[U-Boot] [PATCH] NAND: fixed bugs in nand_[read, write]_skip_bad()

[Scott Wood]

On Tue, Nov 18, 2008 at 07:27:29PM +0200, Valeriy Glushkov wrote:
> - fixed bugs in nand_[read,write]_skip_bad() to double speed

Changelog should specify what the bugs are.

> - added printing of dots to show progress
> 
> Signed-off-by: Valeriy Glushkov <gvv at lstec.com>
> ---
>  drivers/mtd/nand/nand_base.c |   16 ++++++++++++++--
>  drivers/mtd/nand/nand_util.c |    6 ++++--
>  2 files changed, 18 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/mtd/nand/nand_base.c b/drivers/mtd/nand/nand_base.c
> index 00c9d4a..e4da6b1 100644
> --- a/drivers/mtd/nand/nand_base.c
> +++ b/drivers/mtd/nand/nand_base.c
> @@ -1136,8 +1136,11 @@ static int nand_do_read_ops(struct mtd_info *mtd, 
> loff_t from,
>      ret = chip->ecc.read_page_raw(mtd, chip, bufpoi);
>     else
>      ret = chip->ecc.read_page(mtd, chip, bufpoi);
> -   if (ret < 0)
> +   if (ret < 0) {
> +    printf("nand_read_opts: page 0x%x error %d\n",
> +     realpage, ret);
>      break;
> +   }

This patch is whitespace-damaged.

> 
>     /* Transfer not aligned data */
>     if (!aligned) {
> @@ -1190,6 +1193,9 @@ static int nand_do_read_ops(struct mtd_info *mtd, 
> loff_t from,
>    /* Increment page address */
>    realpage++;
> 
> +  if( !((realpage << chip->page_shift) & 0xfffff) )

if (!(xxx)), not if( !(xxx) ).

> +   debug(".");

Should progress meters really be debug()?

Maybe space it out according to percentage, rather than a fixed interval.

>    page = realpage & chip->pagemask;
>    /* Check, if we cross a chip boundary */
>    if (!page) {
> @@ -1828,8 +1834,11 @@ static int nand_do_write_ops(struct mtd_info *mtd, 
> loff_t to,
> 
>    ret = chip->write_page(mtd, chip, wbuf, page, cached,
>             (ops->mode == MTD_OOB_RAW));
> -  if (ret)
> +  if (ret) {
> +   printf("nand_write_opts: page 0x%x error %d\n",
> +    realpage, ret);
>     break;
> +  }
> 
>    writelen -= bytes;
>    if (!writelen)
> @@ -1839,6 +1848,9 @@ static int nand_do_write_ops(struct mtd_info *mtd, 
> loff_t to,
>    buf += bytes;
>    realpage++;
> 
> +  if( !((realpage << chip->page_shift) & 0xfffff) )
> +   debug(".");
> +
>    page = realpage & chip->pagemask;
>    /* Check, if we cross a chip boundary */
>    if (!page) {
> diff --git a/drivers/mtd/nand/nand_util.c b/drivers/mtd/nand/nand_util.c
> index 52b3d21..818a876 100644
> --- a/drivers/mtd/nand/nand_util.c
> +++ b/drivers/mtd/nand/nand_util.c
> @@ -498,8 +498,8 @@ int nand_write_skip_bad(nand_info_t *nand, size_t 
> offset, size_t *length,
>    if (rval != 0) {
>     printf ("NAND write to offset %x failed %d\n",
>      offset, rval);
> -   return rval;
>    }
> +  return rval;
>   }

This bit should be posted in a separate patch, as it's a bugfix that can
go in outside the merge window.

Or, we could just get rid of this altogether and always use the loop.

-Scott