2018-02-06 17:41:31

by Stefan Agner

[permalink] [raw]
Subject: [PATCH 1/2] dt-bindings: mtd: gpmi-nand: specify fsl,use-minimum-ecc behavior

The dt property fsl,use-minimum-ecc requires a NAND chip to
provide a ECC strength/step size, otherwise the driver fails
to probe. This is by design to avoid that the driver uses a
fallback and later changes ECC parameters due to additionion
of a NAND chip driver. Document the current behavior.

Signed-off-by: Stefan Agner <[email protected]>
---
Documentation/devicetree/bindings/mtd/gpmi-nand.txt | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/Documentation/devicetree/bindings/mtd/gpmi-nand.txt b/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
index b289ef3c1b7e..eb2d9919d063 100644
--- a/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
+++ b/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
@@ -34,9 +34,8 @@ Optional properties:
automatically discoverable for some flash
(e.g., according to the ONFI standard).
However, note that if this strength is not
- discoverable or this property is not enabled,
- the software may chooses an implementation-defined
- ECC scheme.
+ discoverable the driver will fail probing with
+ an error.
- fsl,no-blockmark-swap: Don't swap the bad block marker from the OOB
area with the byte in the data area but rely on the
flash based BBT for identifying bad blocks.
--
2.16.1



2018-02-06 17:42:55

by Stefan Agner

[permalink] [raw]
Subject: [PATCH 2/2] mtd: nand: gpmi: add support for specific ECC strength

Add support for specified ECC strength/size using device tree
properties nand-ecc-strength/nand-ecc-step-size.

Signed-off-by: Stefan Agner <[email protected]>
---
.../devicetree/bindings/mtd/gpmi-nand.txt | 5 ++++
drivers/mtd/nand/gpmi-nand/gpmi-nand.c | 29 ++++++++++++++--------
2 files changed, 24 insertions(+), 10 deletions(-)

diff --git a/Documentation/devicetree/bindings/mtd/gpmi-nand.txt b/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
index eb2d9919d063..ea6e9b735160 100644
--- a/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
+++ b/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
@@ -46,6 +46,11 @@ Optional properties:
partitions written from Linux with this feature
turned on may not be accessible by the BootROM
code.
+ - nand-ecc-strength: integer representing the number of bits to correct
+ per ECC step. Needs to be a multiple of 2.
+ - nand-ecc-step-size: integer representing the number of data bytes
+ that are covered by a single ECC step. The driver
+ supports 512 and 1024.

The device tree may optionally contain sub-nodes describing partitions of the
address space. See partition.txt for more detail.
diff --git a/drivers/mtd/nand/gpmi-nand/gpmi-nand.c b/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
index 50f8d4a1b983..8cb378358e11 100644
--- a/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
+++ b/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
@@ -198,17 +198,15 @@ static inline bool gpmi_check_ecc(struct gpmi_nand_data *this)
*
* We may have available oob space in this case.
*/
-static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
+static int set_geometry_by_ecc_info(struct gpmi_nand_data *this,
+ unsigned int ecc_strength, unsigned int ecc_step)
{
struct bch_geometry *geo = &this->bch_geometry;
struct nand_chip *chip = &this->nand;
struct mtd_info *mtd = nand_to_mtd(chip);
unsigned int block_mark_bit_offset;

- if (!(chip->ecc_strength_ds > 0 && chip->ecc_step_ds > 0))
- return -EINVAL;
-
- switch (chip->ecc_step_ds) {
+ switch (ecc_step) {
case SZ_512:
geo->gf_len = 13;
break;
@@ -221,8 +219,8 @@ static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
chip->ecc_strength_ds, chip->ecc_step_ds);
return -EINVAL;
}
- geo->ecc_chunk_size = chip->ecc_step_ds;
- geo->ecc_strength = round_up(chip->ecc_strength_ds, 2);
+ geo->ecc_chunk_size = ecc_step;
+ geo->ecc_strength = round_up(ecc_strength, 2);
if (!gpmi_check_ecc(this))
return -EINVAL;

@@ -230,7 +228,7 @@ static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
if (geo->ecc_chunk_size < mtd->oobsize) {
dev_err(this->dev,
"unsupported nand chip. ecc size: %d, oob size : %d\n",
- chip->ecc_step_ds, mtd->oobsize);
+ ecc_step, mtd->oobsize);
return -EINVAL;
}

@@ -423,9 +421,20 @@ static int legacy_set_geometry(struct gpmi_nand_data *this)

int common_nfc_set_geometry(struct gpmi_nand_data *this)
{
+ struct nand_chip *chip = &this->nand;
+
+ if (chip->ecc.strength > 0 && chip->ecc.size > 0)
+ return set_geometry_by_ecc_info(this, chip->ecc.strength,
+ chip->ecc.size);
+
if ((of_property_read_bool(this->dev->of_node, "fsl,use-minimum-ecc"))
- || legacy_set_geometry(this))
- return set_geometry_by_ecc_info(this);
+ || legacy_set_geometry(this)) {
+ if (!(chip->ecc_strength_ds > 0 && chip->ecc_step_ds > 0))
+ return -EINVAL;
+
+ return set_geometry_by_ecc_info(this, chip->ecc_strength_ds,
+ chip->ecc_step_ds);
+ }

return 0;
}
--
2.16.1


2018-02-12 12:56:50

by Boris Brezillon

[permalink] [raw]
Subject: Re: [PATCH 2/2] mtd: nand: gpmi: add support for specific ECC strength

On Tue, 6 Feb 2018 18:40:21 +0100
Stefan Agner <[email protected]> wrote:

> Add support for specified ECC strength/size using device tree
> properties nand-ecc-strength/nand-ecc-step-size.
>
> Signed-off-by: Stefan Agner <[email protected]>
> ---
> .../devicetree/bindings/mtd/gpmi-nand.txt | 5 ++++
> drivers/mtd/nand/gpmi-nand/gpmi-nand.c | 29 ++++++++++++++--------
> 2 files changed, 24 insertions(+), 10 deletions(-)
>
> diff --git a/Documentation/devicetree/bindings/mtd/gpmi-nand.txt b/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
> index eb2d9919d063..ea6e9b735160 100644
> --- a/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
> +++ b/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
> @@ -46,6 +46,11 @@ Optional properties:
> partitions written from Linux with this feature
> turned on may not be accessible by the BootROM
> code.
> + - nand-ecc-strength: integer representing the number of bits to correct
> + per ECC step. Needs to be a multiple of 2.
> + - nand-ecc-step-size: integer representing the number of data bytes
> + that are covered by a single ECC step. The driver
> + supports 512 and 1024.
>
> The device tree may optionally contain sub-nodes describing partitions of the
> address space. See partition.txt for more detail.
> diff --git a/drivers/mtd/nand/gpmi-nand/gpmi-nand.c b/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
> index 50f8d4a1b983..8cb378358e11 100644
> --- a/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
> +++ b/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
> @@ -198,17 +198,15 @@ static inline bool gpmi_check_ecc(struct gpmi_nand_data *this)
> *
> * We may have available oob space in this case.
> */
> -static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
> +static int set_geometry_by_ecc_info(struct gpmi_nand_data *this,
> + unsigned int ecc_strength, unsigned int ecc_step)

Nitpick: if you send a new version, could you align the second
parameter on the open parenthesis?

> {
> struct bch_geometry *geo = &this->bch_geometry;
> struct nand_chip *chip = &this->nand;
> struct mtd_info *mtd = nand_to_mtd(chip);
> unsigned int block_mark_bit_offset;
>
> - if (!(chip->ecc_strength_ds > 0 && chip->ecc_step_ds > 0))
> - return -EINVAL;
> -
> - switch (chip->ecc_step_ds) {
> + switch (ecc_step) {
> case SZ_512:
> geo->gf_len = 13;
> break;
> @@ -221,8 +219,8 @@ static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
> chip->ecc_strength_ds, chip->ecc_step_ds);
> return -EINVAL;
> }
> - geo->ecc_chunk_size = chip->ecc_step_ds;
> - geo->ecc_strength = round_up(chip->ecc_strength_ds, 2);
> + geo->ecc_chunk_size = ecc_step;
> + geo->ecc_strength = round_up(ecc_strength, 2);
> if (!gpmi_check_ecc(this))
> return -EINVAL;
>
> @@ -230,7 +228,7 @@ static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
> if (geo->ecc_chunk_size < mtd->oobsize) {
> dev_err(this->dev,
> "unsupported nand chip. ecc size: %d, oob size : %d\n",
> - chip->ecc_step_ds, mtd->oobsize);
> + ecc_step, mtd->oobsize);
> return -EINVAL;
> }
>
> @@ -423,9 +421,20 @@ static int legacy_set_geometry(struct gpmi_nand_data *this)
>
> int common_nfc_set_geometry(struct gpmi_nand_data *this)
> {
> + struct nand_chip *chip = &this->nand;
> +
> + if (chip->ecc.strength > 0 && chip->ecc.size > 0)
> + return set_geometry_by_ecc_info(this, chip->ecc.strength,
> + chip->ecc.size);
> +
> if ((of_property_read_bool(this->dev->of_node, "fsl,use-minimum-ecc"))
> - || legacy_set_geometry(this))
> - return set_geometry_by_ecc_info(this);
> + || legacy_set_geometry(this)) {
> + if (!(chip->ecc_strength_ds > 0 && chip->ecc_step_ds > 0))
> + return -EINVAL;
> +
> + return set_geometry_by_ecc_info(this, chip->ecc_strength_ds,
> + chip->ecc_step_ds);
> + }
>
> return 0;
> }



--
Boris Brezillon, Bootlin (formerly Free Electrons)
Embedded Linux and Kernel engineering
http://bootlin.com

2018-02-12 13:14:23

by Boris Brezillon

[permalink] [raw]
Subject: Re: [PATCH 2/2] mtd: nand: gpmi: add support for specific ECC strength

On Tue, 6 Feb 2018 18:40:21 +0100
Stefan Agner <[email protected]> wrote:

> Add support for specified ECC strength/size using device tree
> properties nand-ecc-strength/nand-ecc-step-size.
>
> Signed-off-by: Stefan Agner <[email protected]>
> ---
> .../devicetree/bindings/mtd/gpmi-nand.txt | 5 ++++

I don't care that much, but I know DT maintainers like to have DT
bindings changes placed in a separate patch.

> drivers/mtd/nand/gpmi-nand/gpmi-nand.c | 29 ++++++++++++++--------
> 2 files changed, 24 insertions(+), 10 deletions(-)
>
> diff --git a/Documentation/devicetree/bindings/mtd/gpmi-nand.txt b/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
> index eb2d9919d063..ea6e9b735160 100644
> --- a/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
> +++ b/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
> @@ -46,6 +46,11 @@ Optional properties:
> partitions written from Linux with this feature
> turned on may not be accessible by the BootROM
> code.
> + - nand-ecc-strength: integer representing the number of bits to correct
> + per ECC step. Needs to be a multiple of 2.
> + - nand-ecc-step-size: integer representing the number of data bytes
> + that are covered by a single ECC step. The driver
> + supports 512 and 1024.
>
> The device tree may optionally contain sub-nodes describing partitions of the
> address space. See partition.txt for more detail.
> diff --git a/drivers/mtd/nand/gpmi-nand/gpmi-nand.c b/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
> index 50f8d4a1b983..8cb378358e11 100644
> --- a/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
> +++ b/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
> @@ -198,17 +198,15 @@ static inline bool gpmi_check_ecc(struct gpmi_nand_data *this)
> *
> * We may have available oob space in this case.
> */
> -static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
> +static int set_geometry_by_ecc_info(struct gpmi_nand_data *this,
> + unsigned int ecc_strength, unsigned int ecc_step)
> {
> struct bch_geometry *geo = &this->bch_geometry;
> struct nand_chip *chip = &this->nand;
> struct mtd_info *mtd = nand_to_mtd(chip);
> unsigned int block_mark_bit_offset;
>
> - if (!(chip->ecc_strength_ds > 0 && chip->ecc_step_ds > 0))
> - return -EINVAL;
> -
> - switch (chip->ecc_step_ds) {
> + switch (ecc_step) {
> case SZ_512:
> geo->gf_len = 13;
> break;
> @@ -221,8 +219,8 @@ static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
> chip->ecc_strength_ds, chip->ecc_step_ds);
> return -EINVAL;
> }
> - geo->ecc_chunk_size = chip->ecc_step_ds;
> - geo->ecc_strength = round_up(chip->ecc_strength_ds, 2);
> + geo->ecc_chunk_size = ecc_step;
> + geo->ecc_strength = round_up(ecc_strength, 2);
> if (!gpmi_check_ecc(this))
> return -EINVAL;
>
> @@ -230,7 +228,7 @@ static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
> if (geo->ecc_chunk_size < mtd->oobsize) {
> dev_err(this->dev,
> "unsupported nand chip. ecc size: %d, oob size : %d\n",
> - chip->ecc_step_ds, mtd->oobsize);
> + ecc_step, mtd->oobsize);
> return -EINVAL;
> }
>
> @@ -423,9 +421,20 @@ static int legacy_set_geometry(struct gpmi_nand_data *this)
>
> int common_nfc_set_geometry(struct gpmi_nand_data *this)
> {
> + struct nand_chip *chip = &this->nand;
> +
> + if (chip->ecc.strength > 0 && chip->ecc.size > 0)
> + return set_geometry_by_ecc_info(this, chip->ecc.strength,
> + chip->ecc.size);
> +
> if ((of_property_read_bool(this->dev->of_node, "fsl,use-minimum-ecc"))
> - || legacy_set_geometry(this))
> - return set_geometry_by_ecc_info(this);
> + || legacy_set_geometry(this)) {
> + if (!(chip->ecc_strength_ds > 0 && chip->ecc_step_ds > 0))
> + return -EINVAL;
> +
> + return set_geometry_by_ecc_info(this, chip->ecc_strength_ds,
> + chip->ecc_step_ds);
> + }
>
> return 0;
> }

