This is the mail archive of the
binutils@sources.redhat.com
mailing list for the binutils project.
Re: [RFA:] elflink.h: Don't set linker-script symbols to STT_OBJECT (BSF_OBJECT).
- From: Hans-Peter Nilsson <hans-peter dot nilsson at axis dot com>
- To: ian at airs dot com
- Cc: amodra at bigpond dot net dot au, hans-peter dot nilsson at axis dot com, binutils at sources dot redhat dot com, eric at andante dot org
- Date: Tue, 19 Mar 2002 09:37:22 +0100
- Subject: Re: [RFA:] elflink.h: Don't set linker-script symbols to STT_OBJECT (BSF_OBJECT).
> From: Ian Lance Taylor <ian@airs.com>
> Date: 18 Mar 2002 23:57:08 -0800
> I'm afraid I don't remember why this patch was made. It may have just
> been out of a general idea that every symbol should have a type. Or
> it may been to fix some bug.
Probably not a bug, because then there would have been a
testsuite entry for it. :-]
Speaking of which, here's one for cris-axis-elf:
obj1.s:
.global _start
_start:
nop
obj1.d:
# as: --em=criself
# ld: -mcriself
# objdump: -d
# Note that the linker script symbol __start is set to the same
# value as _start, and will collate before _start and be chosen
# as the presentation symbol at disassembly. Anyway, __start
# shouldn't hinder disassembly by posing as an object symbol.
.*: file format elf32-us-cris
Disassembly of section \.text:
00000006 <__start>:
6: 0f05 nop
\.\.\.
brgds, H-P