Re: [PATCH v2] i2c: i2c-qcom-geni: Serve transfer during early resume stage

[Bjorn Andersson]

On Fri, Oct 11, 2024 at 05:47:57PM +0530, Mukesh Kumar Savaliya wrote:

Your recipients list is broken. Please check go/upstream and adopt b4 to
help avoid such mistakes.

> pm_runtime_get_sync() function fails during PM early resume and returning
> -EACCES because runtime PM for the device is disabled at the early stage
> causing i2c transfer to fail. Make changes to serve transfer with forced
> resume.
> 
> Few i2c clients like PCI OR touch may request i2c transfers during early
> resume stage. In order to serve transfer request do :
> 

This problem description is too generic. I am not aware of any use case
upstream where PCI or touch might need to perform i2c transfers during
early resume; your commit message should educate me.

> 1. Register interrupt with IRQF_EARLY_RESUME and IRQF_NO_SUSPEND flags
>    to avoid timeout of transfer when IRQ is not enabled during early stage.
> 2. Do force resume if pm_runtime_get_sync() is failing after system
>    suspend when runtime PM is not enabled.
> 3. Increment power usage count after forced resume to balance
>    it against regular runtime suspend.
> 

Please avoid the bullet list form technical description of your patch.

> Co-developed-by: Viken Dadhaniya <quic_vdadhani@xxxxxxxxxxx>
> Signed-off-by: Viken Dadhaniya <quic_vdadhani@xxxxxxxxxxx>
> Signed-off-by: Mukesh Kumar Savaliya <quic_msavaliy@xxxxxxxxxxx>
> ---
> v1 -> v2:
> 
> - Changed gi2c->se.dev to dev during dev_dbg() calls.
> - Addressed review comments from Andi and Bjorn.

That's nice, but spell out the changes you're doing so that reviewers
now what you did.

> - Returned 0 instead garbage inside geni_i2c_force_resume().
> - Added comments explaining forced resume transfer when runtime PM
>   remains disabled.
> 
> V1 link: https://patches.linaro.org/project/linux-i2c/patch/20240328123743.1713696-1-quic_msavaliy@xxxxxxxxxxx/
> ---
> ---
>  drivers/i2c/busses/i2c-qcom-geni.c | 61 +++++++++++++++++++++++++-----
>  1 file changed, 51 insertions(+), 10 deletions(-)
> 
> diff --git a/drivers/i2c/busses/i2c-qcom-geni.c b/drivers/i2c/busses/i2c-qcom-geni.c
> index 212336f724a6..e1207f1a3de3 100644
> --- a/drivers/i2c/busses/i2c-qcom-geni.c
> +++ b/drivers/i2c/busses/i2c-qcom-geni.c
> @@ -134,6 +134,8 @@ struct geni_i2c_clk_fld {
>  	u8	t_cycle_cnt;
>  };
>  
> +static int geni_i2c_runtime_resume(struct device *dev);
> +
>  /*
>   * Hardware uses the underlying formula to calculate time periods of
>   * SCL clock cycle. Firmware uses some additional cycles excluded from the
> @@ -675,22 +677,49 @@ static int geni_i2c_fifo_xfer(struct geni_i2c_dev *gi2c,
>  	return num;
>  }
>  
> +static int geni_i2c_force_resume(struct geni_i2c_dev *gi2c)
> +{
> +	struct device *dev = gi2c->se.dev;
> +	int ret;
> +
> +	ret = geni_i2c_runtime_resume(dev);

Wouldn't pm_runtime_force_resume() help you do what you're looking for?

> +	if (ret) {
> +		dev_err(gi2c->se.dev, "Failed to enable SE resources: %d\n", ret);
> +		pm_runtime_put_noidle(dev);
> +		pm_runtime_set_suspended(dev);
> +		return ret;
> +	}
> +	pm_runtime_get_noresume(dev);
> +	pm_runtime_set_active(dev);
> +	return 0;
> +}
> +
>  static int geni_i2c_xfer(struct i2c_adapter *adap,
>  			 struct i2c_msg msgs[],
>  			 int num)
>  {
>  	struct geni_i2c_dev *gi2c = i2c_get_adapdata(adap);
> +	struct device *dev = gi2c->se.dev;
>  	int ret;
>  
>  	gi2c->err = 0;
>  	reinit_completion(&gi2c->done);
> -	ret = pm_runtime_get_sync(gi2c->se.dev);
> -	if (ret < 0) {
> -		dev_err(gi2c->se.dev, "error turning SE resources:%d\n", ret);
> -		pm_runtime_put_noidle(gi2c->se.dev);
> -		/* Set device in suspended since resume failed */
> -		pm_runtime_set_suspended(gi2c->se.dev);
> -		return ret;
> +
> +	/* Serve I2C transfer by forced resume whether Runtime PM is enbled or not */
> +	if (!pm_runtime_enabled(dev) && gi2c->suspended) {
> +		dev_dbg(dev, "Runtime PM is disabled hence force resume, pm_usage_count: %d\n",
> +			atomic_read(&dev->power.usage_count));
> +		ret = geni_i2c_force_resume(gi2c);
> +		if (ret)
> +			return ret;
> +	} else {
> +		ret = pm_runtime_get_sync(dev);
> +		if (ret == -EACCES && gi2c->suspended) {
> +			dev_dbg(dev, "pm_runtime_get_sync() failed-%d, force resume\n", ret);
> +			ret = geni_i2c_force_resume(gi2c);
> +			if (ret)
> +				return ret;
> +		}
>  	}
>  
>  	qcom_geni_i2c_conf(gi2c);
> @@ -700,8 +729,19 @@ static int geni_i2c_xfer(struct i2c_adapter *adap,
>  	else
>  		ret = geni_i2c_fifo_xfer(gi2c, msgs, num);
>  
> -	pm_runtime_mark_last_busy(gi2c->se.dev);
> -	pm_runtime_put_autosuspend(gi2c->se.dev);
> +	/* Does Opposite to Forced Resume when runtime PM was not enabled and served
> +	 * Transfer via forced resume.

Please polish this comment.

Regards,
Bjorn

> +	 */
> +	if (!pm_runtime_enabled(dev) && !gi2c->suspended) {
> +		pm_runtime_put_noidle(dev);
> +		pm_runtime_set_suspended(dev);
> +		/* Reset flag same as runtime suspend, next xfer PM can be enabled */
> +		gi2c->suspended = 0;
> +	} else {
> +		pm_runtime_mark_last_busy(gi2c->se.dev);
> +		pm_runtime_put_autosuspend(gi2c->se.dev);
> +	}
> +
>  	gi2c->cur = NULL;
>  	gi2c->err = 0;
>  	return ret;
> @@ -818,7 +858,8 @@ static int geni_i2c_probe(struct platform_device *pdev)
>  	init_completion(&gi2c->done);
>  	spin_lock_init(&gi2c->lock);
>  	platform_set_drvdata(pdev, gi2c);
> -	ret = devm_request_irq(dev, gi2c->irq, geni_i2c_irq, IRQF_NO_AUTOEN,
> +	ret = devm_request_irq(dev, gi2c->irq, geni_i2c_irq,
> +			       IRQF_NO_AUTOEN | IRQF_EARLY_RESUME | IRQF_NO_SUSPEND,
>  			       dev_name(dev), gi2c);
>  	if (ret) {
>  		dev_err(dev, "Request_irq failed:%d: err:%d\n",
> -- 
> 2.25.1
>