatchLoc =
         translatePreamblePatchLocation(Info.getLocation(), SM);
     D.InsideMainFile = isInsideMainFile(PatchLoc, SM);
-    D.Range = diagnosticRange(Info, *LangOpts);
+    if (auto DRange = diagnosticRange(Info, *LangOpts))
+      D.Range = *DRange;
+    else
+      D.Severity = DiagnosticsEngine::Ignored;
     auto FID = SM.getFileID(Info.getLocation());
     if (const auto FE = SM.getFileEntryRefForID(FID)) {
       D.File = FE->getName().str();
       D.AbsFile = getCanonicalPath(*FE, SM);
     }
     D.ID = Info.getID();
     return D;
   };

clang-tools-extra/clangd/unittests/PreambleTests.cpp

 #define $note[[BAR]] 1
 #define $main[[BAR]] 2)");
     Annotations NewCode(R"(
 #define BAZ 0
 #define BAR 1)");
     auto AST = createPatchedAST(Code.code(), NewCode.code());
     EXPECT_THAT(*AST->getDiagnostics(), IsEmpty());
   }
+  {
+    Annotations Code(R"(
+#ifndef FOO
+#define FOO
+void foo();
+#endif)");
+    // This code will emit a diagnostic for unterminated #ifndef (as stale
+    // preamble has the conditional but main file doesn't terminate it).
+    // We shouldn't emit any diagnotiscs (and shouldn't crash).
+    Annotations NewCode("");
+    auto AST = createPatchedAST(Code.code(), NewCode.code());
+    EXPECT_THAT(*AST->getDiagnostics(), IsEmpty());
+  }
+  {
+    Annotations Code(R"(
+#ifndef FOO
+#define FOO
+void foo();
+#endif)");
+    // This code will emit a diagnostic for unterminated #ifndef (as stale
+    // preamble has the conditional but main file doesn't terminate it).
+    // We shouldn't emit any diagnotiscs (and shouldn't crash).
+    // FIXME: Patch/ignore diagnostics in such cases.
+    Annotations NewCode(R"(
+i[[nt]] xyz;
+    )");
+    auto AST = createPatchedAST(Code.code(), NewCode.code());
+    EXPECT_THAT(
+        *AST->getDiagnostics(),
+        ElementsAre(Diag(NewCode.range(), "pp_unterminated_conditional")));
+  }
 }
 
 MATCHER_P2(Mark, Range, Text, "") {
   return std::tie(arg.Rng, arg.Trivia) == std::tie(Range, Text);
 }
 
 TEST(PreamblePatch, MacroAndMarkHandling) {
   Config Cfg;