Improve "llvm-nm -f sysv" output for Elf files

Specifically, compute and Print Type and Section columns.

Differential Revision: https://reviews.llvm.org/D58263

llvm-svn: 354833

llvm/include/llvm/Object/ELFObjectFile.h

@@ -41,6 +41,8 @@
 namespace llvm {
 namespace object {
 
+extern const llvm::EnumEntry<unsigned> ElfSymbolTypes[8];
+
 class elf_symbol_iterator;
 
 class ELFObjectFileBase : public ObjectFile {
@@ -148,6 +150,17 @@ public:
   uint8_t getELFType() const {
     return getObject()->getSymbolELFType(getRawDataRefImpl());
   }
+
+  void getELFTypeName(StringRef &TN) const {
+    uint8_t et = getELFType();
+    TN = "";
+    for (auto &ee : ElfSymbolTypes) {
+      if (ee.Value == et) {
+        TN = ee.AltName;
+        break;
+      }
+    }
+  }
 };
 
 class elf_symbol_iterator : public symbol_iterator {
@@ -244,6 +257,7 @@ protected:
   Expected<section_iterator> getSymbolSection(const Elf_Sym *Symb,
                                               const Elf_Shdr *SymTab) const;
   Expected<section_iterator> getSymbolSection(DataRefImpl Symb) const override;
+  int getSymbolSectionIndex(DataRefImpl Symb) const override;
 
   void moveSectionNext(DataRefImpl &Sec) const override;
   std::error_code getSectionName(DataRefImpl Sec,
@@ -655,6 +669,13 @@ ELFObjectFile<ELFT>::getSymbolSection(DataRefImpl Symb) const {
   return getSymbolSection(Sym, SymTab);
 }
 
+template <class ELFT>
+int
+ELFObjectFile<ELFT>::getSymbolSectionIndex(DataRefImpl Symb) const {
+  const Elf_Sym *Sym = getSymbol(Symb);
+  return Sym->st_shndx;
+}
+
 template <class ELFT>
 void ELFObjectFile<ELFT>::moveSectionNext(DataRefImpl &Sec) const {
   const Elf_Shdr *ESec = getSection(Sec);

llvm/include/llvm/Object/ObjectFile.h

@@ -174,6 +174,9 @@ public:
   /// end_sections() if it is undefined or is an absolute symbol.
   Expected<section_iterator> getSection() const;
 
+  // Get the section index of the section of this symbol
+  int getSectionIndex() const;
+
   const ObjectFile *getObject() const;
 };
 
@@ -228,6 +231,7 @@ protected:
   virtual Expected<SymbolRef::Type> getSymbolType(DataRefImpl Symb) const = 0;
   virtual Expected<section_iterator>
   getSymbolSection(DataRefImpl Symb) const = 0;
+  virtual int getSymbolSectionIndex(DataRefImpl Symb) const { return -1; }
 
   // Same as above for SectionRef.
   friend class SectionRef;
@@ -375,6 +379,10 @@ inline uint64_t SymbolRef::getCommonSize() const {
   return getObject()->getCommonSymbolSize(getRawDataRefImpl());
 }
 
+inline int SymbolRef::getSectionIndex() const {
+  return getObject()->getSymbolSectionIndex(getRawDataRefImpl());
+}
+
 inline Expected<section_iterator> SymbolRef::getSection() const {
   return getObject()->getSymbolSection(getRawDataRefImpl());
 }

llvm/lib/Object/ELFObjectFile.cpp

@@ -35,6 +35,16 @@
 using namespace llvm;
 using namespace object;
 
+const EnumEntry<unsigned> llvm::object::ElfSymbolTypes[8] = {
+    {"None", "NOTYPE", ELF::STT_NOTYPE},
+    {"Object", "OBJECT", ELF::STT_OBJECT},
+    {"Function", "FUNC", ELF::STT_FUNC},
+    {"Section", "SECTION", ELF::STT_SECTION},
+    {"File", "FILE", ELF::STT_FILE},
+    {"Common", "COMMON", ELF::STT_COMMON},
+    {"TLS", "TLS", ELF::STT_TLS},
+    {"GNU_IFunc", "IFUNC", ELF::STT_GNU_IFUNC}};
+
 ELFObjectFileBase::ELFObjectFileBase(unsigned int Type, MemoryBufferRef Source)
     : ObjectFile(Type, Source) {}
 

llvm/test/tools/llvm-nm/X86/sysv-i386.test

@@ -1,6 +1,6 @@
 RUN: llvm-nm -f sysv %p/Inputs/hello.obj.elf-i386 | FileCheck %s --strict-whitespace
 
 CHECK: Name                  Value   Class        Type         Size     Line  Section
-CHECK: .L.str              |00000000|   r  |                  |0000000f|     |
+CHECK: .L.str              |00000000|   r  |            OBJECT|0000000f|     |.rodata.str1.1
-CHECK: main                |00000000|   T  |                  |00000015|     |
+CHECK: main                |00000000|   T  |              FUNC|00000015|     |.text
-CHECK: puts                |        |   U  |                  |        |     |
+CHECK: puts                |        |   U  |            NOTYPE|        |     |*UND*

llvm/test/tools/llvm-nm/X86/sysv-x86_64.test

@@ -2,8 +2,8 @@ RUN: llvm-nm -f sysv %p/Inputs/hello.obj.elf-x86_64 | FileCheck %s --check-prefi
 RUN: llvm-nm -f sysv %p/Inputs/hello.obj.macho-x86_64 2>&1 | FileCheck %s --check-prefix=MACHO --strict-whitespace
 
