Re: [PATCH 1/3] net/bridge/br_if.c: don't use _WORK_NAR

[Oleg Nesterov <oleg@tv-sign.ru>]

On 02/19, Jarek Poplawski wrote:
>
> On Mon, Feb 19, 2007 at 12:43:59AM +0300, Oleg Nesterov wrote:
> > Afaics, noautorel work_struct buys nothing for "struct net_bridge_port".
> >
> > If del_nbp()->cancel_delayed_work(&p->carrier_check) fails, port_carrier_check
> > may be called later anyway. So the reading of *work in port_carrier_check() is
> > equally unsafe with or without this patch.
> 
> I think this _WORK_NAR is to give some additional
> control, but also is more logical: it lets to decide
> when the work_struct is really release-able

Sadly, it doesn't help here.

                                              (and it's
> definitely not before work function is called, as
> without noautorel).

kfree() doesn't check WORK_STRUCT_PENDING, it makes no
difference if it is set or not when work->func() runs.

> So, even if this functionality isn't used now, I can't
> see what changing this could buy.

We are going to kill _NAR stuff.

Oleg.