Re: [PATCH 09/17] media: camss: csid: allow csid to work without a regulator

[Robert Foss]

On Tue, 11 May 2021 at 20:08, Jonathan Marek <jonathan@xxxxxxxx> wrote:
>
> At least for titan HW, CSID don't have an associated regulator. This change
> is necessary to be able to model this in the CSID resources.
>
> Signed-off-by: Jonathan Marek <jonathan@xxxxxxxx>
> ---
>  drivers/media/platform/qcom/camss/camss-csid.c | 17 +++++++++++------
>  1 file changed, 11 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/media/platform/qcom/camss/camss-csid.c b/drivers/media/platform/qcom/camss/camss-csid.c
> index cc11fbfdae13..528674dea06c 100644
> --- a/drivers/media/platform/qcom/camss/camss-csid.c
> +++ b/drivers/media/platform/qcom/camss/camss-csid.c
> @@ -162,7 +162,7 @@ static int csid_set_power(struct v4l2_subdev *sd, int on)
>                         return ret;
>                 }
>
> -               ret = regulator_enable(csid->vdda);
> +               ret = csid->vdda ? regulator_enable(csid->vdda) : 0;
>                 if (ret < 0) {
>                         pm_runtime_put_sync(dev);
>                         return ret;
> @@ -170,14 +170,16 @@ static int csid_set_power(struct v4l2_subdev *sd, int on)
>
>                 ret = csid_set_clock_rates(csid);
>                 if (ret < 0) {
> -                       regulator_disable(csid->vdda);
> +                       if (csid->vdda)
> +                               regulator_disable(csid->vdda);
>                         pm_runtime_put_sync(dev);
>                         return ret;
>                 }
>
>                 ret = camss_enable_clocks(csid->nclocks, csid->clock, dev);
>                 if (ret < 0) {
> -                       regulator_disable(csid->vdda);
> +                       if (csid->vdda)
> +                               regulator_disable(csid->vdda);
>                         pm_runtime_put_sync(dev);
>                         return ret;
>                 }
> @@ -188,7 +190,8 @@ static int csid_set_power(struct v4l2_subdev *sd, int on)
>                 if (ret < 0) {
>                         disable_irq(csid->irq);
>                         camss_disable_clocks(csid->nclocks, csid->clock);
> -                       regulator_disable(csid->vdda);
> +                       if (csid->vdda)
> +                               regulator_disable(csid->vdda);
>                         pm_runtime_put_sync(dev);
>                         return ret;

Since this & the previous chunks of failure cleanups are growing
larger, maybe it is time to extract all of the failure cleanups into
gotos. That should probably go into a seperate patch though.

>                 }
> @@ -197,7 +200,7 @@ static int csid_set_power(struct v4l2_subdev *sd, int on)
>         } else {
>                 disable_irq(csid->irq);
>                 camss_disable_clocks(csid->nclocks, csid->clock);
> -               ret = regulator_disable(csid->vdda);
> +               ret = csid->vdda ? regulator_disable(csid->vdda) : 0;
>                 pm_runtime_put_sync(dev);
>         }
>
> @@ -634,7 +637,9 @@ int msm_csid_subdev_init(struct camss *camss, struct csid_device *csid,
>
>         /* Regulator */
>
> -       csid->vdda = devm_regulator_get(dev, res->regulator[0]);
> +       csid->vdda = NULL;
> +       if (res->regulator[0])
> +               csid->vdda = devm_regulator_get(dev, res->regulator[0]);
>         if (IS_ERR(csid->vdda)) {
>                 dev_err(dev, "could not get regulator\n");
>                 return PTR_ERR(csid->vdda);

Reviewed-by: Robert Foss <robert.foss@xxxxxxxxxx>