[Intel-gfx] [PATCH] drm/i915/uc: define GuC and HuC FWs for EHL

[Daniele Ceraolo Spurio]

On 8/19/2019 9:25 PM, Summers, Stuart wrote:
> On Mon, 2019-08-19 at 18:23 -0700, Daniele Ceraolo Spurio wrote:
>> First uc firmware release for EHL.
>>
>> Signed-off-by: Daniele Ceraolo Spurio <
>> daniele.ceraolospurio at intel.com>
>> Cc: Matt Roper <matthew.d.roper at intel.com>
>> Cc: Anusha Srivatsa <anusha.srivatsa at intel.com>
>> Cc: Michal Wajdeczko <michal.wajdeczko at intel.com>
>> ---
>>   drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c | 13 +++++++------
>>   1 file changed, 7 insertions(+), 6 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
>> b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
>> index bd22bf11adad..296a82603be0 100644
>> --- a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
>> +++ b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
>> @@ -39,12 +39,13 @@ void intel_uc_fw_change_status(struct intel_uc_fw
>> *uc_fw,
>>    * Must be ordered based on platform + revid, from newer to older.
>>    */
>>   #define INTEL_UC_FIRMWARE_DEFS(fw_def, guc_def, huc_def) \
>> -	fw_def(ICELAKE,    0, guc_def(icl, 33, 0, 0),
>> huc_def(icl,  8,  4, 3238)) \
>> -	fw_def(COFFEELAKE, 0, guc_def(kbl, 33, 0, 0), huc_def(kbl, 02,
>> 00, 1810)) \
>> -	fw_def(GEMINILAKE, 0, guc_def(glk, 33, 0, 0), huc_def(glk, 03,
>> 01, 2893)) \
>> -	fw_def(KABYLAKE,   0, guc_def(kbl, 33, 0, 0), huc_def(kbl, 02,
>> 00, 1810)) \
>> -	fw_def(BROXTON,    0, guc_def(bxt, 33, 0, 0), huc_def(bxt,
>> 01,  8, 2893)) \
>> -	fw_def(SKYLAKE,    0, guc_def(skl, 33, 0, 0), huc_def(skl, 01,
>> 07, 1398))
>> +	fw_def(ELKHARTLAKE, 0, guc_def(ehl, 33, 0, 4),
>> huc_def(ehl,  9,  0,    0)) \
> Is there a reason you are bumping straight to 33.0.4 for EHL rather
> than sticking with the existing firmware version? Or worded
> differently, why don't we bump everything to 33.0.4 if we're adding EHL
> here to stay in sync between the platforms?

33.0.4 is the first release to include an EHL build, so I didn't have 
the choice to stick with 33.0.0 for it, otherwise I would have. As for 
why I didn't update all the other blobs, it was because AFAICS from the 
release notes there are no changes that we need at the moment, mostly 
because the only thing we do with GuC is authenticating HuC and that 
flow is pretty static. All the 33.0.* releases are compatible at the 
interface level so I opted to avoid pushing several more binaries just 
to keep the numbers the same with no real benefit.

As a general point, I think we should expect that the patch number might 
vary across platforms as we get platform-specific features/fixes, but 
major and minor, which indicate the interface version, will be in sync.

Daniele

> Thanks,
> Stuart
>
>> +	fw_def(ICELAKE,     0, guc_def(icl, 33, 0, 0),
>> huc_def(icl,  8,  4, 3238)) \
>> +	fw_def(COFFEELAKE,  0, guc_def(kbl, 33, 0, 0), huc_def(kbl, 02,
>> 00, 1810)) \
>> +	fw_def(GEMINILAKE,  0, guc_def(glk, 33, 0, 0), huc_def(glk, 03,
>> 01, 2893)) \
>> +	fw_def(KABYLAKE,    0, guc_def(kbl, 33, 0, 0), huc_def(kbl, 02,
>> 00, 1810)) \
>> +	fw_def(BROXTON,     0, guc_def(bxt, 33, 0, 0), huc_def(bxt,
>> 01,  8, 2893)) \
>> +	fw_def(SKYLAKE,     0, guc_def(skl, 33, 0, 0), huc_def(skl, 01,
>> 07, 1398))
>>   
>>   #define __MAKE_UC_FW_PATH(prefix_, name_, separator_, major_,
>> minor_, patch_) \
>>   	"i915/" \