Add -stats-stats option

[MatzeB]

Summary

This option behaves in a similar spirit as -save-temps and writes out
llvm statistics in json format.

[aprantl]

include/clang/Basic/DiagnosticFrontendKinds.td

111

statstic"s"? not sure.

lib/Driver/Tools.cpp

6104

ditto :-)

test/Driver/save-stats.c

13

This is up to taste, but just accepting {{.}} as the path separator would be sufficient IMO and might be more readable.

[bruno]

lib/Driver/Tools.cpp

6113

Why removing StatsFile here? IIUC, at this point StatsFile is still the same as the output (if it's a file).

[MatzeB]

lib/Driver/Tools.cpp

6113

a) It behaves like -save-temps=obj (clang -save-temps=obj foo.c -o bar will give you foo.i, foo.o, ...;
b) This makes sense when you have multiple (clang -save-stats=obj foo.c bar.c -o baz) inputs for which you need multiple .stats filenames but you only have 1 output name
c) It magically works for -o - as well

[bruno]

Maybe add some docs to explain the new flags?

lib/Driver/Tools.cpp

6113

Ok, I see now, I misread remove_filename as remove

test/Driver/save-stats.c

2

Is -save-stats == -save-stats=cwd? It doesn't seem so by looking at lib/Driver/Tools.cpp. Need test for the diag::err_drv_invalid_value as well.

13

+1 to Adrian's suggestion

[MatzeB]

Thanks for the reviews.

Addressed comments.

[MatzeB]

test/Driver/save-stats.c

2

Yes, that aspect is handled by the options parser, look at the changes in Options.td

[bruno]

Thanks Matthias

LGTM!

[MatzeB]

test/Driver/save-stats.c

13

For the record: Switching to {{.}} in the final commit broke the windows bots, because they display an escaped slash here (= two backslash characters)...

[aprantl]

Awesome. Sorry!

docs/CommandGuide/clang.rst

 .. option:: -print-search-dirs
 
   Print the paths used for finding libraries and programs.
 
 .. option:: -save-temps
 
   Save intermediate compilation results.
 
+.. option:: -save-stats, -save-stats=cwd, -save-stats=obj
+
+  Save internal code generation (LLVM) statistics to a file in the current
+  directory (:option:`-save-stats`/:option:`-save-stats=cwd`) or the directory
+  of the output file (:option:`-save-state=obj`).
+
 .. option:: -integrated-as, -no-integrated-as
 
   Used to enable and disable, respectively, the use of the integrated
   assembler. Whether the integrated assembler is on by default is target
   dependent.
 
 .. option:: -time
 

include/clang/Basic/DiagnosticFrontendKinds.td

 def err_fe_pth_file_has_no_source_header : Error<
     "PTH file '%0' does not designate an original source header file for -include-pth">;
 def warn_fe_macro_contains_embedded_newline : Warning<
     "macro '%0' contains embedded newline; text after the newline is ignored">;
 def warn_fe_cc_print_header_failure : Warning<
     "unable to open CC_PRINT_HEADERS file: %0 (using stderr)">;
 def warn_fe_cc_log_diagnostics_failure : Warning<
     "unable to open CC_LOG_DIAGNOSTICS file: %0 (using stderr)">;
+def warn_fe_unable_to_open_stats_file : Warning<
+    "unable to open statistics output file '%0': '%1'">;
 def err_fe_no_pch_in_dir : Error<
     "no suitable precompiled header file found in directory '%0'">;
 def err_fe_action_not_available : Error<
     "action %0 not compiled in">;
 
 def warn_fe_serialized_diag_merge_failure : Warning<
     "unable to merge a subprocess's serialized diagnostics">,
     InGroup<SerializedDiagnostics>;

include/clang/Driver/CC1Options.td

def arcmt_check : Flag<["-"], "arcmt-check">,
   HelpText<"Check for ARC migration issues that need manual handling">;
 def arcmt_modify : Flag<["-"], "arcmt-modify">,
   HelpText<"Apply modifications to files to conform to ARC">;
 def arcmt_migrate : Flag<["-"], "arcmt-migrate">,
   HelpText<"Apply modifications and produces temporary files that conform to ARC">;
 
