[PATCH] drm/amd: Check if ASPM is enabled from PCIe subsystem

[Lazar, Lijo]

On 2/17/2022 10:28 AM, Limonciello, Mario wrote:
> [AMD Official Use Only]
> 
>>> commit 0064b0ce85bb ("drm/amd/pm: enable ASPM by default") enabled
>> ASPM
>>> by default but a variety of hardware configurations it turns out that this
>>> caused a regression.
>>>
>>> * PPC64LE hardware does not support ASPM at a hardware level.
>>>     CONFIG_PCIEASPM is often disabled on these architectures.
>>> * Some dGPUs on ALD platforms don't work with ASPM enabled and PCIe
>>>     subsystem disables it
>>>
>>> Check with the PCIe subsystem to see that ASPM has been enabled
>>> or not.
>>>
>>> Fixes: 0064b0ce85bb ("drm/amd/pm: enable ASPM by default")
>>> Link:
>> https://wiki.raptorcs.com/w/images/a/ad/P9_PHB_version1.0_27July2018_p
>> ub.pdf
>>> Link: https://gitlab.freedesktop.org/drm/amd/-/issues/1723
>>> Link: https://gitlab.freedesktop.org/drm/amd/-/issues/1739
>>> Link: https://gitlab.freedesktop.org/drm/amd/-/issues/1885
>>> Link: https://gitlab.freedesktop.org/drm/amd/-/issues/1907
>>> Tested-by: koba.ko at canonical.com
>>> Signed-off-by: Mario Limonciello <mario.limonciello at amd.com>
>>> ---
>>>    drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 3 +++
>>>    1 file changed, 3 insertions(+)
>>>
>>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
>> b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
>>> index 5acd0473ed01..d2548fab51c5 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
>>> @@ -2008,6 +2008,9 @@ static int amdgpu_pci_probe(struct pci_dev
>> *pdev,
>>>    			return -ENODEV;
>>>    	}
>>>
>>> +	if (amdgpu_aspm == -1 && !pcie_aspm_enabled(pdev))
>>> +		amdgpu_aspm = 0;
>>> +
>>
>> Doesn't this disable aspm for all devices even though only one of the
>> bridges doesn't have support?
> 
> At least with how it's a global parameter today I don't think we can do
> much better without a refactor.
> 
> As it's actively breaking people I think I would prefer to get this in as it
> is now (which can go to stable trees too), and then do the refactor as a
> follow up.
> 

There is enable_aspm/program_aspm calls and the check could be in one of 
those places. I agree that for now this is the easier path.

Thanks,
Lijo

>>
>> Thanks,
>> Lijo
>>
>>>    	if (amdgpu_virtual_display ||
>>>    	    amdgpu_device_asic_has_dc_support(flags & AMD_ASIC_MASK))
>>>    		supports_atomic = true;
>>>