From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Mon, 9 Oct 2000 21:27:47 +0100 (BST) From: James Sutherland Subject: RE: [PATCH] VM fix for 2.4.0-test9 & OOM handler In-Reply-To: <32B7FDA9BF4CE64FB10864F5E4972A031DFAD3@cpt-sas-ex01.corptst.amazon.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-mm@kvack.org Return-Path: To: "Wagner, Dave" Cc: mingo@elte.hu, Ed Tomlinson , Mark Hahn , Marco Colombo , Rik van Riel , linux-mm@kvack.org List-ID: On Mon, 9 Oct 2000, Wagner, Dave wrote: > > -----Original Message----- > > From: Ingo Molnar [mailto:mingo@elte.hu] > > Sent: Monday, October 09, 2000 11:02 AM > > To: Ed Tomlinson > > Cc: Mark Hahn; Marco Colombo; Rik van Riel; linux-mm@kvack.org > > Subject: Re: [PATCH] VM fix for 2.4.0-test9 & OOM handler > > > > > > On Mon, 9 Oct 2000, Ed Tomlinson wrote: > > > > > What about the AIX way? When the system is nearly OOM it sends a > > > SIG_DANGER signal to all processes. Those that handle the > > signal are > > > not initial targets for OOM... Also in the SIG_DANGER > > processing they > > > can take there own actions to reduce their memory usage... (we would > > > have to look out for a SIG_DANGER handler that had a memory leak > > > though) > > > > i think 'importance' should be an integer value, not just a > > 'can it handle > > SIG_DANGER' flag. > > > In a perfect world, perhaps. But how many people/systems are going to > have a well-thought out distribution of "importance" values. It's > probably too much to have people even set a single boolean value > reasonably. You could say exactly the same about niceness: how many people have a well-thought out distribution of nice values? Yet the mechanism works reasonably well: important system things get -ve values, normal user processes get the default, CPU hogs get niced down a bit. > How about a bit in the executable to say "unimportant". Netscape, would > of course, have this bit set. ;-) I think an integer, just like nice, would be the way to go. James. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux.eu.org/Linux-MM/