zenpng 0.1.0

PNG encoding and decoding with zencodec trait integration
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218

//! Average filter: `row[i] += floor((left + above) / 2)`
//!
//! Sequential dependency per pixel. For bpp=4, widen left+above to u16,
//! average, narrow back, add. One pixel (4 bytes) per SIMD iteration.

use archmage::prelude::*;
#[cfg(target_arch = "wasm32")]
use safe_unaligned_simd::wasm32::v128_load32_zero;
#[cfg(target_arch = "x86_64")]
use safe_unaligned_simd::x86_64::{_mm_loadu_si32, _mm_storeu_si32};

pub(crate) fn unfilter_avg(row: &mut [u8], prev: &[u8], bpp: usize) {
    match bpp {
        4 => incant!(
            unfilter_avg_bpp4_impl(row, prev),
            [v1, neon, wasm128, scalar]
        ),
        _ => unfilter_avg_scalar_any(row, prev, bpp),
    }
}

// ── Scalar implementation ────────────────────────────────────────────

fn unfilter_avg_scalar_any(row: &mut [u8], prev: &[u8], bpp: usize) {
    let len = row.len();
    for i in 0..bpp.min(len) {
        row[i] = row[i].wrapping_add(prev[i] >> 1);
    }
    for i in bpp..len {
        let avg = ((row[i - bpp] as u16 + prev[i] as u16) >> 1) as u8;
        row[i] = row[i].wrapping_add(avg);
    }
}

// ── SIMD bpp=4 (SSE2 / V1) ──────────────────────────────────────────

#[cfg(target_arch = "x86_64")]
#[arcane]
fn unfilter_avg_bpp4_impl_v1(_token: Sse2Token, row: &mut [u8], prev: &[u8]) {
    let len = row.len();
    if len < 4 {
        return;
    }

    let zero = _mm_setzero_si128();
    let mut a_wide = zero; // left pixel widened to u16

    let mut i = 0;
    while i + 4 <= len {
        // b = above pixel, widened to u16
        let b_raw = _mm_loadu_si32(<&[u8; 4]>::try_from(&prev[i..i + 4]).unwrap());
        let b_wide = _mm_unpacklo_epi8(b_raw, zero);

        // avg = (a + b) >> 1  (u16 arithmetic, no overflow: max 255+255=510)
        let sum = _mm_add_epi16(a_wide, b_wide);
        let avg_wide = _mm_srli_epi16(sum, 1);

        // Narrow avg to u8 (values 0-254, packus won't clamp)
        let avg_narrow = _mm_packus_epi16(avg_wide, zero);

        // Load filtered bytes and add average (wrapping u8 add)
        let filt = _mm_loadu_si32(<&[u8; 4]>::try_from(&row[i..i + 4]).unwrap());
        let result = _mm_add_epi8(filt, avg_narrow);

        // Store 4-byte result
        _mm_storeu_si32(
            <&mut [u8; 4]>::try_from(&mut row[i..i + 4]).unwrap(),
            result,
        );

        // Feedback: a = result widened
        a_wide = _mm_unpacklo_epi8(result, zero);

        i += 4;
    }
}

// ── NEON bpp=4 (aarch64) ─────────────────────────────────────────────

#[cfg(target_arch = "aarch64")]
#[arcane]
fn unfilter_avg_bpp4_impl_neon(_token: NeonToken, row: &mut [u8], prev: &[u8]) {
    let len = row.len();
    if len < 4 {
        return;
    }

    // a_wide = left pixel widened to u16 (starts as zero)
    let mut a_wide = vdup_n_u16(0);

    let mut i = 0;
    while i + 4 <= len {
        // b = above pixel, widened to u16
        let b_bytes = u32::from_le_bytes(<[u8; 4]>::try_from(&prev[i..i + 4]).unwrap());
        let b_raw = vcreate_u8(b_bytes as u64);
        let b_wide = vget_low_u16(vmovl_u8(b_raw));

        // avg = (a + b) >> 1  (u16 arithmetic, no overflow: max 255+255=510)
        let sum = vadd_u16(a_wide, b_wide);
        let avg_wide = vshr_n_u16::<1>(sum);

        // Narrow avg to u8 (values 0-254, vmovn won't saturate)
        let avg_narrow = vmovn_u16(vcombine_u16(avg_wide, vdup_n_u16(0)));

