Re: relevant panic() when combining lagg(4), vlan(4) and bridge(4)

[Sverre Froyen <sverre%viewmark.com@localhost>]

It looks like the changes are not in 10.1.

Sverre

> On Jan 14, 2025, at 22:01, Jean-Yves Migeon (NetBSD) <jym%helkyn.org@localhost> wrote:
> 
> Le 07/01/2025 à 20:47, Sverre Froyen a écrit :
>> Hi Jean-Yves,
>> It looks like bug #57650 is relevant here. I had the same panic and ended up using this configuration successfully
>> # cat /etc/ifconfig.lagg0
>> !ifconfig wm0 up
>> !ifconfig wm1 up
>> create
>> laggproto lacp laggport wm0 laggport wm1
>> 192.168.2.1 netmask 255.255.255.0
>> mtu 1500
>> up
>> Regards,
> 
> So no luck here, the same panic() happens even with MTU set as shown in #57650.
> 
> I think there is a larger locking issue at play in lagg(4) that affects packet handling in untested paths; do you think it should be a separate PR or added to #57650?
> 
> Thanks,
> 
> ===== DDB trace after change =====
> 
> Stopped in pid 0.3 (system) at netbsd:breakpoint+Ox5: leave
> breakpoint() at netbsd:breakpoint+0x5
> vpanic() at netbsd:vpanic+0x183
> panic() at netbsd:panic+0x3c
> lockdebug_abort() at netbsd:lockdebug_abort+0x114
> mutex vector enter() at netbsd:mutex_vector_enter+Ox32b
> bridge_input() at netbsd:bridge_input+0x946
> vlan_input() at netbsd:vlan_input+0x143
> ether_input() at netbsd:ether_input+Ox4c2
> bridge_input() at netbsd:bridge_input+Oxa10
> lagg_input_ethernet() at netbsd:lagg_input_ethernet+Ox2ab
> if_percpuq_softint() at netbsd:if_percpuq_softint+0x8d
> softint_dispatch() at netbsd:softint_dispatch+0x95
> DDB lost frame for netbsd:Xsoftintr+0x4c, trying 0xffffb90939d9d0f0
> Xsoftintr() at netbsd:Xsoftintr+0x4c
> 
> -- 
> jym@