Allow resolving headers from a PCH even after headers+PCH were moved to another path.

Store in PCH the directory that the PCH was originally created in.
If a header file is not found at the path that we expect it to be and the PCH file
was moved from its original location, try to resolve the file by assuming that
header+PCH were moved together and the header is in the same place relative to the PCH.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@125576 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/include/clang/Frontend/FrontendActions.h b/include/clang/Frontend/FrontendActions.h
index 9e68a3c..0f2637f 100644
--- a/include/clang/Frontend/FrontendActions.h
+++ b/include/clang/Frontend/FrontendActions.h
@@ -89,6 +89,7 @@
   static bool ComputeASTConsumerArguments(CompilerInstance &CI,
                                           llvm::StringRef InFile,
                                           std::string &Sysroot,
+                                          std::string &OutputFile,
                                           llvm::raw_ostream *&OS,
                                           bool &Chaining);
 };
diff --git a/include/clang/Serialization/ASTBitCodes.h b/include/clang/Serialization/ASTBitCodes.h
index b117322..5c5c893 100644
--- a/include/clang/Serialization/ASTBitCodes.h
+++ b/include/clang/Serialization/ASTBitCodes.h
@@ -353,7 +353,10 @@
       CUDA_SPECIAL_DECL_REFS = 39,
       
       /// \brief Record code for header search information.
-      HEADER_SEARCH_TABLE = 40
+      HEADER_SEARCH_TABLE = 40,
+
+      /// \brief The directory that the PCH was originally created in.
+      ORIGINAL_PCH_DIR = 41
     };
 
     /// \brief Record types used within a source manager block.
diff --git a/include/clang/Serialization/ASTReader.h b/include/clang/Serialization/ASTReader.h
index 015ccc9..8cb0206 100644
--- a/include/clang/Serialization/ASTReader.h
+++ b/include/clang/Serialization/ASTReader.h
@@ -613,6 +613,13 @@
   /// AST file.
   std::string ActualOriginalFileName;
 
+  /// \brief The directory that the PCH was originally created in. Used to
+  /// allow resolving headers even after headers+PCH was moved to a new path.
+  std::string OriginalDir;
+
+  /// \brief The directory that the PCH we are reading is stored in.
+  std::string CurrentDir;
+
   /// \brief Whether this precompiled header is a relocatable PCH file.
   bool RelocatablePCH;
 
diff --git a/include/clang/Serialization/ASTWriter.h b/include/clang/Serialization/ASTWriter.h
index 5d37eb5..7da0dfe 100644
--- a/include/clang/Serialization/ASTWriter.h
+++ b/include/clang/Serialization/ASTWriter.h
@@ -309,7 +309,8 @@
   void WriteSubStmt(Stmt *S);
 
   void WriteBlockInfoBlock();
-  void WriteMetadata(ASTContext &Context, const char *isysroot);
+  void WriteMetadata(ASTContext &Context, const char *isysroot,
+                     const std::string &OutputFile);
   void WriteLanguageOptions(const LangOptions &LangOpts);
   void WriteStatCache(MemorizeStatCalls &StatCalls);
   void WriteSourceManagerBlock(SourceManager &SourceMgr,
@@ -339,7 +340,7 @@
   void WriteDecl(ASTContext &Context, Decl *D);
 
   void WriteASTCore(Sema &SemaRef, MemorizeStatCalls *StatCalls,
-                    const char* isysroot);
+                    const char* isysroot, const std::string &OutputFile);
   void WriteASTChain(Sema &SemaRef, MemorizeStatCalls *StatCalls,
                      const char* isysroot);
   
@@ -368,6 +369,7 @@
   /// \param PPRec Record of the preprocessing actions that occurred while
   /// preprocessing this file, e.g., macro instantiations
   void WriteAST(Sema &SemaRef, MemorizeStatCalls *StatCalls,
+                const std::string &OutputFile,
                 const char* isysroot);
 
