[Intel-gfx] [PATCH v2 1/6] drm/i915: store all subslice masks

Tvrtko Ursulin tvrtko.ursulin at linux.intel.com
Fri Jan 12 11:05:14 UTC 2018


On 12/01/2018 10:58, Lionel Landwerlin wrote:
> On 12/01/18 10:15, Tvrtko Ursulin wrote:

[snip]

>>
>>> --- a/drivers/gpu/drm/i915/intel_device_info.c
>>> +++ b/drivers/gpu/drm/i915/intel_device_info.c
>>> @@ -80,12 +80,17 @@ void intel_device_info_dump_flags(const struct 
>>> intel_device_info *info,
>>>     static void sseu_dump(const struct sseu_dev_info *sseu, struct 
>>> drm_printer *p)
>>>   {
>>> +    int s;
>>> +
>>>       drm_printf(p, "slice mask: %04x\n", sseu->slice_mask);
>>>       drm_printf(p, "slice total: %u\n", hweight8(sseu->slice_mask));
>>>       drm_printf(p, "subslice total: %u\n", sseu_subslice_total(sseu));
>>> -    drm_printf(p, "subslice mask %04x\n", sseu->subslice_mask);
>>> -    drm_printf(p, "subslice per slice: %u\n",
>>> -           hweight8(sseu->subslice_mask));
>>> +    for (s = 0; s < ARRAY_SIZE(sseu->subslice_mask); s++) {
>>> +        drm_printf(p, "slice%d subslice mask %04x\n",
>>> +               s, sseu->subslice_mask[s]);
>>> +        drm_printf(p, "slice%d subslice per slice: %u\n",
>>> +               s, hweight8(sseu->subslice_mask[s]));
>>
>> Cosmetic only but consider condensing this into one line if you don't 
>> have a preference to either.
> 
> Sure, just conscious about the 80characters.

I wasn't 100% clear here - I meant the kernel log messages, not the 
source code. I think there is no point in logging two lines per slice 
where one only contains a mask, and second a count.

Regards,

Tvrtko


More information about the Intel-gfx mailing list