heic 0.1.4

//! Deblocking filter (H.265 Section 8.7.2)
//!
//! Applies strong/weak filtering at CU and TU boundaries to reduce blocking artifacts.
//! For I-slices, all boundaries have bS=2. For P/B slices, bS depends on
//! prediction mode, reference pictures, and motion vector differences.

use super::inter::PbMotion;
use super::picture::{
    DEBLOCK_FLAG_HORIZ, DEBLOCK_FLAG_VERT, DEBLOCK_PB_EDGE_HORIZ, DEBLOCK_PB_EDGE_VERT,
    DecodedFrame,
};
use super::slice::PredMode;

#[cfg(feature = "std")]
use std::sync::atomic::{AtomicBool, Ordering};

#[cfg(feature = "std")]
static DEBLOCK_TRACE_ENABLED: AtomicBool = AtomicBool::new(false);

/// Enable deblocking trace output (writes to /tmp/our_deblock_trace.txt)
#[cfg(feature = "std")]
pub fn enable_deblock_trace() {
    DEBLOCK_TRACE_ENABLED.store(true, Ordering::Relaxed);
}

#[cfg(feature = "std")]
fn deblock_trace_active() -> bool {
    DEBLOCK_TRACE_ENABLED.load(Ordering::Relaxed)
}

#[cfg(feature = "std")]
#[allow(clippy::too_many_arguments)]
fn trace_edge(
    vertical: bool,
    x: u32,
    y: u32,
    bs: i32,
    qp_p: i32,
    qp_q: i32,
    beta: i32,
    tc: i32,
    de: i32,
    dep: i32,
    deq: i32,
    d: i32,
    p0_0: i32,
    p1_0: i32,
    p2_0: i32,
    p3_0: i32,
    q0_0: i32,
    q1_0: i32,
    q2_0: i32,
    q3_0: i32,
    p0_3: i32,
    p1_3: i32,
    p2_3: i32,
    p3_3: i32,
    q0_3: i32,
    q1_3: i32,
    q2_3: i32,
    q3_3: i32,
) {
    use std::io::Write;
    use std::sync::LazyLock;
    use std::sync::Mutex;
    static TRACE_FILE: LazyLock<Mutex<std::fs::File>> =
        LazyLock::new(|| Mutex::new(std::fs::File::create("/tmp/our_deblock_trace.txt").unwrap()));
    let mut f = TRACE_FILE.lock().unwrap();
    let _ = writeln!(
        f,
        "EDGE {} x={} y={} bS={} QP_P={} QP_Q={} qP_L={} beta={} tc={} dE={} dEp={} dEq={} d={} \
        p[0]={{{},{},{},{}}} q[0]={{{},{},{},{}}} p[3]={{{},{},{},{}}} q[3]={{{},{},{},{}}}",
        if vertical { 'V' } else { 'H' },
        x,
        y,
        bs,
        qp_p,
        qp_q,
        (qp_q + qp_p + 1) >> 1,
        beta,
        tc,
        de,
        dep,
        deq,
        d,
        p0_0,
        p1_0,
        p2_0,
        p3_0,
        q0_0,
        q1_0,
        q2_0,
        q3_0,
        p0_3,
        p1_3,
        p2_3,
        p3_3,
        q0_3,
        q1_3,
        q2_3,
        q3_3
    );
}

/// Beta prime values for deblocking filter (Table 8-12)
/// Index 0-51 maps QP to beta prime threshold
#[rustfmt::skip]
static BETA_PRIME: [u16; 52] = [
     0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
     6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 20, 22, 24,
    26, 28, 30, 32, 34, 36, 38, 40, 42, 44, 46, 48, 50, 52, 54, 56,
    58, 60, 62, 64,
];

/// tC prime values for deblocking filter (Table 8-23)
/// Index 0-53 maps to tC prime threshold
#[rustfmt::skip]
static TC_PRIME: [u16; 54] = [
     0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,
     0,  0,  1,  1,  1,  1,  1,  1,  1,  1,  1,  2,  2,  2,  2,  3,
     3,  3,  3,  4,  4,  4,  5,  5,  6,  6,  7,  8,  9, 10, 11, 13,
    14, 16, 18, 20, 22, 24,
];

