// sha3.c
// 19-Nov-11  Markku-Juhani O. Saarinen <mjos@iki.fi>
// 22-May-25  Kris Kwiatkowski <kris@amongbytes.com>

// Revised 07-Aug-15 to match with official release of FIPS PUB 202 "SHA3"
// Revised 03-Sep-15 for portability + OpenSSL - style API
// Revised 22-May-25 Added bit-interleaved implementation optimized for 32-bit architectures.

#include "sha3.h"

// Interleave even and odd bits into one 64-bit line
uint64_t unshuffle(uint32_t even, uint32_t odd) {
    uint64_t result = 0;
    for (int i = 0; i < 32; i++) {
        result |= ((uint64_t)(even >> i) & 1) << (2 * i);
        result |= ((uint64_t)(odd  >> i) & 1) << (2 * i + 1);
    }
    return result;
}

uint32_t shuffle_even(uint64_t x) {
    x &= 0x5555555555555555ULL;
    x = (x | (x >> 1)) & 0x3333333333333333ULL;
    x = (x | (x >> 2)) & 0x0F0F0F0F0F0F0F0FULL;
    x = (x | (x >> 4)) & 0x00FF00FF00FF00FFULL;
    x = (x | (x >> 8)) & 0x0000FFFF0000FFFFULL;
    x = (x | (x >> 16)) & 0x00000000FFFFFFFFULL;
    return (uint32_t)x;
}

uint32_t shuffle_odd(uint64_t x) {
    return shuffle_even(x >> 1);
}

void sha3_keccakf(uint64_t st[25])
{
    // constants
    const uint64_t keccakf_rndc[24] = {
        0x0000000000000001, 0x0000000000008082, 0x800000000000808a,
        0x8000000080008000, 0x000000000000808b, 0x0000000080000001,
        0x8000000080008081, 0x8000000000008009, 0x000000000000008a,
        0x0000000000000088, 0x0000000080008009, 0x000000008000000a,
        0x000000008000808b, 0x800000000000008b, 0x8000000000008089,
        0x8000000000008003, 0x8000000000008002, 0x8000000000000080,
        0x000000000000800a, 0x800000008000000a, 0x8000000080008081,
        0x8000000000008080, 0x0000000080000001, 0x8000000080008008
    };
    const int keccakf_rotc[24] = {
        1,  3,  6,  10, 15, 21, 28, 36, 45, 55, 2,  14,
        27, 41, 56, 8,  25, 43, 62, 18, 39, 61, 20, 44
    };
    const int keccakf_piln[24] = {
        10, 7,  11, 17, 18, 3, 5,  16, 8,  21, 24, 4,
        15, 23, 19, 13, 12, 2, 20, 14, 22, 9,  6,  1
    };

    // variables
    int i, j, r;
    uint32_t t1, t2;
    uint32_t even[25], odd[25];
    uint32_t bc_even[5], bc_odd[5];

#if __BYTE_ORDER__ != __ORDER_LITTLE_ENDIAN__
    uint8_t *v;

    // endianess conversion. this is redundant on little-endian targets
    for (i = 0; i < 25; i++) {
        v = (uint8_t *) &st[i];
        st[i] = ((uint64_t) v[0])     | (((uint64_t) v[1]) << 8) |
            (((uint64_t) v[2]) << 16) | (((uint64_t) v[3]) << 24) |
            (((uint64_t) v[4]) << 32) | (((uint64_t) v[5]) << 40) |
            (((uint64_t) v[6]) << 48) | (((uint64_t) v[7]) << 56);
    }
#endif

    for (i = 0; i < 25; i++) {
        even[i] = shuffle_even(st[i]);
        odd[i] = shuffle_odd(st[i]);;
    }

    // actual iteration
    for (r = 0; r < KECCAKF_ROUNDS; r++) {
        // Theta
        for (i = 0; i < 5; i++) {
            bc_even[i] = even[i] ^ even[i + 5] ^ even[i + 10] ^ even[i + 15] ^ even[i + 20];
            bc_odd[i] = odd[i] ^ odd[i + 5] ^ odd[i + 10] ^ odd[i + 15] ^ odd[i + 20];
        }

        // Chi
        for (i = 0; i < 5; i++) {

            uint32_t rot32 = ROTL32(bc_odd[(i + 1) % 5], 1);
            t1 = bc_even[(i + 4) % 5] ^ rot32;
            t2 = bc_odd[(i + 4) % 5] ^ bc_even[(i + 1) % 5];

            for (j = 0; j < 25; j += 5) {
                even[j + i] ^= t1;
                odd[j + i] ^= t2;
            }
        }

