[PATCH] D15779: [ELF] - Implemented optimization for R_X86_64_GOTPCREL relocation.

[George Rimar via llvm-commits]

grimar added inline comments.

================
Comment at: ELF/Target.cpp:513
@@ +512,3 @@
+    return false;
+  return (S.getName() != "_DYNAMIC");
+}
----------------
rafael wrote:
> Why is _DYNAMIC special?
> 
> The other checks are not architecture specific and should be in architecture independent code.
That is something directly mentioned in docs I think, but initially in binutils there also were no exception for _DYNAMIC,
and they had to do that finally because it turns out ld.so uses it:
https://sourceware.org/ml/binutils/2012-09/msg00000.html

================
Comment at: ELF/Target.cpp:537
@@ -525,1 +536,3 @@
+  case R_X86_64_REX_GOTPCRELX:
+    return canRelaxGotPCRel(S) ? R_X86_REX_RELAX_GOT_PC : R_GOT_PC;
   case R_X86_64_GOTPCREL:
----------------
rafael wrote:
> I would probably have this return R_GOT_PC and then have a generic logic for "can we optimize this got use". This would be somewhat similar to the fact that we optimize plt acesses to local symbols.
Done, reimplemented in more generic way.


http://reviews.llvm.org/D15779