ic-ec 0.3.2

//! P-256 point formulas, each one `asm!` block over `mulx`/`adcx`/`adox`.
//!
//! GENERATED by `scripts/gen_p256_x86_point.py`; edit that, not this.
//!
//! The field operations are those of [`super::p256_x86`] -- the same
//! multiplication, squaring, addition and subtraction, instruction for
//! instruction -- strung together over a buffer of 32-byte slots instead of
//! called one at a time. What that removes is everything between them: the
//! per-operation feature check, and moving each result out of registers into
//! a Rust value and back into memory for the next. Each formula computes the
//! same field values as its Rust counterpart in `point.rs` -- the doubling's
//! small multiples fused into three operations, the operations in an order
//! the core can overlap -- so the results are equal bit for bit; the tests
//! hold them to that. The doubling is also emitted in place, for the scalar
//! multiplication's loop.
//!
//! Every operation reads its operands before writing its result, so a
//! destination may also be a source. Nothing branches and nothing is indexed
//! by data: the reductions end in `cmovc`, the subtractions in a borrow mask.

use core::arch::asm;
use core::mem::MaybeUninit;

/// Point doubling, `dbl-2001-b` for `a = -3`, as `Point::double`.
///
/// Slots, 32 bytes each, in order: `X`, `Y`, `Z`, `X3`, `Y3`, `Z3`, `delta`, `gamma`, `beta`, `m`, `u`, `m2`, `d`, `w`, `g2`.
///
/// # Safety
///
/// The CPU must support BMI2 and ADX; see `super::p256_x86::has_adx`.
#[inline(never)]
pub unsafe fn double_adx(buf: &mut [MaybeUninit<u64>; 60]) {
    // SAFETY: the caller guarantees the instructions exist. The block reads
    // and writes only the buffer behind `{s}`, a live, exclusive reference
    // of 480 bytes, at offsets below that; it writes only the
    // registers it declares, and does not touch the stack. It reads no slot
    // before writing it but the inputs, which the caller has written.
    unsafe {
        asm!(
            ".p2align 6",
            // delta = sqr(Z)
            "mov rdx, qword ptr [{s} + 64]",
            "mulx {t2}, {t1}, qword ptr [{s} + 72]",
            "mulx {t3}, rax, qword ptr [{s} + 80]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 88]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 72]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 80]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 88]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 80]",
            "mulx {t6}, rax, qword ptr [{s} + 88]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 64]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 72]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 80]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 88]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 192], rax",
            "mov qword ptr [{s} + 200], rdx",
            "mov qword ptr [{s} + 208], {t2}",
            "mov qword ptr [{s} + 216], {t3}",
            // gamma = sqr(Y)
            "mov rdx, qword ptr [{s} + 32]",
            "mulx {t2}, {t1}, qword ptr [{s} + 40]",
            "mulx {t3}, rax, qword ptr [{s} + 48]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 56]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 40]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 48]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 56]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 48]",
            "mulx {t6}, rax, qword ptr [{s} + 56]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 32]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 40]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 48]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 56]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 224], rax",
            "mov qword ptr [{s} + 232], rdx",
            "mov qword ptr [{s} + 240], {t2}",
            "mov qword ptr [{s} + 248], {t3}",
            // m = sub(X, delta)
            "mov {t0}, qword ptr [{s}]",
            "mov {t1}, qword ptr [{s} + 8]",
            "mov {t2}, qword ptr [{s} + 16]",
            "mov {t3}, qword ptr [{s} + 24]",
            "sub {t0}, qword ptr [{s} + 192]",
            "sbb {t1}, qword ptr [{s} + 200]",
            "sbb {t2}, qword ptr [{s} + 208]",
            "sbb {t3}, qword ptr [{s} + 216]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 288], {t0}",
            "mov qword ptr [{s} + 296], {t1}",
            "mov qword ptr [{s} + 304], {t2}",
            "mov qword ptr [{s} + 312], {t3}",
            // u = add(X, delta)
            "mov {t0}, qword ptr [{s}]",
            "mov {t1}, qword ptr [{s} + 8]",
            "mov {t2}, qword ptr [{s} + 16]",
            "mov {t3}, qword ptr [{s} + 24]",
            "xor {c:e}, {c:e}",
            "add {t0}, qword ptr [{s} + 192]",
            "adc {t1}, qword ptr [{s} + 200]",
            "adc {t2}, qword ptr [{s} + 208]",
            "adc {t3}, qword ptr [{s} + 216]",
            "adc {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 320], {t4}",
            "mov qword ptr [{s} + 328], {t5}",
            "mov qword ptr [{s} + 336], {t6}",
            "mov qword ptr [{s} + 344], {t7}",
            // m = mul(m, u)
            "mov rdx, qword ptr [{s} + 320]",
            "mulx {t1}, {t0}, qword ptr [{s} + 288]",
            "mulx {t2}, rax, qword ptr [{s} + 296]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 304]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 312]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 328]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 288]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 296]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 304]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 312]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 336]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 288]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 296]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 304]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 312]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 344]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 288]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 296]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 304]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 312]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 288], rax",
            "mov qword ptr [{s} + 296], rdx",
            "mov qword ptr [{s} + 304], {c}",
            "mov qword ptr [{s} + 312], {t3}",
            // beta = mul(X, gamma)
