From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by kvack.org (8.8.7/8.8.7) with ESMTP id NAA00385 for ; Mon, 25 Jan 1999 13:30:23 -0500 Date: Mon, 25 Jan 1999 10:27:30 -0800 (PST) From: Linus Torvalds Subject: Re: MM deadlock [was: Re: arca-vm-8...] In-Reply-To: <199901251625.QAA04452@dax.scot.redhat.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-mm@kvack.org To: "Stephen C. Tweedie" Cc: Alan Cox , werner@suse.de, andrea@e-mind.com, riel@humbolt.geo.uu.nl, Zlatko.Calusic@CARNet.hr, ebiederm+eric@ccr.net, saw@msu.ru, steve@netplus.net, damonbrent@earthlink.net, reese@isn.net, kalle.andersson@mbox303.swipnet.se, bmccann@indusriver.com, bredelin@ucsd.edu, linux-kernel@vger.rutgers.edu, linux-mm@kvack.org List-ID: On Mon, 25 Jan 1999, Stephen C. Tweedie wrote: > > Regarding the former, is there any chance you'd consider adding a kswapd > wakeup when low_memory gets set in get_free_pages()? Being able to > respond to a burst in network traffic without locking up is not exactly > a minor issue. I did that, only to revert it later, because I didn't think it would make any difference - processes that get to that point will try to free up memory on their own anyway. Note that it wouldn't ever trigger for GFP_ATOMIC allocations, so I suspect you haven't actually _tried_ it? For a machine that gets burst of network traffic with nothing else going on, adding it should essentially amount to a no-op. I'll look at your other patch. Linus -- To unsubscribe, send a message with 'unsubscribe linux-mm my@address' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://humbolt.geo.uu.nl/Linux-MM/