[lld][ELF][SPARC] Emit RELA entries

[LemonBoy]

Summary

Emit RELA entries as specified by the V9 ABI.

[jrtc27]

Can we not have a test that uses textrel; what's wrong with:

.data
x: .xword y

or

.data
x: .xword x

[LemonBoy]

In D102986#2776010, @jrtc27 wrote:

Can we not have a test that uses textrel; what's wrong with:

.data
x: .xword y

or

.data
x: .xword x

You're greeted with an error:
relocation R_SPARC_64 cannot be used against local symbol; recompile with -fPIC

[jrtc27]

I see, because getDynRel isn't implemented in order to return type rather than 0 when passed R_SPARC_64. Why does this backend even exist if it's so utterly broken...

[jrtc27]

Honestly, I just wouldn't bother adding a test for rela then, it'll be implicitly tested by all the tests you add in the following changes

[MaskRay]

(If the idea is to do the test in a merged patch.)

[LemonBoy]

Alright, folding this back into D102986 (with the getDynRel fix?)

lld/ELF/Driver.cpp

if (s == "rel")
       return false;
     if (s == "rela")
       return true;
   }
 
   // Otherwise use the psABI defined relocation entry format.
   uint16_t m = config->emachine;
   return m == EM_AARCH64 || m == EM_AMDGPU || m == EM_HEXAGON || m == EM_PPC ||
-         m == EM_PPC64 || m == EM_RISCV || m == EM_X86_64;
+         m == EM_PPC64 || m == EM_RISCV || m == EM_SPARCV9 || m == EM_X86_64;
 }
 
 static void parseClangOption(StringRef opt, const Twine &msg) {
   std::string err;
   raw_string_ostream os(err);
 
   const char *argv[] = {config->progName.data(), opt.data()};
   if (cl::ParseCommandLineOptions(2, argv, "", &os))

lld/test/ELF/sparcv9-rela.s

+// REQUIRES: sparc
+// RUN: llvm-mc --position-independent -filetype=obj -arch=sparcv9 %s -o %t.o
+// RUN: ld.lld %t.o -o %t.so -shared
+// RUN: llvm-readobj --dynamic-table %t.so | FileCheck %s
+
+sethi %hi(x), %o1
+
+// CHECK:  0x{{0+}}7 RELA