/// Compute boundary strength for an edge (H.265 8.7.2.4)
///
/// For I-slices (inter_ctx is None), bS=2 always.
/// For P/B slices, bS depends on prediction mode, ref pics, and MVs.
///
/// `is_transform_edge`: true if this edge is a transform block boundary (not just PB).
/// The CBF check (bS=1 for non-zero coefficients) only applies at transform block edges.
fn compute_bs(
    x: u32,
    y: u32,
    vertical: bool,
    is_transform_edge: bool,
    inter_ctx: &Option<&InterDeblockCtx<'_>>,
) -> i32 {
    let ctx = match inter_ctx {
        Some(c) => c,
        None => return 2, // I-slice: always bS=2
    };

    // Get positions on P side (before boundary) and Q side (at boundary)
    let (px, py, qx, qy) = if vertical {
        (x.wrapping_sub(1), y, x, y)
    } else {
        (x, y.wrapping_sub(1), x, y)
    };

    // Get prediction modes
    let get_pred = |sx: u32, sy: u32| -> PredMode {
        let idx = (sy / ctx.min_pu_size * ctx.pu_stride + sx / ctx.min_pu_size) as usize;
        if idx < ctx.pred_mode.len() {
            ctx.pred_mode[idx]
        } else {
            PredMode::Intra
        }
    };

    let pred_p = get_pred(px, py);
    let pred_q = get_pred(qx, qy);

    // bS=2: either side is intra
    if pred_p == PredMode::Intra || pred_q == PredMode::Intra {
        return 2;
    }

    // bS=1: CBF check — only at transform block edges (H.265 8.7.2.4)
    // "if the block edge is also a transform block edge and the sample p0 or q0
    //  is in a luma transform block which contains one or more non-zero transform
    //  coefficient levels, bS is set equal to 1"
    if is_transform_edge {
        let get_cbf = |sx: u32, sy: u32| -> bool {
            let idx = (sy / 4 * ctx.cbf_map_stride + sx / 4) as usize;
            if idx < ctx.cbf_map.len() {
                ctx.cbf_map[idx]
            } else {
                false
            }
        };

        if get_cbf(px, py) || get_cbf(qx, qy) {
            return 1;
        }
    }

    // bS=1: different reference pictures or MV difference >= 1 integer pel (4 quarter-pel)
    let get_mv = |sx: u32, sy: u32| -> PbMotion {
        let idx = (sy / ctx.min_pu_size * ctx.pu_stride + sx / ctx.min_pu_size) as usize;
        if idx < ctx.mv_info.len() {
            ctx.mv_info[idx]
        } else {
            PbMotion::UNAVAILABLE
        }
    };

    let mv_p = get_mv(px, py);
    let mv_q = get_mv(qx, qy);

    // H.265 8.7.2.4.5: resolve ref_idx to reference picture (POC)
    // The spec compares actual reference pictures, not raw indices.
    // Two blocks may have different ref_idx values that refer to the same picture.
    let ref_poc = &ctx.ref_poc;
    let resolve_ref = |pred_flag: bool, list: usize, ref_idx: i8| -> i32 {
        if !pred_flag {
            return -1; // No reference in this list
        }
        let idx = ref_idx as usize;
        if idx < super::inter::MAX_NUM_REF_PICS {
            ref_poc[list][idx]
        } else {
            -1
        }
    };

    let ref_pic_p0 = resolve_ref(mv_p.pred_flag[0], 0, mv_p.ref_idx[0]);
    let ref_pic_p1 = resolve_ref(mv_p.pred_flag[1], 1, mv_p.ref_idx[1]);
    let ref_pic_q0 = resolve_ref(mv_q.pred_flag[0], 0, mv_q.ref_idx[0]);
    let ref_pic_q1 = resolve_ref(mv_q.pred_flag[1], 1, mv_q.ref_idx[1]);

    let count_p = mv_p.pred_flag[0] as u8 + mv_p.pred_flag[1] as u8;
    let count_q = mv_q.pred_flag[0] as u8 + mv_q.pred_flag[1] as u8;

    // Different number of motion vectors
    if count_p != count_q {
        return 1;
    }

