i2c: muxes: pca954x: Allow sharing reset GPIO

Some hardware designs with multiple PCA954x devices use a reset GPIO
connected to all the muxes. Support this configuration by making use of
the reset controller framework which can deal with the shared reset
GPIOs. Fall back to the old GPIO descriptor method if the reset
controller framework is not enabled.

Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
---

Notes:
    This patch goes on top of Krzysztof's series adding the GPIO based reset
    controller[1] which will be in linux-6.9. With this I'm able to
    correctly describe my hardware platform in the DTS and have the resets
    appropriately controlled.
    
    [1] - https://lore.kernel.org/all/20240129115216.96479-1-krzysztof.kozlowski@linaro.org/

 drivers/i2c/muxes/i2c-mux-pca954x.c | 46 ++++++++++++++++++++++++-----
 1 file changed, 38 insertions(+), 8 deletions(-)

On 11/03/2024 05:14, Chris Packham wrote:
> Some hardware designs with multiple PCA954x devices use a reset GPIO
> connected to all the muxes. Support this configuration by making use of
> the reset controller framework which can deal with the shared reset
> GPIOs. Fall back to the old GPIO descriptor method if the reset
> controller framework is not enabled.
> 
> Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
> ---
> 
> Notes:
>     This patch goes on top of Krzysztof's series adding the GPIO based reset
>     controller[1] which will be in linux-6.9. With this I'm able to
>     correctly describe my hardware platform in the DTS and have the resets
>     appropriately controlled.
>     
>     [1] - https://lore.kernel.org/all/20240129115216.96479-1-krzysztof.kozlowski@linaro.org/
> 

Where is the changelog? It was v3 or v4 already? Where are the tags?

It looks like you received a tag and forgot to add it.

If you do not know the process, here is a short explanation:
Please add Acked-by/Reviewed-by/Tested-by tags when posting new
versions, under or above your Signed-off-by tag. Tag is "received", when
provided in a message replied to you on the mailing list. Tools like b4
can help here. However, there's no need to repost patches *only* to add
the tags. The upstream maintainer will do that for tags received on the
version they apply.

https://elixir.bootlin.com/linux/v6.5-rc3/source/Documentation/process/submitting-patches.rst#L577

If a tag was not added on purpose, please state why and what changed.

Best regards,
Krzysztof

On 11/03/24 19:25, Krzysztof Kozlowski wrote:
> On 11/03/2024 05:14, Chris Packham wrote:
>> Some hardware designs with multiple PCA954x devices use a reset GPIO
>> connected to all the muxes. Support this configuration by making use of
>> the reset controller framework which can deal with the shared reset
>> GPIOs. Fall back to the old GPIO descriptor method if the reset
>> controller framework is not enabled.
>>
>> Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
>> ---
>>
>> Notes:
>>      This patch goes on top of Krzysztof's series adding the GPIO based reset
>>      controller[1] which will be in linux-6.9. With this I'm able to
>>      correctly describe my hardware platform in the DTS and have the resets
>>      appropriately controlled.
>>      
>>      [1] - https://lore.kernel.org/all/20240129115216.96479-1-krzysztof.kozlowski@linaro.org/
>>
> Where is the changelog? It was v3 or v4 already? Where are the tags?

Since this was spun off the original series I figured it would be best 
to treat it as v1 of a new series. It's unchanged since 
https://lore.kernel.org/lkml/20240112163608.528453-6-krzysztof.kozlowski@linaro.org/ 
so technically this is '[PATCH resend v3]'. I did rebase it on top of 
linux-next just to make sure it still worked.

> It looks like you received a tag and forgot to add it.
>
> If you do not know the process, here is a short explanation:
> Please add Acked-by/Reviewed-by/Tested-by tags when posting new
> versions, under or above your Signed-off-by tag. Tag is "received", when
> provided in a message replied to you on the mailing list. Tools like b4
> can help here. However, there's no need to repost patches *only* to add
> the tags. The upstream maintainer will do that for tags received on the
> version they apply.
>
> https://elixir.bootlin.com/linux/v6.5-rc3/source/Documentation/process/submitting-patches.rst#L577
>
> If a tag was not added on purpose, please state why and what changed.
Oops my bad. I think because by the time v3 rolled round it was part of 
your series I never added the Ack from Peter or your R-by to my local 
commit which is what I sent out. I've added it now if I send a v4 (or is 
it v2 of this?) I'll include the tags.

On 3/10/24 23:14, Chris Packham wrote:
> Some hardware designs with multiple PCA954x devices use a reset GPIO
> connected to all the muxes. Support this configuration by making use of
> the reset controller framework which can deal with the shared reset
> GPIOs. Fall back to the old GPIO descriptor method if the reset
> controller framework is not enabled.


Hello Chris, Krzysztof,


This change makes it so that the reset subsystem reset doesn't behave in 
the same way as the fallback gpio reset. The gpio, as part of acquiring 
it, gets set high, and then set low in the mux driver. So, the device 
reset is toggled. In the case of the reset subsystem option, the reset 
is only de-asserted (so the device is taken out of reset).


I'm interested in preserving the previous behavior but with the shared 
reset line. This can't be done just by doing "assert" first because the 
shared reset subsystem doesn't allow that. So the reset-gpio driver 
would have to implement the reset operation - no problem. However, how 
to specify the wait time for the reset-gpio driver here? Something like 
the simple reset driver maybe? Or a function call from the reset 
consumer driver to specify the wait time for that reset?


