On 8 Mar 2004, at 2:50 PM, Paul Higgins wrote:

> Has any progress been made in order to find a solution to the 'sleep 
> forever' problem?

I don't have any new information on the problem, other than a couple 
data points which may be of interest to those looking into it.

I see this problem with regularity on one machine, an Ultra 30 (both 
running 1.6.2_RC3 and 1.6ZG), but I have never seen it on another 
machine, a single-proc E250 (running various 1.6 release branch 
versions and -current versions up to 1.6ZK). These machines have quite 
a similar architecture. I wonder whether there's a difference here that 
might be helpful in locating the problem.

Aside from the inherent differences in the machines, the only other 
difference I can think of is that I run a serial console on the Ultra 
30, while the E250 is using the openprom-based text on its framebuffer.

- Geoff