Re: [PATCH v2 23/91] reset: simple: Add reset callback

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Maxime,

On Fri, 2020-04-24 at 17:34 +0200, Maxime Ripard wrote:
> The reset-simple code lacks a reset callback that is still pretty easy to
> implement. The only real thing to consider is the delay needed for a device
> to be reset, so let's expose that as part of the reset-simple driver data.
> 
> Cc: Philipp Zabel <p.zabel@xxxxxxxxxxxxxx>
> Signed-off-by: Maxime Ripard <maxime@xxxxxxxxxx>
> ---
>  drivers/reset/reset-simple.c       | 24 ++++++++++++++++++++++++
>  include/linux/reset/reset-simple.h |  6 ++++++
>  2 files changed, 30 insertions(+)
> 
> diff --git a/drivers/reset/reset-simple.c b/drivers/reset/reset-simple.c
> index c854aa351640..602ed972b0a9 100644
> --- a/drivers/reset/reset-simple.c
> +++ b/drivers/reset/reset-simple.c
> @@ -11,6 +11,7 @@
>   * Maxime Ripard <maxime.ripard@xxxxxxxxxxxxxxxxxx>
>   */
>  
> +#include <linux/delay.h>
>  #include <linux/device.h>
>  #include <linux/err.h>
>  #include <linux/io.h>
> @@ -63,6 +64,28 @@ static int reset_simple_deassert(struct reset_controller_dev *rcdev,
>  	return reset_simple_update(rcdev, id, false);
>  }
>  
> +static int reset_simple_reset(struct reset_controller_dev *rcdev,
> +			      unsigned long id)
> +{
> +	struct reset_simple_data *data = to_reset_simple_data(rcdev);
> +	int ret;
> +
> +	if (!data->reset_us)
> +		return -ENOTSUPP;
> +
> +	ret = reset_simple_assert(rcdev, id);
> +	if (ret)
> +		return ret;
> +
> +	usleep_range(data->reset_us, data->reset_us * 2);
> +
> +	ret = reset_simple_deassert(rcdev, id);
> +	if (ret)
> +		return ret;
> +
> +	return 0;

Just
	return reset_simple_deassert(rcdev, id);
here.

> +}
> +
>  static int reset_simple_status(struct reset_controller_dev *rcdev,
>  			       unsigned long id)
>  {
> @@ -80,6 +103,7 @@ static int reset_simple_status(struct reset_controller_dev *rcdev,
>  const struct reset_control_ops reset_simple_ops = {
>  	.assert		= reset_simple_assert,
>  	.deassert	= reset_simple_deassert,
> +	.reset		= reset_simple_reset,
>  	.status		= reset_simple_status,
>  };
>  EXPORT_SYMBOL_GPL(reset_simple_ops);
> diff --git a/include/linux/reset/reset-simple.h b/include/linux/reset/reset-simple.h
> index 08ccb25a55e6..5eb83625a495 100644
> --- a/include/linux/reset/reset-simple.h
> +++ b/include/linux/reset/reset-simple.h
> @@ -27,6 +27,11 @@
>   * @status_active_low: if true, bits read back as cleared while the reset is
>   *                     asserted. Otherwise, bits read back as set while the
>   *                     reset is asserted.
> + * @reset_us: Minimum delay in microseconds needed that needs to be
> + *            waited for between an assert and a deassert to reset the
> + *            device. If multiple consumers with different delay
> + *            requirements are connected to this controller, it must
> + *            be the largest minimum delay.

Please mention that 0 does not meen 0 µs delay, but unknown and thus
reset callback not supported.

With these two issues fixed
Reviewed-by: Philipp Zabel <p.zabel@xxxxxxxxxxxxxx>

regards
Philipp
_______________________________________________
dri-devel mailing list
dri-devel@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/dri-devel




[Index of Archives]     [Linux DRI Users]     [Linux Intel Graphics]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [XFree86]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [XFree86]
  Powered by Linux