[ELF] - Make Bss and BssRelRo sections to be synthetic by nature.

[grimar]

Summary

Previously we had:
static OutputSection *Bss;
static OutputSection *BssRelRo;

That was not very flexible. Sections required addditional handling
and it would be impossible to combine them into single if script
saying something like that:

aabb : { *(.bss.*) }

I think using of approach for synthetics is more natural here.
Patch adds synthetic stubs .bss and .bss.rel.ro from start.
That allows linker to create output sections early.
Later, depending on if we had the copy relocations or not, existing infrastructure
can remove unused output sections.
Patch removes need to add these sections manually.

[grimar]

Will update soon.

[grimar]

• Added proper flag.

[ruiu]

It doesn't seem like an improvement to me. Out::Bss and Out::BssRelRo are containers of synthetic sections, thus they are output sections. You could represent them as input sections, and we want to do if that simplifies code/design, but this patch doesn't seem to simplify it.

[grimar]

D30541 was commited as alternative.

ELF/OutputSections.h

public:
   uint8_t *Loc = nullptr;
 };
 
 // All output sections that are handled by the linker specially are
 // globally accessible. Writer initializes them, so don't use them
 // until Writer is initialized.
 struct Out {
   static uint8_t First;
-  static OutputSection *Bss;
-  static OutputSection *BssRelRo;
   static OutputSection *Opd;
   static uint8_t *OpdBuf;
   static PhdrEntry *TlsPhdr;
   static OutputSection *DebugInfo;
   static OutputSection *ElfHeader;
   static OutputSection *ProgramHeaders;
   static OutputSection *PreinitArray;
   static OutputSection *InitArray;

ELF/OutputSections.cpp

 using namespace llvm::object;
 using namespace llvm::support::endian;
 using namespace llvm::ELF;
 
 using namespace lld;
 using namespace lld::elf;
 
 uint8_t Out::First;
-OutputSection *Out::Bss;
-OutputSection *Out::BssRelRo;
 OutputSection *Out::Opd;
 uint8_t *Out::OpdBuf;
 PhdrEntry *Out::TlsPhdr;
 OutputSection *Out::DebugInfo;
 OutputSection *Out::ElfHeader;
 OutputSection *Out::ProgramHeaders;
 OutputSection *Out::PreinitArray;
 OutputSection *Out::InitArray;

ELF/Relocations.cpp

template <class ELFT> static void addCopyRelSymbol(SharedSymbol *SS) {
   // Copy relocation against zero-sized symbol doesn't make sense.
   uintX_t SymSize = SS->template getSize<ELFT>();
   if (SymSize == 0)
     fatal("cannot create a copy relocation for symbol " + toString(*SS));
 
   // See if this symbol is in a read-only segment. If so, preserve the symbol's
   // memory protection by reserving space in the .bss.rel.ro section.
   bool IsReadOnly = isReadOnly<ELFT>(SS);
-  OutputSection *OSec = IsReadOnly ? Out::BssRelRo : Out::Bss;
-
+  OutputSection *OSec =
+      IsReadOnly ? In<ELFT>::BssRelRo->OutSec : In<ELFT>::Bss->OutSec;
   // Create a SyntheticSection in Out to hold the .bss and the Copy Reloc.
   auto *ISec =
       make<CopyRelSection<ELFT>>(IsReadOnly, SS->getAlignment<ELFT>(), SymSize);
   OSec->addSection(ISec);
 
   // Look through the DSO's dynamic symbol table for aliases and create a
   // dynamic symbol for each one. This causes the copy relocation to correctly
   // interpose any aliases.

ELF/SyntheticSections.h

 // For each copy relocation, we create an instance of this class to
 // reserve space in .bss or .bss.rel.ro.
 template <class ELFT> class CopyRelSection final : public SyntheticSection {
   typedef typename ELFT::uint uintX_t;
 
 public:
   CopyRelSection(bool ReadOnly, uintX_t AddrAlign, size_t Size);
   void writeTo(uint8_t *) override {}
+  bool empty() const override;
   size_t getSize() const override { return Size; }
   size_t Size;
 };
 
 template <class ELFT> class MipsGotSection final : public SyntheticSection {
   typedef typename ELFT::uint uintX_t;
 
 public:

 template <class ELFT>
 SymbolBody *addSyntheticLocal(StringRef Name, uint8_t Type, uint64_t Value,
                               uint64_t Size, InputSectionBase *Section);
 