    // Check if both blocks reference the same set of pictures
    let same_pics = (ref_pic_p0 == ref_pic_q0 && ref_pic_p1 == ref_pic_q1)
        || (ref_pic_p0 == ref_pic_q1 && ref_pic_p1 == ref_pic_q0);

    if !same_pics {
        return 1;
    }

    // Same reference pictures — now check MV differences
    // Get MVs (zero if pred_flag is false, matching libde265 behavior)
    let mv_p0 = if mv_p.pred_flag[0] {
        mv_p.mv[0]
    } else {
        super::inter::MotionVector::ZERO
    };
    let mv_p1 = if mv_p.pred_flag[1] {
        mv_p.mv[1]
    } else {
        super::inter::MotionVector::ZERO
    };
    let mv_q0 = if mv_q.pred_flag[0] {
        mv_q.mv[0]
    } else {
        super::inter::MotionVector::ZERO
    };
    let mv_q1 = if mv_q.pred_flag[1] {
        mv_q.mv[1]
    } else {
        super::inter::MotionVector::ZERO
    };

    // Helper: check if MV difference >= 4 quarter-pel (1 integer sample)
    let mv_diff_ge4 = |a: super::inter::MotionVector, b: super::inter::MotionVector| -> bool {
        let dx = (a.x as i32 - b.x as i32).abs();
        let dy = (a.y as i32 - b.y as i32).abs();
        dx >= 4 || dy >= 4
    };

    if ref_pic_p0 != ref_pic_p1 {
        // Two different reference pictures
        if ref_pic_p0 == ref_pic_q0 {
            // Same order: P0↔Q0, P1↔Q1
            if mv_diff_ge4(mv_p0, mv_q0) || mv_diff_ge4(mv_p1, mv_q1) {
                return 1;
            }
        } else {
            // Cross order: P0↔Q1, P1↔Q0
            if mv_diff_ge4(mv_p0, mv_q1) || mv_diff_ge4(mv_p1, mv_q0) {
                return 1;
            }
        }
    } else {
        // Same reference picture for both lists (refPicP0 == refPicP1)
        // Must check BOTH orderings — bS=1 only if BOTH give MV difference
        let same_order_diff = mv_diff_ge4(mv_p0, mv_q0) || mv_diff_ge4(mv_p1, mv_q1);
        let cross_order_diff = mv_diff_ge4(mv_p0, mv_q1) || mv_diff_ge4(mv_p1, mv_q0);
        if same_order_diff && cross_order_diff {
            return 1;
        }
    }

    0 // bS=0: no filtering needed
}

/// Chroma QP mapping table (Table 8-10) for indices 30-42
#[rustfmt::skip]
static CHROMA_QP_TABLE: [i32; 13] = [
    29, 30, 31, 32, 33, 33, 34, 34, 35, 35, 36, 36, 37,
];

/// Map intermediate chroma QP to actual chroma QP (4:2:0)
fn chroma_qp_mapping(qp_i: i32) -> i32 {
    if qp_i < 30 {
        qp_i
    } else if qp_i >= 43 {
        qp_i - 6
    } else {
        CHROMA_QP_TABLE[(qp_i - 30) as usize]
    }
}

/// Optional inter prediction context for boundary strength derivation
pub struct InterDeblockCtx<'a> {
    /// Prediction mode map (min_pu granularity)
    pub pred_mode: &'a [PredMode],
    /// Motion vector info (min_pu granularity)
    pub mv_info: &'a [PbMotion],
    /// PU map stride
    pub pu_stride: u32,
    /// Minimum PU size in luma samples
    pub min_pu_size: u32,
    /// CBF map at 4x4 granularity (true = has non-zero coefficients)
    pub cbf_map: &'a [bool],
    /// CBF map stride (width / 4)
    pub cbf_map_stride: u32,
    /// Reference picture POCs per list [L0/L1][ref_idx] — needed to resolve
    /// ref_idx to actual reference picture for bS derivation (H.265 8.7.2.4.5)
    pub ref_poc: [[i32; super::inter::MAX_NUM_REF_PICS]; 2],
}