 ELF: Name                  Value           Class        Type         Size             Line  Section
-ELF: main                |0000000000000000|   T  |                  |0000000000000015|     |
+ELF: main                |0000000000000000|   T  |              FUNC|0000000000000015|     |.text
-ELF: puts                |                |   U  |                  |                |     |
+ELF: puts                |                |   U  |            NOTYPE|                |     |*UND*
 
 MACHO: Name                  Value           Class        Type         Size             Line  Section
 MACHO: EH_frame0           |0000000000000068|   s  |                  |0000000000000000|     |

llvm/tools/llvm-nm/llvm-nm.cpp

@@ -262,6 +262,8 @@ struct NMSymbol {
   uint64_t Size;
   char TypeChar;
   StringRef Name;
+  StringRef SectionName;
+  StringRef TypeName;
   BasicSymbolRef Sym;
   // The Sym field above points to the native symbol in the object file,
   // for Mach-O when we are creating symbols from the dyld info the above
@@ -882,8 +884,13 @@ static void sortAndPrintSymbolList(SymbolicFile &Obj, bool printName,
       std::string PaddedName(Name);
       while (PaddedName.length() < 20)
         PaddedName += " ";
+      std::string TNPad = "";
+      int TNL = I->TypeName.size();
+      while ((TNPad.length() + TNL) < 18)
+        TNPad += " ";
       outs() << PaddedName << "|" << SymbolAddrStr << "|   " << I->TypeChar
-             << "  |                  |" << SymbolSizeStr << "|     |\n";
+             << "  |" << TNPad << I->TypeName << "|" << SymbolSizeStr
+             << "|     |" << I->SectionName << "\n";
     }
   }
 
@@ -1078,7 +1085,37 @@ static bool isObject(SymbolicFile &Obj, basic_symbol_iterator I) {
              : elf_symbol_iterator(I)->getELFType() == ELF::STT_OBJECT;
 }
 
-static char getNMTypeChar(SymbolicFile &Obj, basic_symbol_iterator I) {
+// For ELF object files, Set TypeName to the symbol typename, to be printed
+// in the 'Type' column of the SYSV format output.
+static void getNMTypeName(SymbolicFile &Obj, basic_symbol_iterator I,
+                          StringRef &TypeName) {
+  if (isa<ELFObjectFileBase>(&Obj)) {
+    elf_symbol_iterator SymI(I);
+    SymI->getELFTypeName(TypeName);
+  }
+}
+
+// Return Posix nm class type tag (single letter), but also set SecName and
+// section and name, to be used in format=sysv output.
+static char getNMClassTagAndSectionName(SymbolicFile &Obj,
+                                        basic_symbol_iterator I,
+                                        StringRef &SecName) {
+  if (isa<ELFObjectFileBase>(&Obj)) {
+    elf_symbol_iterator SymI(I);
+
+    Expected<elf_section_iterator> SecIOrErr = SymI->getSection();
+    if (!SecIOrErr) {
+      consumeError(SecIOrErr.takeError());
+      return '?';
+    }
+    int SI = SymI->getSectionIndex();
+    if (SI == 0) {
+      SecName = "*UND*";
+    } else {
+      elf_section_iterator secT = *SecIOrErr;
+      secT->getName(SecName);
+    }
+  }
   uint32_t Symflags = I->getFlags();
   if ((Symflags & object::SymbolRef::SF_Weak) && !isa<MachOObjectFile>(Obj)) {
     char Ret = isObject(Obj, I) ? 'v' : 'w';
@@ -1088,8 +1125,10 @@ static char getNMTypeChar(SymbolicFile &Obj, basic_symbol_iterator I) {
   if (Symflags & object::SymbolRef::SF_Undefined)
     return 'U';
 
-  if (Symflags & object::SymbolRef::SF_Common)
+  if (Symflags & object::SymbolRef::SF_Common) {
+    SecName = "*COM*";
     return 'C';
+  }
 
   char Ret = '?';
   if (Symflags & object::SymbolRef::SF_Absolute)
@@ -1201,7 +1240,8 @@ dumpSymbolNamesFromObject(SymbolicFile &Obj, bool printName,
         }
         S.Address = *AddressOrErr;
       }
-      S.TypeChar = getNMTypeChar(Obj, Sym);
+      getNMTypeName(Obj, Sym, S.TypeName);
+      S.TypeChar = getNMClassTagAndSectionName(Obj, Sym, S.SectionName);
       std::error_code EC = Sym.printName(OS);
       if (EC && MachO)
         OS << "bad string index";

llvm/tools/llvm-readobj/ELFDumper.cpp

@@ -1104,16 +1104,6 @@ static const EnumEntry<unsigned> ElfSymbolVisibilities[] = {
     {"HIDDEN",    "HIDDEN",    ELF::STV_HIDDEN},
     {"PROTECTED", "PROTECTED", ELF::STV_PROTECTED}};
 
-static const EnumEntry<unsigned> ElfSymbolTypes[] = {
-    {"None",      "NOTYPE",  ELF::STT_NOTYPE},
-    {"Object",    "OBJECT",  ELF::STT_OBJECT},
-    {"Function",  "FUNC",    ELF::STT_FUNC},
-    {"Section",   "SECTION", ELF::STT_SECTION},
-    {"File",      "FILE",    ELF::STT_FILE},
-    {"Common",    "COMMON",  ELF::STT_COMMON},
-    {"TLS",       "TLS",     ELF::STT_TLS},
-    {"GNU_IFunc", "IFUNC",   ELF::STT_GNU_IFUNC}};
-
 static const EnumEntry<unsigned> AMDGPUSymbolTypes[] = {
   { "AMDGPU_HSA_KERNEL",            ELF::STT_AMDGPU_HSA_KERNEL }
 };