yuv 0.8.13

/*
 * Copyright (c) Radzivon Bartoshyk, 10/2024. All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without modification,
 * are permitted provided that the following conditions are met:
 *
 * 1.  Redistributions of source code must retain the above copyright notice, this
 * list of conditions and the following disclaimer.
 *
 * 2.  Redistributions in binary form must reproduce the above copyright notice,
 * this list of conditions and the following disclaimer in the documentation
 * and/or other materials provided with the distribution.
 *
 * 3.  Neither the name of the copyright holder nor the names of its
 * contributors may be used to endorse or promote products derived from
 * this software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
 * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
 * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
 * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

use crate::internals::ProcessedOffset;
use crate::neon::utils::*;
use crate::yuv_support::{
    CbCrInverseTransform, YuvChromaRange, YuvChromaSubsampling, YuvSourceChannels,
};
use std::arch::aarch64::*;

pub(crate) unsafe fn neon_yuv_p16_to_rgba16_alpha_row<
    const DESTINATION_CHANNELS: u8,
    const SAMPLING: u8,
    const ENDIANNESS: u8,
    const BYTES_POSITION: u8,
    const PRECISION: i32,
    const BIT_DEPTH: usize,
>(
    y_ld_ptr: &[u16],
    u_ld_ptr: &[u16],
    v_ld_ptr: &[u16],
    a_ld_ptr: &[u16],
    rgba: &mut [u16],
    width: u32,
    range: &YuvChromaRange,
    transform: &CbCrInverseTransform<i32>,
    start_cx: usize,
    start_ux: usize,
) -> ProcessedOffset {
    let destination_channels: YuvSourceChannels = DESTINATION_CHANNELS.into();
    assert!(
        destination_channels == YuvSourceChannels::Rgba
            || destination_channels == YuvSourceChannels::Bgra,
        "Cannot call YUV p16 to Rgb8 with alpha without real alpha"
    );
    let channels = destination_channels.get_channels_count();
    let chroma_subsampling: YuvChromaSubsampling = SAMPLING.into();

    let y_corr = vdupq_n_u16(range.bias_y as u16);
    let uv_corr = vdupq_n_s16(range.bias_uv as i16);

    let weights_arr: [i16; 8] = [
        transform.y_coef as i16,
        transform.cr_coef as i16,
        transform.cb_coef as i16,
        -transform.g_coeff_1 as i16,
        -transform.g_coeff_2 as i16,
        0,
        0,
        0,
    ];

    let v_weights = vld1q_s16(weights_arr.as_ptr());

    let mut cx = start_cx;
    let mut ux = start_ux;

    let v_max_values = vdupq_n_u16((1 << BIT_DEPTH) - 1);

    while cx + 16 <= width as usize {
        let y_values0: int16x8_t = vreinterpretq_s16_u16(vqsubq_u16(
            vreinterpretq_u16_s16(vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                y_ld_ptr.get_unchecked(cx..).as_ptr(),
            )),
            y_corr,
        ));
        let y_values1: int16x8_t = vreinterpretq_s16_u16(vqsubq_u16(
            vreinterpretq_u16_s16(vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                y_ld_ptr.get_unchecked((cx + 8)..).as_ptr(),
            )),
            y_corr,
        ));

        let u_values0: int16x8_t;
        let v_values0: int16x8_t;
        let u_values1: int16x8_t;
        let v_values1: int16x8_t;

        if chroma_subsampling == YuvChromaSubsampling::Yuv444 {
            let mut u_values_l0 = vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                u_ld_ptr.get_unchecked(ux..).as_ptr(),
            );
            let mut v_values_l0 = vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                v_ld_ptr.get_unchecked(ux..).as_ptr(),
            );
            let mut u_values_l1 = vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                u_ld_ptr.get_unchecked((ux + 8)..).as_ptr(),
            );
            let mut v_values_l1 = vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                v_ld_ptr.get_unchecked((ux + 8)..).as_ptr(),
            );

            u_values_l0 = vsubq_s16(u_values_l0, uv_corr);
            v_values_l0 = vsubq_s16(v_values_l0, uv_corr);
            u_values_l1 = vsubq_s16(u_values_l1, uv_corr);
            v_values_l1 = vsubq_s16(v_values_l1, uv_corr);

