[uClibc] Re: uclibc and heap randomisation (ET_EXEC)

[Joakim Tjernlund]

> On Monday 30 August 2004 11:44 am, Joakim Tjernlund wrote:
> > Hmm, I removed that code by mistake and that broke procps.
> 
> it also silently killed Gentoo's sandbox ... there is the fact that our 
> current sandbox implementation is an ugly beast, but there is also the fact 
> that it works fine with this code ...

hmm, is it working with current CVS where __curbrk is gone but __environ is still
in there?
I can understand that removing __environ broke stuff, but removing __curbrk should
not break anything. __curbrk is set by the first call to (s)brk()

 Jocke

> 
> theres also the question of what else broke and no one noticed :)  that bit of 
> code seems to provide something vital; but what exactly ?  i have no idea 
> because these details are way above my head :)
> -mike

Time will tell :)