[ELF] Inspect -EL & -EB for OUTPUT_FORMAT(default, big, little)

[MaskRay]

Summary

Choose big if -EB is specified, little if -EL is specified, or default if neither is specified.
The new behavior matches GNU ld.

Fixes: https://github.com/ClangBuiltLinux/linux/issues/1025

[MaskRay]

fix test

[peter.smith]

LGTM, I've made some small suggestions.

lld/ELF/Config.h

157

eb and el are close in name to the isLE option? Perhaps use something like "optLE" or "leSet" to avoid confusion.

lld/ELF/Options.td

162

Worth adding a simple line of text for the help file? Possibly the man page as well.
ld.bfd says "Link little-endian objects" for EL

lld/ELF/ScriptParser.cpp

436

Worth mentioning what happens if second form is used but no choice of -EB or -EL is made.
"choose big if -EB is specified, little if -EL is specified or default if neither is specified."

https://sourceware.org/binutils/docs/ld/Format-Commands.html

[MaskRay]

comments

[MaskRay]

variant->format

[MaskRay]

better diagnostic

lld/ELF/Config.h

struct Configuration {
   bool checkSections;
   bool compressDebugSections;
   bool cref;
   std::vector<std::pair<llvm::GlobPattern, uint64_t>> deadRelocInNonAlloc;
   bool defineCommon;
   bool demangle = true;
   bool dependentLibraries;
   bool disableVerify;
   bool ehFrameHdr;
   bool emitLLVM;
   bool emitRelocs;
   bool enableNewDtags;
   bool executeOnly;
   bool exportDynamic;
   bool fixCortexA53Errata843419;
   bool fixCortexA8;
   bool formatBinary = false;
struct Configuration {
   bool mipsN32Abi = false;
   bool mmapOutputFile;
   bool nmagic;
   bool noDynamicLinker = false;
   bool noinhibitExec;
   bool nostdlib;
   bool oFormatBinary;
   bool omagic;
+  bool optEB = false;
+  bool optEL = false;
   bool optimizeBBJumps;
   bool optRemarksWithHotness;
   bool picThunk;
   bool pie;
   bool printGcSections;
   bool printIcfSections;
   bool relocatable;
   bool relrPackDynRelocs;

lld/ELF/Driver.cpp

static void readConfigs(opt::InputArgList &args) {
   config->zSeparate = getZSeparate(args);
   config->zShstk = hasZOption(args, "shstk");
   config->zStackSize = args::getZOptionValue(args, OPT_z, "stack-size", 0);
   config->zStartStopVisibility = getZStartStopVisibility(args);
   config->zText = getZFlag(args, "text", "notext", true);
   config->zWxneeded = hasZOption(args, "wxneeded");
   setUnresolvedSymbolPolicy(args);
 
+  if (opt::Arg *arg = args.getLastArg(OPT_eb, OPT_el)) {
+    if (arg->getOption().matches(OPT_eb))
+      config->optEB = true;
+    else
+      config->optEL = true;
+  }
+
   for (opt::Arg *arg : args.filtered(OPT_z)) {
     std::pair<StringRef, StringRef> option =
         StringRef(arg->getValue()).split('=');
     if (option.first != "dead-reloc-in-nonalloc")
       continue;
     constexpr StringRef errPrefix = "-z dead-reloc-in-nonalloc=: ";
     std::pair<StringRef, StringRef> kv = option.second.split('=');
     if (kv.first.empty() || kv.second.empty()) {

lld/ELF/Options.td

 
 defm dynamic_list : Eq<"dynamic-list",
    "Read a list of dynamic symbols. (executable) Put matched non-local defined"
    "symbols to the dynamic symbol table. (shared object) References to matched"
    "non-local STV_DEFAULT symbols shouldn't be bound to definitions within the "
    "shared object. Implies -Bsymbolic but does not set DF_SYMBOLIC">,
    MetaVarName<"<file>">;
 
+def eb: F<"EB">, HelpText<"Select the big-endian format in OUTPUT_FORMAT">;
+def el: F<"EL">, HelpText<"Select the little-endian format in OUTPUT_FORMAT">;
+
 defm eh_frame_hdr: B<"eh-frame-hdr",
     "Request creation of .eh_frame_hdr section and PT_GNU_EH_FRAME segment header",
     "Do not create .eh_frame_hdr section">;
 
 def emit_relocs: F<"emit-relocs">, HelpText<"Generate relocations in output">;
 
 def enable_new_dtags: F<"enable-new-dtags">,
   HelpText<"Enable new dynamic tags (default)">;

 def: Separate<["--", "-"], "rpath-link">;
 def: J<"rpath-link=">;
 def: F<"secure-plt">;
 def: F<"sort-common">;
 def: F<"stats">;
 def: F<"warn-execstack">;
 def: F<"warn-once">;
 def: F<"warn-shared-textrel">;
-def: F<"EB">;
-def: F<"EL">;
 def: JoinedOrSeparate<["-"], "G">;
 def: F<"Qy">;
 
 // Hidden option used for testing MIPS multi-GOT implementation.
 defm mips_got_size:
   Eq<"mips-got-size", "Max size of a single MIPS GOT. 0x10000 by default.">,
   Flags<[HelpHidden]>;

lld/ELF/ScriptParser.cpp

return StringSwitch<std::pair<ELFKind, uint16_t>>(s)
       .Case("elf64-tradlittlemips", {ELF64LEKind, EM_MIPS})
       .Case("elf32-littleriscv", {ELF32LEKind, EM_RISCV})
       .Case("elf64-littleriscv", {ELF64LEKind, EM_RISCV})
       .Case("elf64-sparc", {ELF64BEKind, EM_SPARCV9})
       .Case("elf32-msp430", {ELF32LEKind, EM_MSP430})
       .Default({ELFNoneKind, EM_NONE});
 }
 
-// Parse OUTPUT_FORMAT(bfdname) or OUTPUT_FORMAT(bfdname, big, little).
-// Currently we ignore big and little parameters.
+// Parse OUTPUT_FORMAT(bfdname) or OUTPUT_FORMAT(default, big, little). Choose
+// big if -EB is specified, little if -EL is specified, or default if neither is
+// specified.
 void ScriptParser::readOutputFormat() {
   expect("(");
 
+  StringRef s;
   config->bfdname = unquote(next());
-  StringRef s = config->bfdname;
+  if (consume(",")) {
+    s = unquote(next());
+    if (config->optEB)
+      config->bfdname = s;
+    expect(",");
+    s = unquote(next());
+    if (config->optEL)
+      config->bfdname = s;
+  }
+  expect(")");
+  s = config->bfdname;
   if (s.consume_back("-freebsd"))
     config->osabi = ELFOSABI_FREEBSD;
 
   std::tie(config->ekind, config->emachine) = parseBfdName(s);
   if (config->emachine == EM_NONE)
     setError("unknown output format name: " + config->bfdname);
   if (s == "elf32-ntradlittlemips" || s == "elf32-ntradbigmips")
     config->mipsN32Abi = true;
   if (config->emachine == EM_MSP430)
     config->osabi = ELFOSABI_STANDALONE;
-
-  if (consume(")"))
-    return;
-  expect(",");
-  skip();
-  expect(",");
-  skip();
-  expect(")");
 }
 
 void ScriptParser::readPhdrs() {
   expect("{");
 
   while (!errorCount() && !consume("}")) {
     PhdrsCommand cmd;
     cmd.name = next();

lld/docs/ld.lld.1

 .Dv PT_INTERP .
 .It Fl -dynamic-list Ns = Ns Ar file
 Read a list of dynamic symbols from
 .Ar file .
 (executable) Put matched non-local defined symbols to the dynamic symbol table.
 (shared object) References to matched non-local STV_DEFAULT symbols shouldn't be bound to definitions within the shared object. Implies
 .Cm -Bsymbolic
 but does not set DF_SYMBOLIC
+.It Fl -EB
+Select the big-endian variant in the OUTPUT_FORMAT command.
+.It Fl -EL
+Select the little-endian variant in the OUTPUT_FORMAT command.
 .It Fl -eh-frame-hdr
 Request creation of
 .Li .eh_frame_hdr
 section and
 .Dv PT_GNU_EH_FRAME
 segment header.
 .It Fl -emit-relocs , Fl q
 Generate relocations in the output.

