Re: [PATCH v3] mtd: spinand: Fix ECC status and OOB layout for Toshiba TC58CVG2S0H

[Miquel Raynal <miquel.raynal@bootlin.com>]

Hi Schrempf, Clement,

Schrempf Frieder <frieder.schrempf@kontron.de> wrote on Fri, 7 Dec 2018
09:31:47 +0000:

> Hi Miquèl,
> 
> On 07.12.18 09:00, Miquel Raynal wrote:
> > Hi Schrempf, Clément,
> > 
> > Schrempf Frieder <frieder.schrempf@kontron.de> wrote on Mon, 3 Dec 2018
> > 08:01:38 +0000:
> >   
> >> From: Frieder Schrempf <frieder.schrempf@kontron.de>
> >>
> >> When reading the status of the on-chip ECC, the Toshiba chip returns
> >> two different states for reporting corrected bitflips. We should check
> >> for both of them.
> >>
> >> Also return the free OOB bytes as one contiguous area, instead of
> >> multiple sections.
> >>
> >> Suggested-by: Clément Péron <peron.clem@gmail.com>
> >> Signed-off-by: Frieder Schrempf <frieder.schrempf@kontron.de>
> >> Acked-by: Clément Péron <peron.clem@gmail.com>
> >> ---
> >> Changes for v3:
> >> ===============
> >> * Add Acked-by
> >> * Add comment for BBM in OOB layout
> >>
> >> Changes for v2:
> >> ===============
> >> * Remove the changes to the ECC bytes in the OOB layout
> >> ---
> >>   drivers/mtd/nand/spi/toshiba.c | 11 ++++++-----
> >>   1 file changed, 6 insertions(+), 5 deletions(-)
> >>
> >> diff --git a/drivers/mtd/nand/spi/toshiba.c b/drivers/mtd/nand/spi/toshiba.c
> >> index 294bcf6..0812655 100644
> >> --- a/drivers/mtd/nand/spi/toshiba.c
> >> +++ b/drivers/mtd/nand/spi/toshiba.c
> >> @@ -11,6 +11,7 @@
> >>   #include <linux/mtd/spinand.h>
> >>   
> >>   #define SPINAND_MFR_TOSHIBA		0x98
> >> +#define TOSH_STATUS_ECC_HAS_BITFLIPS_T	(3 << 4)
> >>   
> >>   static SPINAND_OP_VARIANTS(read_cache_variants,
> >>   		SPINAND_PAGE_READ_FROM_CACHE_X4_OP(0, 1, NULL, 0),
> >> @@ -33,19 +34,18 @@ static int tc58cvg2s0h_ooblayout_ecc(struct mtd_info *mtd, int section,
> >>   	region->offset = 128 + 16 * section;
> >>   	region->length = 16;
> >>   
> >> -
> >>   	return 0;
> >>   }
> >>   
> >>   static int tc58cvg2s0h_ooblayout_free(struct mtd_info *mtd, int section,
> >>   				      struct mtd_oob_region *region)
> >>   {
> >> -	if (section > 7)
> >> +	if (section > 0)
> >>   		return -ERANGE;
> >>   
> >> -	region->offset = 2 + 16 * section;
> >> -	region->length = 14;
> >> -
> >> +	/* 2 bytes reserved for BBM */
> >> +	region->offset = 2;
> >> +	region->length = 126;
> >>   
> >>   	return 0;
> >>   }
> >> @@ -70,6 +70,7 @@ static int tc58cvg2s0h_ecc_get_status(struct spinand_device *spinand,
> >>   		return -EBADMSG;
> >>   
> >>   	case STATUS_ECC_HAS_BITFLIPS:
> >> +	case TOSH_STATUS_ECC_HAS_BITFLIPS_T:
> >>   		/*
> >>   		 * Let's try to retrieve the real maximum number of bitflips
> >>   		 * in order to avoid forcing the wear-leveling layer to move  
> > 
> > If you don't mind, I would like to squash this with the original patch.
> > What about also adding Clement's Acked-by to it?  
> 
> Sure, you can squash this with the original patch. And I guess adding 
> Clément's Acked-by would be appropriate, too.

Fix squashed, I will push on nand/next today.


Thanks,
Miquèl