 // Linker generated sections which can be used as inputs.
 template <class ELFT> struct In {
   static InputSection *ARMAttributes;
   static BuildIdSection<ELFT> *BuildId;
+  static CopyRelSection<ELFT> *Bss;
+  static CopyRelSection<ELFT> *BssRelRo;
   static InputSection *Common;
   static DynamicSection<ELFT> *Dynamic;
   static StringTableSection<ELFT> *DynStrTab;
   static SymbolTableSection<ELFT> *DynSymTab;
   static EhFrameHeader<ELFT> *EhFrameHdr;
   static GnuHashTableSection<ELFT> *GnuHashTab;
   static GdbIndexSection<ELFT> *GdbIndex;
   static GotSection<ELFT> *Got;
template <class ELFT> struct In {
   static StringTableSection<ELFT> *StrTab;
   static SymbolTableSection<ELFT> *SymTab;
   static VersionDefinitionSection<ELFT> *VerDef;
   static VersionTableSection<ELFT> *VerSym;
   static VersionNeedSection<ELFT> *VerNeed;
 };
 
 template <class ELFT> InputSection *In<ELFT>::ARMAttributes;
+template <class ELFT> CopyRelSection<ELFT> *In<ELFT>::Bss;
+template <class ELFT> CopyRelSection<ELFT> *In<ELFT>::BssRelRo;
 template <class ELFT> BuildIdSection<ELFT> *In<ELFT>::BuildId;
 template <class ELFT> InputSection *In<ELFT>::Common;
 template <class ELFT> DynamicSection<ELFT> *In<ELFT>::Dynamic;
 template <class ELFT> StringTableSection<ELFT> *In<ELFT>::DynStrTab;
 template <class ELFT> SymbolTableSection<ELFT> *In<ELFT>::DynSymTab;
 template <class ELFT> EhFrameHeader<ELFT> *In<ELFT>::EhFrameHdr;
 template <class ELFT> GdbIndexSection<ELFT> *In<ELFT>::GdbIndex;
 template <class ELFT> GnuHashTableSection<ELFT> *In<ELFT>::GnuHashTab;

ELF/SyntheticSections.cpp

forLoop(0, Chunks.size(),
           [&](size_t I) { HashFn(Hashes.data() + I * HashSize, Chunks[I]); });
 
   // Write to the final output buffer.
   HashFn(HashBuf, Hashes);
 }
 
 template <class ELFT>
 CopyRelSection<ELFT>::CopyRelSection(bool ReadOnly, uintX_t AddrAlign, size_t S)
-    : SyntheticSection(SHF_ALLOC, SHT_NOBITS, AddrAlign,
+    : SyntheticSection(SHF_ALLOC | SHF_WRITE, SHT_NOBITS, AddrAlign,
                        ReadOnly ? ".bss.rel.ro" : ".bss"),
       Size(S) {}
 
+template <class ELFT> bool CopyRelSection<ELFT>::empty() const {
+  return !OutSec || OutSec->Sections.size() <= 1;
+}
+
 template <class ELFT>
 void BuildIdSection<ELFT>::writeBuildId(ArrayRef<uint8_t> Buf) {
   switch (Config->BuildId) {
   case BuildIdKind::Fast:
     computeHash(Buf, [](uint8_t *Dest, ArrayRef<uint8_t> Arr) {
       write64le(Dest, xxHash64(toStringRef(Arr)));
     });
     break;

ELF/Writer.cpp

for (StringRef V : {".rel.", ".rela."}) {
       if (Name.startswith(V)) {
         StringRef Inner = getOutputSectionName(Name.substr(V.size() - 1));
         return Saver.save(Twine(V.drop_back()) + Inner);
       }
     }
   }
 
   for (StringRef V :
-       {".text.", ".rodata.", ".data.rel.ro.", ".data.", ".bss.",
-        ".init_array.", ".fini_array.", ".ctors.", ".dtors.", ".tbss.",
+       {".text.", ".rodata.", ".data.rel.ro.", ".data.", ".bss.rel.ro.",
+        ".bss.", ".init_array.", ".fini_array.", ".ctors.", ".dtors.", ".tbss.",
         ".gcc_except_table.", ".tdata.", ".ARM.exidx."}) {
     StringRef Prefix = V.drop_back();
     if (Name.startswith(V) || Name == Prefix)
       return Prefix;
   }
 
   // CommonSection is identified as "COMMON" in linker scripts.
   // By default, it should go to .bss section.

