[lld-macho] Support -add_ast_path

[int3]

Summary

Swift builds seem to use it. All it requires is emitting the
corresponding paths as STABS.

Fixes llvm.org/PR49385.

[clayborg]

The N_AST symbol is used to specify a file on disk that contains the one and only Swift AST blob. This blob of data will be given back to the Swift compiler that is built into LLDB. I don't think this can be specified multiple times. Inline comments ask if

lld/MachO/Options.td

539–542

Does this allow specifying this option more than once? I don't believe this is supported as there should only be one N_AST.

lld/MachO/SyntheticSections.cpp

683

I don't think we can have more than one N_AST values emitted. Does ld64 supports multiple?

685

Does ld64 put the mod time of the AST file in the symbol table value? I can't remember, but it would be worth checking

[int3]

lld/MachO/SyntheticSections.cpp

683

ld64 does support multiple. Do you think we should emit just one value regardless?

685

It doesn't. (add-ast-path.s passes as-is if ld is substituted for %lld).

[clayborg]

Thanks for checking. As long as ld64 allows multiple, I am good.

lld/MachO/Config.h

struct Configuration {
   NamespaceKind namespaceKind = NamespaceKind::twolevel;
   UndefinedSymbolTreatment undefinedSymbolTreatment =
       UndefinedSymbolTreatment::error;
   llvm::MachO::HeaderFileType outputType;
   std::vector<llvm::StringRef> systemLibraryRoots;
   std::vector<llvm::StringRef> librarySearchPaths;
   std::vector<llvm::StringRef> frameworkSearchPaths;
   std::vector<llvm::StringRef> runtimePaths;
+  std::vector<std::string> astPaths;
   std::vector<Symbol *> explicitUndefineds;
   // There are typically very few custom segmentProtections, so use a vector
   // instead of a map.
   std::vector<SegmentProtection> segmentProtections;
 
   llvm::DenseMap<llvm::StringRef, SymbolPriorityEntry> priorities;
   SectionRenameMap sectionRenameMap;
   SegmentRenameMap segmentRenameMap;

lld/MachO/Driver.cpp

config->explicitUndefineds.push_back(symtab->addUndefined(
         arg->getValue(), /*file=*/nullptr, /*isWeakRef=*/false));
   }
 
   for (const Arg *arg : args.filtered(OPT_U))
     symtab->addDynamicLookup(arg->getValue());
 
   config->mapFile = args.getLastArgValue(OPT_map);
   config->outputFile = args.getLastArgValue(OPT_o, "a.out");
+  config->astPaths = args.getAllArgValues(OPT_add_ast_path);
   config->headerPad = args::getHex(args, OPT_headerpad, /*Default=*/32);
   config->headerPadMaxInstallNames =
       args.hasArg(OPT_headerpad_max_install_names);
   config->printEachFile = args.hasArg(OPT_t);
   config->printWhyLoad = args.hasArg(OPT_why_load);
   config->outputType = getOutputType(args);
   if (const Arg *arg = args.getLastArg(OPT_bundle_loader)) {
     if (config->outputType != MH_BUNDLE)

lld/MachO/Options.td

 def dependency_info : Separate<["-"], "dependency_info">,
     MetaVarName<"<path>">,
     HelpText<"Dump dependency info">,
     Group<grp_introspect>;
 def save_temps : Flag<["-"], "save-temps">,
     HelpText<"Save intermediate LTO compilation results">,
     Group<grp_introspect>;
 
-def grp_symtab : OptionGroup<"symtab">, HelpText<"SYMBOL TABLE OPTIMIZATIONS">;
+def grp_symtab : OptionGroup<"symtab">, HelpText<"SYMBOL TABLE">;
 
 def S : Flag<["-"], "S">,
     HelpText<"Strip debug information (STABS or DWARF) from the output">,
     Flags<[HelpHidden]>,
     Group<grp_symtab>;
 def x : Flag<["-"], "x">,
     HelpText<"Exclude non-global symbols from the output symbol table">,
     Flags<[HelpHidden]>,
     Group<grp_symtab>;
 def non_global_symbols_strip_list : Separate<["-"], "non_global_symbols_strip_list">,
     MetaVarName<"<path>">,
     HelpText<"Specify in <path> the non-global symbols that should be removed from the output symbol table">,
     Flags<[HelpHidden]>,
     Group<grp_symtab>;
 def non_global_symbols_no_strip_list : Separate<["-"], "non_global_symbols_no_strip_list">,
     MetaVarName<"<path>">,
     HelpText<"Specify in <path> the non-global symbols that should remain in the output symbol table">,
     Flags<[HelpHidden]>,
     Group<grp_symtab>;
 def oso_prefix : Separate<["-"], "oso_prefix">,
     MetaVarName<"<path>">,
     HelpText<"Remove the prefix <path> from OSO symbols in the debug map">,
     Flags<[HelpHidden]>,
     Group<grp_symtab>;
+def add_ast_path : Separate<["-"], "add_ast_path">,
+    MetaVarName<"<path>">,
+    HelpText<"AST paths will be emitted as STABS">,
+    Group<grp_symtab>;
 