/// Apply the deblocking filter to a decoded frame.
///
/// `beta_offset` and `tc_offset` come from slice header (slice_beta_offset_div2 * 2
/// and slice_tc_offset_div2 * 2).
/// `cb_qp_offset` and `cr_qp_offset` come from PPS (pps_cb_qp_offset / pps_cr_qp_offset).
/// `inter_ctx` is None for I-slices (all bS=2) and Some for P/B slices.
pub fn apply_deblocking_filter(
    frame: &mut DecodedFrame,
    beta_offset: i32,
    tc_offset: i32,
    cb_qp_offset: i32,
    cr_qp_offset: i32,
    inter_ctx: Option<&InterDeblockCtx<'_>>,
) {
    let width = frame.width;
    let height = frame.height;

    // Edge masks: check both transform block edges and prediction block edges
    let vert_edge_mask = DEBLOCK_FLAG_VERT | DEBLOCK_PB_EDGE_VERT;
    let horiz_edge_mask = DEBLOCK_FLAG_HORIZ | DEBLOCK_PB_EDGE_HORIZ;

    // Pass 1: Vertical edges
    // Process at 8-sample intervals in x, 4-sample intervals in y
    let mut x = 8u32;
    while x < width {
        let mut y = 0u32;
        while y < height {
            let bx = x / 4;
            let by = y / 4;
            let idx = (by * frame.deblock_stride + bx) as usize;
            if idx < frame.deblock_flags.len() {
                let flags = frame.deblock_flags[idx];
                if (flags & vert_edge_mask) != 0 {
                    // Is this a transform block edge? (affects CBF check in bS derivation)
                    let is_tb_edge = (flags & DEBLOCK_FLAG_VERT) != 0;

                    // Get QP on both sides
                    let qp_q = frame.qp_map[idx] as i32;
                    let qp_p = if bx > 0 {
                        frame.qp_map[(by * frame.deblock_stride + bx - 1) as usize] as i32
                    } else {
                        qp_q
                    };

                    let bs = compute_bs(x, y, true, is_tb_edge, &inter_ctx);
                    if bs > 0 {
                        filter_edge_luma(frame, x, y, true, qp_p, qp_q, beta_offset, tc_offset, bs);
                    }
                }
            }
            y += 4;
        }
        x += 8;
    }

    // Pass 2: Horizontal edges
    // Process at 4-sample intervals in x, 8-sample intervals in y
    let mut y = 8u32;
    while y < height {
        let mut x = 0u32;
        while x < width {
            let bx = x / 4;
            let by = y / 4;
            let idx = (by * frame.deblock_stride + bx) as usize;
            if idx < frame.deblock_flags.len() {
                let flags = frame.deblock_flags[idx];
                if (flags & horiz_edge_mask) != 0 {
                    let is_tb_edge = (flags & DEBLOCK_FLAG_HORIZ) != 0;

                    let qp_q = frame.qp_map[idx] as i32;
                    let qp_p = if by > 0 {
                        frame.qp_map[((by - 1) * frame.deblock_stride + bx) as usize] as i32
                    } else {
                        qp_q
                    };

                    let bs = compute_bs(x, y, false, is_tb_edge, &inter_ctx);
                    if bs > 0 {
                        filter_edge_luma(
                            frame,
                            x,
                            y,
                            false,
                            qp_p,
                            qp_q,
                            beta_offset,
                            tc_offset,
                            bs,
                        );
                    }
                }
            }
            x += 4;
        }
        y += 8;
    }

    // Chroma deblocking (only at bS=2 boundaries)
    if frame.chroma_format > 0 {
        apply_chroma_deblocking(frame, tc_offset, cb_qp_offset, cr_qp_offset, &inter_ctx);
    }
}

/// Filter a single luma edge (4 samples along the edge)
///
/// For vertical edges: x is the boundary position, filtering samples at x-1..x-4 and x..x+3
/// For horizontal edges: y is the boundary position, filtering samples at y-1..y-4 and y..y+3
///
/// Uses direct plane access with stride-based indexing to avoid per-sample bounds checks.
#[allow(clippy::too_many_arguments)]
fn filter_edge_luma(
    frame: &mut DecodedFrame,
    x: u32,
    y: u32,
    vertical: bool,
    qp_p: i32,
    qp_q: i32,
    beta_offset: i32,
    tc_offset: i32,
    bs: i32,
) {
    let bit_depth = frame.bit_depth as i32;
    let max_val = (1i32 << bit_depth) - 1;