Both patches look good to me. Han, could review them and add your Ack
if you're happy with the changes.

Thanks,

Boris

--
Boris Brezillon, Bootlin (formerly Free Electrons)
Embedded Linux and Kernel engineering
http://bootlin.com

2018-02-14 20:06:49

by Han Xu

[permalink] [raw]
Subject: Re: [PATCH 2/2] mtd: nand: gpmi: add support for specific ECC strength



On 02/06/2018 11:40 AM, Stefan Agner wrote:
> Add support for specified ECC strength/size using device tree
> properties nand-ecc-strength/nand-ecc-step-size.
>
> Signed-off-by: Stefan Agner <[email protected]>
> ---
> .../devicetree/bindings/mtd/gpmi-nand.txt | 5 ++++
> drivers/mtd/nand/gpmi-nand/gpmi-nand.c | 29 ++++++++++++++--------
> 2 files changed, 24 insertions(+), 10 deletions(-)
>
> diff --git a/Documentation/devicetree/bindings/mtd/gpmi-nand.txt b/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
> index eb2d9919d063..ea6e9b735160 100644
> --- a/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
> +++ b/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
> @@ -46,6 +46,11 @@ Optional properties:
> partitions written from Linux with this feature
> turned on may not be accessible by the BootROM
> code.
> + - nand-ecc-strength: integer representing the number of bits to correct
> + per ECC step. Needs to be a multiple of 2.
> + - nand-ecc-step-size: integer representing the number of data bytes
> + that are covered by a single ECC step. The driver
> + supports 512 and 1024.
>
> The device tree may optionally contain sub-nodes describing partitions of the
> address space. See partition.txt for more detail.
> diff --git a/drivers/mtd/nand/gpmi-nand/gpmi-nand.c b/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
> index 50f8d4a1b983..8cb378358e11 100644
> --- a/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
> +++ b/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
> @@ -198,17 +198,15 @@ static inline bool gpmi_check_ecc(struct gpmi_nand_data *this)
> *
> * We may have available oob space in this case.
> */
> -static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
> +static int set_geometry_by_ecc_info(struct gpmi_nand_data *this,
> + unsigned int ecc_strength, unsigned int ecc_step)
> {
> struct bch_geometry *geo = &this->bch_geometry;
> struct nand_chip *chip = &this->nand;
> struct mtd_info *mtd = nand_to_mtd(chip);
> unsigned int block_mark_bit_offset;
>
> - if (!(chip->ecc_strength_ds > 0 && chip->ecc_step_ds > 0))
> - return -EINVAL;
> -
> - switch (chip->ecc_step_ds) {
> + switch (ecc_step) {
> case SZ_512:
> geo->gf_len = 13;
> break;
> @@ -221,8 +219,8 @@ static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
> chip->ecc_strength_ds, chip->ecc_step_ds);
> return -EINVAL;
> }
> - geo->ecc_chunk_size = chip->ecc_step_ds;
> - geo->ecc_strength = round_up(chip->ecc_strength_ds, 2);
> + geo->ecc_chunk_size = ecc_step;
> + geo->ecc_strength = round_up(ecc_strength, 2);
> if (!gpmi_check_ecc(this))
> return -EINVAL;
>
> @@ -230,7 +228,7 @@ static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
> if (geo->ecc_chunk_size < mtd->oobsize) {
> dev_err(this->dev,
> "unsupported nand chip. ecc size: %d, oob size : %d\n",
> - chip->ecc_step_ds, mtd->oobsize);
> + ecc_step, mtd->oobsize);
> return -EINVAL;
> }
>
> @@ -423,9 +421,20 @@ static int legacy_set_geometry(struct gpmi_nand_data *this)
>
> int common_nfc_set_geometry(struct gpmi_nand_data *this)
> {
> + struct nand_chip *chip = &this->nand;
> +
> + if (chip->ecc.strength > 0 && chip->ecc.size > 0)
> + return set_geometry_by_ecc_info(this, chip->ecc.strength,
> + chip->ecc.size);
> +

