Re: [PATCH 1/2] spi: rockchip: add support for "cs-gpios" dts property

[jeffy <jeffy.chen-TNX95d0MmH7DzftRWevZcw@public.gmane.org>]

Hi Shawn,

On 06/12/2017 04:26 PM, jeffy wrote:
> Hi Shawn,
>
> On 06/12/2017 03:15 PM, Shawn Lin wrote:
>> Hi Jeffy,
>>
>> On 2017/6/12 14:14, Jeffy Chen wrote:
>>> Support using "cs-gpios" property to specify cs gpios.
>>>
>>> Signed-off-by: Jeffy Chen <jeffy.chen-TNX95d0MmH7DzftRWevZcw@public.gmane.org>
>>> ---
>>>
>>>   .../devicetree/bindings/spi/spi-rockchip.txt       |  2 +
>>>   drivers/spi/spi-rockchip.c                         | 52
>>> ++++++++++++++++++++++
>>>   2 files changed, 54 insertions(+)
>>>
>>> diff --git a/Documentation/devicetree/bindings/spi/spi-rockchip.txt
>>> b/Documentation/devicetree/bindings/spi/spi-rockchip.txt
>>> index 83da493..02171b2 100644
>>> --- a/Documentation/devicetree/bindings/spi/spi-rockchip.txt
>>> +++ b/Documentation/devicetree/bindings/spi/spi-rockchip.txt
>>
>> The changes for doc should be another patch, and...
>
> but i saw others didn't separate them:
> cf9e478 spi: sh-msiof: Add slave mode support
> 23e291c spi: rockchip: support "sleep" pin configuration
>
>
>>
>>> @@ -17,6 +17,7 @@ Required Properties:
>>>          region.
>>>   - interrupts: The interrupt number to the cpu. The interrupt
>>> specifier format
>>>                 depends on the interrupt controller.
>>> +- cs-gpios : Specifies the gpio pins to be used for chipselects.
>>
>> It's not a required property, otherwise how other boards work as your
>> patch 2 only add this for rk3399-gru.
> oops, i will fix it.
>>
>>>   - clocks: Must contain an entry for each entry in clock-names.
>>>   - clock-names: Shall be "spiclk" for the transfer-clock, and
>>> "apb_pclk" for
>>>                  the peripheral clock.
>>> @@ -48,6 +49,7 @@ Example:
>>>           #address-cells = <1>;
>>>           #size-cells = <0>;
>>>           interrupts = <GIC_SPI 44 IRQ_TYPE_LEVEL_HIGH>;
>>> +        cs-gpios = <&gpio2 23 GPIO_ACTIVE_HIGH>;
>>>           clocks = <&cru SCLK_SPI0>, <&cru PCLK_SPI0>;
>>>           clock-names = "spiclk", "apb_pclk";
>>>           pinctrl-0 = <&spi1_pins>;
>>> diff --git a/drivers/spi/spi-rockchip.c b/drivers/spi/spi-rockchip.c
>>> index acf31f3..04694e1 100644
>>> --- a/drivers/spi/spi-rockchip.c
>>> +++ b/drivers/spi/spi-rockchip.c
>>> @@ -15,6 +15,7 @@
>>>   #include <linux/clk.h>
>>>   #include <linux/dmaengine.h>
>>> +#include <linux/gpio.h>
>>>   #include <linux/module.h>
>>>   #include <linux/of.h>
>>>   #include <linux/pinctrl/consumer.h>
>>> @@ -201,6 +202,10 @@ struct rockchip_spi {
>>>       struct dma_slave_caps dma_caps;
>>>   };
>>> +struct rockchip_spi_data {
>>> +    bool cs_gpio_requested;
>>> +};
>>> +
>>
>> Could you fold cs_gpio_requested into struct rockchip_spi?
we might have multiple children with different cs_gpio, so i think we 
may need a separate data struct for them.
>>
>>>   static inline void spi_enable_chip(struct rockchip_spi *rs, int
>>> enable)
>>>   {
