I can confirm that large packets raise kernel panic.
I was able to raise a panic by doing:
doas nmap -6 -Pn -p22 --data-length 1400
from another host. So potentially anyone can crash the router by sending
large packets.
Details of interfaces:
> em0:
> flags=a48843
> mtu 1500
> description:
>
Hi, I've recorded traffic that was sent from the git server before crashes.
Most likely panic is caused by sending a large packet (1514 bytes) by the
git server to the remote vpn client.
I can privately share the traffic dumps.
пн, 29 нояб. 2021 г. в 18:30, :
> >Synopsis: Kernel panic
Hi Ted,
On Mon, Dec 06, 2021 at 01:23:49PM -0700, Ted Bullock wrote:
| btw, is there a tool in base that I can use to grab a screenshot of
| what X is showing?
Try xwd(1). You can view its output with xwud(1), but there are also
several tools in ports that allow you to read these dumps, I use
On 2021-12-06 1:07 p.m., Mark Kettenis wrote:
> At this point we might be better off using the wsfb driver on sparc64.
> You'll lose modesetting capabilities, but I'm not sure that'd be a big
> loss. Or does the machine come up in an 8bpp instead of 24bpp mode?
It comes up in 24bpp, the display
> Date: Mon, 6 Dec 2021 12:54:52 -0700
> From: Ted Bullock
>
> On 2021-12-05 5:15 p.m., Theo de Raadt wrote:
> > Jonathan Gray wrote:
> >> On Sun, Dec 05, 2021 at 04:54:28PM -0700, Ted Bullock wrote:
> >>> Hey folks,
> >>>
> >>> Looking into another usability fault with the SunBlade 100. This
On 2021-12-05 5:15 p.m., Theo de Raadt wrote:
> Jonathan Gray wrote:
>> On Sun, Dec 05, 2021 at 04:54:28PM -0700, Ted Bullock wrote:
>>> Hey folks,
>>>
>>> Looking into another usability fault with the SunBlade 100. This time
>>> with the onboard video adapter. I'm seeing X segfault when
