[ELF2] - Linker script EXTERN implemented.

[grimar]

Summary

The reason of collecting all undefines in vector is that during reading files we already need to have Symtab created. Or like was done in that patch - to put undefines from scripts somewhere to delay Symtab.addUndefinedOpt() call.

Actually the main problem I see here is that currently ELF type is taken from first file. If we could know it earlier (I am not sure, from command line may be ?) then we would not need such solutions.

[ruiu]

ELF/Driver.cpp

192–194 ↗

(On Diff #37755)

Move this before OPT_z.

test/elf2/linkerscript.s

9–11 ↗

(On Diff #37755)

This test does not actually verify that EXTERN() works in the same way as -u. I'd add one more test to undefined-opt.s.

[grimar]

Review comments addressed.

ELF/Driver.cpp

195–197 ↗

(On Diff #37770)

Done.

test/elf2/linkerscript.s

9–11 ↗

(On Diff #37770)

Done.

[ruiu]

LGTM

lld/trunk/ELF/Config.h

struct Configuration {
   llvm::StringRef Emulation;
   llvm::StringRef Fini;
   llvm::StringRef Init;
   llvm::StringRef OutputFile;
   llvm::StringRef SoName;
   llvm::StringRef Sysroot;
   std::string RPath;
   std::vector<llvm::StringRef> SearchPaths;
+  std::vector<llvm::StringRef> Undefined;
   bool AllowMultipleDefinition;
   bool AsNeeded = false;
   bool Bsymbolic;
   bool DiscardAll;
   bool DiscardLocals;
   bool DiscardNone;
   bool EnableNewDtags;
   bool ExportDynamic;

lld/trunk/ELF/Driver.cpp

void LinkerDriver::createFiles(opt::InputArgList &Args) {
   Config->DynamicLinker = getString(Args, OPT_dynamic_linker);
   Config->Entry = getString(Args, OPT_entry);
   Config->Fini = getString(Args, OPT_fini, "_fini");
   Config->Init = getString(Args, OPT_init, "_init");
   Config->OutputFile = getString(Args, OPT_o);
   Config->SoName = getString(Args, OPT_soname);
   Config->Sysroot = getString(Args, OPT_sysroot);
 
+  for (auto *Arg : Args.filtered(OPT_undefined))
+    Config->Undefined.push_back(Arg->getValue());
+
   for (auto *Arg : Args.filtered(OPT_z))
     if (Arg->getValue() == StringRef("now"))
       Config->ZNow = true;
 
   for (auto *Arg : Args) {
     switch (Arg->getOption().getID()) {
     case OPT_l:
       addFile(searchLibrary(Arg->getValue()));
if (!Config->Shared) {
     // Given that the symbol is effectively unused, we just create a dummy
     // hidden one to avoid the undefined symbol error.
     Symtab.addIgnoredSym("_GLOBAL_OFFSET_TABLE_");
   }
 
   for (std::unique_ptr<InputFile> &F : Files)
     Symtab.addFile(std::move(F));
 
-  for (auto *Arg : Args.filtered(OPT_undefined))
-    Symtab.addUndefinedOpt(Arg->getValue());
+  for (auto &U : Config->Undefined)
+    Symtab.addUndefinedOpt(U);
 
   if (Config->OutputFile.empty())
     Config->OutputFile = "a.out";
 
   // Write the result to the file.
   Symtab.scanShlibUndefined();
   writeResult<ELFT>(&Symtab);
 }

lld/trunk/ELF/LinkerScript.cpp

private:
   StringRef next();
   bool atEOF() { return Tokens.size() == Pos; }
   void expect(StringRef Expect);
 
   void addFile(StringRef Path);
 
   void readAsNeeded();
   void readEntry();
+  void readExtern();
   void readGroup();
   void readInclude();
   void readOutput();
   void readOutputArch();
   void readOutputFormat();
   void readSearchDir();
 
