Re: [V7 14/45] drm/vkms: Add enumerated 1D curve colorop

[Harry Wentland]

On 2025-02-25 06:18, Louis Chauvet wrote:
> Le 20/12/2024 à 05:33, Alex Hung a écrit :
>> From: Harry Wentland <harry.wentl...@amd.com>
>>

(snip)

>> +    { 0xfbfb, 0xfbfb, 0xfbfb, 0 },
>> +    { 0xfcfc, 0xfcfc, 0xfcfc, 0 },
>> +    { 0xfdfd, 0xfdfd, 0xfdfd, 0 },
>> +    { 0xfefe, 0xfefe, 0xfefe, 0 },
>> +    { 0xffff, 0xffff, 0xffff, 0 },
>> +};
>> +
>> +const struct vkms_color_lut linear_eotf = {
>> +    .base = linear_array,
>> +    .lut_length = LUT_SIZE,
>> +    .channel_value2index_ratio = 0xff00ffll
>> +};
>> +
> 
> Can you add a comment to describe the values here?
> Something that could help if we suspect an error in the values of the table, 
> for example:
> 
> This LUT table contains values from <spec/iso/whatever>, mapped linearly from 
> 0.0..1.0 to 0x0..0xFFFF.
> 

The tables were generated with a simple LUT generator
that's based on skia's TF code. You can find it here:
https://gitlab.freedesktop.org/hwentland/lutgen

I'll add a comment at the top of vkms_luts.c to that
regard.

Harry