            u_values0 = u_values_l0;
            v_values0 = v_values_l0;
            u_values1 = u_values_l1;
            v_values1 = v_values_l1;
        } else {
            let mut u_values_l = vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                u_ld_ptr.get_unchecked(ux..).as_ptr(),
            );
            let mut v_values_l = vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                v_ld_ptr.get_unchecked(ux..).as_ptr(),
            );
            u_values_l = vsubq_s16(u_values_l, uv_corr);
            v_values_l = vsubq_s16(v_values_l, uv_corr);

            let u_high = vzip2q_s16(u_values_l, u_values_l);
            let v_high = vzip2q_s16(v_values_l, v_values_l);

            let u_low = vzip1q_s16(u_values_l, u_values_l);
            let v_low = vzip1q_s16(v_values_l, v_values_l);

            u_values0 = u_low;
            v_values0 = v_low;

            u_values1 = u_high;
            v_values1 = v_high;
        }

        let y_high0 = vmull_high_laneq_s16::<0>(y_values0, v_weights);
        let y_high1 = vmull_high_laneq_s16::<0>(y_values1, v_weights);

        let rh0 = vmlal_high_laneq_s16::<1>(y_high0, v_values0, v_weights);
        let bh0 = vmlal_high_laneq_s16::<2>(y_high0, u_values0, v_weights);
        let gh0 = vmlal_high_laneq_s16::<3>(y_high0, v_values0, v_weights);

        let rh1 = vmlal_high_laneq_s16::<1>(y_high1, v_values1, v_weights);
        let bh1 = vmlal_high_laneq_s16::<2>(y_high1, u_values1, v_weights);
        let gh1 = vmlal_high_laneq_s16::<3>(y_high1, v_values1, v_weights);

        let r_high0 = vqrshrun_n_s32::<PRECISION>(rh0);
        let b_high0 = vqrshrun_n_s32::<PRECISION>(bh0);
        let g_high0 =
            vqrshrun_n_s32::<PRECISION>(vmlal_high_laneq_s16::<4>(gh0, u_values0, v_weights));

        let r_high1 = vqrshrun_n_s32::<PRECISION>(rh1);
        let b_high1 = vqrshrun_n_s32::<PRECISION>(bh1);
        let g_high1 =
            vqrshrun_n_s32::<PRECISION>(vmlal_high_laneq_s16::<4>(gh1, u_values1, v_weights));

        let y_low0 = vmull_laneq_s16::<0>(vget_low_s16(y_values0), v_weights);
        let y_low1 = vmull_laneq_s16::<0>(vget_low_s16(y_values1), v_weights);

        let rl0 = vmlal_laneq_s16::<1>(y_low0, vget_low_s16(v_values0), v_weights);
        let bl0 = vmlal_laneq_s16::<2>(y_low0, vget_low_s16(u_values0), v_weights);
        let gl0 = vmlal_laneq_s16::<3>(y_low0, vget_low_s16(v_values0), v_weights);
        let rl1 = vmlal_laneq_s16::<1>(y_low1, vget_low_s16(v_values1), v_weights);
        let bl1 = vmlal_laneq_s16::<2>(y_low1, vget_low_s16(u_values1), v_weights);
        let gl1 = vmlal_laneq_s16::<3>(y_low1, vget_low_s16(v_values1), v_weights);

        let r_low0 = vqrshrun_n_s32::<PRECISION>(rl0);
        let b_low0 = vqrshrun_n_s32::<PRECISION>(bl0);
        let g_low0 = vqrshrun_n_s32::<PRECISION>(vmlal_laneq_s16::<4>(
            gl0,
            vget_low_s16(u_values0),
            v_weights,
        ));

        let r_low1 = vqrshrun_n_s32::<PRECISION>(rl1);
        let b_low1 = vqrshrun_n_s32::<PRECISION>(bl1);
        let g_low1 = vqrshrun_n_s32::<PRECISION>(vmlal_laneq_s16::<4>(
            gl1,
            vget_low_s16(u_values1),
            v_weights,
        ));

        let r_values0 = vminq_u16(vcombine_u16(r_low0, r_high0), v_max_values);
        let g_values0 = vminq_u16(vcombine_u16(g_low0, g_high0), v_max_values);
        let b_values0 = vminq_u16(vcombine_u16(b_low0, b_high0), v_max_values);

        let r_values1 = vminq_u16(vcombine_u16(r_low1, r_high1), v_max_values);
        let g_values1 = vminq_u16(vcombine_u16(g_low1, g_high1), v_max_values);
        let b_values1 = vminq_u16(vcombine_u16(b_low1, b_high1), v_max_values);