 def print_stats : Flag<["-"], "print-stats">,
   HelpText<"Print performance metrics and statistics">;
+def stats_file : Joined<["-"], "stats-file=">,
+  HelpText<"Filename to write statistics to">;
 def fdump_record_layouts : Flag<["-"], "fdump-record-layouts">,
   HelpText<"Dump record layout information">;
 def fdump_record_layouts_simple : Flag<["-"], "fdump-record-layouts-simple">,
   HelpText<"Dump record layout information in a simple form used for testing">;
 def fix_what_you_can : Flag<["-"], "fix-what-you-can">,
   HelpText<"Apply fix-it advice even in the presence of unfixable errors">;
 def fix_only_warnings : Flag<["-"], "fix-only-warnings">,
   HelpText<"Apply fix-it advice only for warnings, not errors">;

include/clang/Driver/Options.td

 def rtlib_EQ : Joined<["-", "--"], "rtlib=">,
   HelpText<"Compiler runtime library to use">;
 def r : Flag<["-"], "r">, Flags<[LinkerInput,NoArgumentUnused]>;
 def save_temps_EQ : Joined<["-", "--"], "save-temps=">, Flags<[DriverOption]>,
   HelpText<"Save intermediate compilation results.">;
 def save_temps : Flag<["-", "--"], "save-temps">, Flags<[DriverOption]>,
   Alias<save_temps_EQ>, AliasArgs<["cwd"]>,
   HelpText<"Save intermediate compilation results">;
+def save_stats_EQ : Joined<["-", "--"], "save-stats=">, Flags<[DriverOption]>,
+  HelpText<"Save llvm statistics.">;
+def save_stats : Flag<["-", "--"], "save-stats">, Flags<[DriverOption]>,
+  Alias<save_stats_EQ>, AliasArgs<["cwd"]>,
+  HelpText<"Save llvm statistics.">;
 def via_file_asm : Flag<["-", "--"], "via-file-asm">, InternalDebugOpt,
   HelpText<"Write assembly to file for input to assemble jobs">;
 def sectalign : MultiArg<["-"], "sectalign", 3>;
 def sectcreate : MultiArg<["-"], "sectcreate", 3>;
 def sectobjectsymbols : MultiArg<["-"], "sectobjectsymbols", 2>;
 def sectorder : MultiArg<["-"], "sectorder", 3>;
 def seg1addr : JoinedOrSeparate<["-"], "seg1addr">;
 def seg__addr__table__filename : Separate<["-"], "seg_addr_table_filename">;

include/clang/Frontend/FrontendOptions.h

public:
 
   /// \brief Auxiliary triple for CUDA compilation.
   std::string AuxTriple;
 
   /// \brief If non-empty, search the pch input file as it was a header
   // included by this file.
   std::string FindPchSource;
 
+  /// Filename to write statistics to.
+  std::string StatsFile;
+
 public:
   FrontendOptions() :
     DisableFree(false), RelocatablePCH(false), ShowHelp(false),
     ShowStats(false), ShowTimers(false), ShowVersion(false),
     FixWhatYouCan(false), FixOnlyWarnings(false), FixAndRecompile(false),
     FixToTemporaries(false), ARCMTMigrateEmitARCErrors(false),
     SkipFunctionBodies(false), UseGlobalModuleIndex(true),
     GenerateGlobalModuleIndex(true), ASTDumpDecls(false), ASTDumpLookups(false),

lib/Driver/Tools.cpp

#endif
 
   // Turn -fplugin=name.so into -load name.so
   for (const Arg *A : Args.filtered(options::OPT_fplugin_EQ)) {
     CmdArgs.push_back("-load");
     CmdArgs.push_back(A->getValue());
     A->claim();
   }
 
+  // Setup statistics file output.
+  if (const Arg *A = Args.getLastArg(options::OPT_save_stats_EQ)) {
+    StringRef SaveStats = A->getValue();
+
+    SmallString<128> StatsFile;
+    bool DoSaveStats = false;
+    if (SaveStats == "obj") {
+      if (Output.isFilename()) {
+        StatsFile.assign(Output.getFilename());
+        llvm::sys::path::remove_filename(StatsFile);
+      }
+      DoSaveStats = true;
+    } else if (SaveStats == "cwd") {
+      DoSaveStats = true;
+    } else {
+      D.Diag(diag::err_drv_invalid_value) << A->getAsString(Args) << SaveStats;
+    }
+
+    if (DoSaveStats) {
+      StringRef BaseName = llvm::sys::path::filename(Input.getBaseInput());
+      llvm::sys::path::append(StatsFile, BaseName);
+      llvm::sys::path::replace_extension(StatsFile, "stats");
+      CmdArgs.push_back(Args.MakeArgString(Twine("-stats-file=") +
+                                           StatsFile));
+    }
+  }
+
   // Forward -Xclang arguments to -cc1, and -mllvm arguments to the LLVM option
   // parser.
   Args.AddAllArgValues(CmdArgs, options::OPT_Xclang);
   for (const Arg *A : Args.filtered(options::OPT_mllvm)) {
     A->claim();
 
