This is the mail archive of the
mailing list for the binutils project.
Re: LD script bug
- From: Nick Clifton <nickc at redhat dot com>
- To: Ankur Arora <ankura at cse dot iitk dot ac dot in>
- Cc: binutils at sources dot redhat dot com, binutils-bug at sources dot redhat dot com
- Date: Fri, 02 Jan 2004 18:13:09 +0000
- Subject: Re: LD script bug
- References: <Pine.LNX.email@example.com>
> ld version: 126.96.36.199.18 20030206 on a vanilla RH 9 system.
Note - this is an old version of the linker. I would recommend that
you download the sources for the latest version (2.14) and find out if
this fixes the problem.
> I am getting this weird behaviour(relevant excerpt from the
> linker map):
> 0x08048527 value = ((. + 0x20) - 0x1)
> 0x08048518 . = (value & 0xffffffe0)
> *fill* 0x08048508 0x10 00
> Fill value should be 0x18 or I am going crazy. Please tell
> me which :).
Well, the *fill* size is correct for the displayed resulting address,
ie 0x08048518 does equal 0x80488508 plus 0x10.
But I agree that the resulting address appears to be wrong.
(0x8048527 & 0xffffffe0) ought to be 0x8048520 rather than 0x8048518,
so the fill size ought to be 0x18.
> In case any more information is required I will be happy
> to provide it.
If switching to the 2.14 release does not fix the problem then please
could you send a *small* reproducible example to the list that
reproduces the problem.