Re: [PATCH 2/2] drm/xe/guc: Update to GuC firmware 70.19.1

[John Harrison <john.c.harrison@intel.com>]

On 1/23/2024 08:41, Lucas De Marchi wrote:
> On Mon, Jan 22, 2024 at 02:14:56PM -0800, Welty, Brian wrote:
>> On 1/22/2024 1:37 PM, John.C.Harrison@Intel.com wrote:
>>> From: John Harrison <John.C.Harrison@Intel.com>
>>>
>>> UAPI compatibility version: 1.8.1
>
> what exactly is this? We don't seem to track it anywhere in the driver.
It is XE_UC_FW_VER_COMPATIBILITY. This is the version number that should 
be looked at for tracking what features are or are not available. The 
raw file version (e.g. 70.19.1) is XE_UC_FW_VER_RELEASE but that is not 
accessible from a VF and cannot be exported to UMDs because again, it 
would break in an SRIOV type scenario. The file version should only be 
used for feature checks when actually loading the GuC itself (which can 
only be done by native/PF), e.g. what to put in the ADS, what load time 
w/a's should be enabled, etc.

>
>>
>> I've wondered for awhile why was lagging behind here....
>>
>> Reviewed-by: Brian Welty <brian.welty@intel.com>
>>
>>>
>>> Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
>>> ---
>>>  drivers/gpu/drm/xe/xe_uc_fw.c | 16 ++++++++--------
>>>  1 file changed, 8 insertions(+), 8 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/xe/xe_uc_fw.c 
>>> b/drivers/gpu/drm/xe/xe_uc_fw.c
>>> index d3f0fe2101a7..8da6cbb8c9e2 100644
>>> --- a/drivers/gpu/drm/xe/xe_uc_fw.c
>>> +++ b/drivers/gpu/drm/xe/xe_uc_fw.c
>>> @@ -103,14 +103,14 @@ struct fw_blobs_by_type {
>>>  };
>>>  #define XE_GUC_FIRMWARE_DEFS(fw_def, mmp_ver, major_ver)            \
>>> -    fw_def(METEORLAKE,    major_ver(i915,    guc,    mtl, 70, 7, 
>>> 0))    \
>>> -    fw_def(DG2,        major_ver(i915,    guc,    dg2,    70, 5, 
>>> 0))    \
>>> -    fw_def(DG1,        major_ver(i915,    guc,    dg1,    70, 5, 
>>> 0))    \
>>> -    fw_def(ALDERLAKE_N,    major_ver(i915,    guc,    tgl, 70, 5, 
>>> 0))    \
>>> -    fw_def(ALDERLAKE_P,    major_ver(i915,    guc, adlp,    70, 5, 
>>> 0))    \
>>> -    fw_def(ALDERLAKE_S,    major_ver(i915,    guc,    tgl, 70, 5, 
>>> 0))    \
>>> -    fw_def(ROCKETLAKE,    major_ver(i915,    guc,    tgl, 70, 5, 
>>> 0))    \
>>> -    fw_def(TIGERLAKE,    major_ver(i915,    guc,    tgl, 70, 5, 0))
>>> +    fw_def(METEORLAKE,    major_ver(i915,    guc,    mtl, 70, 19, 
>>> 1))    \
>>> +    fw_def(DG2,        major_ver(i915,    guc,    dg2,    70, 19, 
>>> 1))    \
>>> +    fw_def(DG1,        major_ver(i915,    guc,    dg1,    70, 19, 
>>> 1))    \
>>> +    fw_def(ALDERLAKE_N,    major_ver(i915,    guc,    tgl, 70, 19, 
>>> 1))    \
>>> +    fw_def(ALDERLAKE_P,    major_ver(i915,    guc, adlp,    70, 19, 
>>> 1))    \
>>> +    fw_def(ALDERLAKE_S,    major_ver(i915,    guc,    tgl, 70, 19, 
>>> 1))    \
>>> +    fw_def(ROCKETLAKE,    major_ver(i915,    guc,    tgl, 70, 19, 
>>> 1))    \
>>> +    fw_def(TIGERLAKE,    major_ver(i915,    guc,    tgl, 70, 19, 1))
>
> are all of these in linux-firmware repo?
Not in the upstream repo yet. I'm about to start that process.

John.

>
> Lucas De Marchi
>
>>>  #define XE_HUC_FIRMWARE_DEFS(fw_def, mmp_ver, no_ver)        \
>>>      fw_def(METEORLAKE,    no_ver(i915,    huc_gsc, mtl))        \