     // We translate this by hand to the -cc1 argument, since nightly test uses
     // it and developers have been trained to spell it with -mllvm.

lib/Frontend/CompilerInstance.cpp

bool CompilerInstance::ExecuteAction(FrontendAction &Act) {
   if (getHeaderSearchOpts().Verbose)
     OS << "clang -cc1 version " CLANG_VERSION_STRING
        << " based upon " << BACKEND_PACKAGE_STRING
        << " default target " << llvm::sys::getDefaultTargetTriple() << "\n";
 
   if (getFrontendOpts().ShowTimers)
     createFrontendTimer();
 
-  if (getFrontendOpts().ShowStats)
+  if (getFrontendOpts().ShowStats || !getFrontendOpts().StatsFile.empty())
     llvm::EnableStatistics();
 
   for (const FrontendInputFile &FIF : getFrontendOpts().Inputs) {
     // Reset the ID tables if we are reusing the SourceManager and parsing
     // regular files.
     if (hasSourceManager() && !Act.isModelParsingAction())
       getSourceManager().clearIDTables();
 
if (getDiagnosticOpts().ShowCarets) {
     if (NumWarnings && NumErrors)
       OS << " and ";
     if (NumErrors)
       OS << NumErrors << " error" << (NumErrors == 1 ? "" : "s");
     if (NumWarnings || NumErrors)
       OS << " generated.\n";
   }
 
-  if (getFrontendOpts().ShowStats && hasFileManager()) {
+  if (getFrontendOpts().ShowStats) {
+    if (hasFileManager()) {
       getFileManager().PrintStats();
-    OS << "\n";
+      OS << '\n';
+    }
+    llvm::PrintStatistics(OS);
+  }
+  StringRef StatsFile = getFrontendOpts().StatsFile;
+  if (!StatsFile.empty()) {
+    std::error_code EC;
+    auto StatS = llvm::make_unique<llvm::raw_fd_ostream>(StatsFile, EC,
+                                                         llvm::sys::fs::F_Text);
+    if (EC) {
+      getDiagnostics().Report(diag::warn_fe_unable_to_open_stats_file)
+          << StatsFile << EC.message();
+    } else {
+      llvm::PrintStatisticsJSON(*StatS);
+    }
   }
 
   return !getDiagnostics().getClient()->getNumErrors();
 }
 
 /// \brief Determine the appropriate source input kind based on language
 /// options.
 static InputKind getSourceInputKindFromOptions(const LangOptions &LangOpts) {

lib/Frontend/CompilerInvocation.cpp

static InputKind ParseFrontendArgs(FrontendOptions &Opts, ArgList &Args,
   Opts.CodeCompleteOpts.IncludeBriefComments
     = Args.hasArg(OPT_code_completion_brief_comments);
 
   Opts.OverrideRecordLayoutsFile
     = Args.getLastArgValue(OPT_foverride_record_layout_EQ);
   Opts.AuxTriple =
       llvm::Triple::normalize(Args.getLastArgValue(OPT_aux_triple));
   Opts.FindPchSource = Args.getLastArgValue(OPT_find_pch_source_EQ);
+  Opts.StatsFile = Args.getLastArgValue(OPT_stats_file);
 
   if (const Arg *A = Args.getLastArg(OPT_arcmt_check,
                                      OPT_arcmt_modify,
                                      OPT_arcmt_migrate)) {
     switch (A->getOption().getID()) {
     default:
       llvm_unreachable("missed a case");
     case OPT_arcmt_check:

lib/StaticAnalyzer/Frontend/AnalysisConsumer.cpp

AnalysisConsumer(const Preprocessor &pp, const std::string &outdir,
     DigestAnalyzerOptions();
     if (Opts->PrintStats) {
       llvm::EnableStatistics();
       TUTotalTimer = new llvm::Timer("Analyzer Total Time");
     }
   }
 
