Subject: Re: 3.0: a summary
To: Izumi Tsutsui <tsutsui@ceres.dti.ne.jp>
From: Christopher SEKIYA <wileyc@rezrov.net>
List: port-sgimips
Date: 12/05/2005 22:25:45
On Mon, Dec 05, 2005 at 10:06:21PM +0900, Izumi Tsutsui wrote:

> (though I'm not sure if it could cause silent reboot. By watchdog?)

My best guess is that the crime ASIC will assert RESET when a) it sees a
threshold number of memory errors within a set amount of time, or b) a memory
error occurs in a memory region that the crime considers critical.

> The real problem is the SGI's design which implements
> non coherent I/O on R10k systems??

As I understand it, yes.  I've been told that there's a fair amount of
workaround in IRIX for the R10k problem.
-- 

-- Chris
	GPG key FEB9DE7F (91AF 4534 4529 4BCC 31A5  938E 023E EEFB FEB9 DE7F)