From louis@mizi.com Tue Apr 08 04:03:12 2008 Received: from mail.mizi.com ([61.107.31.33]) by stoneboat.aleph1.co.uk with esmtps (TLS-1.0:DHE_RSA_3DES_EDE_CBC_SHA1:24) (Exim 4.63) (envelope-from ) id 1Jj46w-0005Vx-V6 for yaffs@lists.aleph1.co.uk; Tue, 08 Apr 2008 04:03:12 +0100 Received: from [61.107.31.74] (peach.mizi.com [61.107.31.74]) by mail.mizi.com (8.12.11/8.12.11) with ESMTP id m3832pU4019010; Tue, 8 Apr 2008 12:02:51 +0900 Message-ID: <47FAE05B.3050303@mizi.com> Date: Tue, 08 Apr 2008 12:02:51 +0900 From: Louis JANG Organization: MIZI Research, Inc. User-Agent: Thunderbird 1.5.0.14pre (X11/20071023) MIME-Version: 1.0 To: Charles Manning References: <47F9C698.5060608@mizi.com> <200804081010.02409.manningc2@actrix.gen.nz> In-Reply-To: <200804081010.02409.manningc2@actrix.gen.nz> Content-Type: text/plain; charset=EUC-KR Content-Transfer-Encoding: quoted-printable X-MIME-Autoconverted: from 8bit to quoted-printable by mail.mizi.com id m3832pU4019010 X-SA-Exim-Connect-IP: 61.107.31.33 X-SA-Exim-Mail-From: louis@mizi.com X-Spam-Checker-Version: SpamAssassin 3.2.3 (2007-08-08) on stoneboat.aleph1.co.uk X-Spam-Level: X-Spam-Status: No, score=-2.6 required=5.0 tests=BAYES_00 autolearn=ham version=3.2.3 X-SA-Exim-Version: 4.2.1 (built Tue, 09 Jan 2007 17:23:22 +0000) X-SA-Exim-Scanned: Yes (on stoneboat.aleph1.co.uk) Cc: yaffs@lists.aleph1.co.uk Subject: Re: [Yaffs] mmap, power-off, and data is lost X-BeenThere: yaffs@lists.aleph1.co.uk X-Mailman-Version: 2.1.9 Precedence: list List-Id: Discussion of YAFFS NAND flash filesystem List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 08 Apr 2008 03:03:13 -0000 Hi Charles, Charles Manning =BE=B4 =B1=DB: > On Monday 07 April 2008 19:00:40 Louis JANG wrote: > =20 >> Hi all, >> >> I'm using yaffs in several mobile platform, and using mmap interface t= o >> read/write contents of file in yaffs. it works well in normal situatio= n, >> but it losts data when I turn off its power without unmount. >> =20 > > With mmapping, the data lives in the page cache until it is flushed to = the fs. > > YAFFS will not see the data until the mmapping is flushed through. This= will=20 > only happen when the pages are sync'ed back to the fs by msync() or w= hen=20 > the page mapping is invalidated (page pushed out or munmap) > > As James says, msync should do what you need. > =20 I added msync before munmap, and data wasn't lost when I use MS_SYNC flag. I also tried with MS_ASYNC and wait about more than 10 seconds, but data was lost in this case. However, the sample code invalidated mmap area by munmap already, and cached data should be flushed, but it wasn't. and I called sync command in shell also. Regards, Louis JANG