        let a_values0 = vld1q_u16(a_ld_ptr.get_unchecked(cx..).as_ptr());
        let a_values1 = vld1q_u16(a_ld_ptr.get_unchecked((cx + 8)..).as_ptr());

        neon_store_rgb16::<DESTINATION_CHANNELS>(
            rgba.get_unchecked_mut(cx * channels..).as_mut_ptr(),
            r_values0,
            g_values0,
            b_values0,
            a_values0,
        );

        neon_store_rgb16::<DESTINATION_CHANNELS>(
            rgba.get_unchecked_mut((cx + 8) * channels..).as_mut_ptr(),
            r_values1,
            g_values1,
            b_values1,
            a_values1,
        );

        cx += 16;

        match chroma_subsampling {
            YuvChromaSubsampling::Yuv420 | YuvChromaSubsampling::Yuv422 => {
                ux += 8;
            }
            YuvChromaSubsampling::Yuv444 => {
                ux += 16;
            }
        }
    }

    while cx + 8 < width as usize {
        let a_values_l = vld1q_u16(a_ld_ptr.get_unchecked(cx..).as_ptr());

        let y_values: int16x8_t = vreinterpretq_s16_u16(vqsubq_u16(
            vreinterpretq_u16_s16(vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                y_ld_ptr.get_unchecked(cx..).as_ptr(),
            )),
            y_corr,
        ));

        let u_high: int16x4_t;
        let v_high: int16x4_t;
        let u_low: int16x4_t;
        let v_low: int16x4_t;

        if chroma_subsampling == YuvChromaSubsampling::Yuv444 {
            let mut u_values_l = vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                u_ld_ptr.get_unchecked(ux..).as_ptr(),
            );
            let mut v_values_l = vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                v_ld_ptr.get_unchecked(ux..).as_ptr(),
            );

            u_values_l = vsubq_s16(u_values_l, uv_corr);
            v_values_l = vsubq_s16(v_values_l, uv_corr);

            u_high = vget_high_s16(u_values_l);
            u_low = vget_low_s16(u_values_l);
            v_high = vget_high_s16(v_values_l);
            v_low = vget_low_s16(v_values_l);
        } else {
            let mut u_values_l = vld_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                u_ld_ptr.get_unchecked(ux..).as_ptr(),
            );
            let mut v_values_l = vld_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                v_ld_ptr.get_unchecked(ux..).as_ptr(),
            );
            u_values_l = vsub_s16(u_values_l, vget_low_s16(uv_corr));
            v_values_l = vsub_s16(v_values_l, vget_low_s16(uv_corr));

            u_high = vzip2_s16(u_values_l, u_values_l);
            v_high = vzip2_s16(v_values_l, v_values_l);

            u_low = vzip1_s16(u_values_l, u_values_l);
            v_low = vzip1_s16(v_values_l, v_values_l);
        }

        let y_high = vmull_high_laneq_s16::<0>(y_values, v_weights);
        let y_low = vmull_laneq_s16::<0>(vget_low_s16(y_values), v_weights);

        let rh = vmlal_laneq_s16::<1>(y_high, v_high, v_weights);
        let bh = vmlal_laneq_s16::<2>(y_high, u_high, v_weights);
        let gh = vmlal_laneq_s16::<3>(y_high, v_high, v_weights);
        let rl = vmlal_laneq_s16::<1>(y_low, v_low, v_weights);
        let bl = vmlal_laneq_s16::<2>(y_low, u_low, v_weights);
        let gl = vmlal_laneq_s16::<3>(y_low, v_low, v_weights);

        let r_high = vqrshrun_n_s32::<PRECISION>(rh);
        let b_high = vqrshrun_n_s32::<PRECISION>(bh);
        let g_high = vqrshrun_n_s32::<PRECISION>(vmlal_laneq_s16::<4>(gh, u_high, v_weights));

        let r_low = vqrshrun_n_s32::<PRECISION>(rl);
        let b_low = vqrshrun_n_s32::<PRECISION>(bl);
        let g_low = vqrshrun_n_s32::<PRECISION>(vmlal_laneq_s16::<4>(gl, u_low, v_weights));

        let r_values = vminq_u16(vcombine_u16(r_low, r_high), v_max_values);
        let g_values = vminq_u16(vcombine_u16(g_low, g_high), v_max_values);
        let b_values = vminq_u16(vcombine_u16(b_low, b_high), v_max_values);

        let v_alpha = a_values_l;

