Bug#1041363: nft BUG: kernel NULL pointer dereference, address: 0000000000000038

[Daniel Gröber <dxld@darkboxed.org>]

Hi Salvatore,

On Sat, Jul 22, 2023 at 10:07:39PM +0200, Salvatore Bonaccorso wrote:
> On Tue, Jul 18, 2023 at 02:35:25AM +0200, Daniel Gröber wrote:
> > I got the following BUG on my router while working on my nftables
> > ruleset. After this happened network connectivity was broken quite severely
> > so some internal state might have gotten messed up too. An attempted reboot
> > never completed and a hard power cut was necessary.
>
> As this is not the newest kernel in bookworm, please test with
> 6.1.38-1. 
> 
> Are you able to reliably reproduce the issue and can share the poc?

Unfortunately this bug only reared it's ugly head once so far. I upgraded
to 6.1-38-1 just after sending this report.

Since that upgrade I have

[   27.057795] WARNING: CPU: 0 PID: 1180 at net/core/flow_dissector.c:1016 __skb_flow_dissect+0xa91/0x1cd0

on every boot on the two of my routers. Is that a known issue? I can't seem
to find any reports but this also happens on my workstation. Let me know if
I should open another bug for that.

The original bug is likely very hard to trigger if it still exists. It's
hard to reconstruct the various changes I was making while testing my
ruleset :/

I'd be happy to have a quick look at the code to see if I can deduce what
triggered it, but I'm not familliar with how to parse these kernel BUG
traces. Could you point me to any docs on how to get some line numbers for
that backtrace?

Thanks,
--Daniel