diff mbox series

pinctrl: ingenic: Fix regmap on X series SoCs

Message ID 20220209230452.19535-1-aidanmacdonald.0x0@gmail.com (mailing list archive)
State Superseded
Headers show
Series pinctrl: ingenic: Fix regmap on X series SoCs | expand

Commit Message

Aidan MacDonald Feb. 9, 2022, 11:04 p.m. UTC
The X series Ingenic SoCs have a shadow GPIO group which
is at a higher offset than the other groups, and is used
for all GPIO configuration. The regmap did not take this
offset into account and set max_register too low. Writes
to the shadow group registers were blocked, which made it
impossible to change any pin configuration.

Fix this by pretending there are at least 8 chips on any
'X' SoC for the purposes of calculating max_register. This
ensures the shadow group is accessible.

Signed-off-by: Aidan MacDonald <aidanmacdonald.0x0@gmail.com>
---
 drivers/pinctrl/pinctrl-ingenic.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

Comments

Zhou Yanjie Feb. 10, 2022, 10:18 a.m. UTC | #1
Hi Aidan,

On 2022/2/10 上午7:04, Aidan MacDonald wrote:
> The X series Ingenic SoCs have a shadow GPIO group which
> is at a higher offset than the other groups, and is used
> for all GPIO configuration. The regmap did not take this
> offset into account and set max_register too low. Writes
> to the shadow group registers were blocked, which made it
> impossible to change any pin configuration.
>
> Fix this by pretending there are at least 8 chips on any
> 'X' SoC for the purposes of calculating max_register. This
> ensures the shadow group is accessible.
>
> Signed-off-by: Aidan MacDonald <aidanmacdonald.0x0@gmail.com>
> ---
>   drivers/pinctrl/pinctrl-ingenic.c | 5 ++++-
>   1 file changed, 4 insertions(+), 1 deletion(-)


The M200 SoC (once planned to be called JZ4785) has a different shadow 
register offset address,

if it needs to be supported in the future, then we need to deal with it 
further, but fortunately pinctrl-ingenic.c

does not involve M200 support yet, so:


Reviewed-by: 周琰杰 (Zhou Yanjie) <zhouyanjie@wanyeetech.com>


Thanks and best regards!


>
> diff --git a/drivers/pinctrl/pinctrl-ingenic.c b/drivers/pinctrl/pinctrl-ingenic.c
> index 2712f51eb238..9d2bccda50f1 100644
> --- a/drivers/pinctrl/pinctrl-ingenic.c
> +++ b/drivers/pinctrl/pinctrl-ingenic.c
> @@ -4168,7 +4168,10 @@ static int __init ingenic_pinctrl_probe(struct platform_device *pdev)
>   		return PTR_ERR(base);
>   
>   	regmap_config = ingenic_pinctrl_regmap_config;
> -	regmap_config.max_register = chip_info->num_chips * chip_info->reg_offset;
> +	if (chip_info->version >= ID_X1000)
> +		regmap_config.max_register = MIN(8, chip_info->num_chips) * chip_info->reg_offset;
> +	else
> +		regmap_config.max_register = chip_info->num_chips * chip_info->reg_offset;
>   
>   	jzpc->map = devm_regmap_init_mmio(dev, base, &regmap_config);
>   	if (IS_ERR(jzpc->map)) {
Paul Cercueil Feb. 10, 2022, 11:03 a.m. UTC | #2
Hi Aidan,

Le mer., févr. 9 2022 at 23:04:54 +0000, Aidan MacDonald 
<aidanmacdonald.0x0@gmail.com> a écrit :
> The X series Ingenic SoCs have a shadow GPIO group which
> is at a higher offset than the other groups, and is used
> for all GPIO configuration. The regmap did not take this
> offset into account and set max_register too low. Writes
> to the shadow group registers were blocked, which made it
> impossible to change any pin configuration.
> 
> Fix this by pretending there are at least 8 chips on any
> 'X' SoC for the purposes of calculating max_register. This
> ensures the shadow group is accessible.

I don't like your solution, it sounds very hacky. I think it would make 
more sense to use a dedicated x1000_pinctrl_regmap_config that would be 
used for the X1000 SoC. That would also allow you to express that there 
are no registers in the 0x400-0x700 range (through 
regmap_config.wr_table / .rd_table).