        neon_store_rgb16::<DESTINATION_CHANNELS>(
            rgba.get_unchecked_mut(cx * channels..).as_mut_ptr(),
            r_values,
            g_values,
            b_values,
            v_alpha,
        );

        cx += 8;

        match chroma_subsampling {
            YuvChromaSubsampling::Yuv420 | YuvChromaSubsampling::Yuv422 => {
                ux += 4;
            }
            YuvChromaSubsampling::Yuv444 => {
                ux += 8;
            }
        }
    }

    if cx < width as usize {
        let diff = width as usize - cx;

        assert!(diff <= 8);

        let mut dst_buffer: [u16; 8 * 4] = [0; 8 * 4];
        let mut y_buffer: [u16; 8] = [0; 8];
        let mut u_buffer: [u16; 8] = [0; 8];
        let mut v_buffer: [u16; 8] = [0; 8];
        let mut a_buffer: [u16; 8] = [0; 8];

        std::ptr::copy_nonoverlapping(
            y_ld_ptr.get_unchecked(cx..).as_ptr(),
            y_buffer.as_mut_ptr().cast(),
            diff,
        );

        let ux_diff = match chroma_subsampling {
            YuvChromaSubsampling::Yuv420 | YuvChromaSubsampling::Yuv422 => diff.div_ceil(2),
            YuvChromaSubsampling::Yuv444 => diff,
        };

        std::ptr::copy_nonoverlapping(
            a_ld_ptr.get_unchecked(cx..).as_ptr(),
            a_buffer.as_mut_ptr().cast(),
            diff,
        );

        std::ptr::copy_nonoverlapping(
            u_ld_ptr.get_unchecked(ux..).as_ptr(),
            u_buffer.as_mut_ptr().cast(),
            ux_diff,
        );

        std::ptr::copy_nonoverlapping(
            v_ld_ptr.get_unchecked(ux..).as_ptr(),
            v_buffer.as_mut_ptr().cast(),
            ux_diff,
        );

        let a_values_l = vld1q_u16(a_buffer.as_ptr().cast());

        let y_values: int16x8_t = vreinterpretq_s16_u16(vqsubq_u16(
            vreinterpretq_u16_s16(vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                y_buffer.as_ptr().cast(),
            )),
            y_corr,
        ));

        let u_high: int16x4_t;
        let v_high: int16x4_t;
        let u_low: int16x4_t;
        let v_low: int16x4_t;

        if chroma_subsampling == YuvChromaSubsampling::Yuv444 {
            let mut u_values_l =
                vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(u_buffer.as_ptr().cast());
            let mut v_values_l =
                vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(v_buffer.as_ptr().cast());

            u_values_l = vsubq_s16(u_values_l, uv_corr);
            v_values_l = vsubq_s16(v_values_l, uv_corr);

            u_high = vget_high_s16(u_values_l);
            u_low = vget_low_s16(u_values_l);
            v_high = vget_high_s16(v_values_l);
            v_low = vget_low_s16(v_values_l);
        } else {
            let mut u_values_l =
                vld_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(u_buffer.as_ptr().cast());
            let mut v_values_l =
                vld_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(v_buffer.as_ptr().cast());
            u_values_l = vsub_s16(u_values_l, vget_low_s16(uv_corr));
            v_values_l = vsub_s16(v_values_l, vget_low_s16(uv_corr));

            u_high = vzip2_s16(u_values_l, u_values_l);
            v_high = vzip2_s16(v_values_l, v_values_l);

            u_low = vzip1_s16(u_values_l, u_values_l);
            v_low = vzip1_s16(v_values_l, v_values_l);
        }

        let y_high = vmull_high_laneq_s16::<0>(y_values, v_weights);
        let y_low = vmull_laneq_s16::<0>(vget_low_s16(y_values), v_weights);

        let rh = vmlal_laneq_s16::<1>(y_high, v_high, v_weights);
        let bh = vmlal_laneq_s16::<2>(y_high, u_high, v_weights);
        let gh = vmlal_laneq_s16::<3>(y_high, v_high, v_weights);
        let rl = vmlal_laneq_s16::<1>(y_low, v_low, v_weights);
        let bl = vmlal_laneq_s16::<2>(y_low, u_low, v_weights);
        let gl = vmlal_laneq_s16::<3>(y_low, v_low, v_weights);

