Nouveau - Oct 2016 - [PATCH] drm/nouveau/acpi: fix check for power resources support

Check whether the kernel really supports power resources for a device,
otherwise the power might not be removed when the device is runtime
suspended (DSM should still work in these cases where PR does not).

Link: https://bugs.freedesktop.org/show_bug.cgi?id=98398
Fixes: 692a17dcc292 ("drm/nouveau/acpi: fix lockup with PCIe runtime
PM")
Signed-off-by: Peter Wu <peter at lekensteyn.nl>
---
Hi,

Maybe Cc: stable (4.8+)?

Compile-tested only. Rick, can you test if this patch fixes the problem for you?

This check was actually in the original patch, but it was changed:
https://lists.freedesktop.org/archives/nouveau/2016-May/025125.html
Re-adding the check as suggested by Mika.

Kind regards,
Peter
---
 drivers/gpu/drm/nouveau/nouveau_acpi.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/nouveau/nouveau_acpi.c
b/drivers/gpu/drm/nouveau/nouveau_acpi.c
index dc57b62..193573d 100644
--- a/drivers/gpu/drm/nouveau/nouveau_acpi.c
+++ b/drivers/gpu/drm/nouveau/nouveau_acpi.c
@@ -240,7 +240,8 @@ static bool nouveau_pr3_present(struct pci_dev *pdev)
 	if (!parent_adev)
 		return false;
 
-	return acpi_has_method(parent_adev->handle, "_PR3");
+	return parent_adev->power.flags.power_resources &&
+		acpi_has_method(parent_adev->handle, "_PR3");
 }
 
 static void nouveau_dsm_pci_probe(struct pci_dev *pdev, acpi_handle
*dhandle_out,
-- 
2.10.1

Op za 29 okt. 2016 om 01:40 schreef Peter Wu <peter at lekensteyn.nl>:
> Check whether the kernel really supports power resources for a device,
> otherwise the power might not be removed when the device is runtime
> suspended (DSM should still work in these cases where PR does not).
>
> Link: https://bugs.freedesktop.org/show_bug.cgi?id=98398
> Fixes: 692a17dcc292 ("drm/nouveau/acpi: fix lockup with PCIe runtime
PM")
> Signed-off-by: Peter Wu <peter at lekensteyn.nl>
> ---
> Hi,
>
> Maybe Cc: stable (4.8+)?
>
> Compile-tested only. Rick, can you test if this patch fixes the problem
> for you?
>
> This check was actually in the original patch, but it was changed:
> https://lists.freedesktop.org/archives/nouveau/2016-May/025125.html
> Re-adding the check as suggested by Mika.
>
> Kind regards,
> Peter
> ---
>  drivers/gpu/drm/nouveau/nouveau_acpi.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/nouveau/nouveau_acpi.c
> b/drivers/gpu/drm/nouveau/nouveau_acpi.c
> index dc57b62..193573d 100644
> --- a/drivers/gpu/drm/nouveau/nouveau_acpi.c
> +++ b/drivers/gpu/drm/nouveau/nouveau_acpi.c
> @@ -240,7 +240,8 @@ static bool nouveau_pr3_present(struct pci_dev *pdev)
>         if (!parent_adev)
>                 return false;
>
> -       return acpi_has_method(parent_adev->handle, "_PR3");
> +       return parent_adev->power.flags.power_resources &&
> +               acpi_has_method(parent_adev->handle, "_PR3");
>  }
>
>  static void nouveau_dsm_pci_probe(struct pci_dev *pdev, acpi_handle
> *dhandle_out,
> --
> 2.10.1
>
> Hi Peter,
This patch indeed does fix the issue for me, thank you.

Kind regards,
Rick

Tested-by: Rick Kerkhof <rick.2889 at gmail.com>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20161029/577c7cfb/attachment-0001.html>

On Sat, Oct 29, 2016 at 01:40:03AM +0200, Peter Wu
wrote:
> Check whether the kernel really supports power resources for a device,
> otherwise the power might not be removed when the device is runtime
> suspended (DSM should still work in these cases where PR does not).
> 
> Link: https://bugs.freedesktop.org/show_bug.cgi?id=98398
> Fixes: 692a17dcc292 ("drm/nouveau/acpi: fix lockup with PCIe runtime
PM")
> Signed-off-by: Peter Wu <peter at lekensteyn.nl>
> ---
> Hi,
> 
> Maybe Cc: stable (4.8+)?
> 
> Compile-tested only. Rick, can you test if this patch fixes the problem for
you?
> 
> This check was actually in the original patch, but it was changed:
> https://lists.freedesktop.org/archives/nouveau/2016-May/025125.html
> Re-adding the check as suggested by Mika.
> 
> Kind regards,
> Peter
> ---
>  drivers/gpu/drm/nouveau/nouveau_acpi.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/nouveau/nouveau_acpi.c
b/drivers/gpu/drm/nouveau/nouveau_acpi.c
> index dc57b62..193573d 100644
> --- a/drivers/gpu/drm/nouveau/nouveau_acpi.c
> +++ b/drivers/gpu/drm/nouveau/nouveau_acpi.c
> @@ -240,7 +240,8 @@ static bool nouveau_pr3_present(struct pci_dev *pdev)
>  	if (!parent_adev)
>  		return false;
>  
> -	return acpi_has_method(parent_adev->handle, "_PR3");
You may want to add comment here telling why we need to check
flags.power_resources. Or alternatively explain this in the changelog
(along the lines that we fail to find PowerResources because ACPICA does
not handle certain conditional things in the same way than Windows 10).

Other than that, looks good

Reviewed-by: Mika Westerberg <mika.westerberg at linux.intel.com>
> +	return parent_adev->power.flags.power_resources &&
> +		acpi_has_method(parent_adev->handle, "_PR3");
>  }
>  
>  static void nouveau_dsm_pci_probe(struct pci_dev *pdev, acpi_handle
*dhandle_out,
> -- 
> 2.10.1