   StringSaver Saver;
   std::vector<StringRef> Tokens;
   size_t Pos = 0;
 };
 }
 
 void LinkerScript::run() {
   while (!atEOF()) {
     StringRef Tok = next();
     if (Tok == ";")
       continue;
     if (Tok == "ENTRY") {
       readEntry();
+    } else if (Tok == "EXTERN") {
+      readExtern();
     } else if (Tok == "GROUP" || Tok == "INPUT") {
       readGroup();
     } else if (Tok == "INCLUDE") {
       readInclude();
     } else if (Tok == "OUTPUT") {
       readOutput();
     } else if (Tok == "OUTPUT_ARCH") {
       readOutputArch();
void LinkerScript::readEntry() {
   // -e <symbol> takes predecence over ENTRY(<symbol>).
   expect("(");
   StringRef Tok = next();
   if (Config->Entry.empty())
     Config->Entry = Tok;
   expect(")");
 }
 
+void LinkerScript::readExtern() {
+  expect("(");
+  for (;;) {
+    StringRef Tok = next();
+    if (Tok == ")")
+      return;
+    Config->Undefined.push_back(Tok);
+  }
+}
+
 void LinkerScript::readGroup() {
   expect("(");
   for (;;) {
     StringRef Tok = next();
     if (Tok == ")")
       return;
     if (Tok == "AS_NEEDED") {
       readAsNeeded();

lld/trunk/test/elf2/linkerscript.s

 # REQUIRES: x86
 # RUN: mkdir -p %t.dir
 # RUN: llvm-mc -filetype=obj -triple=x86_64-unknown-linux %s -o %t
 # RUN: llvm-mc -filetype=obj -triple=x86_64-unknown-linux \
 # RUN:   %p/Inputs/libsearch-st.s -o %t2.o
 # RUN: rm -f %t.dir/libxyz.a
 # RUN: llvm-ar rcs %t.dir/libxyz.a %t2.o
 
+# RUN: echo "EXTERN( undef undef2 )" > %t.script
+# RUN: ld.lld2 %t -o %t2 %t.script
+# RUN: llvm-readobj %t2 > /dev/null
+
 # RUN: echo "GROUP(" %t ")" > %t.script
 # RUN: ld.lld2 -o %t2 %t.script
 # RUN: llvm-readobj %t2 > /dev/null
 
 # RUN: echo "INPUT(" %t ")" > %t.script
 # RUN: ld.lld2 -o %t2 %t.script
 # RUN: llvm-readobj %t2 > /dev/null
 

lld/trunk/test/elf2/undefined-opt.s

 # RUN: ld.lld2 -o %t3 %t.o %tar.a -u bar --undefined=abs
 # RUN: llvm-readobj --symbols %t3 | FileCheck --check-prefix=TWO-UNDEFINED %s
 # TWO-UNDEFINED: Symbols [
 # TWO-UNDEFINED: Name: abs
 # TWO-UNDEFINED: Name: big
 # TWO-UNDEFINED: Name: bar
 # TWO-UNDEFINED: Name: zed
 # TWO-UNDEFINED: ]
+# Now the same logic but linker script is used to set undefines
+# RUN: echo "EXTERN( bar abs )" > %t.script
+# RUN: ld.lld2 -o %t3 %t.o %tar.a %t.script
+# RUN: llvm-readobj --symbols %t3 | FileCheck --check-prefix=TWO-UNDEFINED %s
 
 # Added undefined symbol may be left undefined without error, but
 # shouldn't show up in the dynamic table.
 # RUN: ld.lld2 -shared -o %t4 %t.o %tar.a -u unknown
 # RUN: llvm-readobj --dyn-symbols %t4 | \
 # RUN:     FileCheck --check-prefix=UNK-UNDEFINED-SO %s
 # UNK-UNDEFINED-SO: DynamicSymbols [
 # UNK-UNDEFINED-SO-NOT:     Name: unknown
 # UNK-UNDEFINED-SO: ]
 
 .globl _start;
 _start: