[Ksummit-2012-discuss] [ATTEND] kernel core dump and "dying breath"
dan.carpenter at oracle.com
Wed Jun 27 11:00:12 UTC 2012
On Mon, Jun 25, 2012 at 08:47:27PM -0700, H. Peter Anvin wrote:
> On 06/25/2012 02:56 PM, Jiri Kosina wrote:
> >>That may have been the turning point at which it stopped working, which
> >>would explain it. Shame.
> >There also was a funny patchset which made the kernel output the panic
> >message and stacktrace via morsecode on keyboard LEDs.
> >Same problem -- easy on PS/2 keyboards, but scheduling all the work
> >necessary for completing URBs is just not possible from inside panic.
> Yeah, that's pretty much what it comes down to. Unless we want to
> end up with an entirely separate HID stack just for that... but then
> we'd be better off dumping to a USB stick.
Another idea would be to not scroll the message off the screen in
the first place. Maybe we could look for the [ cut here ] line and
stop when it hits the top of the screen?
I'm not sure if that's something which is even possible.
More information about the Ksummit-2012-discuss