        let r_high = vqrshrun_n_s32::<PRECISION>(rh);
        let b_high = vqrshrun_n_s32::<PRECISION>(bh);
        let g_high = vqrshrun_n_s32::<PRECISION>(vmlal_laneq_s16::<4>(gh, u_high, v_weights));

        let r_low = vqrshrun_n_s32::<PRECISION>(rl);
        let b_low = vqrshrun_n_s32::<PRECISION>(bl);
        let g_low = vqrshrun_n_s32::<PRECISION>(vmlal_laneq_s16::<4>(gl, u_low, v_weights));

        let r_values = vminq_u16(vcombine_u16(r_low, r_high), v_max_values);
        let g_values = vminq_u16(vcombine_u16(g_low, g_high), v_max_values);
        let b_values = vminq_u16(vcombine_u16(b_low, b_high), v_max_values);

        let v_alpha = a_values_l;

        neon_store_rgb16::<DESTINATION_CHANNELS>(
            dst_buffer.as_mut_ptr().cast(),
            r_values,
            g_values,
            b_values,
            v_alpha,
        );

        let dst_shift = cx * channels;
        std::ptr::copy_nonoverlapping(
            dst_buffer.as_mut_ptr().cast(),
            rgba.get_unchecked_mut(dst_shift..).as_mut_ptr(),
            diff * channels,
        );

        cx += diff;
        ux += ux_diff;
    }

    ProcessedOffset { cx, ux }
}

#[cfg(feature = "rdm")]
#[target_feature(enable = "rdm")]
pub(crate) unsafe fn neon_yuv_p16_to_rgba16_alpha_row_rdm<
    const DESTINATION_CHANNELS: u8,
    const SAMPLING: u8,
    const ENDIANNESS: u8,
    const BYTES_POSITION: u8,
    const PRECISION: i32,
    const BIT_DEPTH: usize,
>(
    y_ld_ptr: &[u16],
    u_ld_ptr: &[u16],
    v_ld_ptr: &[u16],
    a_ld_ptr: &[u16],
    rgba: &mut [u16],
    width: u32,
    range: &YuvChromaRange,
    transform: &CbCrInverseTransform<i32>,
    start_cx: usize,
    start_ux: usize,
) -> ProcessedOffset {
    let destination_channels: YuvSourceChannels = DESTINATION_CHANNELS.into();
    assert!(
        destination_channels == YuvSourceChannels::Rgba
            || destination_channels == YuvSourceChannels::Bgra,
        "Cannot call YUV p16 to Rgb8 with alpha without real alpha"
    );
    let channels = destination_channels.get_channels_count();
    let chroma_subsampling: YuvChromaSubsampling = SAMPLING.into();

    let y_corr = vdupq_n_u16(range.bias_y as u16);
    let uv_corr = vdupq_n_s16(range.bias_uv as i16);

    let weights_arr: [i16; 8] = [
        transform.y_coef as i16,
        transform.cr_coef as i16,
        transform.cb_coef as i16,
        -transform.g_coeff_1 as i16,
        -transform.g_coeff_2 as i16,
        0,
        0,
        0,
    ];

    let v_weights = vld1q_s16(weights_arr.as_ptr());

    let zeros = vdupq_n_s16(0);
    let v_max_values = vdupq_n_u16((1 << BIT_DEPTH) - 1);

    let mut cx = start_cx;
    let mut ux = start_ux;

    const SCALE: i32 = 2;

    while cx + 16 < width as usize {
        let y_values0: int16x8_t = vreinterpretq_s16_u16(vqsubq_u16(
            vreinterpretq_u16_s16(vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                y_ld_ptr.get_unchecked(cx..).as_ptr(),
            )),
            y_corr,
        ));
        let y_values1: int16x8_t = vreinterpretq_s16_u16(vqsubq_u16(
            vreinterpretq_u16_s16(vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                y_ld_ptr.get_unchecked((cx + 8)..).as_ptr(),
            )),
            y_corr,
        ));

        let u_values0: int16x8_t;
        let v_values0: int16x8_t;
        let u_values1: int16x8_t;
        let v_values1: int16x8_t;

        if chroma_subsampling == YuvChromaSubsampling::Yuv444 {
            let mut u_values_l0 = vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                u_ld_ptr.get_unchecked(ux..).as_ptr(),
            );
            let mut v_values_l0 = vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                v_ld_ptr.get_unchecked(ux..).as_ptr(),
            );
            let mut u_values_l1 = vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                u_ld_ptr.get_unchecked((ux + 8)..).as_ptr(),
            );
            let mut v_values_l1 = vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                v_ld_ptr.get_unchecked((ux + 8)..).as_ptr(),
            );

