From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Wed, 04 Feb 2004 16:50:31 +0900 From: IWAMOTO Toshihiro Subject: Re: Active Memory Defragmentation: Our implementation & problems In-Reply-To: <1075878652.14155.416.camel@nighthawk> References: <20040204050915.59866.qmail@web9704.mail.yahoo.com> <1075874074.14153.159.camel@nighthawk> <20040204065717.EFB277049E@sv1.valinux.co.jp> <1075878652.14155.416.camel@nighthawk> MIME-Version: 1.0 (generated by SEMI 1.14.3 - "Ushinoya") Content-Type: text/plain; charset=US-ASCII Message-Id: <20040204075031.1BC10704A1@sv1.valinux.co.jp> Sender: owner-linux-mm@kvack.org Return-Path: To: Dave Hansen Cc: IWAMOTO Toshihiro , Alok Mooley , Linux Kernel Mailing List , linux-mm , "Martin J. Bligh" List-ID: At 03 Feb 2004 23:10:52 -0800, Dave Hansen wrote: > remap_onepage() is quite a function. 300 lines. It sure does cover a > lot of ground. :) > > Defragmentation is a bit easier than removal because it isn't as > mandatory. Instead of having to worry about waiting on things like > writeback, the defrag code can just bail. Waiting for !pagewriteback and writing back page at the beginning of remap_onepage() are a sort of "easy part". We need to wait for exclusive access of a page before copying anyway, and interesting things such as vmtruncate can happen while waiting for it. I don't think the code can be much shorter without assuming a single processor !CONFIG_PREEMPT system. -- IWAMOTO Toshihiro -- 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/ . Don't email: aart@kvack.org