Re: xennet bug?

[Manuel Bouyer <bouyer%antioche.eu.org@localhost>]

On Sat, Jan 19, 2019 at 11:03:04AM -0500, Amitai Schleier wrote:
> [...]
> But the third one, just now, looks different. rmind@ suggested this might be
> a bug in the xennet driver:
> 
>     panic: kernel diagnostic assertion "!cpu_intr_p()" failed: file
> "../../../../net/bpf.c", line 1564
>     cpu0: Begin traceback...
>     vpanic() at netbsd:vpanic+0x140
>     kern_assert() at netbsd:kern_assert+0x48
>     _bpf_mtap() at netbsd:_bpf_mtap+0x498
>     xennet_softstart() at netbsd:xennet_softstart+0x3cd
>     xennet_handler.part.3() at netbsd:xennet_handler.part.3+0x1c
>     xennet_handler() at netbsd:xennet_handler+0x18
>     intr_biglock_wrapper() at netbsd:intr_biglock_wrapper+0x1d
>     evtchn_do_event() at netbsd:evtchn_do_event+0x27d
>     do_hypervisor_callback() at netbsd:do_hypervisor_callback+0x17a
>     hypervisor_callback() at netbsd:hypervisor_callback+0xa1
>     idle_loop() at netbsd:idle_loop+0x19c
>     cpu0: End traceback...
> 
> I don't know how to reproduce, but some sort of panic-reboot seems to happen
> multiple times a day now. What can I try to narrow this down?

I don't think it will help with the npf-related one, but the attached patch
should fix this one. Can you try it ?

-- 
Manuel Bouyer <bouyer%antioche.eu.org@localhost>
     NetBSD: 26 ans d'experience feront toujours la difference
--

Index: if_xennet_xenbus.c
===================================================================
RCS file: /cvsroot/src/sys/arch/xen/xen/if_xennet_xenbus.c,v
retrieving revision 1.70.6.1
diff -u -p -u -r1.70.6.1 if_xennet_xenbus.c
--- if_xennet_xenbus.c	26 Jul 2018 23:55:29 -0000	1.70.6.1
+++ if_xennet_xenbus.c	19 Jan 2019 18:52:07 -0000
@@ -953,7 +953,7 @@ again:
 end:
 	if (ifp->if_flags & IFF_OACTIVE) {
 		ifp->if_flags &= ~IFF_OACTIVE;
-		xennet_softstart(sc);
+		softint_schedule(sc->sc_softintr);
 	}
 }