            u_values_l0 = vsubq_s16(u_values_l0, uv_corr);
            v_values_l0 = vsubq_s16(v_values_l0, uv_corr);
            u_values_l1 = vsubq_s16(u_values_l1, uv_corr);
            v_values_l1 = vsubq_s16(v_values_l1, uv_corr);

            u_values0 = vshlq_n_s16::<SCALE>(u_values_l0);
            v_values0 = vshlq_n_s16::<SCALE>(v_values_l0);
            u_values1 = vshlq_n_s16::<SCALE>(u_values_l1);
            v_values1 = vshlq_n_s16::<SCALE>(v_values_l1);
        } else {
            let mut u_values_l = vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                u_ld_ptr.get_unchecked(ux..).as_ptr(),
            );
            let mut v_values_l = vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                v_ld_ptr.get_unchecked(ux..).as_ptr(),
            );
            u_values_l = vsubq_s16(u_values_l, uv_corr);
            v_values_l = vsubq_s16(v_values_l, uv_corr);

            let u_high = vzip2q_s16(u_values_l, u_values_l);
            let v_high = vzip2q_s16(v_values_l, v_values_l);

            let u_low = vzip1q_s16(u_values_l, u_values_l);
            let v_low = vzip1q_s16(v_values_l, v_values_l);

            u_values0 = vshlq_n_s16::<SCALE>(u_low);
            v_values0 = vshlq_n_s16::<SCALE>(v_low);

            u_values1 = vshlq_n_s16::<SCALE>(u_high);
            v_values1 = vshlq_n_s16::<SCALE>(v_high);
        }

        let y_high0 =
            vqrdmulhq_laneq_s16::<0>(vexpand_high_bp_by_2::<BIT_DEPTH>(y_values0), v_weights);
        let y_high1 =
            vqrdmulhq_laneq_s16::<0>(vexpand_high_bp_by_2::<BIT_DEPTH>(y_values1), v_weights);

        let r_vals0 = vqrdmlahq_laneq_s16::<1>(y_high0, v_values0, v_weights);
        let b_vals0 = vqrdmlahq_laneq_s16::<2>(y_high0, u_values0, v_weights);
        let g_vals0 = vqrdmlahq_laneq_s16::<4>(
            vqrdmlahq_laneq_s16::<3>(y_high0, v_values0, v_weights),
            u_values0,
            v_weights,
        );

        let r_vals1 = vqrdmlahq_laneq_s16::<1>(y_high1, v_values1, v_weights);
        let b_vals1 = vqrdmlahq_laneq_s16::<2>(y_high1, u_values1, v_weights);
        let g_vals1 = vqrdmlahq_laneq_s16::<4>(
            vqrdmlahq_laneq_s16::<3>(y_high1, v_values1, v_weights),
            u_values1,
            v_weights,
        );

        let a_values0 = vld1q_u16(a_ld_ptr.get_unchecked(cx..).as_ptr());
        let a_values1 = vld1q_u16(a_ld_ptr.get_unchecked((cx + 8)..).as_ptr());

        let rlv0 = vmaxq_s16(r_vals0, zeros);
        let glv0 = vmaxq_s16(g_vals0, zeros);
        let blv0 = vmaxq_s16(b_vals0, zeros);
        let rlv1 = vmaxq_s16(r_vals1, zeros);
        let glv1 = vmaxq_s16(g_vals1, zeros);
        let blv1 = vmaxq_s16(b_vals1, zeros);

        let r_values0 = vminq_u16(vreinterpretq_u16_s16(rlv0), v_max_values);
        let g_values0 = vminq_u16(vreinterpretq_u16_s16(glv0), v_max_values);
        let b_values0 = vminq_u16(vreinterpretq_u16_s16(blv0), v_max_values);

        let r_values1 = vminq_u16(vreinterpretq_u16_s16(rlv1), v_max_values);
        let g_values1 = vminq_u16(vreinterpretq_u16_s16(glv1), v_max_values);
        let b_values1 = vminq_u16(vreinterpretq_u16_s16(blv1), v_max_values);

        neon_store_rgb16::<DESTINATION_CHANNELS>(
            rgba.get_unchecked_mut(cx * channels..).as_mut_ptr(),
            r_values0,
            g_values0,
            b_values0,
            a_values0,
        );