   /// \brief Emit a source location.
@@ -575,6 +577,7 @@
 /// precompiled header from the parsed source code.
 class PCHGenerator : public SemaConsumer {
   const Preprocessor &PP;
+  std::string OutputFile;
   const char *isysroot;
   llvm::raw_ostream *Out;
   Sema *SemaPtr;
@@ -589,7 +592,7 @@
   const ASTWriter &getWriter() const { return Writer; }
 
 public:
-  PCHGenerator(const Preprocessor &PP, bool Chaining,
+  PCHGenerator(const Preprocessor &PP, const std::string &OutputFile, bool Chaining,
                const char *isysroot, llvm::raw_ostream *Out);
   virtual void InitializeSema(Sema &S) { SemaPtr = &S; }
   virtual void HandleTranslationUnit(ASTContext &Ctx);
diff --git a/lib/Frontend/ASTUnit.cpp b/lib/Frontend/ASTUnit.cpp
index 00902e8..0370c2e 100644
--- a/lib/Frontend/ASTUnit.cpp
+++ b/lib/Frontend/ASTUnit.cpp
@@ -653,7 +653,7 @@
   PrecompilePreambleConsumer(ASTUnit &Unit,
                              const Preprocessor &PP, bool Chaining,
                              const char *isysroot, llvm::raw_ostream *Out)
-    : PCHGenerator(PP, Chaining, isysroot, Out), Unit(Unit) { }
+    : PCHGenerator(PP, "", Chaining, isysroot, Out), Unit(Unit) { }
 
   virtual void HandleTopLevelDecl(DeclGroupRef D) {
     for (DeclGroupRef::iterator it = D.begin(), ie = D.end(); it != ie; ++it) {
@@ -700,9 +700,11 @@
   virtual ASTConsumer *CreateASTConsumer(CompilerInstance &CI,
                                          llvm::StringRef InFile) {
     std::string Sysroot;
+    std::string OutputFile;
     llvm::raw_ostream *OS = 0;
     bool Chaining;
-    if (GeneratePCHAction::ComputeASTConsumerArguments(CI, InFile, Sysroot, 
+    if (GeneratePCHAction::ComputeASTConsumerArguments(CI, InFile, Sysroot,
+                                                       OutputFile,
                                                        OS, Chaining))
       return 0;
     
@@ -2008,7 +2010,7 @@
   std::vector<unsigned char> Buffer;
   llvm::BitstreamWriter Stream(Buffer);
   ASTWriter Writer(Stream);
-  Writer.WriteAST(getSema(), 0, 0);
+  Writer.WriteAST(getSema(), 0, std::string(), 0);
   
   // Write the generated bitstream to "Out".
   if (!Buffer.empty())
diff --git a/lib/Frontend/FrontendActions.cpp b/lib/Frontend/FrontendActions.cpp
index 0ffea20..e454321 100644
--- a/lib/Frontend/FrontendActions.cpp
+++ b/lib/Frontend/FrontendActions.cpp
@@ -83,19 +83,21 @@
 ASTConsumer *GeneratePCHAction::CreateASTConsumer(CompilerInstance &CI,
                                                   llvm::StringRef InFile) {
   std::string Sysroot;
+  std::string OutputFile;
   llvm::raw_ostream *OS = 0;
   bool Chaining;
-  if (ComputeASTConsumerArguments(CI, InFile, Sysroot, OS, Chaining))
+  if (ComputeASTConsumerArguments(CI, InFile, Sysroot, OutputFile, OS, Chaining))
     return 0;
 
   const char *isysroot = CI.getFrontendOpts().RelocatablePCH ?
                              Sysroot.c_str() : 0;  
-  return new PCHGenerator(CI.getPreprocessor(), Chaining, isysroot, OS);
+  return new PCHGenerator(CI.getPreprocessor(), OutputFile, Chaining, isysroot, OS);
 }
 
 bool GeneratePCHAction::ComputeASTConsumerArguments(CompilerInstance &CI,
                                                     llvm::StringRef InFile,
                                                     std::string &Sysroot,
+                                                    std::string &OutputFile,
                                                     llvm::raw_ostream *&OS,
                                                     bool &Chaining) {
   Sysroot = CI.getHeaderSearchOpts().Sysroot;
@@ -111,6 +113,7 @@
   if (!OS)
     return true;
 