 // Initialize Out members.
 template <class ELFT> void Writer<ELFT>::createSyntheticSections() {
   // Initialize all pointers with NULL. This is needed because
   // you can call lld::elf::main more than once as a library.
   memset(&Out::First, 0, sizeof(Out));
 
   auto Add = [](InputSectionBase *Sec) { InputSections.push_back(Sec); };
 
-  // Create singleton output sections.
-  Out::Bss = make<OutputSection>(".bss", SHT_NOBITS, SHF_ALLOC | SHF_WRITE);
-  Out::BssRelRo =
-      make<OutputSection>(".bss.rel.ro", SHT_NOBITS, SHF_ALLOC | SHF_WRITE);
   In<ELFT>::DynStrTab = make<StringTableSection<ELFT>>(".dynstr", true);
   In<ELFT>::Dynamic = make<DynamicSection<ELFT>>();
   In<ELFT>::RelaDyn = make<RelocationSection<ELFT>>(
       Config->Rela ? ".rela.dyn" : ".rel.dyn", Config->ZCombreloc);
   In<ELFT>::ShStrTab = make<StringTableSection<ELFT>>(".shstrtab", false);
 
   Out::ElfHeader = make<OutputSection>("", 0, SHF_ALLOC);
   Out::ElfHeader->Size = sizeof(Elf_Ehdr);
template <class ELFT> void Writer<ELFT>::createSyntheticSections() {
   }
 
   InputSection *Common = createCommonSection<ELFT>();
   if (!Common->Data.empty()) {
     In<ELFT>::Common = Common;
     Add(Common);
   }
 
+  // Add .bss and .bss.rel.ro stubs. These sections are used to
+  // create output sections for copy relocations. Sections has size
+  // of zero, so do not affect output.
+  In<ELFT>::Bss = make<CopyRelSection<ELFT>>(false /*RO*/, 0, 0);
+  Add(In<ELFT>::Bss);
+  In<ELFT>::BssRelRo = make<CopyRelSection<ELFT>>(true /*RO*/, 0, 0);
+  Add(In<ELFT>::BssRelRo);
+
   // Add MIPS-specific sections.
   bool HasDynSymTab =
       !Symtab<ELFT>::X->getSharedFiles().empty() || Config->pic() ||
       Config->ExportDynamic;
   if (Config->EMachine == EM_MIPS) {
     if (!Config->Shared && HasDynSymTab) {
       In<ELFT>::MipsRldMap = make<MipsRldMapSection<ELFT>>();
       Add(In<ELFT>::MipsRldMap);
if (Type == SHT_INIT_ARRAY || Type == SHT_FINI_ARRAY ||
     return true;
 
   if (Sec == In<ELFT>::GotPlt->OutSec)
     return Config->ZNow;
   if (Sec == In<ELFT>::Dynamic->OutSec)
     return true;
   if (In<ELFT>::Got && Sec == In<ELFT>::Got->OutSec)
     return true;
-  if (Sec == Out::BssRelRo)
+  if (Sec == In<ELFT>::BssRelRo->OutSec)
     return true;
 
   StringRef S = Sec->Name;
   return S == ".data.rel.ro" || S == ".ctors" || S == ".dtors" || S == ".jcr" ||
          S == ".eh_frame" || S == ".openbsd.randomdata";
 }
 
 template <class ELFT>
template <class ELFT> void Writer<ELFT>::finalizeSections() {
   // at the end because some tags like RELSZ depend on result
   // of finalizing other sections.
   for (OutputSection *Sec : OutputSections)
     Sec->finalize<ELFT>();
 