        neon_store_rgb16::<DESTINATION_CHANNELS>(
            rgba.get_unchecked_mut((cx + 8) * channels..).as_mut_ptr(),
            r_values1,
            g_values1,
            b_values1,
            a_values1,
        );

        cx += 16;

        match chroma_subsampling {
            YuvChromaSubsampling::Yuv420 | YuvChromaSubsampling::Yuv422 => {
                ux += 8;
            }
            YuvChromaSubsampling::Yuv444 => {
                ux += 16;
            }
        }
    }

    while cx + 8 < width as usize {
        let a_values_l = vld1q_u16(a_ld_ptr.get_unchecked(cx..).as_ptr());

        let y_values: int16x8_t = vreinterpretq_s16_u16(vqsubq_u16(
            vreinterpretq_u16_s16(vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                y_ld_ptr.get_unchecked(cx..).as_ptr(),
            )),
            y_corr,
        ));

        let u_values: int16x8_t;
        let v_values: int16x8_t;

        if chroma_subsampling == YuvChromaSubsampling::Yuv444 {
            let mut u_values_l = vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                u_ld_ptr.get_unchecked(ux..).as_ptr(),
            );
            let mut v_values_l = vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                v_ld_ptr.get_unchecked(ux..).as_ptr(),
            );

            u_values_l = vsubq_s16(u_values_l, uv_corr);
            v_values_l = vsubq_s16(v_values_l, uv_corr);

            u_values = vshlq_n_s16::<SCALE>(u_values_l);
            v_values = vshlq_n_s16::<SCALE>(v_values_l);
        } else {
            let mut u_values_l = vld_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                u_ld_ptr.get_unchecked(ux..).as_ptr(),
            );
            let mut v_values_l = vld_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                v_ld_ptr.get_unchecked(ux..).as_ptr(),
            );
            u_values_l = vsub_s16(u_values_l, vget_low_s16(uv_corr));
            v_values_l = vsub_s16(v_values_l, vget_low_s16(uv_corr));

            let u_high = vzip2_s16(u_values_l, u_values_l);
            let v_high = vzip2_s16(v_values_l, v_values_l);

            let u_low = vzip1_s16(u_values_l, u_values_l);
            let v_low = vzip1_s16(v_values_l, v_values_l);

            u_values = vshlq_n_s16::<SCALE>(vcombine_s16(u_low, u_high));
            v_values = vshlq_n_s16::<SCALE>(vcombine_s16(v_low, v_high));
        }

        let y_high =
            vqrdmulhq_laneq_s16::<0>(vexpand_high_bp_by_2::<BIT_DEPTH>(y_values), v_weights);

        let r_vals = vqrdmlahq_laneq_s16::<1>(y_high, v_values, v_weights);
        let b_vals = vqrdmlahq_laneq_s16::<2>(y_high, u_values, v_weights);
        let g_vals = vqrdmlahq_laneq_s16::<4>(
            vqrdmlahq_laneq_s16::<3>(y_high, v_values, v_weights),
            u_values,
            v_weights,
        );

        let rlv = vmaxq_s16(r_vals, zeros);
        let glv = vmaxq_s16(g_vals, zeros);
        let blv = vmaxq_s16(b_vals, zeros);

        let r_values = vminq_u16(vreinterpretq_u16_s16(rlv), v_max_values);
        let g_values = vminq_u16(vreinterpretq_u16_s16(glv), v_max_values);
        let b_values = vminq_u16(vreinterpretq_u16_s16(blv), v_max_values);

        let v_alpha = a_values_l;

        neon_store_rgb16::<DESTINATION_CHANNELS>(
            rgba.get_unchecked_mut(cx * channels..).as_mut_ptr(),
            r_values,
            g_values,
            b_values,
            v_alpha,
        );

        cx += 8;

        match chroma_subsampling {
            YuvChromaSubsampling::Yuv420 | YuvChromaSubsampling::Yuv422 => {
                ux += 4;
            }
            YuvChromaSubsampling::Yuv444 => {
                ux += 8;
            }
        }
    }

    if cx < width as usize {
        let diff = width as usize - cx;

        assert!(diff <= 8);

        let mut dst_buffer: [u16; 8 * 4] = [0; 8 * 4];
        let mut y_buffer: [u16; 8] = [0; 8];
        let mut u_buffer: [u16; 8] = [0; 8];
        let mut v_buffer: [u16; 8] = [0; 8];
        let mut a_buffer: [u16; 8] = [0; 8];