lld/test/ELF/emulation-aarch64.s

 # RUN: ld.lld -m aarch64linuxb %t.be.o -o %t1.be
 # RUN: llvm-readobj --file-headers %t1.be | FileCheck --check-prefixes=AARCH64,BE %s
 # RUN: ld.lld -m aarch64elfb %t.be.o -o %t2.be
 # RUN: llvm-readobj --file-headers %t2.be | FileCheck --check-prefixes=AARCH64,BE %s
 # RUN: echo 'OUTPUT_FORMAT(elf64-bigaarch64)' > %t.script
 # RUN: ld.lld %t.script %t.be.o -o %t3.be
 # RUN: llvm-readobj --file-headers %t3.be | FileCheck --check-prefixes=AARCH64,BE %s
 
+## Test OUTPUT_FORMAT(default, big, little).
+# RUN: echo 'OUTPUT_FORMAT("elf64-littleaarch64", "elf64-bigaarch64", "elf64-littleaarch64")' > %t.script
+# RUN: ld.lld -EL -T %t.script %t.o -o %t4.le
+# RUN: llvm-readobj --file-headers %t4.le | FileCheck --check-prefixes=AARCH64,LE %s
+# RUN: not ld.lld -EB -T %t.script %t.o -o /dev/null 2>&1 | FileCheck --check-prefix=ERR_BE %s
+
+# RUN: not ld.lld -T %t.script %t.be.o -o /dev/null 2>&1 | FileCheck --check-prefix=ERR_LE %s
+# RUN: not ld.lld -EL -T %t.script %t.be.o -o /dev/null 2>&1 | FileCheck --check-prefix=ERR_LE %s
+# RUN: ld.lld -EB -T %t.script %t.be.o -o %t4.be
+# RUN: llvm-readobj --file-headers %t4.be | FileCheck --check-prefixes=AARCH64,BE %s
+
+# ERR_LE: error: {{.*}}.o is incompatible with elf64-littleaarch64
+# ERR_BE: error: {{.*}}.o is incompatible with elf64-bigaarch64
+
 # AARCH64:      ElfHeader {
 # AARCH64-NEXT:   Ident {
 # AARCH64-NEXT:     Magic: (7F 45 4C 46)
 # AARCH64-NEXT:     Class: 64-bit (0x2)
 # LE-NEXT:          DataEncoding: LittleEndian (0x1)
 # BE-NEXT:          DataEncoding: BigEndian (0x2)
 # AARCH64-NEXT:     FileVersion: 1
 # AARCH64-NEXT:     OS/ABI: SystemV (0x0)

lld/test/ELF/invalid-linkerscript.test

 # RUN: echo "INCLUDE /no/such/file" > %t7
 # RUN: not ld.lld %t7 no-such-file 2>&1 | FileCheck -check-prefix=ERR7 %s
 # ERR7: cannot find linker script /no/such/file
 # ERR7: cannot open no-such-file:
 
 # RUN: echo "OUTPUT_FORMAT(x y z)" > %t8
 # RUN: not ld.lld %t8 no-such-file 2>&1 | FileCheck -check-prefix=ERR8 %s
 # RUN: not ld.lld -m elf_amd64 %t8 no-such-file 2>&1 | FileCheck -check-prefix=ERR8 %s
-# ERR8: unknown output format name: x
+# ERR8: ) expected, but got y
 # ERR8: cannot open no-such-file:
 
 # RUN: echo "OUTPUT_FORMAT(elf64-x86-64 y z)" > %t9
 # RUN: not ld.lld %t9 no-such-file 2>&1 | FileCheck -check-prefix=ERR9 %s
-# ERR9: , expected, but got y
+# ERR9: ) expected, but got y
 # ERR9: cannot open no-such-file: