From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from lazybastard.de ([212.112.238.170] helo=longford.logfs.org) by bombadil.infradead.org with esmtps (Exim 4.68 #1 (Red Hat Linux)) id 1K2rfp-0006bl-Bd for linux-mtd@lists.infradead.org; Sun, 01 Jun 2008 17:48:53 +0000 Date: Sun, 1 Jun 2008 19:48:42 +0200 From: =?utf-8?B?SsO2cm4=?= Engel To: Alexey Korolev Subject: Re: [RFC/PATCH 2/3] NAND multiple plane feature Message-ID: <20080601174841.GH13094@logfs.org> References: MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: Cc: linux-mtd@lists.infradead.org, tglx@linutronix.de, dwmw2@infradead.org, vasiliy.leonenko@gmail.com List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Wed, 28 May 2008 14:08:01 +0100, Alexey Korolev wrote: > > As NAND multiple plane architecture assumes simultaneous write/erase of > several pages/blocks at the same time, we have to modify page/eraseblock > sizes and report modified size to upper layers. In other words physical > erase block size/ page size != reported erase block size/ page size. > For example if we have dual plane device we have to extend erase block > size and page size in 2 times. Before actually reading the datasheets (just now) I had hoped that manufacturers would provide us several independent read/write/erase units per chip and allow software to deal with each plane as if it was a seperate chip. _That_ would have been really useful. And for NOR flashes, Intel has already shown how to do it. But hoping for manufacturers to get it right rarely works - it certainly didn't work in this case. As it seems, we can either program two planes in a weird lock-step process or ignore the feature. And the lock-step variant isn't useful for much more than doubling/quadrupling the erasesize and writesize. With all the disadvantages that brings. :( Speaking about the disadvantages, if the dual plane feature is enabled/disabled across reboots and erase size or write size changes, we're in for a lot of fun from the filesystem size. F.e. JFFS2 will experience data loss when erase size isn't stable. Jörn -- All models are wrong. Some models are useful. -- George Box