RE: [PATCH i-g-t v2 2/2] HAX: kms_cursor_legacy: Use IGT_FORMAT_MOD_PREFERRED for cursor FBs

[Jani Nikula <jani.nikula@intel.com>]

On Wed, 13 May 2026, "B, Jeevan" <jeevan.b@intel.com> wrote:
>> -----Original Message-----
>> From: Nikula, Jani <jani.nikula@intel.com>
>> Sent: Wednesday, May 13, 2026 2:06 PM
>> To: B, Jeevan <jeevan.b@intel.com>; igt-dev@lists.freedesktop.org
>> Cc: Manna, Animesh <animesh.manna@intel.com>; Dibin Moolakadan
>> Subrahmanian <dibin.moolakadan.subrahmanian@intel.com>; Thasleem,
>> Mohammed <mohammed.thasleem@intel.com>; Naladala, Ramanaidu
>> <ramanaidu.naladala@intel.com>; B, Jeevan <jeevan.b@intel.com>
>> Subject: Re: [PATCH i-g-t v2 2/2] HAX: kms_cursor_legacy: Use
>> IGT_FORMAT_MOD_PREFERRED for cursor FBs
>> 
>> On Wed, 13 May 2026, Jeevan B <jeevan.b@intel.com> wrote:
>> > Switch all cursor-related framebuffer creations in kms_cursor_legacy
>> > to use IGT_FORMAT_MOD_PREFERRED instead of
>> DRM_FORMAT_MOD_LINEAR.
>>
>> Why?
>
> Well, the plan is to remove this hardcoded modifier usage. Since these modifiers keep changing from platform to platform,
> we currently have to update them manually or ensure that only supported tile formats are used.
> Therefore, we are considering introducing something like IGT_FORMAT_MOD_PREFERRED, 
> which would automatically select the best possible modifier for the given platform.

My point is, that information belongs in the commit message.

If the commit message only explains the "what", the reviewer can only
cross check that the commit message matches the code changes, but still
has no idea why the changes are being made.

BR,
Jani.


-- 
Jani Nikula, Intel