Subject: Re: saslpasswd2 and virtdomains
From: Igor Brezac (igor at ipass dot net)
Date: Wed May 21 2003 - 09:09:56 EDT
On Wed, 21 May 2003, Rob Siemborski wrote:
> On Tue, 20 May 2003, Ken Murchison wrote:
> > > You are correct, I am trying to debug too many things at the same time. ;(
> > > cmd_login() calls canonify_userid() first.
> > >
> > > How about this?
> > I don't see a problem with this if it solves your problem. It doesn't seem to
> > break anything in my testing.
The patch is no good becuase canon_user points to the same string as user.
canon_userid() needs use a different buffer to generate canon_user or
functions calling canon_userid() needs to pass a copy of user to the
> This looks like its just masking a problem to me. Username
> canonicalization should really be idempotent (otherwise, what is a
> cannonical userid?).
The idea here is to pass raw userid to sasl_checkpass, otherwise
sasl_checkpass canonifies already canonified user.
What do you think?