Cheers,
-Paul

> Signed-off-by: Aidan MacDonald <aidanmacdonald.0x0@gmail.com>
> ---
>  drivers/pinctrl/pinctrl-ingenic.c | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/pinctrl/pinctrl-ingenic.c 
> b/drivers/pinctrl/pinctrl-ingenic.c
> index 2712f51eb238..9d2bccda50f1 100644
> --- a/drivers/pinctrl/pinctrl-ingenic.c
> +++ b/drivers/pinctrl/pinctrl-ingenic.c
> @@ -4168,7 +4168,10 @@ static int __init ingenic_pinctrl_probe(struct 
> platform_device *pdev)
>  		return PTR_ERR(base);
> 
>  	regmap_config = ingenic_pinctrl_regmap_config;
> -	regmap_config.max_register = chip_info->num_chips * 
> chip_info->reg_offset;
> +	if (chip_info->version >= ID_X1000)
> +		regmap_config.max_register = MIN(8, chip_info->num_chips) * 
> chip_info->reg_offset;
> +	else
> +		regmap_config.max_register = chip_info->num_chips * 
> chip_info->reg_offset;
> 
>  	jzpc->map = devm_regmap_init_mmio(dev, base, &regmap_config);
>  	if (IS_ERR(jzpc->map)) {
> --
> 2.34.1
>
Aidan MacDonald Feb. 10, 2022, 11:10 p.m. UTC | #3
On Thu, Feb 10, 2022 at 11:03:13AM +0000, Paul Cercueil wrote:
> Hi Aidan,
> 
> Le mer., févr. 9 2022 at 23:04:54 +0000, Aidan MacDonald 
> <aidanmacdonald.0x0@gmail.com> a écrit :
> > The X series Ingenic SoCs have a shadow GPIO group which
> > is at a higher offset than the other groups, and is used
> > for all GPIO configuration. The regmap did not take this
> > offset into account and set max_register too low. Writes
> > to the shadow group registers were blocked, which made it
> > impossible to change any pin configuration.
> > 
> > Fix this by pretending there are at least 8 chips on any
> > 'X' SoC for the purposes of calculating max_register. This
> > ensures the shadow group is accessible.
> 
> I don't like your solution, it sounds very hacky. I think it would make 
> more sense to use a dedicated x1000_pinctrl_regmap_config that would be 
> used for the X1000 SoC. That would also allow you to express that there 
> are no registers in the 0x400-0x700 range (through 
> regmap_config.wr_table / .rd_table).
> 
> Cheers,
> -Paul
> 

That's fine, I'll put together a v2 patch using the approach you suggest.
I think all the 'X' SoCs will require a similar regmap as they're using
ingenic_gpio_shadow_* functions too, but I'll check their datasheets to
be sure.

> > Signed-off-by: Aidan MacDonald <aidanmacdonald.0x0@gmail.com>
> > ---
> >  drivers/pinctrl/pinctrl-ingenic.c | 5 ++++-
> >  1 file changed, 4 insertions(+), 1 deletion(-)
> > 
> > diff --git a/drivers/pinctrl/pinctrl-ingenic.c 
> > b/drivers/pinctrl/pinctrl-ingenic.c
> > index 2712f51eb238..9d2bccda50f1 100644
> > --- a/drivers/pinctrl/pinctrl-ingenic.c
> > +++ b/drivers/pinctrl/pinctrl-ingenic.c
> > @@ -4168,7 +4168,10 @@ static int __init ingenic_pinctrl_probe(struct 
> > platform_device *pdev)
> >  		return PTR_ERR(base);
> > 
> >  	regmap_config = ingenic_pinctrl_regmap_config;
> > -	regmap_config.max_register = chip_info->num_chips * 
> > chip_info->reg_offset;
> > +	if (chip_info->version >= ID_X1000)
> > +		regmap_config.max_register = MIN(8, chip_info->num_chips) * 
> > chip_info->reg_offset;
> > +	else
> > +		regmap_config.max_register = chip_info->num_chips * 
> > chip_info->reg_offset;
> > 
> >  	jzpc->map = devm_regmap_init_mmio(dev, base, &regmap_config);
> >  	if (IS_ERR(jzpc->map)) {
> > --
> > 2.34.1
> > 
> 
>
diff mbox series

Patch

diff --git a/drivers/pinctrl/pinctrl-ingenic.c b/drivers/pinctrl/pinctrl-ingenic.c
index 2712f51eb238..9d2bccda50f1 100644
--- a/drivers/pinctrl/pinctrl-ingenic.c
+++ b/drivers/pinctrl/pinctrl-ingenic.c
@@ -4168,7 +4168,10 @@  static int __init ingenic_pinctrl_probe(struct platform_device *pdev)
 		return PTR_ERR(base);
 
 	regmap_config = ingenic_pinctrl_regmap_config;
-	regmap_config.max_register = chip_info->num_chips * chip_info->reg_offset;
+	if (chip_info->version >= ID_X1000)
+		regmap_config.max_register = MIN(8, chip_info->num_chips) * chip_info->reg_offset;
+	else
+		regmap_config.max_register = chip_info->num_chips * chip_info->reg_offset;
 
 	jzpc->map = devm_regmap_init_mmio(dev, base, &regmap_config);
 	if (IS_ERR(jzpc->map)) {