    // Compute thresholds
    let qp_l = (qp_q + qp_p + 1) >> 1;
    let q_beta = (qp_l + beta_offset).clamp(0, 51);
    let beta = (BETA_PRIME[q_beta as usize] as i32) << (bit_depth - 8);
    let q_tc = (qp_l + 2 * (bs - 1) + tc_offset).clamp(0, 53);
    let tc = (TC_PRIME[q_tc as usize] as i32) << (bit_depth - 8);

    if tc == 0 {
        return;
    }

    let stride = frame.y_stride();
    let plane = &mut frame.y_plane;

    // Compute stride-based addressing:
    // - step_along: stride between adjacent samples along the edge (k direction)
    // - step_across: stride between adjacent samples perpendicular to edge (p/q direction)
    // - base_q: offset of q[0][0] in plane
    let (step_along, step_across, base_q) = if vertical {
        // Vertical edge: k steps in y (stride), p/q steps in x (1)
        (stride, 1usize, y as usize * stride + x as usize)
    } else {
        // Horizontal edge: k steps in x (1), p/q steps in y (stride)
        (1usize, stride, y as usize * stride + x as usize)
    };
    // base_p = one step before the boundary on the p side
    let base_p = base_q - step_across;

    // Bounds check: ensure all 4 samples on both sides are in-bounds
    // q side extends 3 steps across from base_q, plus 3 steps along
    // p side extends 3 steps across back from base_p (= base_q - step_across)
    if base_p < 3 * step_across {
        return;
    }
    let last_q = base_q + 3 * step_along + 3 * step_across;
    if last_q >= plane.len() {
        return;
    }

    // Read samples for edge decision at k=0 and k=3
    let k3 = 3 * step_along;
    let p0_0 = plane[base_p] as i32;
    let p1_0 = plane[base_p - step_across] as i32;
    let p2_0 = plane[base_p - 2 * step_across] as i32;
    let p3_0 = plane[base_p - 3 * step_across] as i32;
    let q0_0 = plane[base_q] as i32;
    let q1_0 = plane[base_q + step_across] as i32;
    let q2_0 = plane[base_q + 2 * step_across] as i32;
    let q3_0 = plane[base_q + 3 * step_across] as i32;

    let p0_3 = plane[base_p + k3] as i32;
    let p1_3 = plane[base_p + k3 - step_across] as i32;
    let p2_3 = plane[base_p + k3 - 2 * step_across] as i32;
    let p3_3 = plane[base_p + k3 - 3 * step_across] as i32;
    let q0_3 = plane[base_q + k3] as i32;
    let q1_3 = plane[base_q + k3 + step_across] as i32;
    let q2_3 = plane[base_q + k3 + 2 * step_across] as i32;
    let q3_3 = plane[base_q + k3 + 3 * step_across] as i32;

    // Edge decision (H.265 8.7.2.5.3)
    let dp0 = (p2_0 - 2 * p1_0 + p0_0).abs();
    let dp3 = (p2_3 - 2 * p1_3 + p0_3).abs();
    let dq0 = (q2_0 - 2 * q1_0 + q0_0).abs();
    let dq3 = (q2_3 - 2 * q1_3 + q0_3).abs();

    let dpq0 = dp0 + dq0;
    let dpq3 = dp3 + dq3;
    let dp = dp0 + dp3;
    let dq = dq0 + dq3;
    let d = dpq0 + dpq3;

    if d >= beta {
        #[cfg(feature = "std")]
        if deblock_trace_active() {
            trace_edge(
                vertical, x, y, bs, qp_p, qp_q, beta, tc, 0, 0, 0, d, p0_0, p1_0, p2_0, p3_0, q0_0,
                q1_0, q2_0, q3_0, p0_3, p1_3, p2_3, p3_3, q0_3, q1_3, q2_3, q3_3,
            );
        }
        return;
    }

    // Determine filter strength
    let d_sam0 = 2 * dpq0 < (beta >> 2)
        && (p3_0 - p0_0).abs() + (q0_0 - q3_0).abs() < (beta >> 3)
        && (p0_0 - q0_0).abs() < ((5 * tc + 1) >> 1);

    let d_sam3 = 2 * dpq3 < (beta >> 2)
        && (p3_3 - p0_3).abs() + (q0_3 - q3_3).abs() < (beta >> 3)
        && (p0_3 - q0_3).abs() < ((5 * tc + 1) >> 1);

    let strong = d_sam0 && d_sam3;
    let d_ep = dp < ((beta + (beta >> 1)) >> 3);
    let d_eq = dq < ((beta + (beta >> 1)) >> 3);

    #[allow(unused_variables)]
    let de = if strong { 2 } else { 1 };

    #[cfg(feature = "std")]
    if deblock_trace_active() {
        trace_edge(
            vertical,
            x,
            y,
            bs,
            qp_p,
            qp_q,
            beta,
            tc,
            de,
            d_ep as i32,
            d_eq as i32,
            d,
            p0_0,
            p1_0,
            p2_0,
            p3_0,
            q0_0,
            q1_0,
            q2_0,
            q3_0,
            p0_3,
            p1_3,
            p2_3,
            p3_3,
            q0_3,
            q1_3,
            q2_3,
            q3_3,
        );
    }

    // Apply filter for all 4 samples along the edge
    for k in 0..4usize {
        let k_off = k * step_along;

        // Read p[0..3] and q[0..3] for this sample
        let p0 = plane[base_p + k_off] as i32;
        let p1 = plane[base_p + k_off - step_across] as i32;
        let p2 = plane[base_p + k_off - 2 * step_across] as i32;
        let q0 = plane[base_q + k_off] as i32;
        let q1 = plane[base_q + k_off + step_across] as i32;
        let q2 = plane[base_q + k_off + 2 * step_across] as i32;

        if strong {
            let p3 = plane[base_p + k_off - 3 * step_across] as i32;
            let q3 = plane[base_q + k_off + 3 * step_across] as i32;
            let tc2 = 2 * tc;

            // Strong filter (H.265 8.7.2.5.7)
            let p0_f = ((p2 + 2 * p1 + 2 * p0 + 2 * q0 + q1 + 4) >> 3)
                .clamp(p0 - tc2, p0 + tc2)
                .clamp(0, max_val);
            let p1_f = ((p2 + p1 + p0 + q0 + 2) >> 2)
                .clamp(p1 - tc2, p1 + tc2)
                .clamp(0, max_val);
            let p2_f = ((2 * p3 + 3 * p2 + p1 + p0 + q0 + 4) >> 3)
                .clamp(p2 - tc2, p2 + tc2)
                .clamp(0, max_val);
            let q0_f = ((p1 + 2 * p0 + 2 * q0 + 2 * q1 + q2 + 4) >> 3)
                .clamp(q0 - tc2, q0 + tc2)
                .clamp(0, max_val);
            let q1_f = ((p0 + q0 + q1 + q2 + 2) >> 2)
                .clamp(q1 - tc2, q1 + tc2)
                .clamp(0, max_val);
            let q2_f = ((p0 + q0 + q1 + 3 * q2 + 2 * q3 + 4) >> 3)
                .clamp(q2 - tc2, q2 + tc2)
                .clamp(0, max_val);

            plane[base_p + k_off] = p0_f as u16;
            plane[base_p + k_off - step_across] = p1_f as u16;
            plane[base_p + k_off - 2 * step_across] = p2_f as u16;
            plane[base_q + k_off] = q0_f as u16;
            plane[base_q + k_off + step_across] = q1_f as u16;
            plane[base_q + k_off + 2 * step_across] = q2_f as u16;
        } else {
            // Weak filter
            let delta = (9 * (q0 - p0) - 3 * (q1 - p1) + 8) >> 4;

            if delta.abs() < 10 * tc {
                let delta = delta.clamp(-tc, tc);

                plane[base_p + k_off] = (p0 + delta).clamp(0, max_val) as u16;
                plane[base_q + k_off] = (q0 - delta).clamp(0, max_val) as u16;

                if d_ep {
                    let delta_p =
                        ((((p2 + p0 + 1) >> 1) - p1 + delta) >> 1).clamp(-(tc >> 1), tc >> 1);
                    plane[base_p + k_off - step_across] = (p1 + delta_p).clamp(0, max_val) as u16;
                }
                if d_eq {
                    let delta_q =
                        ((((q2 + q0 + 1) >> 1) - q1 - delta) >> 1).clamp(-(tc >> 1), tc >> 1);
                    plane[base_q + k_off + step_across] = (q1 + delta_q).clamp(0, max_val) as u16;
                }
            }
        }
    }
}

