From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from n1.cetrtapot.si ([212.30.80.17]) by canuck.infradead.org with esmtp (Exim 4.43 #1 (Red Hat Linux)) id 1DbJyU-0005OR-8w for linux-mtd@lists.infradead.org; Thu, 26 May 2005 11:08:45 -0400 Message-ID: <4295E673.3050704@cetrtapot.si> Date: Thu, 26 May 2005 17:08:35 +0200 From: Hinko Kocevar MIME-Version: 1.0 To: Damian Slee References: In-Reply-To: Content-Type: text/plain; charset=windows-1250; format=flowed Content-Transfer-Encoding: 7bit Cc: linux-mtd@lists.infradead.org Subject: Re: Serial flash memory support List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Damian Slee wrote: > Jffs2 does support atmel serial dataflash, as far as the driver for spi, > that is more hardware dependent. They recently changed the conditional > compiles to be CONFIG_JFFS2_FS_WRITEBUFFER for nand and dataflash. > > BTW JFFS2 takes ages to do the startup scan of the flash with the 528 or > 1056 block size of the dataflash. Our serial flash uses 256 byte pages and is forming sectors with them, depending on the whole flash size. But I guess that is no case for accessing it quicker. > I was evalutating the latest cvs JFFS2 last week upon eCos. I think it > was about 7 seconds for 64kbytes, with the SPI clocked at 15MHz. That is too slow for our application. My conclusion would be that JFFS2 is too much for 'slow' SPI serial flash. I guess will just use raw access to the data stored there and achieve maximum speed. Can mtd raw access driver/module be used for this purpose or is that driver meant for a another stuff? regards, hinko k -- ..because under Linux "if something is possible in principle, then it is already implemented or somebody is working on it". --LKI