[PATCH 0/6] mtd: teach mxc-nand about ONFI probing

[Uwe Kleine-König]

Hello,

On Tue, Feb 10, 2015 at 07:59:54PM +0100, Uwe Kleine-König wrote:
> this series implements support for ONFI probing in mxc-nand.
> Before a PARAM command was just ignored and the following read_bytes
> read whatever then happened to be available from the controller's RAM.
> But actually this wasn't that bad because the driver already failed to
> read the ONFI marker at offset 32 in response to the READID command and
> so the PARAM command wasn't issued at all :-)
> 
> Patch 1 is a robustness fix. Patch 2 is necessary to not make the
> controller stuck when the READPAGE command is issued before the flash
> chip is identified. Patch 3 is needed to be able to read more than 6
> bytes in the read_byte callback. Finally patch 4 fixes the driver to
> allow reading out the READID ONFI marker and patch 5 implements support
> for the PARAM command.
> 
> Patch 6 implements a WARN that triggers when the core requests an
> unknown command from the driver. In the case of missing PARAM support it
> wouldn't have triggered because READID at offset 0x20 failed, too, but
> it might be worthwhile anyhow; I'm not sure though, so I marked it RFC.
> Not sure the set of commands will grow in the near future ...
> 
> Best regards
> Uwe
> 
> Uwe Kleine-König (6):
>   mtd: mxc-nand: Add a timeout when waiting for interrupt
>   mtd: mxc-nand: Only enable hardware checksumming for fully detected
>     flashes
>   mtd: mxc-nand: Do the word to byte mangling in the read_byte callback
>   mtd: mxc-nand: Allow to use column addresses different from 0
>   mtd: mxc-nand: Implement support for PARAM command
>   [RFC] mtd: mxc-nand: Warn on unimplemented commands
I think this series if fine to go into 4.1-rc1. If so it would be great
to get it into next in the near future. David? Brian? Any feedback?

Thanks
Uwe

-- 
Pengutronix e.K.                           | Uwe Kleine-König            |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |