[Object, MachO] Mark symbols from DATA and BSS sections as ST_Data

[kubamracek]

Summary

In MachOObjectFile::getSymbolType we currently always return SymbolRef::ST_Function for symbols from any section. In order for llvm-symbolizer to correctly symbolize Mach-O globals, symbols from data and BSS sections should return SymbolRef::ST_Data.

[samsonov]

LGTM

lib/Object/MachOObjectFile.cpp

404 ↗

(On Diff #39938)

Do you actually use this variable?

llvm/trunk/lib/Object/MachOObjectFile.cpp

SymbolRef::Type MachOObjectFile::getSymbolType(DataRefImpl Symb) const {
   // If this is a STAB debugging symbol, we can do nothing more.
   if (n_type & MachO::N_STAB)
     return SymbolRef::ST_Debug;
 
   switch (n_type & MachO::N_TYPE) {
     case MachO::N_UNDF :
       return SymbolRef::ST_Unknown;
     case MachO::N_SECT :
+      section_iterator Sec = *getSymbolSection(Symb);
+      if (Sec->isData() || Sec->isBSS())
+        return SymbolRef::ST_Data;
       return SymbolRef::ST_Function;
   }
   return SymbolRef::ST_Other;
 }
 
 uint32_t MachOObjectFile::getSymbolFlags(DataRefImpl DRI) const {
   MachO::nlist_base Entry = getSymbolTableEntryBase(this, DRI);
 

llvm/trunk/test/tools/llvm-objdump/X86/macho-symbol-table.test

 RUN: llvm-objdump -macho -t %p/Inputs/hello.obj.macho-x86_64 | FileCheck %s
 
 CHECK: SYMBOL TABLE:
-CHECK: 000000000000003b l     F __TEXT,__cstring  L_.str
-CHECK: 0000000000000068 l     F __TEXT,__eh_frame EH_frame0
+CHECK: 000000000000003b l       __TEXT,__cstring  L_.str
+CHECK: 0000000000000068 l       __TEXT,__eh_frame EH_frame0
 CHECK: 0000000000000000 g     F __TEXT,__text         _main
-CHECK: 0000000000000080 g     F __TEXT,__eh_frame _main.eh
+CHECK: 0000000000000080 g       __TEXT,__eh_frame _main.eh
 CHECK: 0000000000000000         *UND*                 _printf

llvm/trunk/tools/llvm-objdump/MachODump.cpp

for (unsigned SectIdx = 0; SectIdx != Sections.size(); SectIdx++) {
     // Disassemble symbol by symbol.
     for (unsigned SymIdx = 0; SymIdx != Symbols.size(); SymIdx++) {
       ErrorOr<StringRef> SymNameOrErr = Symbols[SymIdx].getName();
       if (std::error_code EC = SymNameOrErr.getError())
         report_fatal_error(EC.message());
       StringRef SymName = *SymNameOrErr;
 
       SymbolRef::Type ST = Symbols[SymIdx].getType();
-      if (ST != SymbolRef::ST_Function)
+      if (ST != SymbolRef::ST_Function && ST != SymbolRef::ST_Data)
         continue;
 
       // Make sure the symbol is defined in this section.
       bool containsSym = Sections[SectIdx].containsSymbol(Symbols[SymIdx]);
       if (!containsSym)
         continue;
 
       // If we are only disassembling one symbol see if this is that symbol.