Re: movidis x16 hard lockup using 2.6.33

[Peter 'p2' De Schrijver <p2@debian.org>]

Hi,

> * David Daney (ddaney@caviumnetworks.com) [100329 18:54]:
> > [...]
> 
> I need to admit I'm sorry but I can't read enough from the log file.
> Serial console output is http://alius.ayous.org/~aba/screenlog.0
> 
> This kernel is with the NMI-handler but without the "don't reboot"
> patch:
> Linux version 2.6.34-rc2-dsa-octeon (aba@gabrielli) (gcc version 4.4.3
> (Debian 4.4.3-3) ) #5 SMP Sat Mar 27 10:16:03 UTC 2010
> 
> and this one has NMI-handler plus the small loop:
> Linux version 2.6.34-rc2-dsa-octeon (aba@gabrielli) (gcc version 4.4.3
> (Debian 4.4.3-3) ) #6 SMP Mon Mar 29 22:23:26 UTC 2010
> 
> 
> Unfortunatly it looks like we get neither the information nor don't
> reboot, but adding the NMU handler changed behaviour from "machine
> freezes" to "machine reboots".
> 

http://zobel.ftbfs.de/.x/lucatelli-nmi-watchdog-output.txt 
Dump of one of those hangs. Most cores seem to be stuck in wait 
(0xffffffff81100b80), except for core 1 which is in octeon_irq_ciu0_ack 
(octeon_irq_ciu0_ack).

Any insight welcome. 

Thanks,

Peter.