From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail191.messagelabs.com (mail191.messagelabs.com [216.82.242.19]) by kanga.kvack.org (Postfix) with SMTP id 56D0A8D0039 for ; Mon, 28 Feb 2011 18:07:42 -0500 (EST) Date: Tue, 1 Mar 2011 00:07:12 +0100 From: Andrea Arcangeli Subject: Re: [PATCH 2/2] mm: compaction: Minimise the time IRQs are disabled while isolating pages for migration Message-ID: <20110228230712.GR22700@random.random> References: <1298664299-10270-1-git-send-email-mel@csn.ul.ie> <1298664299-10270-3-git-send-email-mel@csn.ul.ie> <20110228230131.GB1896@barrios-desktop> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20110228230131.GB1896@barrios-desktop> Sender: owner-linux-mm@kvack.org List-ID: To: Minchan Kim Cc: Mel Gorman , Andrew Morton , Arthur Marsh , Clemens Ladisch , Linux-MM , Linux Kernel Mailing List On Tue, Mar 01, 2011 at 08:01:31AM +0900, Minchan Kim wrote: > I am not sure it's good if we release the lock whenever lru->lock was contended > unconditionally? There are many kinds of lru_lock operations(add to lru, > del from lru, isolation, reclaim, activation, deactivation and so on). This is mostly to mirror cond_resched_lock (which actually uses spin_needbreak but it's ok to have it also when preempt is off). I doubt it makes a big difference but I tried to mirror cond_resched_lock. > Do we really need to release the lock whenever all such operations were contened? > I think what we need is just spin_is_contended_irqcontext. > Otherwise, please write down the comment for justifying for it. What is spin_is_contended_irqcontext? > This patch is for reducing for irq latency but do we have to check signal > in irq hold time? I think it's good idea to check the signal in case the loop is very long and this is run in direct compaction context. -- 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-mm.org/ . Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/ Don't email: email@kvack.org