l4-hurd
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: currency based virtual memory


From: Neal H. Walfield
Subject: Re: currency based virtual memory
Date: 07 Nov 2002 16:51:03 -0500
User-agent: Gnus/5.0808 (Gnus v5.8.8) Emacs/21.2

> > Well, it should not swap everything out.  But if you really way to,
> > then I imagine you would have to ask someone else to wake you up.
> 
> That brings up another issue I wanted to ask about before. What if a
> process doesn't give back RAM to the system when the memory manager
> asks it to? Your slides say that the memory manager can decide to
> revoke all mappings (which is more or less the same thing as killing
> the process, right?).

Almost: the VMM does not know how to kill a process gracefully.

> I find it a little scary that processes that miss a memory-return
> deadline will be killed with no mercy. Will they even get the chance
> to dump core?

The problem is the VMM does not know about the underlying OS
personalities.  One approach would be to send a notification to a task
in the OS personality that could handle the situation gracefully.  It
could decide to force a core dump or send it to swap or whatever.

> A better alternative might be to swap out the process.
[ snip ]

I just think that the VMM is the wrong place for this.  The OS
personality should worry about these things since it is the one
setting the policy anyway.




reply via email to

[Prev in Thread] Current Thread [Next in Thread]