+  OutputFile = CI.getFrontendOpts().OutputFile;
   Chaining = CI.getInvocation().getFrontendOpts().ChainedPCH &&
              !CI.getPreprocessorOpts().ImplicitPCHInclude.empty();
   return false;
diff --git a/lib/Serialization/ASTReader.cpp b/lib/Serialization/ASTReader.cpp
index 5e3fa13..3dcc62e 100644
--- a/lib/Serialization/ASTReader.cpp
+++ b/lib/Serialization/ASTReader.cpp
@@ -1181,6 +1181,39 @@
   }
 }
 
+/// \brief If a header file is not found at the path that we expect it to be
+/// and the PCH file was moved from its original location, try to resolve the
+/// file by assuming that header+PCH were moved together and the header is in
+/// the same place relative to the PCH.
+static std::string
+resolveFileRelativeToOriginalDir(const std::string &Filename,
+                                 const std::string &OriginalDir,
+                                 const std::string &CurrDir) {
+  assert(OriginalDir != CurrDir &&
+         "No point trying to resolve the file if the PCH dir didn't change");
+  using namespace llvm::sys;
+  llvm::SmallString<128> filePath(Filename);
+  fs::make_absolute(filePath);
+  assert(path::is_absolute(OriginalDir));
+  llvm::SmallString<128> currPCHPath(CurrDir);
+
+  path::const_iterator fileDirI = path::begin(path::parent_path(filePath)),
+                       fileDirE = path::end(path::parent_path(filePath));
+  path::const_iterator origDirI = path::begin(OriginalDir),
+                       origDirE = path::end(OriginalDir);
+  // Skip the common path components from filePath and OriginalDir.
+  while (fileDirI != fileDirE && origDirI != origDirE &&
+         *fileDirI == *origDirI) {
+    ++fileDirI;
+    ++origDirI;
+  }
+  for (; origDirI != origDirE; ++origDirI)
+    path::append(currPCHPath, "..");
+  path::append(currPCHPath, fileDirI, fileDirE);
+  path::append(currPCHPath, path::filename(Filename));
+  return currPCHPath.str();
+}
+
 /// \brief Get a cursor that's correctly positioned for reading the source
 /// location entry with the given ID.
 ASTReader::PerFileData *ASTReader::SLocCursorForID(unsigned ID) {
@@ -1235,6 +1268,14 @@
     std::string Filename(BlobStart, BlobStart + BlobLen);
     MaybeAddSystemRootToFilename(Filename);
     const FileEntry *File = FileMgr.getFile(Filename);
+    if (File == 0 && !OriginalDir.empty() && !CurrentDir.empty() &&
+        OriginalDir != CurrentDir) {
+      std::string resolved = resolveFileRelativeToOriginalDir(Filename,
+                                                              OriginalDir,
+                                                              CurrentDir);
+      if (!resolved.empty())
+        File = FileMgr.getFile(resolved);
+    }
     if (File == 0)
       File = FileMgr.getVirtualFile(Filename, (off_t)Record[4],
                                     (time_t)Record[5]);
@@ -2179,6 +2220,12 @@
       MaybeAddSystemRootToFilename(OriginalFileName);
       break;
 
+    case ORIGINAL_PCH_DIR:
+      // The primary AST will be the last to get here, so it will be the one
+      // that's used.
+      OriginalDir.assign(BlobStart, BlobLen);
+      break;
+
     case VERSION_CONTROL_BRANCH_REVISION: {
       const std::string &CurBranch = getClangFullRepositoryVersion();
       llvm::StringRef ASTBranch(BlobStart, BlobLen);
@@ -2393,6 +2440,11 @@
   // Set the AST file name.
   F.FileName = FileName;
 
