From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Mon, 25 Sep 2000 18:11:21 +0200 From: Andrea Arcangeli Subject: Re: the new VMt Message-ID: <20000925181121.A27023@athlon.random> References: <20000925174138.D25814@athlon.random> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: ; from mingo@elte.hu on Mon, Sep 25, 2000 at 06:02:18PM +0200 Sender: owner-linux-mm@kvack.org Return-Path: To: Ingo Molnar Cc: Alan Cox , Marcelo Tosatti , Linus Torvalds , Rik van Riel , Roger Larsson , MM mailing list , linux-kernel@vger.kernel.org List-ID: On Mon, Sep 25, 2000 at 06:02:18PM +0200, Ingo Molnar wrote: > Frankly, how often do we allocate multi-order pages? I've just made quick The deadlock Alan pointed out can happen also with single page allocation if we in 2.4.x-current put a loop in GFP_KERNEL. > ie. 99.45% of all allocations are single-page! 0.50% is the 8kb You're right. That's why it's a waste to have so many order in the buddy allocator. Even more now that the hashtables should be allocated with the bootmem allocator! :) Chuck seen the slowdown of increasing the highest order allocation in his bench. But of course in 2.2.x we can't avoid that. Andrea -- 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/