Merge "[Lut] Explain flatten policy for 3d lut in NDK" into main

 * \a ADisplayLutsEntry_create.

 *

 * 1D Lut(s) are treated as gain curves.

 *

 * 3D Lut(s) are used for direct color manipulations.

 * The values of 3D Lut(s) data should be normalized to the range 0.0f

 * to 1.0f, inclusive. And 1.0f is the maximum panel luminance.

 * And 3D Lut(s) data is organized in RGB order

 * (R0, R1, R2, ..., RN, G0, G1, G2, ..., GN, B0, B1, B2, ..., BN) if N is the dimension.

 * For 3D Lut(s), the values should be normalized to the range 0.0

 * to 1.0 , inclusive. And 1.0 is the maximum panel luminance.

 * And If N is the size of each dimension, the data is arranged in RGB order:

 * R(0, 0, 0), R(0, 0, 1), ..., R(0, 0, N - 1),

 * R(0, 1, 0), ..., R(0, 1, N - 1), ..., R(0, N - 1, N - 1),

 * R(1, 0, 0), ..., R(1, 0, N - 1), ..., R(1, N - 1, N - 1), ..., R(N - 1, N - 1, N - 1),

 * G(0, 0, 0), ..., G(N - 1, N - 1, N - 1),

 * B(0, 0, 0), ..., B(N - 1, N - 1, N - 1).

 *

 * When a GPU shader samples 3D Lut data, it's accessed in a flat, one-dimensional arrangement.

 * Assuming that we have a 3D array ORIGINAL[N][N][N],

 * then ORIGINAL[x][y][z] is mapped to FLAT[z + N * (y + N * x)].

 *

 * @param buffer The lut raw buffer. The function creates a copy of it and does not need to

 * outlive the life of the ADisplayLutsEntry.