[PATCH] Please review my session-info branch

Nadezhda Ivanova nivanova at samba.org
Sun Aug 15 08:45:48 MDT 2010


Hi Andrew,
I do not understand enough the purpose of your patch, but replacing
anonymous_session with system_session in acl.py is not acceptable - recently
I added some tests for denying anonymous access, which I will activate when
I am able to get this feature in master. The tests do work against windows
and are tested :). If you tell me what else I can use to establish an
anonymous connection, then I'll use it instead and the patch would be OK,
otherwise I need this binding.

Also, what is the purpose of
http://gitweb.samba.org/?p=abartlet/samba.git/.git;a=commitdiff;h=080d1a8e1c4f7ff1cb9c484df81cfd4cd6636726?
Are you sure it doesn't change the resulting token? We use this session
during provision to make sure the object's descriptors are generated
properly. I'll compile your branch and take a look at the resulting
descriptors to make sure they look as they should...

Regards,
Nadya

On Sun, Aug 15, 2010 at 4:37 AM, Andrew Bartlett <abartlet at samba.org> wrote:

> I've been working to rationalise the 'session_info' code in Samba4,
> partly to try and fix the issue that I introduced a while back, where
> changes to Samba4's smbtorture caused failures in 'make test' in Samba3.
>
> The problem was that smbtorture in this case runs an un-provisioned
> Samba4 server in the smbtorture binanry.  This change causes the
> anonymous case to be just hard-coded again, rather than referring to
> sam.ldb or privilages.ldb.
>
> Guenther,
>
> My tests now show that RPC-SPOOLSS-NOTIFY passes, but other spoolss
> tests still fail.  Is there anything more I need to to show that I've
> fixed https://bugzilla.samba.org/show_bug.cgi?id=7486 ?  The example
> shown can never have worked - we have always required a provision to
> launch the 'samba' server, so can you give me a more specific test that
> would work enough for what RPC-SPOOLSS-NOTIFY needs?
>
> Nadya,
>
> The question I have for you is:  Does this change break some use of the
> acl.py test that isn't picked up by make test?
>
> The changes are in my 'session-info' branch:
>
> http://gitweb.samba.org/?p=abartlet/samba.git/.git;a=shortlog;h=refs/heads/session-info
>
> Andrew Bartlett
> --
> Andrew Bartlett                                http://samba.org/~abartlet/<http://samba.org/%7Eabartlet/>
> Authentication Developer, Samba Team           http://samba.org
> Samba Developer, Cisco Inc.
>


More information about the samba-technical mailing list