/// Apply chroma deblocking filter
///
/// For I-slices, all edges have bS=2, so chroma deblocking applies everywhere.
/// Chroma deblocking only modifies p0 and q0 (one sample on each side).
fn apply_chroma_deblocking(
    frame: &mut DecodedFrame,
    tc_offset: i32,
    cb_qp_offset: i32,
    cr_qp_offset: i32,
    inter_ctx: &Option<&InterDeblockCtx<'_>>,
) {
    let width = frame.width;
    let height = frame.height;
    let bit_depth_c = frame.bit_depth as i32; // Same as luma for typical HEIC
    let max_val = (1i32 << bit_depth_c) - 1;

    // Chroma subsampling factors
    let (sub_x, sub_y) = match frame.chroma_format {
        1 => (2u32, 2u32),
        2 => (2, 1),
        3 => (1, 1),
        _ => return,
    };

    let c_stride = frame.c_stride();
    let c_height = height / sub_y;
    let c_width = width / sub_x;

    // For 4:2:0: chroma edges are at 8-chroma-pixel intervals (16 luma pixels).
    // Per H.265 8.7.2, chroma deblocking requires both sides to have width/height >= 8
    // in chroma samples. The edge processing unit is 4 chroma samples along the edge.
    //
    // Matching libde265: xIncr = 2*SubWidthC (in 4-luma-pixel deblock grid units),
    // yIncr = SubHeightC for vertical, yIncr = 2*SubHeightC for horizontal.
    let x_step_vert = 8 * sub_x; // luma x step for vertical edges (16 for 4:2:0)
    let y_step_vert = 4 * sub_y; // luma y step for vertical edges (8 for 4:2:0)
    let x_step_horiz = 4 * sub_x; // luma x step for horizontal edges (8 for 4:2:0)
    let y_step_horiz = 8 * sub_y; // luma y step for horizontal edges (16 for 4:2:0)

    let vert_edge_mask = DEBLOCK_FLAG_VERT | DEBLOCK_PB_EDGE_VERT;
    let horiz_edge_mask = DEBLOCK_FLAG_HORIZ | DEBLOCK_PB_EDGE_HORIZ;

    // Pass 1: Vertical edges
    let mut x = x_step_vert;
    while x < width {
        let mut y = 0u32;
        while y < height {
            let bx = x / 4;
            let by = y / 4;
            let idx = (by * frame.deblock_stride + bx) as usize;
            if idx < frame.deblock_flags.len() && (frame.deblock_flags[idx] & vert_edge_mask) != 0 {
                let is_tb_edge = (frame.deblock_flags[idx] & DEBLOCK_FLAG_VERT) != 0;
                // Chroma only filters at bS=2
                let bs = compute_bs(x, y, true, is_tb_edge, inter_ctx);
                if bs < 2 {
                    y += y_step_vert;
                    continue;
                }

                let qp_q = frame.qp_map[idx] as i32;
                let qp_p = if bx > 0 {
                    frame.qp_map[(by * frame.deblock_stride + bx - 1) as usize] as i32
                } else {
                    qp_q
                };

                let cx = x / sub_x;
                let cy = y / sub_y;

                for c_idx in 0..2 {
                    let qp_offset = if c_idx == 0 {
                        cb_qp_offset
                    } else {
                        cr_qp_offset
                    };
                    let qp_i = ((qp_q + qp_p + 1) >> 1) + qp_offset;
                    let qp_c = chroma_qp_mapping(qp_i);
                    let q_tc = (qp_c + 2 + tc_offset).clamp(0, 53);
                    let tc = (TC_PRIME[q_tc as usize] as i32) << (bit_depth_c - 8);

                    if tc == 0 {
                        continue;
                    }

                    let plane = if c_idx == 0 {
                        &mut frame.cb_plane
                    } else {
                        &mut frame.cr_plane
                    };