+  if (FileName != "-") {
+    CurrentDir = llvm::sys::path::parent_path(FileName);
+    if (CurrentDir.empty()) CurrentDir = ".";
+  }
+
   // Open the AST file.
   //
   // FIXME: This shouldn't be here, we should just take a raw_ostream.
diff --git a/lib/Serialization/ASTWriter.cpp b/lib/Serialization/ASTWriter.cpp
index f1af835..cb29973 100644
--- a/lib/Serialization/ASTWriter.cpp
+++ b/lib/Serialization/ASTWriter.cpp
@@ -899,7 +899,8 @@
 }
 
 /// \brief Write the AST metadata (e.g., i686-apple-darwin9).
-void ASTWriter::WriteMetadata(ASTContext &Context, const char *isysroot) {
+void ASTWriter::WriteMetadata(ASTContext &Context, const char *isysroot,
+                              const std::string &OutputFile) {
   using namespace llvm;
 
   // Metadata
@@ -947,6 +948,23 @@
     Stream.EmitRecordWithBlob(FileAbbrevCode, Record, MainFileNameStr);
   }
 
+  // Original PCH directory
+  if (!OutputFile.empty() && OutputFile != "-") {
+    BitCodeAbbrev *Abbrev = new BitCodeAbbrev();
+    Abbrev->Add(BitCodeAbbrevOp(ORIGINAL_PCH_DIR));
+    Abbrev->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Blob)); // File name
+    unsigned AbbrevCode = Stream.EmitAbbrev(Abbrev);
+
+    llvm::SmallString<128> OutputPath(OutputFile);
+
+    llvm::sys::fs::make_absolute(OutputPath);
+    StringRef origDir = llvm::sys::path::parent_path(OutputPath);
+
+    RecordData Record;
+    Record.push_back(ORIGINAL_PCH_DIR);
+    Stream.EmitRecordWithBlob(AbbrevCode, Record, origDir);
+  }
+
   // Repository branch/version information.
   BitCodeAbbrev *RepoAbbrev = new BitCodeAbbrev();
   RepoAbbrev->Add(BitCodeAbbrevOp(VERSION_CONTROL_BRANCH_REVISION));
@@ -2568,6 +2586,7 @@
 }
 
 void ASTWriter::WriteAST(Sema &SemaRef, MemorizeStatCalls *StatCalls,
+                         const std::string &OutputFile,
                          const char *isysroot) {
   // Emit the file header.
   Stream.Emit((unsigned)'C', 8);
@@ -2580,11 +2599,12 @@
   if (Chain)
     WriteASTChain(SemaRef, StatCalls, isysroot);
   else
-    WriteASTCore(SemaRef, StatCalls, isysroot);
+    WriteASTCore(SemaRef, StatCalls, isysroot, OutputFile);
 }
 
 void ASTWriter::WriteASTCore(Sema &SemaRef, MemorizeStatCalls *StatCalls,
-                             const char *isysroot) {
+                             const char *isysroot,
+                             const std::string &OutputFile) {
   using namespace llvm;
 
   ASTContext &Context = SemaRef.Context;
@@ -2692,7 +2712,7 @@
   // Write the remaining AST contents.
   RecordData Record;
   Stream.EnterSubblock(AST_BLOCK_ID, 5);
-  WriteMetadata(Context, isysroot);
+  WriteMetadata(Context, isysroot, OutputFile);
   WriteLanguageOptions(Context.getLangOptions());
   if (StatCalls && !isysroot)
     WriteStatCache(*StatCalls);
@@ -2827,7 +2847,7 @@
 
   RecordData Record;
   Stream.EnterSubblock(AST_BLOCK_ID, 5);
-  WriteMetadata(Context, isysroot);
+  WriteMetadata(Context, isysroot, "");
   if (StatCalls && !isysroot)
     WriteStatCache(*StatCalls);
   // FIXME: Source manager block should only write new stuff, which could be
diff --git a/lib/Serialization/GeneratePCH.cpp b/lib/Serialization/GeneratePCH.cpp
index 3b4f869..b8833ce 100644
--- a/lib/Serialization/GeneratePCH.cpp
+++ b/lib/Serialization/GeneratePCH.cpp
@@ -27,10 +27,11 @@
 using namespace clang;
 