I was wondering how to keep, let's say u-boot, and kernel ecc setting
aligned, if users can specify the strength and step_ds in DT? Did u-boot
enable to get these parameters from DT?
> if ((of_property_read_bool(this->dev->of_node, "fsl,use-minimum-ecc"))
> - || legacy_set_geometry(this))
> - return set_geometry_by_ecc_info(this);
> + || legacy_set_geometry(this)) {
> + if (!(chip->ecc_strength_ds > 0 && chip->ecc_step_ds > 0))
> + return -EINVAL;
> +
> + return set_geometry_by_ecc_info(this, chip->ecc_strength_ds,
> + chip->ecc_step_ds);
> + }
>
> return 0;
> }

2018-02-14 20:13:13

by Boris Brezillon

[permalink] [raw]
Subject: Re: [PATCH 2/2] mtd: nand: gpmi: add support for specific ECC strength

On Wed, 14 Feb 2018 16:28:36 +0000
Han Xu <[email protected]> wrote:

> On 02/06/2018 11:40 AM, Stefan Agner wrote:
> > Add support for specified ECC strength/size using device tree
> > properties nand-ecc-strength/nand-ecc-step-size.
> >
> > Signed-off-by: Stefan Agner <[email protected]>
> > ---
> > .../devicetree/bindings/mtd/gpmi-nand.txt | 5 ++++
> > drivers/mtd/nand/gpmi-nand/gpmi-nand.c | 29 ++++++++++++++--------
> > 2 files changed, 24 insertions(+), 10 deletions(-)
> >
> > diff --git a/Documentation/devicetree/bindings/mtd/gpmi-nand.txt b/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
> > index eb2d9919d063..ea6e9b735160 100644
> > --- a/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
> > +++ b/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
> > @@ -46,6 +46,11 @@ Optional properties:
> > partitions written from Linux with this feature
> > turned on may not be accessible by the BootROM
> > code.
> > + - nand-ecc-strength: integer representing the number of bits to correct
> > + per ECC step. Needs to be a multiple of 2.
> > + - nand-ecc-step-size: integer representing the number of data bytes
> > + that are covered by a single ECC step. The driver
> > + supports 512 and 1024.
> >
> > The device tree may optionally contain sub-nodes describing partitions of the
> > address space. See partition.txt for more detail.
> > diff --git a/drivers/mtd/nand/gpmi-nand/gpmi-nand.c b/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
> > index 50f8d4a1b983..8cb378358e11 100644
> > --- a/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
> > +++ b/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
> > @@ -198,17 +198,15 @@ static inline bool gpmi_check_ecc(struct gpmi_nand_data *this)
> > *
> > * We may have available oob space in this case.
> > */
> > -static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
> > +static int set_geometry_by_ecc_info(struct gpmi_nand_data *this,
> > + unsigned int ecc_strength, unsigned int ecc_step)
> > {
> > struct bch_geometry *geo = &this->bch_geometry;
> > struct nand_chip *chip = &this->nand;
> > struct mtd_info *mtd = nand_to_mtd(chip);
> > unsigned int block_mark_bit_offset;
> >
> > - if (!(chip->ecc_strength_ds > 0 && chip->ecc_step_ds > 0))
> > - return -EINVAL;
> > -
> > - switch (chip->ecc_step_ds) {
> > + switch (ecc_step) {
> > case SZ_512:
> > geo->gf_len = 13;
> > break;
> > @@ -221,8 +219,8 @@ static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
> > chip->ecc_strength_ds, chip->ecc_step_ds);
> > return -EINVAL;
> > }
> > - geo->ecc_chunk_size = chip->ecc_step_ds;
> > - geo->ecc_strength = round_up(chip->ecc_strength_ds, 2);
> > + geo->ecc_chunk_size = ecc_step;
> > + geo->ecc_strength = round_up(ecc_strength, 2);
> > if (!gpmi_check_ecc(this))
> > return -EINVAL;
> >
> > @@ -230,7 +228,7 @@ static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
> > if (geo->ecc_chunk_size < mtd->oobsize) {
> > dev_err(this->dev,
> > "unsupported nand chip. ecc size: %d, oob size : %d\n",
> > - chip->ecc_step_ds, mtd->oobsize);
> > + ecc_step, mtd->oobsize);
> > return -EINVAL;
> > }
> >
> > @@ -423,9 +421,20 @@ static int legacy_set_geometry(struct gpmi_nand_data *this)
> >
> > int common_nfc_set_geometry(struct gpmi_nand_data *this)
> > {
> > + struct nand_chip *chip = &this->nand;
> > +
> > + if (chip->ecc.strength > 0 && chip->ecc.size > 0)
> > + return set_geometry_by_ecc_info(this, chip->ecc.strength,
> > + chip->ecc.size);
> > +
>
> I was wondering how to keep, let's say u-boot, and kernel ecc setting
> aligned, if users can specify the strength and step_ds in DT? Did u-boot
> enable to get these parameters from DT?