        // Load filtered bytes and add average (wrapping u8 add)
        let filt_bytes = u32::from_le_bytes(<[u8; 4]>::try_from(&row[i..i + 4]).unwrap());
        let filt = vcreate_u8(filt_bytes as u64);
        let result = vadd_u8(filt, avg_narrow);

        // Store 4-byte result
        let result_u32 = vget_lane_u32::<0>(vreinterpret_u32_u8(result));
        row[i..i + 4].copy_from_slice(&result_u32.to_le_bytes());

        // Feedback: a = result widened
        a_wide = vget_low_u16(vmovl_u8(result));

        i += 4;
    }
}

// ── WASM SIMD128 bpp=4 ──────────────────────────────────────────────

#[cfg(target_arch = "wasm32")]
#[arcane]
fn unfilter_avg_bpp4_impl_wasm128(_token: Wasm128Token, row: &mut [u8], prev: &[u8]) {
    let len = row.len();
    if len < 4 {
        return;
    }

    let zero = i16x8_splat(0);
    let mut a_wide = zero; // left pixel widened to u16

    let mut i = 0;
    while i + 4 <= len {
        // b = above pixel, widened to u16
        let b_raw = v128_load32_zero(<&[u8; 4]>::try_from(&prev[i..i + 4]).unwrap());
        let b_wide = i16x8_extend_low_u8x16(b_raw);

        // avg = (a + b) >> 1  (u16 arithmetic, no overflow: max 255+255=510)
        let sum = i16x8_add(a_wide, b_wide);
        let avg_wide = u16x8_shr(sum, 1);

        // Narrow avg to u8 (values 0-254, saturating narrow won't clamp)
        let avg_narrow = u8x16_narrow_i16x8(avg_wide, zero);

        // Load filtered bytes and add average (wrapping u8 add)
        let filt = v128_load32_zero(<&[u8; 4]>::try_from(&row[i..i + 4]).unwrap());
        let result = i8x16_add(filt, avg_narrow);

        // Store 4-byte result
        let val = (i32x4_extract_lane::<0>(result) as u32).to_le_bytes();
        row[i..i + 4].copy_from_slice(&val);

        // Feedback: a = result widened
        a_wide = i16x8_extend_low_u8x16(result);

        i += 4;
    }
}

// Scalar fallback for incant! dispatch
fn unfilter_avg_bpp4_impl_scalar(_token: ScalarToken, row: &mut [u8], prev: &[u8]) {
    unfilter_avg_scalar_any(row, prev, 4);
}

#[cfg(test)]
mod tests {
    use archmage::testing::{CompileTimePolicy, for_each_token_permutation};

    fn scalar_avg(row: &mut [u8], prev: &[u8], bpp: usize) {
        let len = row.len();
        for i in 0..bpp.min(len) {
            row[i] = row[i].wrapping_add(prev[i] >> 1);
        }
        for i in bpp..len {
            let avg = ((row[i - bpp] as u16 + prev[i] as u16) >> 1) as u8;
            row[i] = row[i].wrapping_add(avg);
        }
    }

    #[test]
    fn avg_bpp4_all_tiers() {
        let report = for_each_token_permutation(CompileTimePolicy::Warn, |_perm| {
            for &len in &[0, 4, 8, 12, 100, 4096, 65536] {
                let prev: Vec<u8> = (0..len).map(|i| (i * 7 + 13) as u8).collect();
                let filtered: Vec<u8> = (0..len).map(|i| (i * 3 + 5) as u8).collect();

                let mut expected = filtered.clone();
                scalar_avg(&mut expected, &prev, 4);

                let mut actual = filtered.clone();
                super::unfilter_avg(&mut actual, &prev, 4);

                assert_eq!(actual, expected, "avg bpp=4 mismatch at len={len}");
            }
        });
        eprintln!("avg bpp=4: {report}");
    }

    #[test]
    fn avg_other_bpp_unchanged() {
        for &bpp in &[1, 2, 3, 6, 8] {
            for &len in &[0, bpp, bpp * 4, bpp * 100] {
                let prev: Vec<u8> = (0..len).map(|i| (i * 11 + 3) as u8).collect();
                let filtered: Vec<u8> = (0..len).map(|i| (i * 5 + 7) as u8).collect();

                let mut expected = filtered.clone();
                scalar_avg(&mut expected, &prev, bpp);

                let mut actual = filtered.clone();
                super::unfilter_avg(&mut actual, &prev, bpp);

                assert_eq!(actual, expected, "avg bpp={bpp} mismatch at len={len}");
            }
        }
    }
}