> what causes pages to be paged out is not that processes are dormant,
> it is that memory is low.  pages which have been used "recently" are
> not paged out but rather put back at the end of the list to consider.

Hmm. But if the "frozen X server" isn't getting paged out, then what is
keeping it from running?  Is there any chance that intensive UBC activity
could cause an I/O bound process to obtain an artificially high scheduling
priority (thus hogging the CPU)?

> when we're looking at a given page, we don't know what process or processes
> have it mapped, or even whether it's mapped at all.  it would be possible to
> add a bunch of code to keep track of this, but it would be prohibitively
> expensive in terms of cpu usage.  so how would we know whether any process
> is likely to need this page again in the next few seconds?

Suppose I postulate an additional algorithm that watches for dormant
processes and pages them out.

Then the simple fact that the page is still in memory means that we must
presume it might be wanted again in the next few seconds.

Todd Whitesel
toddpw @ best.com