From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from [83.222.19.8] (helo=mail.asoft.ru) by linuxtogo.org with esmtp (Exim 4.69) (envelope-from ) id 1Lq1vU-0003kU-Lb for openembedded-devel@openembedded.org; Sat, 04 Apr 2009 11:12:34 +0200 Received: from localhost ([95.72.100.66]) (authenticated bits=0) by mail.asoft.ru (8.13.4/8.13.4) with ESMTP id n3499q03020186 for ; Sat, 4 Apr 2009 13:09:55 +0400 Date: Sat, 4 Apr 2009 13:09:46 +0400 From: Dmitry Artamonow To: openembedded-devel@openembedded.org Message-ID: <20090404090946.GA13413@rainbow> References: <1237150071-24115-1-git-send-email-mad_soft@inbox.ru> MIME-Version: 1.0 In-Reply-To: User-Agent: Mutt/1.5.16 (2007-06-09) X-Greylist: Sender succeeded SMTP AUTH authentication, not delayed by milter-greylist-2.0.2 (mail.asoft.ru [0.0.0.0]); Sat, 04 Apr 2009 13:09:56 +0400 (MSD) X-SA-Exim-Connect-IP: 83.222.19.8 X-SA-Exim-Mail-From: mad@mail.asoft.ru X-Spam-Checker-Version: SpamAssassin 3.2.5 (2008-06-10) on serenity X-Spam-Level: X-Spam-Status: No, score=-1.2 required=5.0 tests=AWL,BAYES_00,RDNS_NONE autolearn=no version=3.2.5 X-SA-Exim-Version: 4.2.1 (built Wed, 25 Jun 2008 17:14:11 +0000) X-SA-Exim-Scanned: Yes (on linuxtogo.org) Subject: Re: [PATCH/RFC] base-files: add noatime to rootfs mount options X-BeenThere: openembedded-devel@lists.openembedded.org X-Mailman-Version: 2.1.11 Precedence: list Reply-To: openembedded-devel@lists.openembedded.org List-Id: Using the OpenEmbedded metadata to build Distributions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 04 Apr 2009 09:12:34 -0000 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline On 13:48 Sun 29 Mar , Koen Kooi wrote: > On 15-03-09 21:47, Dmitry Artamonow wrote: >> In most machines' fstabs rootfs mounted with 'defaults' option. >> This implies that atime logging is on - i.e. on every read from >> filesystem kernel writes atime stamp to files' metadata. >> Here in embedded world, where rootfs is usually stored on some >> kind of flash (internal NAND/NOR, or external SD/MMC/CF card) >> it is a real problem, because higher number of writes causes >> flash wearing, higher system load and higher chance of filesystem >> corruption. >> >> Fix it by adding noatime to rootfs mount options in fstab. > > Isn't 'relatime' a better option? Maybe yes, but we need more care in this case, because relatime has been introduced in 2.6.20 [1], and there's still can be machines using older kernels (judging solely by looking at the recipes in recipes/linux). I fear that we may break booting on these machines by passing unsupported option to mount. [1] http://kernelnewbies.org/Linux_2_6_20#head-6804816ab4f07548635032afbedd3e1f4eb3d1d7 > regards, > > Koen > -- Best regards, Dmitry "MAD" Artamonow