            "mov rdx, qword ptr [{s} + 224]",
            "mulx {t1}, {t0}, qword ptr [{s}]",
            "mulx {t2}, rax, qword ptr [{s} + 8]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 16]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 24]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 232]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s}]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 8]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 16]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 24]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 240]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s}]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 8]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 16]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 24]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 248]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s}]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 8]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 16]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 24]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 256], rax",
            "mov qword ptr [{s} + 264], rdx",
            "mov qword ptr [{s} + 272], {c}",
            "mov qword ptr [{s} + 280], {t3}",
            // m2 = sqr(m)
            "mov rdx, qword ptr [{s} + 288]",
            "mulx {t2}, {t1}, qword ptr [{s} + 296]",
            "mulx {t3}, rax, qword ptr [{s} + 304]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 312]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 296]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 304]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 312]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 304]",
            "mulx {t6}, rax, qword ptr [{s} + 312]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 288]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 296]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 304]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 312]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 352], rax",
            "mov qword ptr [{s} + 360], rdx",
            "mov qword ptr [{s} + 368], {t2}",
            "mov qword ptr [{s} + 376], {t3}",
            // w = add(Y, Z)
            "mov {t0}, qword ptr [{s} + 32]",
            "mov {t1}, qword ptr [{s} + 40]",
            "mov {t2}, qword ptr [{s} + 48]",
            "mov {t3}, qword ptr [{s} + 56]",
            "xor {c:e}, {c:e}",
            "add {t0}, qword ptr [{s} + 64]",
            "adc {t1}, qword ptr [{s} + 72]",
            "adc {t2}, qword ptr [{s} + 80]",
            "adc {t3}, qword ptr [{s} + 88]",
            "adc {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 416], {t4}",
            "mov qword ptr [{s} + 424], {t5}",
            "mov qword ptr [{s} + 432], {t6}",
            "mov qword ptr [{s} + 440], {t7}",
            // w = sqr(w)
            "mov rdx, qword ptr [{s} + 416]",
            "mulx {t2}, {t1}, qword ptr [{s} + 424]",
            "mulx {t3}, rax, qword ptr [{s} + 432]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 440]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 424]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 432]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 440]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 432]",
            "mulx {t6}, rax, qword ptr [{s} + 440]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 416]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 424]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 432]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 440]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 416], rax",
            "mov qword ptr [{s} + 424], rdx",
            "mov qword ptr [{s} + 432], {t2}",
            "mov qword ptr [{s} + 440], {t3}",
            // d = cms(beta, m2, 12, 9)
            "mov {t0}, -1",
            "sub {t0}, qword ptr [{s} + 352]",
            "mov {t1:e}, 0xffffffff",
            "sbb {t1}, qword ptr [{s} + 360]",
            "mov {t2:e}, 0",
            "sbb {t2}, qword ptr [{s} + 368]",
            "mov {t3}, 0xffffffff00000001",
            "sbb {t3}, qword ptr [{s} + 376]",
            "mov edx, 9",
            "mulx {c}, {t0}, {t0}",
            "mulx {t5}, {t1}, {t1}",
            "add {t1}, {c}",
            "mulx {c}, {t2}, {t2}",
            "adc {t2}, {t5}",
            "mulx {t4}, {t3}, {t3}",
            "adc {t3}, {c}",
            "adc {t4}, 0",
            "mov edx, 12",
            "xor {t6:e}, {t6:e}",
            "mulx {c}, rax, qword ptr [{s} + 256]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 264]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 272]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx rdx, rax, qword ptr [{s} + 280]",
            "adcx {t3}, rax",
            "adox rdx, {t4}",
            "adcx rdx, {t6}",
            "add rdx, 1",
            "mov rax, rdx",
            "shl rax, 32",
            "mov {c}, rax",
            "neg rax",
            "sub {c}, rdx",
            "sub {c}, 1",
            "add {t0}, rdx",
            "adc {t1}, rax",
            "adc {t2}, -1",
            "adc {t3}, {c}",
            "cmc",
            "sbb rax, rax",
            "mov rdx, rax",
            "shr rdx, 32",
            "mov {c}, 0xffffffff00000001",
            "and {c}, rax",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, 0",
            "adc {t3}, {c}",
            "mov qword ptr [{s} + 384], {t0}",
            "mov qword ptr [{s} + 392], {t1}",
            "mov qword ptr [{s} + 400], {t2}",
            "mov qword ptr [{s} + 408], {t3}",
            // g2 = sqr(gamma)
            "mov rdx, qword ptr [{s} + 224]",
            "mulx {t2}, {t1}, qword ptr [{s} + 232]",
            "mulx {t3}, rax, qword ptr [{s} + 240]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 248]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 232]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 240]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 248]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 240]",
            "mulx {t6}, rax, qword ptr [{s} + 248]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 224]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 232]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 240]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 248]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 448], rax",
            "mov qword ptr [{s} + 456], rdx",
            "mov qword ptr [{s} + 464], {t2}",
