From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ladislav Michl Date: Mon, 25 Jan 2016 16:56:51 +0100 Subject: [U-Boot] [PATCHv7 6/7] igep00x0: UBIize In-Reply-To: References: <20160117030929.GA28493@localhost.localdomain> <20160117031603.GG28493@localhost.localdomain> <56A0759F.4050509@denx.de> <56A5C30E.5070401@denx.de> Message-ID: <20160125155651.GA2245@localhost.localdomain> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de Hi Enric, On Mon, Jan 25, 2016 at 08:26:23AM +0100, Enric Balletbo Serra wrote: > The ROM boot on OMAP reads the first 4 blocks searching for the SPL, > in production is a practice flash the SPL 4 times. OneNAND/NAND > devices can have different block sizes and the OMAP ROM boot supports > block sizes from 64KB to 512K. For IGEP boards in particular, at least > there are boards that have block size of 128K and 256K. What I would > meant here is set as default the mtdparts variable to reserve 2M for > SPL, just to cover all the cases. > > mtdparts=mtdparts=gpmc-nand.0:2m(SPL),-(UBI)\0 So far there was no ack or nack to yesterday's proposal making that dynamic; Both boot ROM and ubispl code thinks in terms of eraseblocks, only mtd needs fixed offset. So I'd like to see this offset calculated as 4*block_size, not some "worst case" value... ladis