From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <20000927182028.B14797@saw.sw.com.sg> Date: Wed, 27 Sep 2000 18:20:28 +0800 From: Andrey Savochkin Subject: Re: the new VMt References: <20000925192453.R2615@redhat.com> <20000925123456.A16612@hq.fsmlabs.com> <20000925202549.V2615@redhat.com> <20000925140419.A18243@hq.fsmlabs.com> <20000925171411.A2397@codepoet.org> <20000926091744.A25214@hq.fsmlabs.com> <20000926170406.C1343@redhat.com> <20000926110247.A4698@codepoet.org> <20000926180820.E1343@redhat.com> <20000926114501.B4780@codepoet.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <20000926114501.B4780@codepoet.org>; from "Erik Andersen" on Tue, Sep 26, 2000 at 11:45:02AM Sender: owner-linux-mm@kvack.org Return-Path: To: Erik Andersen Cc: "Stephen C. Tweedie" , yodaiken@fsmlabs.com, MM mailing list , linux-kernel@vger.kernel.org List-ID: On Tue, Sep 26, 2000 at 11:45:02AM -0600, Erik Andersen wrote: [snip] > "Overcommit" to me is the same things as Mark Hemment stated earlier in this > thread -- the "fact that the system has over committed its memory resources. > ie. it has sold too many tickets for the number of seats in the plane, and all > the passengers have turned up." Basically any case where too many tickets > have been sold (applied to the entire system, and all subsystems). [snip] > If the Beancounter patch lets the kernel count "passengers", classify them > (with user hinting) so the pilot and flight attendants (init, X, or whatever) > always stay on the plane, and has some sane predictable mechanism for booting > non-priveledged passengers, then I am all for it. That's exactly what I'm doing. > How does one provide the kernel with hints as to which processes are sacred? > Where does one find this beancounter patch? How much weight does it add to > the kernel? ftp://ftp.sw.com.sg/pub/Linux/people/saw/kernel/user_beancounter/UserBeancounter.html The current version has some drawbacks, and one of them is the performance. Memory accounting is implemented as a kernel thread which goes through page tables of processes (similar to kswapd), and it appears to consume 1-5% of CPU (depending on number of processes). I consider it unacceptable, and have started reimplementation of the process memory accounting from the beginning. Best regards Andrey -- 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/