From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <19851.474.327087.194875@ipc1.ka-ro> Date: Thu, 24 Mar 2011 09:33:30 +0100 From: =?utf-8?Q?Lothar_Wa=C3=9Fmann?= To: Jason Liu Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit Subject: Re: [PATCH 5/7] add GPMI support for imx28 In-Reply-To: References: <1300239773-4222-1-git-send-email-b32955@freescale.com> <1300239773-4222-6-git-send-email-b32955@freescale.com> <19848.39439.380861.195051@ipc1.ka-ro> <4D8964F4.3070104@freescale.com> <19850.2584.566366.251874@ipc1.ka-ro> <4D8AB490.30408@freescale.com> <19850.62451.678409.328798@ipc1.ka-ro> Cc: linux-arm-kernel@lists.infradead.org, Huang Shijie , linux@arm.linux.org.uk, dwmw2@infradead.org, linux-mtd@lists.infradead.org List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Hi, Jason Liu writes: > 2011/3/24 Lothar Waßmann : > > Hi, > [...] > >> Most of the time, the timing is really based on a safe timing setting. > >> But in the original GPMI driver in the FREESCALE BSP, there exits some > >> nands need to be set with their own timing setting. > >> > >> So I do not use the safe timing for _ALL_ the nand, and i'd better get > >> it from the > >> database. > >> > > It should be sufficient to provide timing info from platform_data in > > special cases instead of bloating the nand id database with that > > stuff. Platforms might need to adjust the timing because of > > peculiarities in the HW. Thus the timing info should be provided from > > there, not from the chip database. > > No, we can't. since some boards only provide one NAND socket, customer > can place any NAND flash chip they want. And with the NAND tech evolution, some > NAND flash will end-of-line and customer need change another new NAND. > As we all know that RAW NAND flash vendor does not follow the same > rule to encode > ID related information such as page size/oob size/block size etc. Even > with same vendor, > take Sumsung as example, there is not one generic rule to encode the > id information. > If you do NAND support with many customers, you will find that > currenly linux NAND driver > (nand_base.c/nand_ids.c) provide very poor support for that. > That's a reason to improve that support, not to invent private fixes for each chip driver. Lothar Waßmann -- ___________________________________________________________ Ka-Ro electronics GmbH | Pascalstraße 22 | D - 52076 Aachen Phone: +49 2408 1402-0 | Fax: +49 2408 1402-10 Geschäftsführer: Matthias Kaussen Handelsregistereintrag: Amtsgericht Aachen, HRB 4996 www.karo-electronics.de | info@karo-electronics.de ___________________________________________________________