Thanks,

Eddie


>
> Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
> ---
>
> Notes:
>      This patch goes on top of Krzysztof's series adding the GPIO based reset
>      controller[1] which will be in linux-6.9. With this I'm able to
>      correctly describe my hardware platform in the DTS and have the resets
>      appropriately controlled.
>      
>      [1] - https://lore.kernel.org/all/20240129115216.96479-1-krzysztof.kozlowski@linaro.org/
>
>   drivers/i2c/muxes/i2c-mux-pca954x.c | 46 ++++++++++++++++++++++++-----
>   1 file changed, 38 insertions(+), 8 deletions(-)
>
> diff --git a/drivers/i2c/muxes/i2c-mux-pca954x.c b/drivers/i2c/muxes/i2c-mux-pca954x.c
> index f5dfc33b97c0..c3f4ff08ac38 100644
> --- a/drivers/i2c/muxes/i2c-mux-pca954x.c
> +++ b/drivers/i2c/muxes/i2c-mux-pca954x.c
> @@ -49,6 +49,7 @@
>   #include <linux/pm.h>
>   #include <linux/property.h>
>   #include <linux/regulator/consumer.h>
> +#include <linux/reset.h>
>   #include <linux/slab.h>
>   #include <linux/spinlock.h>
>   #include <dt-bindings/mux/mux.h>
> @@ -116,6 +117,9 @@ struct pca954x {
>   	unsigned int irq_mask;
>   	raw_spinlock_t lock;
>   	struct regulator *supply;
> +
> +	struct gpio_desc *reset_gpio;
> +	struct reset_control *reset_cont;
>   };
>   
>   /* Provide specs for the MAX735x, PCA954x and PCA984x types we know about */
> @@ -518,6 +522,35 @@ static int pca954x_init(struct i2c_client *client, struct pca954x *data)
>   	return ret;
>   }
>   
> +static int pca954x_get_reset(struct device *dev, struct pca954x *data)
> +{
> +	data->reset_cont = devm_reset_control_get_optional_shared(dev, NULL);
> +	if (IS_ERR(data->reset_cont))
> +		return dev_err_probe(dev, PTR_ERR(data->reset_cont),
> +				     "Failed to get reset\n");
> +	else if (data->reset_cont)
> +		return 0;
> +
> +	/*
> +	 * fallback to legacy reset-gpios
> +	 */
> +	data->reset_gpio = devm_gpiod_get_optional(dev, "reset", GPIOD_OUT_HIGH);
> +	if (IS_ERR(data->reset_gpio)) {
> +		return dev_err_probe(dev, PTR_ERR(data->reset_gpio),
> +				     "Failed to get reset gpio");
> +	}
> +
> +	return 0;
> +}
> +
> +static void pca954x_reset_deassert(struct pca954x *data)
> +{
> +	if (data->reset_cont)
> +		reset_control_deassert(data->reset_cont);
> +	else
> +		gpiod_set_value_cansleep(data->reset_gpio, 0);
> +}
> +
>   /*
>    * I2C init/probing/exit functions
>    */
> @@ -526,7 +559,6 @@ static int pca954x_probe(struct i2c_client *client)
>   	const struct i2c_device_id *id = i2c_client_get_device_id(client);
>   	struct i2c_adapter *adap = client->adapter;
>   	struct device *dev = &client->dev;
> -	struct gpio_desc *gpio;
>   	struct i2c_mux_core *muxc;
>   	struct pca954x *data;
>   	int num;
> @@ -554,15 +586,13 @@ static int pca954x_probe(struct i2c_client *client)
>   		return dev_err_probe(dev, ret,
>   				     "Failed to enable vdd supply\n");
>   
> -	/* Reset the mux if a reset GPIO is specified. */
> -	gpio = devm_gpiod_get_optional(dev, "reset", GPIOD_OUT_HIGH);
> -	if (IS_ERR(gpio)) {
> -		ret = PTR_ERR(gpio);
> +	ret = pca954x_get_reset(dev, data);
> +	if (ret)
>   		goto fail_cleanup;
> -	}
> -	if (gpio) {
> +
> +	if (data->reset_cont || data->reset_gpio) {
>   		udelay(1);
> -		gpiod_set_value_cansleep(gpio, 0);
> +		pca954x_reset_deassert(data);
>   		/* Give the chip some time to recover. */
>   		udelay(1);
>   	}

On 06/06/2024 23:58, Eddie James wrote:
> 
> On 3/10/24 23:14, Chris Packham wrote:
>> Some hardware designs with multiple PCA954x devices use a reset GPIO
>> connected to all the muxes. Support this configuration by making use of
>> the reset controller framework which can deal with the shared reset
>> GPIOs. Fall back to the old GPIO descriptor method if the reset
>> controller framework is not enabled.
> 
> 
> Hello Chris, Krzysztof,
> 
> 
> This change makes it so that the reset subsystem reset doesn't behave in 
> the same way as the fallback gpio reset. The gpio, as part of acquiring 
> it, gets set high, and then set low in the mux driver. So, the device 
> reset is toggled. In the case of the reset subsystem option, the reset 
> is only de-asserted (so the device is taken out of reset).
> 
> 
> I'm interested in preserving the previous behavior but with the shared 
> reset line. This can't be done just by doing "assert" first because the 
> shared reset subsystem doesn't allow that. So the reset-gpio driver 
> would have to implement the reset operation - no problem. However, how 
> to specify the wait time for the reset-gpio driver here? Something like 
> the simple reset driver maybe? Or a function call from the reset 
> consumer driver to specify the wait time for that reset?

You can check my slides from EOSS/OSSNA this year:
https://ossna2024.sched.com/event/1aPvr?iframe=no
maybe links to original work will help you.

Best regards,
Krzysztof