> 
> This patch *does* seem to avoid the complete freezes.  Instead of
> dirtying every page in the active list, exhasting the free list, and
> then freezing whilst the dirty pages are dribbled out to disk at
> v_min_free/sec, the VM system is now cleaning the pages continuously.
> Pages are being written back to disk at a steady rate of about 5% of
> the pageout rate the hardware I'm using can sustain (e.g., what it
> reaches during the free-running part of the cyclic, pathological
> case.)
> 
> 
Hmmm... Does the port that you are using support pageout clustering?
If it doesn't, then you are getting about the kind of perf that I
would expect.  If your disk drive does not write-behind buffer, then
you will get only one page per rotation.  This is the reason that we
(FreeBSD) implemented write clustering of multiple pages long ago.
Read clustering is also helpful, but not nearly as much in a paging
situation.  (Read clustering is most useful in startup of large
programs, and secondarily useful in a low-memory paging situation.)

I would expect on a 3600 rpm drive to get about 60 pages/sec without
clustering...

John