Final fixes for bug #7119.

[Jeremy Allison]

On Wed, Feb 10, 2010 at 07:27:36PM +0100, Stefan (metze) Metzmacher wrote:
> Jeremy Allison schrieb:
> > On Wed, Feb 10, 2010 at 05:37:42PM +0100, Stefan (metze) Metzmacher wrote:
> >> Jeremy Allison schrieb:
> >>> On Tue, Feb 09, 2010 at 10:05:22AM -0800, Jeremy Allison wrote:
> >>>> On Tue, Feb 09, 2010 at 12:00:55PM -0600, Stefan Metzmacher wrote:
> >>>>> The branch, master has been updated
> >>>>>        via  1686a5e... s3:smbd: Fix really ugly bool vs. int bug!!!
> >>>> Oh I missed that in the review, sorry :-(. We'll need that
> >>>> in 3.5.0. Can you attach to the bug report and get Karolin
> >>>> to push ?
> >>> Karolin,
> >>>
> >>> 	I've updated:
> >>>
> >>> https://bugzilla.samba.org/show_bug.cgi?id=7119
> >>>
> >>> with the fixes metze just added to master. We'll need
> >>> these to fully fix this for 3.5.0.
> >> I just found why the transs continue messages isn't always generated...
> >>
> >> We don't respect the clients buffer size (from the first session setup)
> >> when doing extended security... Is that correct?
> > 
> > Yes, I think that's correct. I recall this code being added
> > initially via Isilon, on testing with Win2K3 servers.
> 
> It's there since the first support of extended security.
> Are you sure it wasn't just forgotten in
> b728042334f67738fd1a6fdd03e619bdb78fe06a
> 
> Tridge: do you remember this?
> 
> We seem to behave differently in source4/smb_server/smb/sesssetup.c
> 
> >> My w2k3 test servers is happy to get the large responses in one trans
> >> response. So I'm not sure if it's a bug or a feature...
> 
> I guess we should better test it...


Yes, a torture test here would be really valuable.

Jeremy.