 PCHGenerator::PCHGenerator(const Preprocessor &PP,
+                           const std::string &OutputFile,
                            bool Chaining,
                            const char *isysroot,
                            llvm::raw_ostream *OS)
-  : PP(PP), isysroot(isysroot), Out(OS), SemaPtr(0),
+  : PP(PP), OutputFile(OutputFile), isysroot(isysroot), Out(OS), SemaPtr(0),
     StatCalls(0), Stream(Buffer), Writer(Stream), Chaining(Chaining) {
   // Install a stat() listener to keep track of all of the stat()
   // calls.
@@ -50,7 +51,7 @@
   
   // Emit the PCH file
   assert(SemaPtr && "No Sema?");
-  Writer.WriteAST(*SemaPtr, StatCalls, isysroot);
+  Writer.WriteAST(*SemaPtr, StatCalls, OutputFile, isysroot);
 
   // Write the generated bitstream to "Out".
   Out->write((char *)&Buffer.front(), Buffer.size());
diff --git a/test/PCH/headersearch.cpp b/test/PCH/headersearch.cpp
new file mode 100644
index 0000000..c7849dc
--- /dev/null
+++ b/test/PCH/headersearch.cpp
@@ -0,0 +1,43 @@
+// Test reading of PCH with changed location of original input files,
+// i.e. invoking header search.
+
+// Generate the original files:
+// RUN: mkdir -p %t_orig/sub %t_orig/sub2
+// RUN: echo 'struct orig_sub{char c; int i; };' > %t_orig/sub/orig_sub.h
+// RUN: echo 'void orig_sub2_1();' > %t_orig/sub2/orig_sub2_1.h
+// RUN: echo '#include "orig_sub2_1.h"' > %t_orig/sub2/orig_sub2.h
+// RUN: echo 'template <typename T> void tf() { orig_sub2_1(); T::foo(); }' >> %t_orig/sub2/orig_sub2.h
+// RUN: echo 'void foo() {}' > %t_orig/tmp2.h
+// RUN: echo '#include "tmp2.h"' > %t_orig/all.h
+// RUN: echo '#include "sub/orig_sub.h"' >> %t_orig/all.h
+// RUN: echo '#include "orig_sub2.h"' >> %t_orig/all.h
+// RUN: echo 'int all();' >> %t_orig/all.h
+
+// Generate the PCH:
+// RUN: cd %t_orig && %clang_cc1 -x c++ -emit-pch -o all.h.pch -Isub2 all.h
+// RUN: rm -rf %t_moved
+// RUN: mv %t_orig %t_moved
+
+// Check diagnostic with location in original source:
+// RUN: %clang_cc1 -include-pch all.h.pch -I%t_moved -I%t_moved/sub2 -Wpadded -emit-obj -o %t.o %s 2> %t.stderr
+// RUN: grep 'struct orig_sub' %t.stderr
+
+// Check diagnostic with 2nd location in original source:
+// RUN: not %clang_cc1 -DREDECL -include-pch all.h.pch -I%t_moved -I%t_moved/sub2 -emit-obj -o %t.o %s 2> %t.stderr
+// RUN: grep 'void foo' %t.stderr
+
+// Check diagnostic with instantiation location in original source:
+// RUN: not %clang_cc1 -DINSTANTIATION -include-pch all.h.pch -I%t_moved -I%t_moved/sub2 -emit-obj -o %t.o %s 2> %t.stderr
+// RUN: grep 'orig_sub2_1' %t.stderr
+
+void qq(orig_sub*) {all();}
+
+#ifdef REDECL
+float foo() {return 0;}
+#endif
+
+#ifdef INSTANTIATION
+void f() {
+  tf<int>();
+}
+#endif