            "mov qword ptr [{s} + 472], {t3}",
            // u = mul(d, m)
            "mov rdx, qword ptr [{s} + 288]",
            "mulx {t1}, {t0}, qword ptr [{s} + 384]",
            "mulx {t2}, rax, qword ptr [{s} + 392]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 400]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 408]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 296]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 384]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 392]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 400]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 408]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 304]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 384]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 392]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 400]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 408]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 312]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 384]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 392]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 400]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 408]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 320], rax",
            "mov qword ptr [{s} + 328], rdx",
            "mov qword ptr [{s} + 336], {c}",
            "mov qword ptr [{s} + 344], {t3}",
            // X3 = cms(beta, d, 4, 1)
            "xor {t4:e}, {t4:e}",
            "mov {t0}, -1",
            "sub {t0}, qword ptr [{s} + 384]",
            "mov {t1:e}, 0xffffffff",
            "sbb {t1}, qword ptr [{s} + 392]",
            "mov {t2:e}, 0",
            "sbb {t2}, qword ptr [{s} + 400]",
            "mov {t3}, 0xffffffff00000001",
            "sbb {t3}, qword ptr [{s} + 408]",
            "mov edx, 4",
            "xor {t6:e}, {t6:e}",
            "mulx {c}, rax, qword ptr [{s} + 256]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 264]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 272]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx rdx, rax, qword ptr [{s} + 280]",
            "adcx {t3}, rax",
            "adox rdx, {t4}",
            "adcx rdx, {t6}",
            "add rdx, 1",
            "mov rax, rdx",
            "shl rax, 32",
            "mov {c}, rax",
            "neg rax",
            "sub {c}, rdx",
            "sub {c}, 1",
            "add {t0}, rdx",
            "adc {t1}, rax",
            "adc {t2}, -1",
            "adc {t3}, {c}",
            "cmc",
            "sbb rax, rax",
            "mov rdx, rax",
            "shr rdx, 32",
            "mov {c}, 0xffffffff00000001",
            "and {c}, rax",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, 0",
            "adc {t3}, {c}",
            "mov qword ptr [{s} + 96], {t0}",
            "mov qword ptr [{s} + 104], {t1}",
            "mov qword ptr [{s} + 112], {t2}",
            "mov qword ptr [{s} + 120], {t3}",
            // w = sub(w, gamma)
            "mov {t0}, qword ptr [{s} + 416]",
            "mov {t1}, qword ptr [{s} + 424]",
            "mov {t2}, qword ptr [{s} + 432]",
            "mov {t3}, qword ptr [{s} + 440]",
            "sub {t0}, qword ptr [{s} + 224]",
            "sbb {t1}, qword ptr [{s} + 232]",
            "sbb {t2}, qword ptr [{s} + 240]",
            "sbb {t3}, qword ptr [{s} + 248]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 416], {t0}",
            "mov qword ptr [{s} + 424], {t1}",
            "mov qword ptr [{s} + 432], {t2}",
            "mov qword ptr [{s} + 440], {t3}",
            // Z3 = sub(w, delta)
            "mov {t0}, qword ptr [{s} + 416]",
            "mov {t1}, qword ptr [{s} + 424]",
            "mov {t2}, qword ptr [{s} + 432]",
            "mov {t3}, qword ptr [{s} + 440]",
            "sub {t0}, qword ptr [{s} + 192]",
            "sbb {t1}, qword ptr [{s} + 200]",
            "sbb {t2}, qword ptr [{s} + 208]",
            "sbb {t3}, qword ptr [{s} + 216]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 160], {t0}",
            "mov qword ptr [{s} + 168], {t1}",
            "mov qword ptr [{s} + 176], {t2}",
            "mov qword ptr [{s} + 184], {t3}",
            // Y3 = cms(u, g2, 3, 8)
            "mov {t0}, -1",
            "sub {t0}, qword ptr [{s} + 448]",
            "mov {t1:e}, 0xffffffff",
            "sbb {t1}, qword ptr [{s} + 456]",
            "mov {t2:e}, 0",
            "sbb {t2}, qword ptr [{s} + 464]",
            "mov {t3}, 0xffffffff00000001",
            "sbb {t3}, qword ptr [{s} + 472]",
            "mov {t4}, {t3}",
            "shr {t4}, 61",
            "shld {t3}, {t2}, 3",
            "shld {t2}, {t1}, 3",
            "shld {t1}, {t0}, 3",
            "shl {t0}, 3",
            "mov edx, 3",
            "xor {t6:e}, {t6:e}",
            "mulx {c}, rax, qword ptr [{s} + 320]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 328]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 336]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx rdx, rax, qword ptr [{s} + 344]",
            "adcx {t3}, rax",
            "adox rdx, {t4}",
            "adcx rdx, {t6}",
            "add rdx, 1",
            "mov rax, rdx",
            "shl rax, 32",
            "mov {c}, rax",
            "neg rax",
            "sub {c}, rdx",
            "sub {c}, 1",
            "add {t0}, rdx",
            "adc {t1}, rax",
            "adc {t2}, -1",
            "adc {t3}, {c}",
            "cmc",
            "sbb rax, rax",
            "mov rdx, rax",
            "shr rdx, 32",
            "mov {c}, 0xffffffff00000001",
            "and {c}, rax",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, 0",
            "adc {t3}, {c}",
            "mov qword ptr [{s} + 128], {t0}",
            "mov qword ptr [{s} + 136], {t1}",
            "mov qword ptr [{s} + 144], {t2}",
            "mov qword ptr [{s} + 152], {t3}",
            s = in(reg) buf.as_mut_ptr(),
            t0 = out(reg) _,
            t1 = out(reg) _,
            t2 = out(reg) _,
            t3 = out(reg) _,
            t4 = out(reg) _,
            t5 = out(reg) _,
            t6 = out(reg) _,
            t7 = out(reg) _,
            c = out(reg) _,
            out("rax") _,
            out("rdx") _,
            options(nostack),
        );
    }
}
#[allow(dead_code)]
pub const DOUBLE_X3: usize = 3;
#[allow(dead_code)]
pub const DOUBLE_Y3: usize = 4;
#[allow(dead_code)]
pub const DOUBLE_Z3: usize = 5;

/// `double_adx` with X3, Y3 and Z3 written over X, Y and Z: the scalar multiplication's doubling, on a buffer whose first slots are its accumulator.
