[PATCH v4 0/6] Export the ECC step size to user applications

[Brian Norris]

Hi Huang,

On Fri, Aug 16, 2013 at 10:10:03AM +0800, Huang Shijie wrote:
> In order to implement the NAND boot for some Freescale's chips, such as
> imx23/imx28/imx50/imx6, we use a tool (called kobs-ng) to burn the uboot
> and some metadata to nand chip. And the ROM code will use the metadata to
> configrate the BCH, and to find the uboot.
>
> The ECC information(ecc step size, ecc strength) which is used to configrate
> the BCH is part of the metadata. The kobs-ng can gets the ecc strength from
> the sys node /sys/*/mtdX/ecc_strength now. But it can't gets the ecc step size.
>
> This patch set is used to export the ecc step size to user applications.
> With this patch set, the kobs-ng can gets the ecc step size now.
>
> v3 --> v4:
> [1] rename the ecc_step to ecc_step_size.
>
> v2 --> v3:
> [1] replace the nand_scan() with nand_scan_ident()/nand_scan_tail(),
>    Let the MTD layer to do the initialization for the ECC info.
>    removed some hack code.
> [2] move the ONFI nand's hooks in nand_set_defaults().
> [3] change the comments.
> [4] misc
>
> v1 --> v2:
> [1] rename the ecc_size to ecc_step.
> [2] rebase on the latest l2-mtd.
>
> Huang Shijie (6):
>   mtd: add a new field to mtd_info{}
>   mtd: add a new sys node to show the ecc step size
>   mtd: set the ecc step size for master/slave mtd_info
>   mtd: set ONFI nand's default hooks in nand_set_defaults()
>   mtd: gpmi: remove the nand_scan()
>   mtd: update the ABI document about the ecc step size
>
>  Documentation/ABI/testing/sysfs-class-mtd |   10 ++++
>  drivers/mtd/mtdcore.c                     |   11 +++++
>  drivers/mtd/mtdpart.c                     |    1 +
>  drivers/mtd/nand/gpmi-nand/gpmi-nand.c    |   65 +++++++++++++++-------------
>  drivers/mtd/nand/nand_base.c              |   13 +++---
>  include/linux/mtd/mtd.h                   |    3 +
>  6 files changed, 67 insertions(+), 36 deletions(-)

I think we're still missing updates for a few drivers, which probably
shouldn't have mtd->ecc_step_size == 0. For instance,
nand/alauda.c and maybe a few others. Can you take a pass at them,
Huang? Perhaps take a guess at their step size and CC relevant author(s).
If no one cares, then maybe they can stay as zero.

Anyway, I pushed the whole series, including my edits to patch 6. I will
amend if there are more comments.

Thanks,
Brian