Solved winbind problems on Solaris x86 (was RE: winbind authenticator patches)

Hi,

Il 19.29 28/06/2002 Jerry Murdock ha scritto:
>Attached are patches for the below approach. Seem OK on both FBSD and
>RedHat 7.3.
>
>Please review.
>
>Thanks,
>Jerry

I have just made a test on a x86 Solaris where ntlm winbind authenticators
never worked, and now with the patches it works fine.

Regards

Guido

>----- Original Message -----
>From: "Jerry Murdock" <jmurdock@itraktech.com>
>To: "Squid Developers" <squid-dev@squid-cache.org>
>Cc: "Henrik Nordstrom" <hno@marasystems.com>; "Chemolli Francesco (USI)"
><ChemolliF@GruppoCredit.it>
>Sent: Friday, June 28, 2002 11:59 AM
>Subject: Re: RE: winbind authenticator patches
>
>
> > ----- Original Message -----
> > From: "Henrik Nordstrom" <hno@marasystems.com>
> > To: "Jerry Murdock" <jmurdock@itraktech.com>
> > Sent: Friday, June 28, 2002 8:25 AM
> > Subject: Fwd: RE: winbind authenticator patches
> >
> > > Subject: RE: winbind authenticator patches
> > > Date: Fri, 28 Jun 2002 12:08:27 +0200
> > > From: "Chemolli Francesco (USI)" <ChemolliF@GruppoCredit.it>
> > > To: "'Henrik Nordstrom'" <hno@marasystems.com>,
>squid-dev@squid-cache.org
> > >
> > > No, they have a problem (checked this with abartlet).
> > >
> > > They don't check the return code, only the struct. If the return code is
> > > false, then the result may very well be senseless garbage. Thus they
> > > both have to be checked, return code first and result then.
> > >
> > Please bear with me, I'm looking for a little guidance. I'm very willing
>to
> > track this down, but saying my C is a little rusty would be flattery.
> >
> > I think the real problem is that the return code is currently being
>compared
> > to WINBINDD_OK, when it should be compared to NSS_STATUS_SUCCESS.
> >
> > Under FBSD, these are not the same value(WINBINDD_OK=1,
> > NSS_STATUS_SUCCESS=0).
> >
> > I'm loading up a Linux session in VMware now for comparison, but I assume
> > they are the same value under Linux, otherwise I don't see how the code
> > could work there.
> >
> > If this sounds reasonable, is there any need to also check response.result
> > in the struct?
> >
> > Thanks,
> > Jerry
> >
>
>

-
=======================================================
Serassio Guido
Via Albenga, 11/4 10134 - Torino - ITALY
E-mail: guido.serassio@serassio.it
WWW: http://www.serassio.it
Received on Mon Jul 01 2002 - 03:27:24 MDT