Re: unconfiguring swap at shutdown

[Jason Thorpe <thorpej%shagadelic.org@localhost>]

On Sep 2, 2008, at 9:16 AM, Steven M. Bellovin wrote:

> In principle, you're right; in practice, it's all just signal  
> delivery,
> sometimes to processes that appear to be blocked waiting for something
> important.  We'd need a different mechanism to terminate processes  
> with
> very extreme prejudice.

No, I'm not talking about signal delivery.  Signal delivery is a very  
specific code path that only executes when a process is about to  
return to userspace.

Yes, I'm talking about a different mechanism.

-- thorpej