Somebody else saw processes take a long time to start; I see the
opposite: sometimes large processes take a very long time to exit, while
using all cpu. Here is a snapshot from top, several minutes after pan
started to exit:

124 processes: 8 runnable, 113 sleeping, 2 stopped, 1 on processor
CPU states:  0.9% user,  0.0% nice, 99.1% system,  0.0% interrupt,  0.0% idle
Memory: 405M Act, 213M Inact, 5408K Wired, 41M Exec, 85M File, 264M Free
Swap: 2028M Total, 979M Used, 1049M Free

  PID USERNAME PRI NICE   SIZE   RES STATE      TIME   WCPU    CPU COMMAND
27613 rhialto   64    0     0K   76M RUN        7:40 98.97% 98.97% pan
  505 rhialto    2    0    80M 2056K poll       8:05  0.05%  0.05% xconsole

and after this it went on for a couple of minutes more.
It's not only pan, I have also seen it with firefox, I think. In this
case pan actually crashed because it ran out of memory (ulimit -d was
set to unlimited, with 1GB physical memory which is the hard limit). I'm
not sure if that is relevant.

This is with NetBSD/amd64 2.0.2. I haven't seen this with NetBSD/alpha
2.0.

-Olaf.
-- 
___ Olaf 'Rhialto' Seibert      -- You author it, and I'll reader it.
\X/ rhialto/at/xs4all.nl        -- Cetero censeo "authored" delendum esse.