Re: 4.0 chroots, threading, and builds

[Manuel Bouyer <bouyer%antioche.eu.org@localhost>]

On Mon, Sep 29, 2008 at 09:35:36AM +0000, Andrew Doran wrote:
> On Mon, Sep 29, 2008 at 11:25:25AM +0200, Manuel Bouyer wrote:
> 
> > > On Thu, Sep 25, 2008 at 04:58:46PM +0200, Manuel Bouyer wrote:
> > > 
> > > > > - Problems identified years ago do not seem to have been fixed.
> > > > 
> > > > We have been living with them for years, I can't see why it would now
> > > > be a showstopper.
> > > 
> > > We are not living with them any more, because a massive engineering effort
> > > was made to fix them. They're fixed with the new threading model in 5.0.
> > 
> > Sure, but we're talking about pre-5.0 binairies, that use the old
> > threading model.
> 
> Those binaries will continue to work with a 5.0 kernel and libraries.

*and libraries* is the point that cause problem. We don't want to change
libraries.


> Please stop propogating disinformation.

I don't think it's disinformation. We're just looking at things from a
different point of view.


> 
> > So it's not a 4.x userland anymore, it won't work with a 4.x kernel.
> > Unacceptable for me.
> 
> If a one line addition to /etc/ld.so.conf is acceptable for you, then it
> fits your use case nicely (regularly swapping back and forth between 4.0 and
> 5.0 kernels).

If threaded application built using this trick can run on a vanilla 4.x
system, then it would be acceptable for me (others may dissagree).
But I still can't see how mergin -revivesa would hurt so badly.

-- 
Manuel Bouyer, LIP6, Universite Paris VI.           
Manuel.Bouyer%lip6.fr@localhost
     NetBSD: 26 ans d'experience feront toujours la difference
--