///
/// Slots, 32 bytes each, in order: `X`, `Y`, `Z`, `delta`, `gamma`, `beta`, `m`, `u`, `m2`, `d`, `w`, `g2`.
///
/// # Safety
///
/// The CPU must support BMI2 and ADX; see `super::p256_x86::has_adx`.
#[inline(never)]
pub unsafe fn double_in_place_adx(buf: &mut [MaybeUninit<u64>; 48]) {
    // SAFETY: the caller guarantees the instructions exist. The block reads
    // and writes only the buffer behind `{s}`, a live, exclusive reference
    // of 384 bytes, at offsets below that; it writes only the
    // registers it declares, and does not touch the stack. It reads no slot
    // before writing it but the inputs, which the caller has written.
    unsafe {
        asm!(
            ".p2align 6",
            // delta = sqr(Z)
            "mov rdx, qword ptr [{s} + 64]",
            "mulx {t2}, {t1}, qword ptr [{s} + 72]",
            "mulx {t3}, rax, qword ptr [{s} + 80]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 88]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 72]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 80]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 88]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 80]",
            "mulx {t6}, rax, qword ptr [{s} + 88]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 64]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 72]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 80]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 88]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 96], rax",
            "mov qword ptr [{s} + 104], rdx",
            "mov qword ptr [{s} + 112], {t2}",
            "mov qword ptr [{s} + 120], {t3}",
            // gamma = sqr(Y)
            "mov rdx, qword ptr [{s} + 32]",
            "mulx {t2}, {t1}, qword ptr [{s} + 40]",
            "mulx {t3}, rax, qword ptr [{s} + 48]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 56]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 40]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 48]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 56]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 48]",
            "mulx {t6}, rax, qword ptr [{s} + 56]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 32]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 40]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 48]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 56]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 128], rax",
            "mov qword ptr [{s} + 136], rdx",
            "mov qword ptr [{s} + 144], {t2}",
            "mov qword ptr [{s} + 152], {t3}",
            // m = sub(X, delta)
            "mov {t0}, qword ptr [{s}]",
            "mov {t1}, qword ptr [{s} + 8]",
            "mov {t2}, qword ptr [{s} + 16]",
            "mov {t3}, qword ptr [{s} + 24]",
            "sub {t0}, qword ptr [{s} + 96]",
            "sbb {t1}, qword ptr [{s} + 104]",
            "sbb {t2}, qword ptr [{s} + 112]",
            "sbb {t3}, qword ptr [{s} + 120]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 192], {t0}",
            "mov qword ptr [{s} + 200], {t1}",
            "mov qword ptr [{s} + 208], {t2}",
            "mov qword ptr [{s} + 216], {t3}",
            // u = add(X, delta)
            "mov {t0}, qword ptr [{s}]",
            "mov {t1}, qword ptr [{s} + 8]",
            "mov {t2}, qword ptr [{s} + 16]",
            "mov {t3}, qword ptr [{s} + 24]",
            "xor {c:e}, {c:e}",
            "add {t0}, qword ptr [{s} + 96]",
            "adc {t1}, qword ptr [{s} + 104]",
            "adc {t2}, qword ptr [{s} + 112]",
            "adc {t3}, qword ptr [{s} + 120]",
            "adc {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 224], {t4}",
            "mov qword ptr [{s} + 232], {t5}",
            "mov qword ptr [{s} + 240], {t6}",
            "mov qword ptr [{s} + 248], {t7}",
            // m = mul(m, u)
            "mov rdx, qword ptr [{s} + 224]",
            "mulx {t1}, {t0}, qword ptr [{s} + 192]",
            "mulx {t2}, rax, qword ptr [{s} + 200]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 208]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 216]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 232]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 192]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 200]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 208]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 216]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 240]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 192]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 200]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 208]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 216]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 248]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 192]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 200]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 208]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 216]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 192], rax",
            "mov qword ptr [{s} + 200], rdx",
            "mov qword ptr [{s} + 208], {c}",
            "mov qword ptr [{s} + 216], {t3}",
            // beta = mul(X, gamma)
            "mov rdx, qword ptr [{s} + 128]",
            "mulx {t1}, {t0}, qword ptr [{s}]",
            "mulx {t2}, rax, qword ptr [{s} + 8]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 16]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 24]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 136]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s}]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 8]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 16]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 24]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 144]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s}]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 8]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 16]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 24]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 152]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s}]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 8]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 16]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 24]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 160], rax",