Both u-boot and Linux have to support using strength/step information
passed through the DT before you can start patching dts files. Anyway,
users will have to change their DT (add the nand-ecc-strength/step-size
properties) to use this feature, so they should quickly notice that
something is going wrong and update their bootloader or drop those
nand-ecc-xxx props.

> > if ((of_property_read_bool(this->dev->of_node, "fsl,use-minimum-ecc"))
> > - || legacy_set_geometry(this))
> > - return set_geometry_by_ecc_info(this);
> > + || legacy_set_geometry(this)) {
> > + if (!(chip->ecc_strength_ds > 0 && chip->ecc_step_ds > 0))
> > + return -EINVAL;
> > +
> > + return set_geometry_by_ecc_info(this, chip->ecc_strength_ds,
> > + chip->ecc_step_ds);
> > + }
> >
> > return 0;
> > }



--
Boris Brezillon, Bootlin (formerly Free Electrons)
Embedded Linux and Kernel engineering
http://bootlin.com

2018-02-14 21:37:29

by Stefan Agner

[permalink] [raw]
Subject: Re: [PATCH 2/2] mtd: nand: gpmi: add support for specific ECC strength

On 14.02.2018 20:05, Boris Brezillon wrote:
> On Wed, 14 Feb 2018 16:28:36 +0000
> Han Xu <[email protected]> wrote:
>
>> On 02/06/2018 11:40 AM, Stefan Agner wrote:
>> > Add support for specified ECC strength/size using device tree
>> > properties nand-ecc-strength/nand-ecc-step-size.
>> >
>> > Signed-off-by: Stefan Agner <[email protected]>
>> > ---
>> > .../devicetree/bindings/mtd/gpmi-nand.txt | 5 ++++
>> > drivers/mtd/nand/gpmi-nand/gpmi-nand.c | 29 ++++++++++++++--------
>> > 2 files changed, 24 insertions(+), 10 deletions(-)
>> >
>> > diff --git a/Documentation/devicetree/bindings/mtd/gpmi-nand.txt b/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
>> > index eb2d9919d063..ea6e9b735160 100644
>> > --- a/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
>> > +++ b/Documentation/devicetree/bindings/mtd/gpmi-nand.txt
>> > @@ -46,6 +46,11 @@ Optional properties:
>> > partitions written from Linux with this feature
>> > turned on may not be accessible by the BootROM
>> > code.
>> > + - nand-ecc-strength: integer representing the number of bits to correct
>> > + per ECC step. Needs to be a multiple of 2.
>> > + - nand-ecc-step-size: integer representing the number of data bytes
>> > + that are covered by a single ECC step. The driver
>> > + supports 512 and 1024.
>> >
>> > The device tree may optionally contain sub-nodes describing partitions of the
>> > address space. See partition.txt for more detail.
>> > diff --git a/drivers/mtd/nand/gpmi-nand/gpmi-nand.c b/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
>> > index 50f8d4a1b983..8cb378358e11 100644
>> > --- a/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
>> > +++ b/drivers/mtd/nand/gpmi-nand/gpmi-nand.c
>> > @@ -198,17 +198,15 @@ static inline bool gpmi_check_ecc(struct gpmi_nand_data *this)
>> > *
>> > * We may have available oob space in this case.
>> > */
