[uClibc] Re: implementing __libc_stack_end

[Peter S. Mazinger]

On Mon, 20 Jun 2005, Mark S. Mathews wrote:

> 
> Is your test for a powerpc target?
> 
> The problem isn't in the apps that I'm using, the reference to 
> __libc_stack_end is in libgcc.  I don't have the code in front of me, but 
> it comes from a macro expansion in the config/rs6000 stuff.
> 
> The first failure I get is when buildroot is compiling/linking ldd.
> 
> Thanks for the reply,

I have checked the gcc sources and this is used only for rs6000.

I think you should tell upstream about it, because this is glibc 
dependency.

As of gcc-cvs the linux-unwind.h file should have
#ifndef inhibit_libc at the beginning and #endif at the end
(this would ensure, that cross-compiling will work)

currently (gcc-cvs) has this only for i386 and mips, but should have it 
probably for all archs

Peter

> 
> -Mark
> 
> 
> On Sun, 19 Jun 2005, Peter S. Mazinger wrote:
> 
> > On Sun, 19 Jun 2005, Joakim Tjernlund wrote:
> >
> >>>
> >>> I don't think this has to be implemented. As of the ~300 apps I have built
> >>> against uclibc, only kaffe was using this explicitely.
> >>
> >> Great, have you used gcc 3.4.4 as well?
> >
> > I have rebuilt all 300 apps (switching to uclibc-svn) w/ gcc-3.4.4
> >
> > I have only seen mentioning some backtrace possibility in gdb related to
> > __libc_stack_end (sjhill could tell more about it maybe, the comment is
> > somewhere in the nptl port, can't recall)
> >
> > Peter
> >
> >
> 
> 

-- 
Peter S. Mazinger <ps dot m at gmx dot net>           ID: 0xA5F059F2
Key fingerprint = 92A4 31E1 56BC 3D5A 2D08  BB6E C389 975E A5F0 59F2