Kernel panics after upgrade to R1

[newsense]

_15 was really broken on the 3rd FW, couldn't curl the kernel, opnsense-patch would timeout eventually complaining it cannot verify the sig - which was bonkers.

Managed to winscp the kernel and sig file and then I installed it with -zkr 24.7.b_7 -l /foldername and will see what happens.

The other two are happy for now on _7, with 60' and 80' uptime respectively.

[franco]

Let's just try this one:

# opnsense-update -zkr 24.7.r1_2

I placed a bet...


Cheers,
Franco

[computeralex92]

Let's just try this one:

# opnsense-update -zkr 24.7.r1_2

I placed a bet...


Cheers,
Franco

Just installed it, until now no problems or panic.

[Vasco]

Let's just try this one:

# opnsense-update -zkr 24.7.r1_2

I placed a bet...


Cheers,
Franco

Sorry for partially missing the tests. Submited a crash report after boot with this one but didn't see a "panic" in dmesg-

[franco]

I've replaced the original kernel by including this https://github.com/opnsense/src/commit/de60ffe06fd6

It may or may not be the right one, but it looks promising and I want to avoid people catching the bad one as best we can.


Cheers,
Franco

[newsense]

Just finished moving all 3 boxes to r1_2.

_b7 was there on all 3 with an ~8 hour uptime.

[planetf1]

Just to say I've not had any kernel panics with the initial 24.7.r1 - uptime 14 hours
n100 miniPC running Proxmox 8.2.4 16GB ram
opnsense 24.7.r1 in an 8GB VM with 1xintel i226v passthrough (wan) and 1 proxmox/linux bridge
Connection is pppoe, dual stack ipv4/v6
Simple config - using unbound, suricata (lan), crowdsec

All is 'just working'. Not seeing any unexpected kernel issues.

Nice job :-)

[newsense]

The kernel panics only happened on bare metal, virtualized worked ok.

Just a heads up for the other kernel testers, if you're on r1_2  from snapshots and check for updates the 24.7.r1 kernel will be installed and cause a reboot just because of the name change, otherwise it is the same kernel.

Code: [Select]

14.1-RELEASE-p2 FreeBSD 14.1-RELEASE-p2 stable/24.7-n267750-de60ffe06fd6 SMP amd64

[franco]

de60ffe06fd6 is the relevant part of the hotfixed kernel, yep

Since _15 was bad and _7 was good it was just a matter of an educated guess and it looks like we found it. Thanks all for the help!


Cheers,
Franco

[computeralex92]

The new kernel is working for me without any issue.
Thanks all for the testing and debugging this problem!

[franco]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=279899#c15

Of course, what looks like a proper fix found its way to FreeBSD's stable/14 branch yesterday afternoon


Cheers,
Franco

[ProximusAl]

I love your comment on there Franco.......

A man after my own heart!!!

Excellent job!

[franco]

Thanks, but it's safe to assume the people that matter in this won't appreciate the candidness. Still how does the old saying go? "Do good things and talk about it" is what I'd like to see.

Here's an amended kernel with the proper fix. I also have it on my box so fingers crossed.

# opnsense-update -zkr 24.7.r1_5


Cheers,
Franco

[newsense]

Moved the fleet to the 5th amendment

[newsense]

With regards to the bug, and after reading the thread on bugs.freebsd.org, I still can't say I understood why it appeared to work just fine on multiple virtual environments but trigger relatively quickly on bare metal... Given its nature I would have expected a similar and consistent crash regardless of where it was running