[PATCH v2 5/5] drm/amdgpu: do not resume device in thaw for normal hibernation

[Mario Limonciello]

On 7/4/2025 6:12 AM, Samuel Zhang wrote:
> For normal hibernation, GPU do not need to be resumed in thaw since it
> is not involved in writing the hibernation image. Skip resume in this
> case can reduce the hibernation time.

Since you have the measurements would you mind including them in the 
commit message for reference?

> 
> For cancelled hibernation, GPU need to be resumed.

If I'm following right you are actually handling two different things in 
this patch aren't you?

1) A change in thaw() to only resume on aborted hibernation
2) A change in shutdown() to skip running if the in s4 when shutdown() 
is called.

So I think it would be more logical to split this into two patches.

> 
> Signed-off-by: Samuel Zhang <guoqing.zhang at amd.com>
> ---
>   drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 8 ++++++++
>   1 file changed, 8 insertions(+)
> 
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
> index 4f8632737574..e064816aae4d 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
> @@ -2541,6 +2541,10 @@ amdgpu_pci_shutdown(struct pci_dev *pdev)
>   	if (amdgpu_ras_intr_triggered())
>   		return;
>   
> +	/* device maybe not resumed here, return immediately in this case */
> +	if (adev->in_s4 && adev->in_suspend)
> +		return;
> +
>   	/* if we are running in a VM, make sure the device
>   	 * torn down properly on reboot/shutdown.
>   	 * unfortunately we can't detect certain
> @@ -2655,6 +2659,10 @@ static int amdgpu_pmops_thaw(struct device *dev)
>   {
>   	struct drm_device *drm_dev = dev_get_drvdata(dev);
>   
> +	/* do not resume device for normal hibernation */
> +	if (pm_transition.event == PM_EVENT_THAW)
> +		return 0;
> +

Without digging into pm.h documentation I think it's not going to be 
very obvious next time we look at this code that amdgpu_device_resume() 
is only intended for the aborted case.

How would you feel about a switch/case?

Something like this:

switch (pm_transition.event) {
/* normal hibernation */
case PM_EVENT_THAW:
	return 0;
/* for aborted hibernation */
case PM_EVENT_RECOVER:
	return amdgpu_device_resume(drm_dev, true);
default:
	return -EOPNOTSUP;
}


>   	return amdgpu_device_resume(drm_dev, true);
>   }
>