[PATCH] memberof: check for empty arrays to avoid segfaults

[Sumit Bose]

Hi,

I'm quite confident that this patch can fix
https://fedorahosted.org/sssd/ticket/2430 . Unfortunately I have no idea
how to reproduce a condition which would trigger this issue, so I'm on
completely sure. Any ideas for reproducers are welcome.

bye,
Sumit

[Jakub Hrozek]

On Wed, Oct 29, 2014 at 03:26:19PM +0100, Sumit Bose wrote:
> Hi,
>
> I'm quite confident that this patch can fix
> https://fedorahosted.org/sssd/ticket/2430 . Unfortunately I have no idea
> how to reproduce a condition which would trigger this issue, so I'm on
> completely sure. Any ideas for reproducers are welcome.
>
> bye,
> Sumit

I agree that according to the traceback, this change would prevent us
from reaching "mbof_del_execute_op(first->children[0]);", so it's a good
change.

I haven't found a way to reproduce the issue and IIRC neither did Lukas.

The original reporter seems to be satisfied with a workaround and
doesn't reply to our requests for testing, so my say is that we push the
patch to master anyway.

ACK

[Jakub Hrozek]

On Tue, Nov 04, 2014 at 12:28:49PM +0100, Jakub Hrozek wrote:
> On Wed, Oct 29, 2014 at 03:26:19PM +0100, Sumit Bose wrote:
> > Hi,
> >
> > I'm quite confident that this patch can fix
> > https://fedorahosted.org/sssd/ticket/2430 . Unfortunately I have no idea
> > how to reproduce a condition which would trigger this issue, so I'm on
> > completely sure. Any ideas for reproducers are welcome.
> >
> > bye,
> > Sumit
>
> I agree that according to the traceback, this change would prevent us
> from reaching "mbof_del_execute_op(first->children[0]);", so it's a good
> change.
>
> I haven't found a way to reproduce the issue and IIRC neither did Lukas.
>
> The original reporter seems to be satisfied with a workaround and
> doesn't reply to our requests for testing, so my say is that we push the
> patch to master anyway.
>
> ACK

* master: a1bd8bc666df7fa696523ec8ec1dfe3d79780588