   // Dynamic section must be the last one in this list and dynamic
   // symbol table section (DynSymTab) must be the first one.
   finalizeSynthetic<ELFT>(
-      {In<ELFT>::DynSymTab, In<ELFT>::GnuHashTab, In<ELFT>::HashTab,
-       In<ELFT>::SymTab,    In<ELFT>::ShStrTab,   In<ELFT>::StrTab,
-       In<ELFT>::VerDef,    In<ELFT>::DynStrTab,  In<ELFT>::GdbIndex,
-       In<ELFT>::Got,       In<ELFT>::MipsGot,    In<ELFT>::IgotPlt,
-       In<ELFT>::GotPlt,    In<ELFT>::RelaDyn,    In<ELFT>::RelaIplt,
-       In<ELFT>::RelaPlt,   In<ELFT>::Plt,        In<ELFT>::Iplt,
-       In<ELFT>::Plt,       In<ELFT>::EhFrameHdr, In<ELFT>::VerSym,
-       In<ELFT>::VerNeed,   In<ELFT>::Dynamic});
+      {In<ELFT>::Bss,        In<ELFT>::BssRelRo, In<ELFT>::DynSymTab,
+       In<ELFT>::GnuHashTab, In<ELFT>::HashTab,  In<ELFT>::SymTab,
+       In<ELFT>::ShStrTab,   In<ELFT>::StrTab,   In<ELFT>::VerDef,
+       In<ELFT>::DynStrTab,  In<ELFT>::GdbIndex, In<ELFT>::Got,
+       In<ELFT>::MipsGot,    In<ELFT>::IgotPlt,  In<ELFT>::GotPlt,
+       In<ELFT>::RelaDyn,    In<ELFT>::RelaIplt, In<ELFT>::RelaPlt,
+       In<ELFT>::Plt,        In<ELFT>::Iplt,     In<ELFT>::Plt,
+       In<ELFT>::EhFrameHdr, In<ELFT>::VerSym,   In<ELFT>::VerNeed,
+       In<ELFT>::Dynamic});
 }
 
 template <class ELFT> void Writer<ELFT>::addPredefinedSections() {
-  // Add BSS sections.
-  auto Add = [=](OutputSection *Sec) {
-    if (!Sec->Sections.empty()) {
-      Sec->assignOffsets<ELFT>();
-      OutputSections.push_back(Sec);
-    }
-  };
-  Add(Out::Bss);
-  Add(Out::BssRelRo);
-
   // ARM ABI requires .ARM.exidx to be terminated by some piece of data.
   // We have the terminater synthetic section class. Add that at the end.
   auto *OS = dyn_cast_or_null<OutputSection>(findSection(".ARM.exidx"));
   if (OS && !OS->Sections.empty() && !Config->Relocatable)
     OS->addSection(make<ARMExidxSentinelSection<ELFT>>());
 }
 
 // The linker is expected to define SECNAME_start and SECNAME_end

test/ELF/map-file.s

 // CHECK-NEXT: 0000000000201012 0000000000000000     0                         zed
 // CHECK-NEXT: 0000000000201014 0000000000000000     4         .text
 // CHECK-NEXT: 0000000000201014 0000000000000000     4                 {{.*}}{{/|\\}}map-file.s.tmp3.o
 // CHECK-NEXT: 0000000000201014 0000000000000000     0                         bah
 // CHECK-NEXT: 0000000000201014 0000000000000001     4                 {{.*}}{{/|\\}}map-file.s.tmp4.a(map-file.s.tmp4.o)
 // CHECK-NEXT: 0000000000201014 0000000000000000     0                         baz
 // CHECK-NEXT: 0000000000202000 0000000000000004    16 .bss
 // CHECK-NEXT: 0000000000202000 0000000000000004    16         COMMON
+// CHECK-NEXT: 0000000000202004 0000000000000000     1         .bss
 // CHECK-NEXT: 0000000000000000 0000000000000008     1 .comment
 // CHECK-NEXT: 0000000000000000 0000000000000008     1         .comment
 // CHECK-NEXT: 0000000000000000 00000000000000f0     8 .symtab
 // CHECK-NEXT: 0000000000000000 00000000000000f0     8         .symtab
 // CHECK-NEXT: 0000000000000000 0000000000000039     1 .shstrtab
 // CHECK-NEXT: 0000000000000000 0000000000000039     1         .shstrtab
 // CHECK-NEXT: 0000000000000000 000000000000002f     1 .strtab
 // CHECK-NEXT: 0000000000000000 000000000000002f     1         .strtab
 
 // RUN: not ld.lld %t1.o %t2.o %t3.o %t4.a -o %t -Map=/ 2>&1 \
 // RUN:  | FileCheck -check-prefix=FAIL %s
 // FAIL: cannot open /