Re: [PATCHv2] drm/amdgpu: disable ASPM on Intel AlderLake based systems

From: "Gong, Richard" <richard.gong@amd.com>
To: Paul Menzel <pmenzel@molgen.mpg.de>
Cc: mario.limonciello@amd.com, dri-devel@lists.freedesktop.org,
	amd-gfx@lists.freedesktop.org, linux-kernel@vger.kernel.org,
	alexander.deucher@amd.com, christian.koenig@amd.com,
	xinhui.pan@amd.com, airlied@linux.ie, daniel@ffwll.ch
Subject: Re: [PATCHv2] drm/amdgpu: disable ASPM on Intel AlderLake based systems
Date: Sun, 10 Apr 2022 19:37:34 -0500	[thread overview]
Message-ID: <cacb177f-20e2-b50a-806f-777837fba693@amd.com> (raw)
In-Reply-To: <44354d78-b340-fbc4-fd6c-060d7ad3404e@molgen.mpg.de>


On 4/8/2022 7:19 PM, Paul Menzel wrote:
> Dear Richard,
>
>
> Thank you for your patch.
>
> Am 08.04.22 um 21:05 schrieb Richard Gong:
>> Active State Power Management (ASPM) feature is enabled since kernel 
>> 5.14.
>> There are some AMD GFX cards (such as WX3200 and RX640) that cannot be
>> used with Intel AlderLake based systems to enable ASPM. Using these GFX
>
> Alder Lake
Actually there are 2 formats (one with space, another is w/o space) in 
the upstream sources, so I will keep that unchanged and use the format 
w/o space.
>
>> cards as video/display output, Intel Alder Lake based systems will hang
>> during suspend/resume.
>
> Please reflow for 75 characters per line.
>
> Also please mention the exact system you had problems with (also 
> firmware versions).
>
>>
>> Add extra check to disable ASPM on Intel AlderLake based systems.
>
> Is that a problem with Intel Alder Lake or the Dell system? Shouldn’t 
> ASPM just be disabled for the problematic cards for the Dell system. 
> You write newer cards worked fine.
>
>> Fixes: 0064b0ce85bb ("drm/amd/pm: enable ASPM by default")
>> Link: 
>> https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgitlab.freedesktop.org%2Fdrm%2Famd%2F-%2Fissues%2F1885&amp;data=04%7C01%7Crichard.gong%40amd.com%7C6b94ff2249244c04974e08da19bea71b%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637850604066094079%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=%2FiF%2Bnwzm5RlClT%2Fv%2B0RJvmVwsc%2FiwV3jCiFq7PB84wM%3D&amp;reserved=0
>> Signed-off-by: Richard Gong <richard.gong@amd.com>
>> ---
>> v2: correct commit description
>>      move the check from chip family to problematic platform
>> ---
>>   drivers/gpu/drm/amd/amdgpu/vi.c | 17 ++++++++++++++++-
>>   1 file changed, 16 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c 
>> b/drivers/gpu/drm/amd/amdgpu/vi.c
>> index 039b90cdc3bc..8b4eaf54b23e 100644
>> --- a/drivers/gpu/drm/amd/amdgpu/vi.c
>> +++ b/drivers/gpu/drm/amd/amdgpu/vi.c
>> @@ -81,6 +81,10 @@
>>   #include "mxgpu_vi.h"
>>   #include "amdgpu_dm.h"
>>   +#if IS_ENABLED(CONFIG_X86_64)
>> +#include <asm/intel-family.h>
>> +#endif
>> +
>>   #define ixPCIE_LC_L1_PM_SUBSTATE    0x100100C6
>>   #define PCIE_LC_L1_PM_SUBSTATE__LC_L1_SUBSTATES_OVERRIDE_EN_MASK 
>> 0x00000001L
>>   #define PCIE_LC_L1_PM_SUBSTATE__LC_PCI_PM_L1_2_OVERRIDE_MASK 
>> 0x00000002L
>> @@ -1134,13 +1138,24 @@ static void vi_enable_aspm(struct 
>> amdgpu_device *adev)
>>           WREG32_PCIE(ixPCIE_LC_CNTL, data);
>>   }
>>   +static bool intel_core_apsm_chk(void)
>
> aspm
>
>> +{
>> +#if IS_ENABLED(CONFIG_X86_64)
>> +    struct cpuinfo_x86 *c = &cpu_data(0);
>> +
>> +    return (c->x86 == 6 && c->x86_model == INTEL_FAM6_ALDERLAKE);
>> +#else
>> +    return false;
>> +#endif
>
> Please do the check in C code and not the preprocessor.
>
>> +}
>> +
>>   static void vi_program_aspm(struct amdgpu_device *adev)
>>   {
>>       u32 data, data1, orig;
>>       bool bL1SS = false;
>>       bool bClkReqSupport = true;
>>   -    if (!amdgpu_device_should_use_aspm(adev))
>> +    if (!amdgpu_device_should_use_aspm(adev) || intel_core_apsm_chk())
>>           return;
>>         if (adev->flags & AMD_IS_APU ||
>
>
> Kind regards,
>
> Paul