From mboxrd@z Thu Jan 1 00:00:00 1970 From: Simon Schwarz Date: Fri, 29 Jul 2011 14:17:36 +0200 Subject: [U-Boot] [PATCH V6 3/5] nand spl: add NAND Library to new SPL In-Reply-To: <4E318BD7.7010102@ti.com> References: <1311771039-31691-1-git-send-email-simonschwarzcor@gmail.com> <1311842291-24837-1-git-send-email-simonschwarzcor@gmail.com> <1311842291-24837-4-git-send-email-simonschwarzcor@gmail.com> <4E314DF1.3090007@ti.com> <4E316C53.4000701@gmail.com> <4E318BD7.7010102@ti.com> Message-ID: <4E32A4E0.4080702@gmail.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de Hi Aneesh, On 07/28/2011 06:18 PM, Aneesh V wrote: [snip] >> >> Oh. There I played around with the Access Size Adaptation of the GPMC - >> It is still a x16 interface - this is what the 16 refers to IMHO. But > > Ok. I have to admit that I am not a NAND expert and I do not understand > this code well. > >> for sake of simplicity I will change this back to 16bit access - I don't >> think that there is a big performance impact although I didn't measure >> it. > > No. If it's an OMAP specific optimization, I don't see a reason to > remove it. Looks like that may actually improve performance. However, > you may have to take into account of the alignment of buffer, the size > requested etc. Please have a look at the implementation in drivers/mtd > /nand/davinci_nand.c(although the implementation here seems to be for > 8-bit devices, something similar may be possible for 16-bit) I literally just played around with that ;) I will add it in the standard version here. For my BA I will have to evaluate if it has a performance impact anyway - if it has I will send a patch later. > >> >> I cloned them because the functions in nand_base.c are static. >> >> My solution: deleted the cloned functions - use these from nand_base by >> removing the static modifier and add them to nand.h > > I hope there won't be any name-space conflict due to this. Me too. For devkit8000 there is none - BUILDALL will show. > > best regards, > Aneesh Regards Simon