            "mov qword ptr [{s} + 168], rdx",
            "mov qword ptr [{s} + 176], {c}",
            "mov qword ptr [{s} + 184], {t3}",
            // m2 = sqr(m)
            "mov rdx, qword ptr [{s} + 192]",
            "mulx {t2}, {t1}, qword ptr [{s} + 200]",
            "mulx {t3}, rax, qword ptr [{s} + 208]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 216]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 200]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 208]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 216]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 208]",
            "mulx {t6}, rax, qword ptr [{s} + 216]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 192]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 200]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 208]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 216]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 256], rax",
            "mov qword ptr [{s} + 264], rdx",
            "mov qword ptr [{s} + 272], {t2}",
            "mov qword ptr [{s} + 280], {t3}",
            // w = add(Y, Z)
            "mov {t0}, qword ptr [{s} + 32]",
            "mov {t1}, qword ptr [{s} + 40]",
            "mov {t2}, qword ptr [{s} + 48]",
            "mov {t3}, qword ptr [{s} + 56]",
            "xor {c:e}, {c:e}",
            "add {t0}, qword ptr [{s} + 64]",
            "adc {t1}, qword ptr [{s} + 72]",
            "adc {t2}, qword ptr [{s} + 80]",
            "adc {t3}, qword ptr [{s} + 88]",
            "adc {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 320], {t4}",
            "mov qword ptr [{s} + 328], {t5}",
            "mov qword ptr [{s} + 336], {t6}",
            "mov qword ptr [{s} + 344], {t7}",
            // w = sqr(w)
            "mov rdx, qword ptr [{s} + 320]",
            "mulx {t2}, {t1}, qword ptr [{s} + 328]",
            "mulx {t3}, rax, qword ptr [{s} + 336]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 344]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 328]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 336]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 344]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 336]",
            "mulx {t6}, rax, qword ptr [{s} + 344]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 320]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 328]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 336]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 344]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 320], rax",
            "mov qword ptr [{s} + 328], rdx",
            "mov qword ptr [{s} + 336], {t2}",
            "mov qword ptr [{s} + 344], {t3}",
            // d = cms(beta, m2, 12, 9)
            "mov {t0}, -1",
            "sub {t0}, qword ptr [{s} + 256]",
            "mov {t1:e}, 0xffffffff",
            "sbb {t1}, qword ptr [{s} + 264]",
            "mov {t2:e}, 0",
            "sbb {t2}, qword ptr [{s} + 272]",
            "mov {t3}, 0xffffffff00000001",
            "sbb {t3}, qword ptr [{s} + 280]",
            "mov edx, 9",
            "mulx {c}, {t0}, {t0}",
            "mulx {t5}, {t1}, {t1}",
            "add {t1}, {c}",
            "mulx {c}, {t2}, {t2}",
            "adc {t2}, {t5}",
            "mulx {t4}, {t3}, {t3}",
            "adc {t3}, {c}",
            "adc {t4}, 0",
            "mov edx, 12",
            "xor {t6:e}, {t6:e}",
            "mulx {c}, rax, qword ptr [{s} + 160]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 168]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 176]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx rdx, rax, qword ptr [{s} + 184]",
            "adcx {t3}, rax",
            "adox rdx, {t4}",
            "adcx rdx, {t6}",
            "add rdx, 1",
            "mov rax, rdx",
            "shl rax, 32",
            "mov {c}, rax",
            "neg rax",
            "sub {c}, rdx",
            "sub {c}, 1",
            "add {t0}, rdx",
            "adc {t1}, rax",
            "adc {t2}, -1",
            "adc {t3}, {c}",
            "cmc",
            "sbb rax, rax",
            "mov rdx, rax",
            "shr rdx, 32",
            "mov {c}, 0xffffffff00000001",
            "and {c}, rax",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, 0",
            "adc {t3}, {c}",
            "mov qword ptr [{s} + 288], {t0}",
            "mov qword ptr [{s} + 296], {t1}",
            "mov qword ptr [{s} + 304], {t2}",
            "mov qword ptr [{s} + 312], {t3}",
            // g2 = sqr(gamma)
            "mov rdx, qword ptr [{s} + 128]",
            "mulx {t2}, {t1}, qword ptr [{s} + 136]",
            "mulx {t3}, rax, qword ptr [{s} + 144]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 152]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 136]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 144]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 152]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 144]",
            "mulx {t6}, rax, qword ptr [{s} + 152]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 128]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 136]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 144]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 152]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 352], rax",
            "mov qword ptr [{s} + 360], rdx",
            "mov qword ptr [{s} + 368], {t2}",
            "mov qword ptr [{s} + 376], {t3}",
            // u = mul(d, m)
            "mov rdx, qword ptr [{s} + 192]",
            "mulx {t1}, {t0}, qword ptr [{s} + 288]",
