Hi Joe,
On Tue, 2020-02-11 at 13:40 -0500, Joe Korty wrote:
> On Tue, Feb 11, 2020 at 06:49:15PM +0100, Sebastian Andrzej Siewior
> wrote:
> > On 2020-02-11 09:00:07 [-0500], Joe Korty wrote:
> > > 4.14-rt: Fix wrong-variable use in irq_set_affinity_notifier.
> > >
> > > The bug was introduced in the 4.14-rt patch
> > >
> > > 0461-genirq-Handle-missing-work_struct-in-
> > > irq_set_affinit.patch
> > >
> > > The symptom is a NULL pointer panic in the i40e driver on
> > > system shutdown.
> >
> > ???
> > > 4.19-rt and above do not have this problem due to a refactoring.
> >
> > That would be Tom's to pick. Is v4.14-RT the only one affected? I
> > was
> > under the impression that we fixed it already in each stable rt
> > tree.
>
> A quick grep of all the -rt patch files in kernel.org/../projects/rt,
> newer
> than May, 2019, shows that 4.14 is the only one needing a fix.
>
Yeah, what happened here was that I screwed up when fixing a merge
conflict, and inadvertently changed old_notify->work to notify->work in
the fix.
Thanks for pointing this out - I'll include your patch doing the right
thing in the next update, coming up shortly..
Tom
> patch-4.14.170-rt74.patch.xz:
> + kthread_cancel_work_sync(¬ify->work);
> patch-4.4.208-rt191.patch.xz:
> patch-3.18.18-rt15.patch.xz:
> patch-5.0.21-rt16.patch.xz:
> patch-5.2.21-rt14.patch.xz:
> patch-5.4.17-rt9.patch.xz:
> patch-4.9.201-rt134.patch.xz:
>
> >
> > > Signed-off-by: Joe Korty <Joe.Korty@xxxxxxxxxxxxxxxxx>
> > >
> > > Index: b/kernel/irq/manage.c
> > > =================================================================
> > > ==
> > > --- a/kernel/irq/manage.c
> > > +++ b/kernel/irq/manage.c
> > > @@ -353,7 +353,7 @@ irq_set_affinity_notifier(unsigned int i
> > >
> > > if (old_notify) {
> > > #ifdef CONFIG_PREEMPT_RT_BASE
> > > - kthread_cancel_work_sync(¬ify->work);
> > > + kthread_cancel_work_sync(&old_notify->work);
> > > #else
> > > cancel_work_sync(&old_notify->work);
> > > #endif
> >
> > Sebastian
>
>
