Create a metadata file named as the artifact id of the output file

Signed-off-by: Bharathi Seshadri <bharathi.seshadri@gmail.com>

llvm-project/clang/include/clang/Basic/CodeGenOptions.h

@@ -183,6 +183,10 @@ class CodeGenOptions : public CodeGenOptionsBase {
   /// The string containing the gitoid of the omnibor file.
   std::string RecordOmniBor;
 
+  /// The string containing the commandline for the Omnibor metadata,
+  /// if non-empty.
+  std::string OmniborCommandLine;
+
   /// List of dependent source/header files.
   /// This is shared with DependencyOuputOptions.
   /// This has same contents as Dependencies in DependencyCollector.

llvm-project/clang/include/clang/Driver/Options.td

@@ -4991,6 +4991,9 @@ def record_command_line : Separate<["-"], "record-command-line">,
 def record_omnibor: Separate<["-"], "record-omnibor">,
   HelpText<"The string to embed in the .note.omnibor section.">,
   MarshallingInfoString<CodeGenOpts<"RecordOmniBor">>;
+def omnibor_command_line : Separate<["-"], "omnibor-command-line">,
+  HelpText<"The command line string to be recorded in Omnibor metadata file.">,
+  MarshallingInfoString<CodeGenOpts<"OmniborCommandLine">>;
 def compress_debug_sections_EQ : Joined<["-", "--"], "compress-debug-sections=">,
     HelpText<"DWARF debug sections compression type">, Values<"none,zlib">,
     NormalizedValuesScope<"llvm::DebugCompressionType">, NormalizedValues<["None", "Z"]>,

llvm-project/clang/lib/CodeGen/CodeGenModule.cpp

@@ -6452,6 +6452,7 @@ static std::string ComputeSHA1OmniBorData(CodeGenModule &CGM,
   gitoid = Result.str();
 
   SmallString<128> gitoidPath(CGM.getCodeGenOpts().RecordOmniBor);
+  llvm::sys::path::append(gitoidPath, "objects");
   llvm::sys::path::append(gitoidPath, "gitoid_blob_sha1");
   llvm::sys::path::append(gitoidPath, gitoidHex.substr(0, 2));
   std::error_code EC;
@@ -6496,6 +6497,7 @@ static std::string ComputeSHA256OmniBorData(CodeGenModule &CGM,
   gitoid = Result.str();
 
   SmallString<128> gitoidPath(CGM.getCodeGenOpts().RecordOmniBor);
+  llvm::sys::path::append(gitoidPath, "objects");
   llvm::sys::path::append(gitoidPath, "gitoid_blob_sha256");
   llvm::sys::path::append(gitoidPath, gitoidHex.substr(0, 2));
   std::error_code EC;

llvm-project/clang/lib/Driver/ToolChains/Clang.cpp

@@ -6822,10 +6822,26 @@ void Clang::ConstructJob(Compilation &C, const JobAction &JA,
     OutputPath = env;
   }
   if (!OutputPath.str().empty()) {
-    llvm::sys::path::append(OutputPath, "objects");
     CmdArgs.push_back("-record-omnibor");
     CmdArgs.push_back(Args.MakeArgString(OutputPath));
-    // Create the .bom/objects directory
+
+    // Collect command line arguments as metadata
+    ArgStringList OriginalArgs;
+    for (const auto &Arg : Args)
+      Arg->render(Args, OriginalArgs);
+
+    SmallString<256> Flags;
+    EscapeSpacesAndBackslashes(Exec, Flags);
+    for (const char *OriginalArg : OriginalArgs) {
+      SmallString<128> EscapedArg;
+      EscapeSpacesAndBackslashes(OriginalArg, EscapedArg);
+      Flags += " ";
+      Flags += EscapedArg;
+    }
+    auto FlagsArgString = Args.MakeArgString(Flags);
+    CmdArgs.push_back("-omnibor-command-line");
+    CmdArgs.push_back(FlagsArgString);
+
     auto EC =
         llvm::sys::fs::create_directories(OutputPath, /*IgnoreExisting=*/true);
     if (EC)