 def grp_bitcode : OptionGroup<"bitcode">, HelpText<"BITCODE BUILD FLOW">;
 
 def bitcode_bundle : Flag<["-"], "bitcode_bundle">,
     HelpText<"Generate an embedded bitcode bundle in the __LLVM,__bundle section of the output">,
     Flags<[HelpHidden]>,
     Group<grp_bitcode>;
 def bitcode_hide_symbols : Flag<["-"], "bitcode_hide_symbols">,
def twolevel_namespace_hints : Flag<["-"], "twolevel_namespace_hints">,
     Group<grp_obsolete>;
 def slow_stubs : Flag<["-"], "slow_stubs">,
     HelpText<"This option is obsolete in ld64">,
     Flags<[HelpHidden]>,
     Group<grp_obsolete>;
 
 def grp_undocumented : OptionGroup<"undocumented">, HelpText<"UNDOCUMENTED">;
 
-def add_ast_path : Flag<["-"], "add_ast_path">,
-    HelpText<"This option is undocumented in ld64">,
-    Flags<[HelpHidden]>,
-    Group<grp_undocumented>;
 def add_linker_option : Flag<["-"], "add_linker_option">,
     HelpText<"This option is undocumented in ld64">,
     Flags<[HelpHidden]>,
     Group<grp_undocumented>;
 def add_source_version : Flag<["-"], "add_source_version">,
     HelpText<"This option is undocumented in ld64">,
     Flags<[HelpHidden]>,
     Group<grp_undocumented>;

lld/MachO/SyntheticSections.cpp

 
 void SymtabSection::emitEndFunStab(Defined *defined) {
   StabsEntry stab(N_FUN);
   stab.value = defined->size;
   stabs.emplace_back(std::move(stab));
 }
 
 void SymtabSection::emitStabs() {
+  for (const std::string &s : config->astPaths) {
+    StabsEntry astStab(N_AST);
+    astStab.strx = stringTableSection.addString(s);
+    stabs.emplace_back(std::move(astStab));
+  }
+
   std::vector<Defined *> symbolsNeedingStabs;
   for (const SymtabEntry &entry :
        concat<SymtabEntry>(localSymbols, externalSymbols)) {
     Symbol *sym = entry.sym;
     if (auto *defined = dyn_cast<Defined>(sym)) {
       if (defined->isAbsolute())
         continue;
       InputSection *isec = defined->isec;
void SymtabSection::emitStabs() {
   });
 
   // Emit STABS symbols so that dsymutil and/or the debugger can map address
   // regions in the final binary to the source and object files from which they
   // originated.
   InputFile *lastFile = nullptr;
   for (Defined *defined : symbolsNeedingStabs) {
     InputSection *isec = defined->isec;
-    ObjFile *file = dyn_cast<ObjFile>(isec->file);
-    assert(file);
+    ObjFile *file = cast<ObjFile>(isec->file);
 
     if (lastFile == nullptr || lastFile != file) {
       if (lastFile != nullptr)
         emitEndSourceStab();
       lastFile = file;
 
       emitBeginSourceStab(file->compileUnit);
       emitObjectFileStab(file);

lld/test/MachO/add-ast-path.s

+# REQUIRES: x86
+# RUN: llvm-mc -filetype=obj -triple=x86_64-apple-darwin %s -o %t.o
+# RUN: %lld -lSystem %t.o -o %t -add_ast_path asdf -add_ast_path fdsa
+# RUN: dsymutil -s %t | FileCheck %s
+# CHECK:      [     0] {{[0-9a-f]+}} 32 (N_AST        ) 00     0000   0000000000000000 'asdf'
+# CHECK-NEXT: [     1] {{[0-9a-f]+}} 32 (N_AST        ) 00     0000   0000000000000000 'fdsa'
+
+.globl _main
+
+_main:
+  ret