[Samba] Bug 6870 resurfaced in Samba 4.2.10

Rebecca Gellman rebecca at starfleet-net.co.uk
Tue Oct 18 09:06:34 UTC 2016


 

On 2016-10-17 19:37, Jeremy Allison wrote: 

> On Mon, Oct 17, 2016 at 11:27:12AM -0700, Jeremy Allison wrote: 
> 
>> No, it's not so bad. libcli/cldap/cldap.c is clearly supposed to cope
>> with caller.cldap == NULL - you can see this in the function
>> cldap_search_state_destructor() which checks:
>> 
>> if (s->caller.cldap)
>> 
>> before doing anything. Can you try this (amended) patch ?
> 
> Scratch that - EBADPATCH (sorry :-).
> 
> Here's something I think might actually work :-).
> 
> Sorry,
> 
> Jeremy.

Glad you checked that... my next response would be "I have a gdb
backtrace here that disagrees with you...." ;) 

This patch looks good. net ads workgroup did indeed give the right
output, and allowed the rest of the steps in our sequence to complete
with no segfauls. 

This is part of the Smoothwall domain join mechanism, so there was
somewhat of an urgency on this one :) 

Thank you very much for your assistance; let me know if you need any
further patches testing on this bug. 

Regards 

Rebecca Gellman 
  


More information about the samba mailing list