Re: OpenSSL and Bug #1716

Hi Robert,

At 12.21 27/08/2006, Robert Collins wrote:

>On Sun, 2006-08-27 at 12:17 +0200, Guido Serassio wrote:
> > Hi Henrik,
> >
> > At 00.22 27/08/2006, Guido Serassio wrote:
> > >Tomorrow morning I will try to define a complete test suite, still I
> > >don't have found a rule of how this diabolic function prototype
> "works" ... :-)
> >
> > I have completed the tests, and this is the result:
> > http://www.squid-cache.org/bugs/attachment.cgi?id=1197
> >
> > So:
> >
> > - When using OpenSSL 9.8.x, the current code is fine
> > - When using OpenSSL 9.7.x, if the compiler is gcc 3.4.x, the
> > behaviour is predictable
> > - But when using OpenSSL 9.7.x with gcc 3.3.x things become
> complicated ....
> >
> > I think that the only resolution for gcc 3.3.x problem is a test into
> > configure.
>
>Or we could require openSSL 9.8.x?

I think that is not feasible.

On some Unix platforms (AIX, Irix and HP Tru64) is not so easy to
find recent binary packages and is not so simple build an own package
from sources: sometimes even the compiler is not available ...

An example: one month ago I have must build Samba 3 with Kerberos
support on an AIX machine: nearly 2 days of build before having a
working Samba .... :-(

And just a not so complex test into configure should be able to
address the problem.

Regards

Guido

-
========================================================
Guido Serassio
Acme Consulting S.r.l. - Microsoft Certified Partner
Via Lucia Savarino, 1 10098 - Rivoli (TO) - ITALY
Tel. : +39.011.9530135 Fax. : +39.011.9781115
Email: guido.serassio@acmeconsulting.it
WWW: http://www.acmeconsulting.it/
Received on Sun Aug 27 2006 - 04:35:13 MDT