>> > -static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
>> > +static int set_geometry_by_ecc_info(struct gpmi_nand_data *this,
>> > + unsigned int ecc_strength, unsigned int ecc_step)
>> > {
>> > struct bch_geometry *geo = &this->bch_geometry;
>> > struct nand_chip *chip = &this->nand;
>> > struct mtd_info *mtd = nand_to_mtd(chip);
>> > unsigned int block_mark_bit_offset;
>> >
>> > - if (!(chip->ecc_strength_ds > 0 && chip->ecc_step_ds > 0))
>> > - return -EINVAL;
>> > -
>> > - switch (chip->ecc_step_ds) {
>> > + switch (ecc_step) {
>> > case SZ_512:
>> > geo->gf_len = 13;
>> > break;
>> > @@ -221,8 +219,8 @@ static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
>> > chip->ecc_strength_ds, chip->ecc_step_ds);
>> > return -EINVAL;
>> > }
>> > - geo->ecc_chunk_size = chip->ecc_step_ds;
>> > - geo->ecc_strength = round_up(chip->ecc_strength_ds, 2);
>> > + geo->ecc_chunk_size = ecc_step;
>> > + geo->ecc_strength = round_up(ecc_strength, 2);
>> > if (!gpmi_check_ecc(this))
>> > return -EINVAL;
>> >
>> > @@ -230,7 +228,7 @@ static int set_geometry_by_ecc_info(struct gpmi_nand_data *this)
>> > if (geo->ecc_chunk_size < mtd->oobsize) {
>> > dev_err(this->dev,
>> > "unsupported nand chip. ecc size: %d, oob size : %d\n",
>> > - chip->ecc_step_ds, mtd->oobsize);
>> > + ecc_step, mtd->oobsize);
>> > return -EINVAL;
>> > }
>> >
>> > @@ -423,9 +421,20 @@ static int legacy_set_geometry(struct gpmi_nand_data *this)
>> >
>> > int common_nfc_set_geometry(struct gpmi_nand_data *this)
>> > {
>> > + struct nand_chip *chip = &this->nand;
>> > +
>> > + if (chip->ecc.strength > 0 && chip->ecc.size > 0)
>> > + return set_geometry_by_ecc_info(this, chip->ecc.strength,
>> > + chip->ecc.size);
>> > +
>>
>> I was wondering how to keep, let's say u-boot, and kernel ecc setting
>> aligned, if users can specify the strength and step_ds in DT? Did u-boot
>> enable to get these parameters from DT?
>
> Both u-boot and Linux have to support using strength/step information
> passed through the DT before you can start patching dts files. Anyway,
> users will have to change their DT (add the nand-ecc-strength/step-size
> properties) to use this feature, so they should quickly notice that
> something is going wrong and update their bootloader or drop those
> nand-ecc-xxx props.
>

Yes, the MTD stack has support for those properties also in U-Boot.

But currently the GPMI NAND driver (mxs_nand) in U-Boot has no DT
support.

But I do have a preliminary patch set for this. Will send it out soon.

--
Stefan


>> > if ((of_property_read_bool(this->dev->of_node, "fsl,use-minimum-ecc"))
>> > - || legacy_set_geometry(this))
>> > - return set_geometry_by_ecc_info(this);
>> > + || legacy_set_geometry(this)) {
>> > + if (!(chip->ecc_strength_ds > 0 && chip->ecc_step_ds > 0))
>> > + return -EINVAL;
>> > +
>> > + return set_geometry_by_ecc_info(this, chip->ecc_strength_ds,
>> > + chip->ecc_step_ds);
>> > + }
>> >
>> > return 0;
>> > }