This is the mail archive of the
libc-alpha@sourceware.cygnus.com
mailing list for the glibc project.
Re: [Michael Deutschmann <michael@talamasca.wkpowerlink.com>]
- To: Franz dot Sirl-kernel at lauterbach dot com (Franz Sirl)
- Subject: Re: [Michael Deutschmann <michael@talamasca.wkpowerlink.com>]
- From: hjl at lucon dot org (H.J. Lu)
- Date: Mon, 27 Sep 1999 11:09:18 -0700 (PDT)
- Cc: aj at suse dot de (Andreas Jaeger),libc-alpha at sourceware dot cygnus dot com (GNU C. library Alpha Tester List),michael at talamasca dot wkpowerlink dot com (Michael Deutschmann)
>
> At 17:50 27.09.99 , Andreas Jaeger wrote:
>
> >Hi,
> >
> >Michael reported a problem with a broken ld.so in glibc 2.1.2. He
> >finally nailed this down to LDFLAGS=-s in his environment. I can
> >reproduce the problem with binutils 2.9.1.0.25 (he used 2.9.5.0.13).
> >
> >Is this a binutils bug or is there another reason not to use `-s' for
> >linking?
>
> Hmm, there were (are?) some mystery problems (corrupted binaries) with
> strip and/or LDFLAGS=-s on Linux/PPC as well. That's why we still ship an
> unstripped glibc. It tends to happen especially with large binaries, but we
> never found why, cause nobody could reproduce it reliably. I _thought_ it
> was fixed in binutils-2.9.[45].*, but I wasn't brave enough yet to try it
> on the whole distribution...
>
>
I will make a new binutils shortly to fix it. BTW, it is a very old
bug. It has been in there for a longf time.
--
H.J. Lu (hjl@gnu.org)