            "mulx {t2}, rax, qword ptr [{s} + 296]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 304]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 312]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 200]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 288]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 296]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 304]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 312]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 208]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 288]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 296]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 304]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 312]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 216]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 288]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 296]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 304]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 312]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 224], rax",
            "mov qword ptr [{s} + 232], rdx",
            "mov qword ptr [{s} + 240], {c}",
            "mov qword ptr [{s} + 248], {t3}",
            // X = cms(beta, d, 4, 1)
            "xor {t4:e}, {t4:e}",
            "mov {t0}, -1",
            "sub {t0}, qword ptr [{s} + 288]",
            "mov {t1:e}, 0xffffffff",
            "sbb {t1}, qword ptr [{s} + 296]",
            "mov {t2:e}, 0",
            "sbb {t2}, qword ptr [{s} + 304]",
            "mov {t3}, 0xffffffff00000001",
            "sbb {t3}, qword ptr [{s} + 312]",
            "mov edx, 4",
            "xor {t6:e}, {t6:e}",
            "mulx {c}, rax, qword ptr [{s} + 160]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 168]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 176]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx rdx, rax, qword ptr [{s} + 184]",
            "adcx {t3}, rax",
            "adox rdx, {t4}",
            "adcx rdx, {t6}",
            "add rdx, 1",
            "mov rax, rdx",
            "shl rax, 32",
            "mov {c}, rax",
            "neg rax",
            "sub {c}, rdx",
            "sub {c}, 1",
            "add {t0}, rdx",
            "adc {t1}, rax",
            "adc {t2}, -1",
            "adc {t3}, {c}",
            "cmc",
            "sbb rax, rax",
            "mov rdx, rax",
            "shr rdx, 32",
            "mov {c}, 0xffffffff00000001",
            "and {c}, rax",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, 0",
            "adc {t3}, {c}",
            "mov qword ptr [{s}], {t0}",
            "mov qword ptr [{s} + 8], {t1}",
            "mov qword ptr [{s} + 16], {t2}",
            "mov qword ptr [{s} + 24], {t3}",
            // w = sub(w, gamma)
            "mov {t0}, qword ptr [{s} + 320]",
            "mov {t1}, qword ptr [{s} + 328]",
            "mov {t2}, qword ptr [{s} + 336]",
            "mov {t3}, qword ptr [{s} + 344]",
            "sub {t0}, qword ptr [{s} + 128]",
            "sbb {t1}, qword ptr [{s} + 136]",
            "sbb {t2}, qword ptr [{s} + 144]",
            "sbb {t3}, qword ptr [{s} + 152]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 320], {t0}",
            "mov qword ptr [{s} + 328], {t1}",
            "mov qword ptr [{s} + 336], {t2}",
            "mov qword ptr [{s} + 344], {t3}",
            // Z = sub(w, delta)
            "mov {t0}, qword ptr [{s} + 320]",
            "mov {t1}, qword ptr [{s} + 328]",
            "mov {t2}, qword ptr [{s} + 336]",
            "mov {t3}, qword ptr [{s} + 344]",
            "sub {t0}, qword ptr [{s} + 96]",
            "sbb {t1}, qword ptr [{s} + 104]",
            "sbb {t2}, qword ptr [{s} + 112]",
            "sbb {t3}, qword ptr [{s} + 120]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 64], {t0}",
            "mov qword ptr [{s} + 72], {t1}",
            "mov qword ptr [{s} + 80], {t2}",
            "mov qword ptr [{s} + 88], {t3}",
            // Y = cms(u, g2, 3, 8)
            "mov {t0}, -1",
            "sub {t0}, qword ptr [{s} + 352]",
            "mov {t1:e}, 0xffffffff",
            "sbb {t1}, qword ptr [{s} + 360]",
            "mov {t2:e}, 0",
            "sbb {t2}, qword ptr [{s} + 368]",
            "mov {t3}, 0xffffffff00000001",
            "sbb {t3}, qword ptr [{s} + 376]",
            "mov {t4}, {t3}",
            "shr {t4}, 61",
            "shld {t3}, {t2}, 3",
            "shld {t2}, {t1}, 3",
            "shld {t1}, {t0}, 3",
            "shl {t0}, 3",
            "mov edx, 3",
            "xor {t6:e}, {t6:e}",
            "mulx {c}, rax, qword ptr [{s} + 224]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 232]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 240]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx rdx, rax, qword ptr [{s} + 248]",
            "adcx {t3}, rax",
            "adox rdx, {t4}",
            "adcx rdx, {t6}",
            "add rdx, 1",
            "mov rax, rdx",
            "shl rax, 32",
            "mov {c}, rax",
            "neg rax",
            "sub {c}, rdx",
            "sub {c}, 1",
            "add {t0}, rdx",
            "adc {t1}, rax",
            "adc {t2}, -1",
            "adc {t3}, {c}",
            "cmc",
            "sbb rax, rax",
            "mov rdx, rax",
            "shr rdx, 32",
            "mov {c}, 0xffffffff00000001",
            "and {c}, rax",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, 0",
            "adc {t3}, {c}",
            "mov qword ptr [{s} + 32], {t0}",
            "mov qword ptr [{s} + 40], {t1}",
            "mov qword ptr [{s} + 48], {t2}",
            "mov qword ptr [{s} + 56], {t3}",
            s = in(reg) buf.as_mut_ptr(),
            t0 = out(reg) _,
            t1 = out(reg) _,
            t2 = out(reg) _,
            t3 = out(reg) _,
            t4 = out(reg) _,
            t5 = out(reg) _,
            t6 = out(reg) _,
            t7 = out(reg) _,
            c = out(reg) _,
            out("rax") _,
            out("rdx") _,
            options(nostack),
        );
    }
}
#[allow(dead_code)]
pub const DOUBLE_IN_PLACE_X: usize = 0;
#[allow(dead_code)]
pub const DOUBLE_IN_PLACE_Y: usize = 1;
#[allow(dead_code)]
pub const DOUBLE_IN_PLACE_Z: usize = 2;

/// Jacobian addition, `add-2007-bl`, as `Point::add_formula`; also leaves `H` and `R`.