llvm-project/clang/test/CodeGen/omnibor_test.c

@@ -1,5 +1,12 @@
 // RUN: rm -rf %t && mkdir %t
 
+// RUN: rm -f objects/gitoid_blob_sha1/0b/e9b91c3e456c910b32afc862ffd073b7c61d5f
+// RUN: env OMNIBOR_DIR= %clang -c -frecord-omnibor=%t %S/Inputs/omnibor.c -I%S/Inputs/omnibor.h
+// RUN: llvm-readelf -n omnibor.o | FileCheck --check-prefix=BOM_NOTE_SECTION %s
+// RUN: cat %t/objects/gitoid_blob_sha1/0b/e9b91c3e456c910b32afc862ffd073b7c61d5f | FileCheck --check-prefix=BOM_FILE_SHA1_CONTENTS %s
+// RUN: cat %t/objects/gitoid_blob_sha256/58/ed318e337ad6260f471c086e9c9985543c5fbd507c3b44924f5af37bd0ea96 | FileCheck --check-prefix=BOM_FILE_SHA256_CONTENTS %s
+// RUN: cat %t/metadata/llvm/17274786fd44c95cae7ccb2d0b29ca1738c3cbb1 | FileCheck --check-prefix=METADATA_CONTENTS %s
+
 // RUN: %clang -c -frecord-omnibor=%t -o %t/omnibor_1.o %S/Inputs/omnibor.c -I%S/Inputs/omnibor.h
 // RUN: llvm-readelf -n %t/omnibor_1.o | FileCheck --check-prefix=BOM_NOTE_SECTION %s
 // RUN: cat %t/objects/gitoid_blob_sha1/0b/e9b91c3e456c910b32afc862ffd073b7c61d5f | FileCheck --check-prefix=BOM_FILE_SHA1_CONTENTS %s
@@ -27,12 +34,6 @@
 // RUN: cat %t/objects/gitoid_blob_sha1/0b/e9b91c3e456c910b32afc862ffd073b7c61d5f | FileCheck --check-prefix=BOM_FILE_SHA1_CONTENTS %s
 // RUN: cat %t/objects/gitoid_blob_sha256/58/ed318e337ad6260f471c086e9c9985543c5fbd507c3b44924f5af37bd0ea96 | FileCheck --check-prefix=BOM_FILE_SHA256_CONTENTS %s
 
-// RUN: rm -f objects/gitoid_blob_sha1/0b/e9b91c3e456c910b32afc862ffd073b7c61d5f
-// RUN: env OMNIBOR_DIR= %clang -c -frecord-omnibor=%t %S/Inputs/omnibor.c -I%S/Inputs/omnibor.h
-// RUN: llvm-readelf -n omnibor.o | FileCheck --check-prefix=BOM_NOTE_SECTION %s
-// RUN: cat %t/objects/gitoid_blob_sha1/0b/e9b91c3e456c910b32afc862ffd073b7c61d5f | FileCheck --check-prefix=BOM_FILE_SHA1_CONTENTS %s
-// RUN: cat %t/objects/gitoid_blob_sha256/58/ed318e337ad6260f471c086e9c9985543c5fbd507c3b44924f5af37bd0ea96 | FileCheck --check-prefix=BOM_FILE_SHA256_CONTENTS %s
-
 // BOM_NOTE_SECTION: Displaying notes found in: .note.omnibor
 // BOM_NOTE_SECTION: Owner                Data size        Description
 // BOM_NOTE_SECTION: OMNIBOR               0x00000014       NT_GITOID_SHA1
@@ -47,3 +48,7 @@
 // BOM_FILE_SHA256_CONTENTS: gitoid:blob:sha256
 // BOM_FILE_SHA256_CONTENTS: blob 465be75836446b37f31c5db1f29a8689f37cd5625d4b30237877703fc1070f5e
 // BOM_FILE_SHA256_CONTENTS: blob 8ad020236bd23736a75699ba4e83f52593d61c4c9f8d5932b57fce011f832bf8
+
+// METADATA_CONTENTS: output: omnibor.o
+// METADATA_CONTENTS_NEXT: input {{.*}}/omnibor.c
+// METADATA_CONTENTS_NEXT: input {{.*}}/omnibor.h

llvm-project/clang/tools/driver/cc1_main.cpp

@@ -38,6 +38,8 @@
 #include "llvm/Support/ManagedStatic.h"
 #include "llvm/Support/Path.h"
 #include "llvm/Support/Process.h"
+#include "llvm/Support/SHA1.h"
+#include "llvm/Support/SHA256.h"
 #include "llvm/Support/Signals.h"
 #include "llvm/Support/TargetSelect.h"
 #include "llvm/Support/TimeProfiler.h"
@@ -181,6 +183,38 @@ static int PrintSupportedCPUs(std::string TargetStr) {
   return 0;
 }
 
