Re: [PATCH 5/8] mhi: pci_generic: Add suspend/resume/recovery procedure

[Hemant Kumar <hemantk@codeaurora.org>]

Hi Loic,

On 11/13/20 6:59 AM, Loic Poulain wrote:
> Add support for system wide suspend/resume. During suspend, MHI
> device controller must be put in M3 state and PCI bus in D3 state.
> 
> Add a recovery procedure allowing to reinitialize the device in case
> of error during resume steps, which can happen if device loses power
> (and so its context) while system suspend.
> 
> Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
> ---
[..]
>   
> +static void mhi_pci_recovery_work(struct work_struct *work)
> +{
> +	struct mhi_pci_device *mhi_pdev = container_of(work, struct mhi_pci_device,
> +						       recovery_work);
> +	struct mhi_controller *mhi_cntrl = &mhi_pdev->mhi_cntrl;
> +	struct pci_dev *pdev = to_pci_dev(mhi_cntrl->cntrl_dev);
> +	int err;
> +
> +	dev_warn(&pdev->dev, "device recovery started\n");
> +
> +	/* Clean up MHI state */
> +	if (test_and_clear_bit(MHI_PCI_DEV_STARTED, &mhi_pdev->status)) {
> +		mhi_power_down(mhi_cntrl, false);
> +		mhi_unprepare_after_power_down(mhi_cntrl);
> +	}
> +
> +	/* Check if we can recover without full reset */
> +	pci_set_power_state(pdev, PCI_D0);
> +	pci_load_saved_state(pdev, mhi_pdev->pci_state);
> +	pci_restore_state(pdev);
> +
> +	if (!mhi_pci_is_alive(mhi_cntrl))
> +		goto err_try_reset;
> +
> +	err = mhi_prepare_for_power_up(mhi_cntrl);
> +	if (err)
> +		goto err_try_reset;
> +
> +	err = mhi_sync_power_up(mhi_cntrl);
> +	if (err)
> +		goto err_unprepare;
> +
> +	set_bit(MHI_PCI_DEV_STARTED, &mhi_pdev->status);
> +	return;
> +
> +err_unprepare:
> +	mhi_unprepare_after_power_down(mhi_cntrl);
> +err_try_reset:
> +	if (pci_reset_function(pdev))
> +		dev_err(&pdev->dev, "Recovery failed\n");
> +}
> +
>   static int mhi_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
>   {
>   	const struct mhi_pci_dev_info *info = (struct mhi_pci_dev_info *) id->driver_data;
> @@ -333,6 +377,8 @@ static int mhi_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
>   	if (!mhi_pdev)
>   		return -ENOMEM;
>   
> +	INIT_WORK(&mhi_pdev->recovery_work, mhi_pci_recovery_work);
> +
>   	mhi_cntrl_config = info->config;
>   	mhi_cntrl = &mhi_pdev->mhi_cntrl;
>   	mhi_cntrl->cntrl_dev = &pdev->dev;
> @@ -395,6 +441,8 @@ static void mhi_pci_remove(struct pci_dev *pdev)
>   	struct mhi_pci_device *mhi_pdev = pci_get_drvdata(pdev);
>   	struct mhi_controller *mhi_cntrl = &mhi_pdev->mhi_cntrl;
>   
> +	cancel_work_sync(&mhi_pdev->recovery_work);
> +
>   	if (test_and_clear_bit(MHI_PCI_DEV_STARTED, &mhi_pdev->status)) {
>   		mhi_power_down(mhi_cntrl, true);
>   		mhi_unprepare_after_power_down(mhi_cntrl);
> @@ -463,12 +511,64 @@ static const struct pci_error_handlers mhi_pci_err_handler = {
>   	.reset_done = mhi_pci_reset_done,
>   };
>   
> +int  __maybe_unused mhi_pci_suspend(struct device *dev)
> +{
> +	struct pci_dev *pdev = to_pci_dev(dev);
> +	struct mhi_pci_device *mhi_pdev = dev_get_drvdata(dev);
> +	struct mhi_controller *mhi_cntrl = &mhi_pdev->mhi_cntrl;
> +
is it possible we come here and work still did not get a chance to run ? 
If so we can flush it before going through suspend.
> +	/* Transition to M3 state */
> +	mhi_pm_suspend(mhi_cntrl);
> +
> +	pci_save_state(pdev);
> +	pci_disable_device(pdev);
> +	pci_wake_from_d3(pdev, true);
> +	pci_set_power_state(pdev, PCI_D3hot);
are you planning to park the device in D3hot all the time ?
Is there a plan to move to D3Cold in your use case ?
> +
> +	return 0;
> +}
> +
> +static int __maybe_unused mhi_pci_resume(struct device *dev)
> +{
> +	struct pci_dev *pdev = to_pci_dev(dev);
> +	struct mhi_pci_device *mhi_pdev = dev_get_drvdata(dev);
> +	struct mhi_controller *mhi_cntrl = &mhi_pdev->mhi_cntrl;
> +	int err;
> +
> +	pci_set_power_state(pdev, PCI_D0);
> +	pci_restore_state(pdev);
> +	pci_set_master(pdev);
> +
> +	err = pci_enable_device(pdev);
> +	if (err)
> +		goto err_recovery;
> +
> +	/* Exit M3, transition to M0 state */
> +	err = mhi_pm_resume(mhi_cntrl);
> +	if (err) {
> +		dev_err(&pdev->dev, "failed to resume device: %d\n", err);
> +		goto err_recovery;
> +	}
> +
> +	return 0;
> +
> +err_recovery:
> +	/* The device may have loose power or crashed, try recovering it */
> +	queue_work(system_long_wq, &mhi_pdev->recovery_work);
> +	return 0;
> +}
> +
[..]

Thanks,
Hemant

-- 
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project