///
/// Slots, 32 bytes each, in order: `X1`, `Y1`, `Z1`, `X2`, `Y2`, `Z2`, `X3`, `Y3`, `Z3`, `H`, `R`, `z1z1`, `z2z2`, `u1`, `u2`, `s1`, `s2`, `i`, `j`, `v`, `t1`, `t2`, `t3`, `t4`.
///
/// # Safety
///
/// The CPU must support BMI2 and ADX; see `super::p256_x86::has_adx`.
#[inline(never)]
pub unsafe fn add_adx(buf: &mut [MaybeUninit<u64>; 96]) {
    // SAFETY: the caller guarantees the instructions exist. The block reads
    // and writes only the buffer behind `{s}`, a live, exclusive reference
    // of 768 bytes, at offsets below that; it writes only the
    // registers it declares, and does not touch the stack. It reads no slot
    // before writing it but the inputs, which the caller has written.
    unsafe {
        asm!(
            ".p2align 6",
            // z1z1 = sqr(Z1)
            "mov rdx, qword ptr [{s} + 64]",
            "mulx {t2}, {t1}, qword ptr [{s} + 72]",
            "mulx {t3}, rax, qword ptr [{s} + 80]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 88]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 72]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 80]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 88]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 80]",
            "mulx {t6}, rax, qword ptr [{s} + 88]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 64]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 72]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 80]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 88]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 352], rax",
            "mov qword ptr [{s} + 360], rdx",
            "mov qword ptr [{s} + 368], {t2}",
            "mov qword ptr [{s} + 376], {t3}",
            // z2z2 = sqr(Z2)
            "mov rdx, qword ptr [{s} + 160]",
            "mulx {t2}, {t1}, qword ptr [{s} + 168]",
            "mulx {t3}, rax, qword ptr [{s} + 176]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 184]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 168]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 176]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 184]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 176]",
            "mulx {t6}, rax, qword ptr [{s} + 184]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 160]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 168]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 176]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 184]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 384], rax",
            "mov qword ptr [{s} + 392], rdx",
            "mov qword ptr [{s} + 400], {t2}",
            "mov qword ptr [{s} + 408], {t3}",
            // t1 = mul(Y1, Z2)
            "mov rdx, qword ptr [{s} + 160]",
            "mulx {t1}, {t0}, qword ptr [{s} + 32]",
            "mulx {t2}, rax, qword ptr [{s} + 40]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 48]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 56]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 168]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 32]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 40]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 48]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 56]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 176]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 32]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 40]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 48]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 56]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 184]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 32]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 40]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 48]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 56]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 640], rax",
            "mov qword ptr [{s} + 648], rdx",
            "mov qword ptr [{s} + 656], {c}",
            "mov qword ptr [{s} + 664], {t3}",
            // u2 = mul(X2, z1z1)
            "mov rdx, qword ptr [{s} + 352]",
            "mulx {t1}, {t0}, qword ptr [{s} + 96]",
            "mulx {t2}, rax, qword ptr [{s} + 104]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 112]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 120]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 360]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 96]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 104]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 112]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 120]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 368]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 96]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 104]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 112]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 120]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 376]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 96]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 104]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 112]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 120]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 448], rax",
            "mov qword ptr [{s} + 456], rdx",
            "mov qword ptr [{s} + 464], {c}",
            "mov qword ptr [{s} + 472], {t3}",
            // u1 = mul(X1, z2z2)
            "mov rdx, qword ptr [{s} + 384]",
            "mulx {t1}, {t0}, qword ptr [{s}]",
            "mulx {t2}, rax, qword ptr [{s} + 8]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 16]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 24]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 392]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s}]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 8]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 16]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 24]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 400]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s}]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 8]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 16]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 24]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 408]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s}]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 8]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 16]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 24]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 416], rax",
            "mov qword ptr [{s} + 424], rdx",
            "mov qword ptr [{s} + 432], {c}",
            "mov qword ptr [{s} + 440], {t3}",
            // t2 = mul(Y2, Z1)
            "mov rdx, qword ptr [{s} + 64]",
            "mulx {t1}, {t0}, qword ptr [{s} + 128]",
            "mulx {t2}, rax, qword ptr [{s} + 136]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 144]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 152]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 72]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 128]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 136]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 144]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 152]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 80]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 128]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 136]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 144]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 152]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 88]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 128]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 136]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 144]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 152]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 672], rax",
