public inbox for linux-mtd@lists.infradead.org
 help / color / mirror / Atom feed
* [PATCH] MTD: NAND: Fix bad-block detection for 8-bit NAND
@ 2007-09-26 14:10 Frank Haverkamp
  2007-09-26 15:10 ` Thomas Gleixner
  0 siblings, 1 reply; 2+ messages in thread
From: Frank Haverkamp @ 2007-09-26 14:10 UTC (permalink / raw)
  To: Thomas Gleixner; +Cc: David Woodhouse, haver, MTD Mailinglist

[-- Attachment #1: Type: text/plain, Size: 2187 bytes --]

Hi Thomas,

I found the following block in our 128MiB NAND flash:

hexdump vic.img.460 
0000000 0000 0000 0000 0000 0000 0000 0000 0000
*
0000800 00ff 0000 0000 0000 0000 0000 0000 0000
0000810 0000 0000 0000 0000 0000 0000 0000 0000
*
0001040 00ff 0000 0000 0000 0000 0000 0000 0000
0001050 0000 0000 0000 0000 0000 0000 0000 0000
*
0021000

According to the specification this is a good block, because the bytes
in the 1st two OOB areas at offset 0 are 0xff. Nevertheless it was
detected bad because:

static struct nand_bbt_descr largepage_memorybased = {
	.options = 0,
	.offs = 0,
	.len = 2,                    /* <<<< bytes to check!!! */
	.pattern = scan_ff_pattern
};

used a len of 2 instead of 1, which I consider to be correct. Here a
potential fix for the problem. Please have a look if it matches your
understanding of the specification.

Signed-off-by: Frank Haverkamp <haver@vnet.ibm.com>
---
 drivers/mtd/nand/nand_bbt.c |   20 +++++++++++++++++---
 1 file changed, 17 insertions(+), 3 deletions(-)

--- ubi-2.6.git.orig/drivers/mtd/nand/nand_bbt.c
+++ ubi-2.6.git/drivers/mtd/nand/nand_bbt.c
@@ -1081,7 +1081,14 @@ static struct nand_bbt_descr smallpage_m
 	.pattern = scan_ff_pattern
 };
 
-static struct nand_bbt_descr largepage_memorybased = {
+static struct nand_bbt_descr largepage_memorybased_8bit = {
+	.options = 0,
+	.offs = 0,
+	.len = 1,
+	.pattern = scan_ff_pattern
+};
+
+static struct nand_bbt_descr largepage_memorybased_16bit = {
 	.options = 0,
 	.offs = 0,
 	.len = 2,
@@ -1179,8 +1186,15 @@ int nand_default_bbt(struct mtd_info *mt
 		this->bbt_td = NULL;
 		this->bbt_md = NULL;
 		if (!this->badblock_pattern) {
-			this->badblock_pattern = (mtd->writesize > 512) ?
-			    &largepage_memorybased : &smallpage_memorybased;
+			if (mtd->writesize > 512) {
+				if (this->options & NAND_BUSWIDTH_16)
+					this->badblock_pattern =
+						&largepage_memorybased_16bit;
+				else
+					this->badblock_pattern =
+					       &largepage_memorybased_8bit;
+			} else
+				this->badblock_pattern = &smallpage_memorybased;
 		}
 	}
 	return nand_scan_bbt(mtd, this->badblock_pattern);


[-- Attachment #2: smime.p7s --]
[-- Type: application/x-pkcs7-signature, Size: 5269 bytes --]

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [PATCH] MTD: NAND: Fix bad-block detection for 8-bit NAND
  2007-09-26 14:10 [PATCH] MTD: NAND: Fix bad-block detection for 8-bit NAND Frank Haverkamp
@ 2007-09-26 15:10 ` Thomas Gleixner
  0 siblings, 0 replies; 2+ messages in thread
From: Thomas Gleixner @ 2007-09-26 15:10 UTC (permalink / raw)
  To: haver; +Cc: MTD Mailinglist, David Woodhouse


On Wed, 2007-09-26 at 16:10 +0200, Frank Haverkamp wrote:
> Hi Thomas,
> 
> I found the following block in our 128MiB NAND flash:
> 
> hexdump vic.img.460 
> 0000000 0000 0000 0000 0000 0000 0000 0000 0000
> *
> 0000800 00ff 0000 0000 0000 0000 0000 0000 0000
> 0000810 0000 0000 0000 0000 0000 0000 0000 0000
> *
> 0001040 00ff 0000 0000 0000 0000 0000 0000 0000
> 0001050 0000 0000 0000 0000 0000 0000 0000 0000
> *
> 0021000
> 
> According to the specification this is a good block, because the bytes
> in the 1st two OOB areas at offset 0 are 0xff. Nevertheless it was
> detected bad because:
> 
> static struct nand_bbt_descr largepage_memorybased = {
> 	.options = 0,
> 	.offs = 0,
> 	.len = 2,                    /* <<<< bytes to check!!! */
> 	.pattern = scan_ff_pattern
> };
> 
> used a len of 2 instead of 1, which I consider to be correct. Here a
> potential fix for the problem. Please have a look if it matches your
> understanding of the specification.
> 
> Signed-off-by: Frank Haverkamp <haver@vnet.ibm.com>

Acked-by: Thomas Gleixner <tglx@linutronix.de>

> ---
>  drivers/mtd/nand/nand_bbt.c |   20 +++++++++++++++++---
>  1 file changed, 17 insertions(+), 3 deletions(-)
> 
> --- ubi-2.6.git.orig/drivers/mtd/nand/nand_bbt.c
> +++ ubi-2.6.git/drivers/mtd/nand/nand_bbt.c
> @@ -1081,7 +1081,14 @@ static struct nand_bbt_descr smallpage_m
>  	.pattern = scan_ff_pattern
>  };
>  
> -static struct nand_bbt_descr largepage_memorybased = {
> +static struct nand_bbt_descr largepage_memorybased_8bit = {
> +	.options = 0,
> +	.offs = 0,
> +	.len = 1,
> +	.pattern = scan_ff_pattern
> +};
> +
> +static struct nand_bbt_descr largepage_memorybased_16bit = {
>  	.options = 0,
>  	.offs = 0,
>  	.len = 2,
> @@ -1179,8 +1186,15 @@ int nand_default_bbt(struct mtd_info *mt
>  		this->bbt_td = NULL;
>  		this->bbt_md = NULL;
>  		if (!this->badblock_pattern) {
> -			this->badblock_pattern = (mtd->writesize > 512) ?
> -			    &largepage_memorybased : &smallpage_memorybased;
> +			if (mtd->writesize > 512) {
> +				if (this->options & NAND_BUSWIDTH_16)
> +					this->badblock_pattern =
> +						&largepage_memorybased_16bit;
> +				else
> +					this->badblock_pattern =
> +					       &largepage_memorybased_8bit;
> +			} else
> +				this->badblock_pattern = &smallpage_memorybased;
>  		}
>  	}
>  	return nand_scan_bbt(mtd, this->badblock_pattern);
> 

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2007-09-26 15:11 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2007-09-26 14:10 [PATCH] MTD: NAND: Fix bad-block detection for 8-bit NAND Frank Haverkamp
2007-09-26 15:10 ` Thomas Gleixner

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox