Re: [PATCH v2 3/3] mfd: atmel-flexcom: Add support for lan966x flexcom chip-select configuration

[Kavyasree.Kotagiri]

>>>>> LAN966x SoC have 5 flexcoms. Each flexcom has 2 chip-selects.
>>>>> For each chip select of each flexcom there is a configuration
>>>>> register FLEXCOM_SHARED[0-4]:SS_MASK[0-1]. The width of
>>>>> configuration register is 21 because there are 21 shared pins
>>>>> on each of which the chip select can be mapped. Each bit of the
>>>>> register represents a different FLEXCOM_SHARED pin.
>>>>> Signed-off-by: Kavyasree Kotagiri <kavyasree.kotagiri@xxxxxxxxxxxxx>
>>>>> ---
>>>>> v1 -> v2:
>>>>> - use GENMASK for mask, macros for maximum allowed values.
>>>>> - use u32 values for flexcom chipselects instead of strings.
>>>>> - disable clock in case of errors.
>>>>> drivers/mfd/atmel-flexcom.c | 93
>>>> ++++++++++++++++++++++++++++++++++++-
>>>>> 1 file changed, 92 insertions(+), 1 deletion(-)
>>>>> diff --git a/drivers/mfd/atmel-flexcom.c b/drivers/mfd/atmel-flexcom.c
>>>>> index 33caa4fba6af..ac700a85b46f 100644
>>>>> --- a/drivers/mfd/atmel-flexcom.c
>>>>> +++ b/drivers/mfd/atmel-flexcom.c
>>>>> @@ -28,15 +28,68 @@
>>>>> #define FLEX_MR_OPMODE(opmode)    (((opmode) <<
>>>> FLEX_MR_OPMODE_OFFSET) &    \
>>>>>                FLEX_MR_OPMODE_MASK)
>>>>> +/* LAN966x flexcom shared register offsets */
>>>>> +#define FLEX_SHRD_SS_MASK_0    0x0
>>>>> +#define FLEX_SHRD_SS_MASK_1    0x4
>>>>> +#define FLEX_SHRD_PIN_MAX    20
>>>>> +#define FLEX_CS_MAX        1
>>>>> +#define FLEX_SHRD_MASK        GENMASK(20, 0)
>>>>> +
>>>>> +struct atmel_flex_caps {
>>>>> +    bool has_flx_cs;
>>>>> +};
>>>>> +
>>>>> struct atmel_flexcom {
>>>>>   void __iomem *base;
>>>>> +    void __iomem *flexcom_shared_base;
>>>>>   u32 opmode;
>>>>>   struct clk *clk;
>>>>> };
>>>>> +static int atmel_flexcom_lan966x_cs_config(struct platform_device
>> *pdev)
>>>>> +{
>>>>> +    struct atmel_flexcom *ddata = dev_get_drvdata(&pdev->dev);
>>>>> +    struct device_node *np = pdev->dev.of_node;
>>>>> +    u32 flx_shrd_pins[2], flx_cs[2], val;
>>>>> +    int err, i, count;
>>>>> +
>>>>> +    count = of_property_count_u32_elems(np, "microchip,flx-shrd-
>>>> pins");
>>>>> +    if (count <= 0 || count > 2) {
>>>>> +        dev_err(&pdev->dev, "Invalid %s property (%d)\n", "flx-shrd-
>>>> pins",
>>>>> +                count);
>>>>> +        return -EINVAL;
>>>>> +    }
>>>>> +
>>>>> +    err = of_property_read_u32_array(np, "microchip,flx-shrd-pins",
>>>> flx_shrd_pins, count);
>>>>> +    if (err)
>>>>> +        return err;
>>>>> +
>>>>> +    err = of_property_read_u32_array(np, "microchip,flx-cs", flx_cs,
>>>> count);
>>>>> +    if (err)
>>>>> +        return err;
>>>>> +
>>>>> +    for (i = 0; i < count; i++) {
>>>>> +        if (flx_shrd_pins[i] > FLEX_SHRD_PIN_MAX)
>>>>> +            return -EINVAL;
>>>>> +
>>>>> +        if (flx_cs[i] > FLEX_CS_MAX)
>>>>> +            return -EINVAL;
>>>>> +
>>>>> +        val = ~(1 << flx_shrd_pins[i]) & FLEX_SHRD_MASK;
>>>>> +
>>>>> +        if (flx_cs[i] == 0)
>>>>> +            writel(val, ddata->flexcom_shared_base +
>>>> FLEX_SHRD_SS_MASK_0);
>>>>> +        else
>>>>> +            writel(val, ddata->flexcom_shared_base +
>>>> FLEX_SHRD_SS_MASK_1);
>>>> There is still an open question on this topic from previous version.
>>> https://lore.kernel.org/linux-arm-
>> kernel/PH0PR11MB48724DE09A50D67F1EA9FBE092D89@PH0PR11MB4872.n
>> amprd11.prod.outlook.com/
>> "previous version" meant for me this the one at [1]... Another point that
>> the versioning of this series is bad.
>> The question was the following:
>> "I may miss something but I don't see here the approach you introduced in
>> [1]:
>> +            err = mux_control_select(flx_mux, args.args[0]);
>> +            if (!err) {
>> +                mux_control_deselect(flx_mux);
>> "
>> As I had in mind that you said you need mux_control_deselect() because
>> your
>> serial remain blocked otherwise (but I don't find that in the comments of
>> [1]). And I don't see something similar to mux_control_deselect() being
>> called in this patch.
>> [1]
>> https://lore.kernel.org/linux-arm-kernel/5f9fcc33-cc0f-c404-cf7f-
>> cb73f60154ff@xxxxxxxxxxxxx/
>>> As part of comments from Peter Rosin - Instead of using mux driver, This
>> patch is introducing
>>> new dt-properties in atmel-flexom driver itlself to configure Flexcom
>> shared registers.
>>> Based on the chip-select(0 or 1) to be mapped to flexcom shared pin, write
>> to the
>>> respective register.
>>> If you still have any questions, please comment.
> https://lore.kernel.org/linux-arm-kernel/PH0PR11MB48724DE09A50D67F1EA9FBE092D89@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx/
> To avoid confusion, I stopped continuing with above patch versioning(mux driver approach).
> I started new patch series in which I am configuring FLEXCOM_SHARED[0-4]:SS_MASK[0-1]
> registers in atmel-flexcom.c driver using new DT-properties, mux driver approach is no more followed
> as suggested by Peter Rosin:
> "
>> If you are content with just programming a fixed set of values to
>> a couple of registers depending on how the board is wired, some
>> extra DT property on some node related to the flexcom seems like
>> a better fit than a mux driver.
> Based on your inputs, I planned to send a new patch with new DT properties
> introduced in atmel-flexcom.c driver rather than mux driver.
> 
> Thanks,
> Kavya
> "
> 
> Thanks,
> Kavya

Hi Claudiu,

Please let me know if you still have any comments. If not, I will send my v3 with clk_disable_unprepare moved to goto and some minor fixes(irq flags) in dt-bindings.

>>>>> +    }
>>>>> +
>>>>> +    return 0;
>>>>> +}
>>>>> +
>>>>> static int atmel_flexcom_probe(struct platform_device *pdev)
>>>>> {
>>>>>   struct device_node *np = pdev->dev.of_node;
>>>>> +    const struct atmel_flex_caps *caps;
>>>>>   struct resource *res;
>>>>>   struct atmel_flexcom *ddata;
>>>>>   int err;
>>>>> @@ -76,13 +129,51 @@ static int atmel_flexcom_probe(struct
>>>> platform_device *pdev)
>>>>>    */
>>>>>   writel(FLEX_MR_OPMODE(ddata->opmode), ddata->base +
>>>> FLEX_MR);
>>>>> +    caps = of_device_get_match_data(&pdev->dev);
>>>>> +    if (!caps) {
>>>>> +        dev_err(&pdev->dev, "Could not retrieve flexcom caps\n");
>>>>> +        clk_disable_unprepare(ddata->clk);
>>>> Could you keep a common path to disable the clock? A goto label
>> something
>>>> like this:
>>>>       ret = -EINVAL;
>>>>       got clk_disable_unprepare;
>>>>> +        return -EINVAL;
>>>>> +    }
>>>>> +
>>>>> +    if (caps->has_flx_cs) {
>>>>> +        ddata->flexcom_shared_base =
>>>> devm_platform_get_and_ioremap_resource(pdev, 1, NULL);
>>>>> +        if (IS_ERR(ddata->flexcom_shared_base)) {
>>>>> +            clk_disable_unprepare(ddata->clk);
>>>>> +            return dev_err_probe(&pdev->dev,
>>>>> +                    PTR_ERR(ddata-
>>>>> flexcom_shared_base),
>>>>> +                    "failed to get flexcom shared base
>>>> address\n");
>>>>           ret = dev_err_probe(...);
>>>>           goto clk_disable_unprepare;
>>>>> +        }
>>>>> +
>>>>> +        err = atmel_flexcom_lan966x_cs_config(pdev);
>>>>> +        if (err) {
>>>>> +            clk_disable_unprepare(ddata->clk);
>>>>> +            return err;
>>>>           goto clk_disable_unprepare;
>>>>> +        }
>>>>> +    }
>>>>> +
>>>> clk_unprepare:
>>>>>   clk_disable_unprepare(ddata->clk);
>>>>   if (ret)
>>>>       return ret;
>>>>>   return devm_of_platform_populate(&pdev->dev);
>>>>> }
>>>>> +static const struct atmel_flex_caps atmel_flexcom_caps = {};
>>>>> +
>>>>> +static const struct atmel_flex_caps lan966x_flexcom_caps = {
>>>>> +    .has_flx_cs = true,
>>>>> +};
>>>>> +
>>>>> static const struct of_device_id atmel_flexcom_of_match[] = {
>>>>> -    { .compatible = "atmel,sama5d2-flexcom" },
>>>>> +    {
>>>>> +        .compatible = "atmel,sama5d2-flexcom",
>>>>> +        .data = &atmel_flexcom_caps,
>>>>> +    },
>>>>> +
>>>>> +    {
>>>>> +        .compatible = "microchip,lan966x-flexcom",
>>>>> +        .data = &lan966x_flexcom_caps,
>>>>> +    },
>>>>> +
>>>>>   { /* sentinel */ }
>>>>> };
>>>>> MODULE_DEVICE_TABLE(of, atmel_flexcom_of_match);