When running the ld testsuite with -melf_i386_sol2 instead of -melf_i386
on Solaris, one test regresses:
FAIL: Build libgot-1-i386.so
The issue is that one line in the --got-contents output differs
unexpectedly:
Global Offset Table '.got.plt' contains 4 entries:
Index: Address Reloc Sym. Name + Addend/Value
- 0:
00200200 20016c
- 1:
00200204 0
- 2:
00200208 0
- 3:
0020020c R_386_JUMP_SLOT bar + 156
+ 0:
00200270 2001dc
+ 1:
00200274 0
+ 2:
00200278 0
+ 3:
0020027c R_386_JUMP_SLOT bar + 1c6
While ld-i386/libgot-1.rd already allows for differences in the
addresses, the addend is assumed to be fixed. However, this is not the
case with -melf_i386_sol2. The difference is that .hash, .dynsym, and
.symtab have additional entries as required by the Solaris ABI:
* In .dynsym and .symtab, _DYNAMIC, _GLOBAL_OFFSET_TABLE_, and
_PROCEDURE_LINKAGE_TABLE_ are added.
* .symtab also gains _END_ and _START_.
This explains the differences in addresses and addends, but they are
completely benign.
This patch thus allows for arbitrary addends.
Tested on {i386,amd64}-pc-solaris2.11 with both -melf_i386_sol2 and
-melf_i386, and {i686,x86_64}-pc-linux-gnu.
2026-01-23 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
ld:
PR ld/33350
* testsuite/ld-i386/libgot-1.rd: Allow for different
R_386_JUMP_SLOT addends.