Solaris 2.6 and latest NTDOM code

[Gerald W. Carter]

On Fri, 26 Jun 1998, Jeremy Allison wrote:

> Bernie Kirby wrote:
> 
> >         Same problem here.
> >         core dump in 'get_nb_flags' called from line
> > 515 in nmbd_winsserver.c.
> > 
> >         nmb->additional data is Null.
> 
> This looks like memory corruption problems.
> In the function wins_process_name_registration_request()
> the same pointer (p->packet.nmb->additional) is used
> heavily. It is the contents of the pointer p that
> are copied into the userdata pointer, after the
> packet has been locked (line 807) so that the
> contents of the original packet are still kept
> for the wins_register_query_fail() function
> to use. If the nmb->additional pointer is non-zero
> in the wins_process_name_registration_request() then
> it should be non-zero in wins_register_query_fail().

This is where it showed up on me.
 
> If not - then someone is scribbling on memory.
> 
> Now Chris has been adding some changes to the
> HEAD branch nmbd to make the transition to a
> sensible WINS back-end (gdbm) easier, and at
> SGI they've been running with the HEAD branch
> wins server code for a while, so I doubt that
> any of Chris's changes will have done this,
> but I'm CC:ing him just to be sure :-).
> 
> To track this down you could try purify (if you
> have a copy) or compile Samba with -DMEM_MAN
> to link in the (simple) malloc/free checker
> that Andrew wrote. With this compiled in you
> can hit nmbd with a SIGUSR1 to have it check
> all malloced areas for corruption.
> 

Will try the -DMEM_MAN compile and see what happens.




Thanks Jeremy.

j-
________________________________________________________________________
                            Gerald ( Jerry ) Carter	
Engineering Network Services                           Auburn University 
jerry at eng.auburn.edu             http://www.eng.auburn.edu/users/cartegw

       "...a hundred billion castaways looking for a home."
                                  - Sting "Message in a Bottle" ( 1979 )