                    // Process 4 chroma samples along the edge
                    let num_samples = 4u32.min(c_height.saturating_sub(cy));
                    for k in 0..num_samples {
                        let row = (cy + k) as usize;
                        if cx < 2 || cx as usize >= c_stride || row >= plane.len() / c_stride {
                            continue;
                        }
                        let base = row * c_stride;
                        let ci = cx as usize;
                        if ci + 1 >= c_stride {
                            continue;
                        }
                        let p1 = plane[base + ci - 2] as i32;
                        let p0 = plane[base + ci - 1] as i32;
                        let q0 = plane[base + ci] as i32;
                        let q1 = plane[base + ci + 1] as i32;

                        let delta = (((q0 - p0) * 4 + p1 - q1 + 4) >> 3).clamp(-tc, tc);
                        plane[base + ci - 1] = (p0 + delta).clamp(0, max_val) as u16;
                        plane[base + ci] = (q0 - delta).clamp(0, max_val) as u16;
                    }
                }
            }
            y += y_step_vert;
        }
        x += x_step_vert;
    }

    // Pass 2: Horizontal edges
    let mut y = y_step_horiz;
    while y < height {
        let mut x = 0u32;
        while x < width {
            let bx = x / 4;
            let by = y / 4;
            let idx = (by * frame.deblock_stride + bx) as usize;
            if idx < frame.deblock_flags.len() && (frame.deblock_flags[idx] & horiz_edge_mask) != 0
            {
                let is_tb_edge = (frame.deblock_flags[idx] & DEBLOCK_FLAG_HORIZ) != 0;
                let bs = compute_bs(x, y, false, is_tb_edge, inter_ctx);
                if bs < 2 {
                    x += x_step_horiz;
                    continue;
                }

                let qp_q = frame.qp_map[idx] as i32;
                let qp_p = if by > 0 {
                    frame.qp_map[((by - 1) * frame.deblock_stride + bx) as usize] as i32
                } else {
                    qp_q
                };

                let cx = x / sub_x;
                let cy = y / sub_y;

                for c_idx in 0..2 {
                    let qp_offset = if c_idx == 0 {
                        cb_qp_offset
                    } else {
                        cr_qp_offset
                    };
                    let qp_i = ((qp_q + qp_p + 1) >> 1) + qp_offset;
                    let qp_c = chroma_qp_mapping(qp_i);
                    let q_tc = (qp_c + 2 + tc_offset).clamp(0, 53);
                    let tc = (TC_PRIME[q_tc as usize] as i32) << (bit_depth_c - 8);

                    if tc == 0 {
                        continue;
                    }

                    let plane = if c_idx == 0 {
                        &mut frame.cb_plane
                    } else {
                        &mut frame.cr_plane
                    };

                    // Process 4 chroma samples along the edge
                    let num_samples = 4u32.min(c_width.saturating_sub(cx));
                    for k in 0..num_samples {
                        let col = (cx + k) as usize;
                        if cy < 2 || col >= c_stride {
                            continue;
                        }
                        let row_q = cy as usize;
                        let row_p = row_q - 1;
                        if row_q + 1 >= plane.len() / c_stride || row_p < 1 {
                            continue;
                        }

                        let p1 = plane[(row_p - 1) * c_stride + col] as i32;
                        let p0 = plane[row_p * c_stride + col] as i32;
                        let q0 = plane[row_q * c_stride + col] as i32;
                        let q1 = plane[(row_q + 1) * c_stride + col] as i32;

                        let delta = (((q0 - p0) * 4 + p1 - q1 + 4) >> 3).clamp(-tc, tc);
                        plane[row_p * c_stride + col] = (p0 + delta).clamp(0, max_val) as u16;
                        plane[row_q * c_stride + col] = (q0 - delta).clamp(0, max_val) as u16;
                    }
                }
            }
            x += x_step_horiz;
        }
        y += y_step_horiz;
    }
}