>>>       writel_relaxed((enable ? 1 : 0), rs->regs + ROCKCHIP_SPI_SSIENR);
>>> @@ -297,6 +302,50 @@ static void rockchip_spi_set_cs(struct spi_device
>>> *spi, bool enable)
>>>       pm_runtime_put_sync(rs->dev);
>>>   }
>>> +static int rockchip_spi_setup(struct spi_device *spi)
>>> +{
>>> +    int ret = 0;
>>> +    unsigned long flags = (spi->mode & SPI_CS_HIGH) ?
>>> +                  GPIOF_OUT_INIT_LOW : GPIOF_OUT_INIT_HIGH;
>>> +    struct rockchip_spi_data *data = spi_get_ctldata(spi);
>>> +
>>> +    if (!gpio_is_valid(spi->cs_gpio))
>>> +        return 0;
>>
>> return -EINVAL?
> i think we can still support the original way, which means no "cs-gpios"
> and do nothing in setup.
>>
>>> +
>>> +    if (!data) {
>>> +        data = kzalloc(sizeof(*data), GFP_KERNEL);
>>> +        if (!data)
>>> +            return -ENOMEM;
>>> +        spi_set_ctldata(spi, data);
>>> +    }
>>> +
>>> +    if (!data->cs_gpio_requested) {
>>> +        ret = gpio_request_one(spi->cs_gpio, flags,
>>> +                       dev_name(&spi->dev));
>>> +        if (!ret)
>>> +            data->cs_gpio_requested = 1;
>>> +    } else
>>> +        ret = gpio_direction_output(spi->cs_gpio, flags);
>>
>> need brace around 'else' statement. Also I don't see data used
>> elsewhere, so you need these code above.
> ok.
> and the cs_gpio_requested is to mark cs_gpio requested, because the
> setup func might be called multiple times, we only need to request gpio
> at the first time.
>>
>>> +
>>> +    if (ret < 0)
>>> +        dev_err(&spi->dev, "Failed to setup cs gpio(%d): %d\n",
>>> +            spi->cs_gpio, ret);
>>> +
>>> +    return ret;
>>> +}
>>> +
>>> +static void rockchip_spi_cleanup(struct spi_device *spi)
>>> +{
>>> +    struct rockchip_spi_data *data = spi_get_ctldata(spi);
>>> +
>>> +    if (data) {
>>> +        if (data->cs_gpio_requested)
>>> +            gpio_free(spi->cs_gpio);
>>> +        kfree(data);
>>> +        spi_set_ctldata(spi, NULL);
>>> +    }
>>> +}
>>> +
>>>   static int rockchip_spi_prepare_message(struct spi_master *master,
>>>                       struct spi_message *msg)
>>>   {
>>> @@ -744,11 +793,14 @@ static int rockchip_spi_probe(struct
>>> platform_device *pdev)
>>>       master->bits_per_word_mask = SPI_BPW_MASK(16) | SPI_BPW_MASK(8);
>>>       master->set_cs = rockchip_spi_set_cs;
>>> +    master->setup = rockchip_spi_setup;
>>> +    master->cleanup = rockchip_spi_cleanup;
>>>       master->prepare_message = rockchip_spi_prepare_message;
>>>       master->unprepare_message = rockchip_spi_unprepare_message;
>>>       master->transfer_one = rockchip_spi_transfer_one;
>>>       master->max_transfer_size = rockchip_spi_max_transfer_size;
>>>       master->handle_err = rockchip_spi_handle_err;
>>> +    master->flags = SPI_MASTER_GPIO_SS;
>>>       rs->dma_tx.ch = dma_request_chan(rs->dev, "tx");
>>>       if (IS_ERR(rs->dma_tx.ch)) {
>>>
>>
>>
>>
>


--
To unsubscribe from this list: send the line "unsubscribe linux-spi" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html