[PATCHSET] Samba AD with MIT Kerberos

[Andrew Bartlett]

On Mon, 2017-03-13 at 08:29 +0100, Andreas Schneider via samba-
technical wrote:
> Hello,
> 
> after more than 3 years of work I finally got this:
> 
> 
> 	ALL OK (14658 tests in 2030 testsuites)
> 
> 
> The testsuite completed for the first time!

Sadly back on internal Heimdal it fails with:

../source4/kdc/pac-glue.c: In function ‘samba_make_krb5_pac’:
../source4/kdc/pac-glue.c:425:22: error: expected ‘=’, ‘,’, ‘;’, ‘asm’ or ‘__attribute__’ before ‘{’ token
  krb5_data null_data {
                      ^
../source4/kdc/pac-glue.c:426:3: error: expected expression before ‘.’ token
   .length = 0,
   ^
../source4/kdc/pac-glue.c:426:14: error: left-hand operand of comma expression has no effect [-Werror=unused-value]
   .length = 0,
              ^
../source4/kdc/pac-glue.c:427:15: error: left-hand operand of comma expression has no effect [-Werror=unused-value]
   .data = NULL,
               ^
../source4/kdc/pac-glue.c:519:8: error: ‘null_data’ undeclared (first use in this function)
       &null_data);
        ^~~~~~~~~
../source4/kdc/pac-glue.c:519:8: note: each undeclared identifier is reported only once for each function it appears in
cc1: all warnings being treated as errors
Waf: Leaving directory `/data/samba/git/samba/bin'
Build failed:  -> task failed (err #1): 
	{task: cc pac-glue.c -> pac-glue_13.o}


Now I'm sure I could fix that up pretty quickly, but presumably there
are some other niggles along the way.  If you could get me a series
that passes a full autobuild that would be great, as I would like to
push it soon.

Beyond that:

* Can you please look carefully at the interaction of the docs test
with the new parameter?  I think you need to add mit kdc command to the
special_cases in docs.py:SmbDotConfTests(), or pass in the default via
an XML entity (see the dynconfig code for how that is done). 

* "param: Add 'mit kdc config' option to smb.conf" adds a new parameter
to the same file as an existing parameter.  The current practice is one
per file

* "s4-torture: Fix reauth tests with smaller clockskew grace time"
Why does this patch undo "selftest: Set clowskew grace time to 1
second"?

* For "waf: Create kerberos_implementation.py for provisioning" I
wonder if getting a variable from the C code might be more robust, like
we use for detecting if we have the NTVFS file server?

Thanks for all your hard work on this, and I look forward to looking at
this again soon!

Andrew Bartlett