named kills raspberry pi

[Steve O'Hara-Smith]

On Fri, 8 Feb 2013 04:58:27 -0700
Warner Losh <imp at bsdimp.com> wrote:

> 
> On Feb 8, 2013, at 12:57 AM, Steve O'Hara-Smith wrote:
> 
> > On Thu, 7 Feb 2013 22:32:40 -0800
> > Tim Kientzle <tim at kientzle.com> wrote:
> > 
> >> 
> >> On Feb 7, 2013, at 2:30 PM, Steve O'Hara-Smith wrote:
> >> 
> >>> 	Hi,
> >>> 
> >>> 	As soon as I start named either by rc.d script or directly,
> >>> whether in an ssh session or directly on the video console, with or
> >>> without chroot, my early model 256MB model B pi locks up solid, no
> >>> response on console, all ssh sessions freeze, no response to ping,
> >>> nothing until power cycle. I can build ports all day long, but if I
> >>> start named everything stops dead.
> >> 
> >> $ truss named -f
> > 
> > 	Yep, nice idea, that or ktrace. Sadly there's no truss built
> > for the rpi, as for ktrace there was no sign of the output file after
> > reboot.
> > 
> >> If you're lucky, you might catch some of the last system calls
> >> before it hangs, which might give a clue.
> > 
> > 	I think Warner has pointed me in the right direction, apart from
> > anything else it looks like there's a three year old fix for the
> > ras_start address that's never been committed - I'll be trying that
> > today.
> 
> I thought that fix had been committed...  But the issue isn't that fix,
> but rather the difference between armv6 atomics and armv[45] atomics. We
> shouldn't be using ras* anything at all on armv6.

	I had a nasty feeling it might be, I was just hoping it might still
work anyway. Failing that I'll look into an armv6 implementation of
cpmxchg. Either way if I get it to work I'll post a patch to this list.

-- 
Steve O'Hara-Smith                          |   Directable Mirror Arrays
C:>WIN                                      | A better way to focus the sun
The computer obeys and wins.                |    licences available see
You lose and Bill collects.                 |    http://www.sohara.org/