        // Rho Pi
        t1 = even[1]; t2 = odd[1];
        for (i = 0; i < 24; i++) {
            j = keccakf_piln[i];
            bc_even[0] = even[j]; bc_odd[0] = odd[j];

            int half = keccakf_rotc[i] >> 1;
            if (keccakf_rotc[i]&1) {
                // U0 = ROT32(U1, tau)
                odd[j] = ROTL32(t1, half);
                // U1 = ROT32(U0, tau + 1)
                even[j] = ROTL32(t2, half + 1);
            } else {
                // U0 = ROT32(U0, tau)
                odd[j] = ROTL32(t2, half);
                // U1 = ROT32(U1, tau)
                even[j] = ROTL32(t1, half);
            }

            t1 = bc_even[0]; t2 = bc_odd[0];
        }

        //  Chi
        for (j = 0; j < 25; j += 5) {
            for (i = 0; i < 5; i++) {
                bc_even[i] = even[j + i];
                bc_odd[i] = odd[j + i];
            }
            for (i = 0; i < 5; i++) {
                even[j + i] ^= (~bc_even[(i + 1) % 5]) & bc_even[(i + 2) % 5];
                odd[j + i] ^= (~bc_odd[(i + 1) % 5]) & bc_odd[(i + 2) % 5];
            }
        }

        //  Iota (can be precomputed)
        even[0] ^= shuffle_even(keccakf_rndc[r]);
        odd[0] ^= shuffle_odd(keccakf_rndc[r]);
    }

    for (i = 0; i < 25; i++) {
        st[i] = unshuffle(even[i], odd[i]);
    }

#if __BYTE_ORDER__ != __ORDER_LITTLE_ENDIAN__
    // endianess conversion. this is redundant on little-endian targets
    for (i = 0; i < 25; i++) {
        v = (uint8_t *) &st[i];
        t = st[i];
        v[0] = t & 0xFF;
        v[1] = (t >> 8) & 0xFF;
        v[2] = (t >> 16) & 0xFF;
        v[3] = (t >> 24) & 0xFF;
        v[4] = (t >> 32) & 0xFF;
        v[5] = (t >> 40) & 0xFF;
        v[6] = (t >> 48) & 0xFF;
        v[7] = (t >> 56) & 0xFF;
    }
#endif
}

// Initialize the context for SHA3

int sha3_init(sha3_ctx_t *c, int mdlen)
{
    int i;

    for (i = 0; i < 25; i++)
        c->st.q[i] = 0;
    c->mdlen = mdlen;
    c->rsiz = 200 - 2 * mdlen;
    c->pt = 0;

    return 1;
}

// update state with more data

int sha3_update(sha3_ctx_t *c, const void *data, size_t len)
{
    size_t i;
    int j;

    j = c->pt;
    for (i = 0; i < len; i++) {
        c->st.b[j++] ^= ((const uint8_t *) data)[i];
        if (j >= c->rsiz) {
            sha3_keccakf(c->st.q);
            j = 0;
        }
    }
    c->pt = j;

    return 1;
}

// finalize and output a hash

int sha3_final(void *md, sha3_ctx_t *c)
{
    int i;

    c->st.b[c->pt] ^= 0x06;
    c->st.b[c->rsiz - 1] ^= 0x80;
    sha3_keccakf(c->st.q);

    for (i = 0; i < c->mdlen; i++) {
        ((uint8_t *) md)[i] = c->st.b[i];
    }

    return 1;
}

// compute a SHA-3 hash (md) of given byte length from "in"

void *sha3(const void *in, size_t inlen, void *md, int mdlen)
{
    sha3_ctx_t sha3;

    sha3_init(&sha3, mdlen);
    sha3_update(&sha3, in, inlen);
    sha3_final(md, &sha3);

    return md;
}

// SHAKE128 and SHAKE256 extensible-output functionality

void shake_xof(sha3_ctx_t *c)
{
    c->st.b[c->pt] ^= 0x1F;
    c->st.b[c->rsiz - 1] ^= 0x80;
    sha3_keccakf(c->st.q);
    c->pt = 0;
}

void shake_out(sha3_ctx_t *c, void *out, size_t len)
{
    size_t i;
    int j;

    j = c->pt;
    for (i = 0; i < len; i++) {
        if (j >= c->rsiz) {
            sha3_keccakf(c->st.q);
            j = 0;
        }
        ((uint8_t *) out)[i] = c->st.b[j++];
    }
    c->pt = j;
}