Re: [PATCH v10 04/10] mtd: nand: omap: fix device scan: NAND_CMD_READID, NAND_CMD_RESET, CMD_CMD_PARAM use only x8 bus

[Ezequiel Garcia <ezequiel.garcia-wi1+55ScJUtKEb57/3fJTNBPR1lH4CV8@public.gmane.org>]

Hi Gupta,

On Wed, Oct 23, 2013 at 01:15:20PM +0000, Gupta, Pekon wrote:
> Hi,
> 
> > From: Ezequiel Garcia [mailto:ezequiel.garcia-wi1+55ScJUtKEb57/3fJTNBPR1lH4CV8@public.gmane.org]
> [...]
> > FWIW, I have a Beaglebone with a 16-bit bus NAND attached to it.
> > 
> > Coincidentally, yesterday I was doing some tests as I'm ramping up the
> > NAND and I found that weird double nand_scan_ident() call.
> > The whole thing looks buggy to me, so I'm happy to help, review, test
> > and patches to take care of this.
> > 
> Yes, thanks .. that would be of great help.. 
> And may be your experience of Atmel drivers would help me here..
> 

It's not Atmel, but Marvell :-)

> *Correct, should not be double calls to nand_scan_ident()..*
> But there is a constrain in nand_base.c, that it does not allow ONFI
> page reading in x16 mode.. So how to overcome that..
> 
> I see the similar implementation in your ATMEL driver, it does not use
> NAND_BUSWIDTH_AUTO so how do you perform ONFI read
> for x16 devices ?
> drivers/mtd/nand/atmel_nand.c @@atmel_nand_probe()
> /* here you move to x16 mode based on your DT or platform data */
> 	if (host->board.bus_width_16)	/* 16-bit bus width */
> 		nand_chip->options |= NAND_BUSWIDTH_16;
> /* And then you call nand_scan_ident */
> /* first scan to find the device and get the page size */
> 	if (nand_scan_ident(mtd, 1, NULL)) {
> 		res = -ENXIO;
> 		goto err_scan_ident;
> 	}
> 
> Wouldn't this fail, _unless_ your device is listed in nand_flash_id[] ?
> because it would not be able to read ONFI params.. 
> Refer below commit.. 
> commit 0ce82b7f7b7373b16ecf7b5725e21e2975204500
> Author:     Matthieu CASTET <matthieu.castet-ITF29qwbsa/QT0dZR+AlfA@public.gmane.org>
> AuthorDate: 2013-01-16
> 
> 

Not my driver, but I'm taking a look at it now. Not sure if I'll get
into something here.

> 
> 
> > I'm using some TI SDK with some ancient v3.2.x (with no git history!),
> > but from this discussion it seems the issue is still present in
> > mainline.
> > 
> Aah sorry, then you might have some problem here in rebasing the
> patches. But still if you can, thanks much ..
> 

I'm currently trying mainline (just for this issue not for my product).
I just need some time to prepare the bootargs and write a DT node for
the NAND cape.

Again, not sure if I'll make some progress, but I'll give it a shot :-)
-- 
Ezequiel García, Free Electrons
Embedded Linux, Kernel and Android Engineering
http://free-electrons.com
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html