samba - May 2023 - On Debian 12: nsupdate (as called from samba_dnsupdate) crashes named/bind9_dlz

Steven Monai via samba <samba at lists.samba.org>
wrote:
> On 2023-05-20 1:43 a.m., Rowland Penny via samba wrote:
>> On 20/05/2023 04:44, Steven Monai via samba wrote:
>>> I found an open bug in bugzilla that reports a very similar
assertion
>>> failure: "Bug 14030 - named crashes on DLZ zone update" 
>>> (https://bugzilla.samba.org/show_bug.cgi?id=14030). Any chance this
>>> Bug is related to what I'm seeing?
>> 
>> That appears to be a Samba problem, whilst yours appears to possibly be
>> a Bind9 problem.
It is a samba problem that causes bind9 to crash.
> Maybe. But nevermind that Bug; it appears to be FreeBSD-specific, and I
> am using Linux (more specifically Debian, on amd64 architecture).
I use the patch from the bug on debian and the chrashes of named I had
constantly are gone.

Rob
-- 
In C we had to code our own bugs. 
In C++ we can inherit them.

On Mon, 2023-05-22 at 11:15 +0200, Robert Grimm via samba
wrote:
> Steven Monai via samba <samba at lists.samba.org> wrote:
> > On 2023-05-20 1:43 a.m., Rowland Penny via samba wrote:
> > > On 20/05/2023 04:44, Steven Monai via samba wrote:
> 
> > > > I found an open bug in bugzilla that reports a very similar
assertion
> > > > failure: "Bug 14030 - named crashes on DLZ zone
update"
> > > > (https://bugzilla.samba.org/show_bug.cgi?id=14030). Any
chance this
> > > > Bug is related to what I'm seeing?
> > > 
> > > That appears to be a Samba problem, whilst yours appears to
possibly be
> > > a Bind9 problem.
> 
> It is a samba problem that causes bind9 to crash.
I totally agree. ?Our plugin is not well-restrained in the C symbol
names it exports. ?Longer term we should restrict the symbols the
plugin exports, like we do for libnss_winbind etc.
> > Maybe. But nevermind that Bug; it appears to be FreeBSD-specific,
> > and I
> > am using Linux (more specifically Debian, on amd64 architecture).
> 
> I use the patch from the bug on debian and the chrashes of named I
> had
> constantly are gone.
I'm very sorry that this bug hasn't got the final push into our
releases, particularly given all the work and diagnostics that have
gone into it.

I've reviewed the patch and assigned it to our release manager, it
should end up in new releases soon. ?In the meantime you can file a
debian bug to get the patch into those packages sooner, as they have
now been vetted upstream.

Andrew Bartlett
> Rob
> -- 
> In C we had to code our own bugs. 
> In C++ we can inherit them.
I love it!

-- 
Andrew Bartlett (he/him)        https://samba.org/~abartlet/
Samba Team Member (since 2001)  https://samba.org
Samba Developer, Catalyst IT    https://catalyst.net.nz/services/samba