[PATCH] drm/amdgpu: Add "harvest" to IP discovery sysfs

Christian König ckoenig.leichtzumerken at gmail.com
Tue Feb 15 14:37:02 UTC 2022



Am 15.02.22 um 15:33 schrieb Luben Tuikov:
> On 2022-02-15 09:25, Christian König wrote:
>> Am 15.02.22 um 00:21 schrieb Luben Tuikov:
>>> Add the "harvest" field to the IP attributes in
>>> the IP discovery sysfs visualization, as this
>>> field is present in the binary data.
>>>
>>> Cc: Alex Deucher <Alexander.Deucher at amd.com>
>>> Signed-off-by: Luben Tuikov <luben.tuikov at amd.com>
>>> ---
>>>    drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c | 8 ++++++++
>>>    1 file changed, 8 insertions(+)
>>>
>>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
>>> index c8dbdb78988ce0..0496d369504641 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
>>> @@ -393,6 +393,7 @@ struct ip_hw_instance {
>>>    	int hw_id;
>>>    	u8  num_instance;
>>>    	u8  major, minor, revision;
>>> +	u8  harvest;
>> Should we maybe use bool here instead?
>>
>> Apart from that looks good to me.
> Thanks Christian.
>
> I don't mind using bool here.
>
> I saw the field in the binary data is 4 bits and represented that.
>
> Is the field actually bool in the IP binary data?

I'm not sure either.

I would have expected it to be a single bit flag in the binary, but 
4bits sounds like it serves some more purpose.

Probably best to stick to u8 for now in this case.

Regards,
Christian.

> I can change the patch and resubmit.
>
> Regards,
> Luben
>
>> Regards,
>> Christian.
>>
>>>    
>>>    	int num_base_addresses;
>>>    	u32 base_addr[];
>>> @@ -440,6 +441,11 @@ static ssize_t revision_show(struct ip_hw_instance *ip_hw_instance, char *buf)
>>>    	return sysfs_emit(buf, "%d\n", ip_hw_instance->revision);
>>>    }
>>>    
>>> +static ssize_t harvest_show(struct ip_hw_instance *ip_hw_instance, char *buf)
>>> +{
>>> +	return sysfs_emit(buf, "0x%01X\n", ip_hw_instance->harvest);
>>> +}
>>> +
>>>    static ssize_t num_base_addresses_show(struct ip_hw_instance *ip_hw_instance, char *buf)
>>>    {
>>>    	return sysfs_emit(buf, "%d\n", ip_hw_instance->num_base_addresses);
>>> @@ -471,6 +477,7 @@ static struct ip_hw_instance_attr ip_hw_attr[] = {
>>>    	__ATTR_RO(major),
>>>    	__ATTR_RO(minor),
>>>    	__ATTR_RO(revision),
>>> +	__ATTR_RO(harvest),
>>>    	__ATTR_RO(num_base_addresses),
>>>    	__ATTR_RO(base_addr),
>>>    };
>>> @@ -708,6 +715,7 @@ static int amdgpu_discovery_sysfs_ips(struct amdgpu_device *adev,
>>>    			ip_hw_instance->major = ip->major;
>>>    			ip_hw_instance->minor = ip->minor;
>>>    			ip_hw_instance->revision = ip->revision;
>>> +			ip_hw_instance->harvest = ip->harvest;
>>>    			ip_hw_instance->num_base_addresses = ip->num_base_address;
>>>    
>>>    			for (kk = 0; kk < ip_hw_instance->num_base_addresses; kk++)
>>>
>>> base-commit: d8604f1d237a145db48bae4ea60b85a5875df307
> Regards,



More information about the amd-gfx mailing list