        std::ptr::copy_nonoverlapping(
            y_ld_ptr.get_unchecked(cx..).as_ptr(),
            y_buffer.as_mut_ptr().cast(),
            diff,
        );

        let ux_diff = match chroma_subsampling {
            YuvChromaSubsampling::Yuv420 | YuvChromaSubsampling::Yuv422 => diff.div_ceil(2),
            YuvChromaSubsampling::Yuv444 => diff,
        };

        std::ptr::copy_nonoverlapping(
            a_ld_ptr.get_unchecked(cx..).as_ptr(),
            a_buffer.as_mut_ptr().cast(),
            diff,
        );

        std::ptr::copy_nonoverlapping(
            u_ld_ptr.get_unchecked(ux..).as_ptr(),
            u_buffer.as_mut_ptr().cast(),
            ux_diff,
        );

        std::ptr::copy_nonoverlapping(
            v_ld_ptr.get_unchecked(ux..).as_ptr(),
            v_buffer.as_mut_ptr().cast(),
            ux_diff,
        );

        let a_values_l = vld1q_u16(a_buffer.as_ptr().cast());

        let y_values: int16x8_t = vreinterpretq_s16_u16(vqsubq_u16(
            vreinterpretq_u16_s16(vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(
                y_buffer.as_ptr().cast(),
            )),
            y_corr,
        ));

        let u_values: int16x8_t;
        let v_values: int16x8_t;

        if chroma_subsampling == YuvChromaSubsampling::Yuv444 {
            let mut u_values_l =
                vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(u_buffer.as_ptr().cast());
            let mut v_values_l =
                vldq_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(v_buffer.as_ptr().cast());

            u_values_l = vsubq_s16(u_values_l, uv_corr);
            v_values_l = vsubq_s16(v_values_l, uv_corr);

            u_values = vshlq_n_s16::<SCALE>(u_values_l);
            v_values = vshlq_n_s16::<SCALE>(v_values_l);
        } else {
            let mut u_values_l =
                vld_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(u_buffer.as_ptr().cast());
            let mut v_values_l =
                vld_s16_endian::<ENDIANNESS, BYTES_POSITION, BIT_DEPTH>(v_buffer.as_ptr().cast());
            u_values_l = vsub_s16(u_values_l, vget_low_s16(uv_corr));
            v_values_l = vsub_s16(v_values_l, vget_low_s16(uv_corr));

            let u_high = vzip2_s16(u_values_l, u_values_l);
            let v_high = vzip2_s16(v_values_l, v_values_l);

            let u_low = vzip1_s16(u_values_l, u_values_l);
            let v_low = vzip1_s16(v_values_l, v_values_l);

            u_values = vshlq_n_s16::<SCALE>(vcombine_s16(u_low, u_high));
            v_values = vshlq_n_s16::<SCALE>(vcombine_s16(v_low, v_high));
        }

        let y_high =
            vqrdmulhq_laneq_s16::<0>(vexpand_high_bp_by_2::<BIT_DEPTH>(y_values), v_weights);

        let r_vals = vqrdmlahq_laneq_s16::<1>(y_high, v_values, v_weights);
        let b_vals = vqrdmlahq_laneq_s16::<2>(y_high, u_values, v_weights);
        let g_vals = vqrdmlahq_laneq_s16::<4>(
            vqrdmlahq_laneq_s16::<3>(y_high, v_values, v_weights),
            u_values,
            v_weights,
        );

        let rlv = vmaxq_s16(r_vals, zeros);
        let glv = vmaxq_s16(g_vals, zeros);
        let blv = vmaxq_s16(b_vals, zeros);

        let r_values = vminq_u16(vreinterpretq_u16_s16(rlv), v_max_values);
        let g_values = vminq_u16(vreinterpretq_u16_s16(glv), v_max_values);
        let b_values = vminq_u16(vreinterpretq_u16_s16(blv), v_max_values);

        let v_alpha = a_values_l;

        neon_store_rgb16::<DESTINATION_CHANNELS>(
            dst_buffer.as_mut_ptr().cast(),
            r_values,
            g_values,
            b_values,
            v_alpha,
        );

        let dst_shift = cx * channels;
        std::ptr::copy_nonoverlapping(
            dst_buffer.as_mut_ptr().cast(),
            rgba.get_unchecked_mut(dst_shift..).as_mut_ptr(),
            diff * channels,
        );

        cx += diff;
        ux += ux_diff;
    }

    ProcessedOffset { cx, ux }
}