   ~AnalysisConsumer() override {
-    if (Opts->PrintStats)
+    if (Opts->PrintStats) {
       delete TUTotalTimer;
+      llvm::PrintStatistics();
+    }
   }
 
   void DigestAnalyzerOptions() {
     if (Opts->AnalysisDiagOpt != PD_NONE) {
       // Create the PathDiagnosticConsumer.
       ClangDiagPathDiagConsumer *clangDiags =
           new ClangDiagPathDiagConsumer(PP.getDiagnostics());
       PathConsumers.push_back(clangDiags);

test/Driver/save-stats.c

+// RUN: %clang -target x86_64-apple-darwin -save-stats %s -### 2>&1 | FileCheck %s
+// RUN: %clang -target x86_64-apple-darwin -save-stats=cwd %s -### 2>&1 | FileCheck %s
+// CHECK: "-stats-file=save-stats.stats"
+// CHECK: "{{.*}}save-stats.c"
+
+// RUN: %clang -target x86_64-apple-darwin -S %s -### 2>&1 | FileCheck %s -check-prefix=NO-STATS
+// NO-STATS-NO: -stats-file
+// NO-STATS: "{{.*}}save-stats.c"
+// NO-STATS-NO: -stats-file
+
+// RUN: %clang -target x86_64-apple-darwin -save-stats=obj -c -o obj/dir/save-stats.o %s -### 2>&1 | FileCheck %s -check-prefix=CHECK-OBJ
+// CHECK-OBJ: "-stats-file=obj/dir{{.}}save-stats.stats"
+// CHECK-OBJ: "-o" "obj/dir{{.}}save-stats.o"
+
+// RUN: %clang -target x86_64-apple-darwin -save-stats=obj -c %s -### 2>&1 | FileCheck %s -check-prefix=CHECK-OBJ-NOO
+// CHECK-OBJ-NOO: "-stats-file=save-stats.stats"
+// CHECK-OBJ-NOO: "-o" "save-stats.o"
+
+// RUN: %clang -target x86_64-apple-darwin -save-stats=bla -c %s -### 2>&1 | FileCheck %s -check-prefix=CHECK-INVALID
+// CHECK-INVALID: invalid value 'bla' in '-save-stats=bla'

test/Frontend/stats-file.c

+// RUN: %clang_cc1 -emit-llvm -o /dev/null -stats-file=%t %s
+// RUN: FileCheck -input-file=%t %s
+// CHECK: {
+//  ... here come some json values ...
+// CHECK: }
+
+// RUN: %clang_cc1 -emit-llvm -o %t -stats-file=%S/doesnotexist/bla %s 2>&1 | FileCheck -check-prefix=OUTPUTFAIL %s
+// OUTPUTFAIL: warning: unable to open statistic output file '{{.*}}{{[/\\]}}doesnotexist{{[/\\]}}bla': '{{[Nn]}}o such file or directory'

test/Misc/warning-flags.c

(2) It prevents us adding new warnings to Clang that have no -W flag.  All
     new warnings should have -W flags.
 
     If you add a new warning without a flag, this test will fail.  To fix
     this test, simply add a warning group to that warning.
 
 
 The list of warnings below should NEVER grow.  It should gradually shrink to 0.
 
-CHECK: Warnings without flags (84):
+CHECK: Warnings without flags (85):
 CHECK-NEXT:   ext_excess_initializers
 CHECK-NEXT:   ext_excess_initializers_in_char_array_initializer
 CHECK-NEXT:   ext_expected_semi_decl_list
 CHECK-NEXT:   ext_explicit_specialization_storage_class
 CHECK-NEXT:   ext_initializer_string_for_char_array_too_long
 CHECK-NEXT:   ext_missing_declspec
 CHECK-NEXT:   ext_missing_whitespace_after_macro_name
 CHECK-NEXT:   ext_new_paren_array_nonconst

 CHECK-NEXT:   warn_dup_category_def
 CHECK-NEXT:   warn_duplicate_protocol_def
 CHECK-NEXT:   warn_enum_value_overflow
 CHECK-NEXT:   warn_expected_qualified_after_typename
 CHECK-NEXT:   warn_extraneous_char_constant
 CHECK-NEXT:   warn_fe_cc_log_diagnostics_failure
 CHECK-NEXT:   warn_fe_cc_print_header_failure
 CHECK-NEXT:   warn_fe_macro_contains_embedded_newline
+CHECK-NEXT:   warn_fe_unable_to_open_stats_file
 CHECK-NEXT:   warn_file_asm_volatile
 CHECK-NEXT:   warn_ignoring_ftabstop_value
 CHECK-NEXT:   warn_implements_nscopying
 CHECK-NEXT:   warn_incompatible_qualified_id
 CHECK-NEXT:   warn_inline_namespace_reopened_noninline
 CHECK-NEXT:   warn_invalid_asm_cast_lvalue
 CHECK-NEXT:   warn_maynot_respond
 CHECK-NEXT:   warn_method_param_redefinition