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

use hashbrown::HashMap;
use itertools::Itertools;
use p3_field::PrimeField32;
use p3_matrix::dense::RowMajorMatrix;
use p3_matrix::Matrix;
use p3_maybe_rayon::prelude::{ParallelIterator, ParallelSlice};
use std::borrow::BorrowMut;

use crate::{
    air::MachineAir,
    bytes::{event::ByteRecord, ByteLookupEvent},
    runtime::{ExecutionRecord, Program},
};

use super::{ShaExtendChip, ShaExtendCols, ShaExtendEvent, NUM_SHA_EXTEND_COLS};

impl<F: PrimeField32> MachineAir<F> for ShaExtendChip {
    type Record = ExecutionRecord;

    type Program = Program;

    fn name(&self) -> String {
        "ShaExtend".to_string()
    }

    fn generate_trace(
        &self,
        input: &ExecutionRecord,
        _: &mut ExecutionRecord,
    ) -> RowMajorMatrix<F> {
        let rows = Vec::new();

        let mut new_byte_lookup_events = Vec::new();
        let mut wrapped_rows = Some(rows);
        for i in 0..input.sha_extend_events.len() {
            self.event_to_rows(
                &input.sha_extend_events[i],
                &mut wrapped_rows,
                &mut new_byte_lookup_events,
            );
        }

        let mut rows = wrapped_rows.unwrap();
        let nb_rows = rows.len();
        let mut padded_nb_rows = nb_rows.next_power_of_two();
        if padded_nb_rows == 2 || padded_nb_rows == 1 {
            padded_nb_rows = 4;
        }
        for i in nb_rows..padded_nb_rows {
            let mut row = [F::zero(); NUM_SHA_EXTEND_COLS];
            let cols: &mut ShaExtendCols<F> = row.as_mut_slice().borrow_mut();
            cols.populate_flags(i);
            rows.push(row);
        }

        // Convert the trace to a row major matrix.
        let mut trace = RowMajorMatrix::new(
            rows.into_iter().flatten().collect::<Vec<_>>(),
            NUM_SHA_EXTEND_COLS,
        );

        // Write the nonces to the trace.
        for i in 0..trace.height() {
            let cols: &mut ShaExtendCols<F> =
                trace.values[i * NUM_SHA_EXTEND_COLS..(i + 1) * NUM_SHA_EXTEND_COLS].borrow_mut();
            cols.nonce = F::from_canonical_usize(i);
        }

        trace
    }

    fn generate_dependencies(&self, input: &Self::Record, output: &mut Self::Record) {
        let chunk_size = std::cmp::max(input.sha_extend_events.len() / num_cpus::get(), 1);

        let blu_batches = input
            .sha_extend_events
            .par_chunks(chunk_size)
            .map(|events| {
                let mut blu: HashMap<u32, HashMap<ByteLookupEvent, usize>> = HashMap::new();
                events.iter().for_each(|event| {
                    self.event_to_rows::<F>(event, &mut None, &mut blu);
                });
                blu
            })
            .collect::<Vec<_>>();

        output.add_sharded_byte_lookup_events(blu_batches.iter().collect_vec());
    }

    fn included(&self, shard: &Self::Record) -> bool {
        !shard.sha_extend_events.is_empty()
    }
}

impl ShaExtendChip {
    fn event_to_rows<F: PrimeField32>(
        &self,
        event: &ShaExtendEvent,
        rows: &mut Option<Vec<[F; NUM_SHA_EXTEND_COLS]>>,
        blu: &mut impl ByteRecord,
    ) {
        let shard = event.shard;
        for j in 0..48usize {
            let mut row = [F::zero(); NUM_SHA_EXTEND_COLS];
            let cols: &mut ShaExtendCols<F> = row.as_mut_slice().borrow_mut();
            cols.is_real = F::one();
            cols.populate_flags(j);
            cols.shard = F::from_canonical_u32(event.shard);
            cols.channel = F::from_canonical_u8(event.channel);
            cols.clk = F::from_canonical_u32(event.clk);
            cols.w_ptr = F::from_canonical_u32(event.w_ptr);

            cols.w_i_minus_15
                .populate(event.channel, event.w_i_minus_15_reads[j], blu);
            cols.w_i_minus_2
                .populate(event.channel, event.w_i_minus_2_reads[j], blu);
            cols.w_i_minus_16
                .populate(event.channel, event.w_i_minus_16_reads[j], blu);
            cols.w_i_minus_7
                .populate(event.channel, event.w_i_minus_7_reads[j], blu);

            // `s0 := (w[i-15] rightrotate 7) xor (w[i-15] rightrotate 18) xor (w[i-15] rightshift 3)`.
            let w_i_minus_15 = event.w_i_minus_15_reads[j].value;
            let w_i_minus_15_rr_7 =
                cols.w_i_minus_15_rr_7
                    .populate(blu, shard, event.channel, w_i_minus_15, 7);
            let w_i_minus_15_rr_18 =
                cols.w_i_minus_15_rr_18
                    .populate(blu, shard, event.channel, w_i_minus_15, 18);
            let w_i_minus_15_rs_3 =
                cols.w_i_minus_15_rs_3
                    .populate(blu, shard, event.channel, w_i_minus_15, 3);
            let s0_intermediate = cols.s0_intermediate.populate(
                blu,
                shard,
                event.channel,
                w_i_minus_15_rr_7,
                w_i_minus_15_rr_18,
            );
            let s0 = cols.s0.populate(
                blu,
                shard,
                event.channel,
                s0_intermediate,
                w_i_minus_15_rs_3,
            );

            // `s1 := (w[i-2] rightrotate 17) xor (w[i-2] rightrotate 19) xor (w[i-2] rightshift 10)`.
            let w_i_minus_2 = event.w_i_minus_2_reads[j].value;
            let w_i_minus_2_rr_17 =
                cols.w_i_minus_2_rr_17
                    .populate(blu, shard, event.channel, w_i_minus_2, 17);
            let w_i_minus_2_rr_19 =
                cols.w_i_minus_2_rr_19
                    .populate(blu, shard, event.channel, w_i_minus_2, 19);
            let w_i_minus_2_rs_10 =
                cols.w_i_minus_2_rs_10
                    .populate(blu, shard, event.channel, w_i_minus_2, 10);
            let s1_intermediate = cols.s1_intermediate.populate(
                blu,
                shard,
                event.channel,
                w_i_minus_2_rr_17,
                w_i_minus_2_rr_19,
            );
            let s1 = cols.s1.populate(
                blu,
                shard,
                event.channel,
                s1_intermediate,
                w_i_minus_2_rs_10,
            );

            // Compute `s2`.
            let w_i_minus_7 = event.w_i_minus_7_reads[j].value;
            let w_i_minus_16 = event.w_i_minus_16_reads[j].value;
            cols.s2
                .populate(blu, shard, event.channel, w_i_minus_16, s0, w_i_minus_7, s1);

            cols.w_i.populate(event.channel, event.w_i_writes[j], blu);

            if rows.as_ref().is_some() {
                rows.as_mut().unwrap().push(row);
            }
        }
    }
}