+static std::string convertToHex(StringRef Input) {
+  static const char *const LUT = "0123456789abcdef";
+  size_t Length = Input.size();
+
+  std::string Output;
+  Output.reserve(2 * Length);
+  for (size_t i = 0; i < Length; ++i) {
+    const unsigned char c = Input[i];
+    Output.push_back(LUT[c >> 4]);
+    Output.push_back(LUT[c & 15]);
+  }
+  return Output;
+}
+
+static std::string getSHA1Hash(std::string Filename) {
+  llvm::ErrorOr<std::unique_ptr<llvm::MemoryBuffer>> fileBuf =
+      llvm::MemoryBuffer::getFile(Filename, /*IsText=*/true);
+  if (!fileBuf) {
+    // Diags.Report(diag::err_fe_error_opening) << Filename;
+    llvm::errs() << "\n error opening " << Filename;
+    return std::string();
+  }
+
+  llvm::SHA1 Hash;
+  std::string initData =
+      "blob " + std::to_string(fileBuf.get()->getBufferSize()) + '\0';
+  Hash.update(StringRef(initData));
+  Hash.update(fileBuf.get()->getBuffer());
+  auto Result = Hash.final();
+  return convertToHex(Result);
+}
+
 int cc1_main(ArrayRef<const char *> Argv, const char *Argv0, void *MainAddr) {
   ensureSufficientStack();
 
@@ -272,6 +306,50 @@ int cc1_main(ArrayRef<const char *> Argv, const char *Argv0, void *MainAddr) {
   // later errors use the default handling behavior instead.
   llvm::remove_fatal_error_handler();
 
+  if (!Clang->getCodeGenOpts().RecordOmniBor.empty()) {
+    SmallString<128> gitoidPath(Clang->getCodeGenOpts().RecordOmniBor);
+    llvm::sys::path::append(gitoidPath, "metadata/llvm");
+    auto EC =
+        llvm::sys::fs::create_directories(gitoidPath, /*IgnoreExisting=*/true);
+    if (EC)
+      llvm::errs() << "\nCannot create metadata file ";
+
+    std::vector<std::string> MetadataLines;
+    std::string outLine("\noutput: ");
+    outLine.append(Clang->getFrontendOpts().OutputFile);
+    MetadataLines.push_back(outLine);
+
+    if (Clang->getCodeGenOpts().BomDependencies->size()) {
+      std::vector<std::string> &Deps =
+          *(Clang->getCodeGenOpts().BomDependencies);
+      for (auto file : Deps) {
+        std::string Line = "\ninput " + file;
+        MetadataLines.push_back(Line);
+      }
+    }
+
+    // Metadata contents
+    std::string MetadataContents;
+    for (auto line : MetadataLines)
+      MetadataContents.append(line);
+
+    // Write command line
+    if (!Clang->getCodeGenOpts().OmniborCommandLine.empty()) {
+      MetadataContents.append("\nbuild_cmd: ");
+      MetadataContents.append(Clang->getCodeGenOpts().OmniborCommandLine);
+    }
+
+    // Write metadata
+    std::string artifact_id = getSHA1Hash(Clang->getFrontendOpts().OutputFile);
+    llvm::sys::path::append(gitoidPath, artifact_id);
+    std::string MetadataFile(gitoidPath);
+    llvm::raw_fd_ostream OSM(MetadataFile, EC, llvm::sys::fs::OF_TextWithCRLF);
+    if (EC) {
+      llvm::errs() << MetadataFile << EC.message();
+    }
+    OSM << MetadataContents;
+  }
+
   // When running with -disable-free, don't do any destruction or shutdown.
   if (Clang->getFrontendOpts().DisableFree) {
     llvm::BuryPointer(std::move(Clang));