            "mov qword ptr [{s} + 680], rdx",
            "mov qword ptr [{s} + 688], {c}",
            "mov qword ptr [{s} + 696], {t3}",
            // H = sub(u2, u1)
            "mov {t0}, qword ptr [{s} + 448]",
            "mov {t1}, qword ptr [{s} + 456]",
            "mov {t2}, qword ptr [{s} + 464]",
            "mov {t3}, qword ptr [{s} + 472]",
            "sub {t0}, qword ptr [{s} + 416]",
            "sbb {t1}, qword ptr [{s} + 424]",
            "sbb {t2}, qword ptr [{s} + 432]",
            "sbb {t3}, qword ptr [{s} + 440]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 288], {t0}",
            "mov qword ptr [{s} + 296], {t1}",
            "mov qword ptr [{s} + 304], {t2}",
            "mov qword ptr [{s} + 312], {t3}",
            // i = add(H, H)
            "mov {t0}, qword ptr [{s} + 288]",
            "mov {t1}, qword ptr [{s} + 296]",
            "mov {t2}, qword ptr [{s} + 304]",
            "mov {t3}, qword ptr [{s} + 312]",
            "xor {c:e}, {c:e}",
            "add {t0}, qword ptr [{s} + 288]",
            "adc {t1}, qword ptr [{s} + 296]",
            "adc {t2}, qword ptr [{s} + 304]",
            "adc {t3}, qword ptr [{s} + 312]",
            "adc {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 544], {t4}",
            "mov qword ptr [{s} + 552], {t5}",
            "mov qword ptr [{s} + 560], {t6}",
            "mov qword ptr [{s} + 568], {t7}",
            // i = sqr(i)
            "mov rdx, qword ptr [{s} + 544]",
            "mulx {t2}, {t1}, qword ptr [{s} + 552]",
            "mulx {t3}, rax, qword ptr [{s} + 560]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 568]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 552]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 560]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 568]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 560]",
            "mulx {t6}, rax, qword ptr [{s} + 568]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 544]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 552]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 560]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 568]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 544], rax",
            "mov qword ptr [{s} + 552], rdx",
            "mov qword ptr [{s} + 560], {t2}",
            "mov qword ptr [{s} + 568], {t3}",
            // s1 = mul(t1, z2z2)
            "mov rdx, qword ptr [{s} + 384]",
            "mulx {t1}, {t0}, qword ptr [{s} + 640]",
            "mulx {t2}, rax, qword ptr [{s} + 648]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 656]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 664]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 392]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 640]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 648]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 656]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 664]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 400]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 640]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 648]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 656]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 664]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 408]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 640]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 648]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 656]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 664]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 480], rax",
            "mov qword ptr [{s} + 488], rdx",
            "mov qword ptr [{s} + 496], {c}",
            "mov qword ptr [{s} + 504], {t3}",
            // v = mul(u1, i)
            "mov rdx, qword ptr [{s} + 544]",
            "mulx {t1}, {t0}, qword ptr [{s} + 416]",
            "mulx {t2}, rax, qword ptr [{s} + 424]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 432]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 440]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 552]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 416]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 424]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 432]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 440]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 560]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 416]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 424]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 432]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 440]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 568]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 416]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 424]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 432]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 440]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 608], rax",
            "mov qword ptr [{s} + 616], rdx",
            "mov qword ptr [{s} + 624], {c}",
            "mov qword ptr [{s} + 632], {t3}",
            // s2 = mul(t2, z1z1)
            "mov rdx, qword ptr [{s} + 352]",
            "mulx {t1}, {t0}, qword ptr [{s} + 672]",
            "mulx {t2}, rax, qword ptr [{s} + 680]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 688]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 696]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 360]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 672]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 680]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 688]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 696]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 368]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 672]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 680]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 688]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 696]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 376]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 672]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 680]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 688]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 696]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 512], rax",
            "mov qword ptr [{s} + 520], rdx",
            "mov qword ptr [{s} + 528], {c}",
            "mov qword ptr [{s} + 536], {t3}",
            // j = mul(H, i)
            "mov rdx, qword ptr [{s} + 544]",
            "mulx {t1}, {t0}, qword ptr [{s} + 288]",
            "mulx {t2}, rax, qword ptr [{s} + 296]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 304]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 312]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 552]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 288]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 296]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 304]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 312]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 560]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 288]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 296]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 304]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 312]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 568]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 288]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 296]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 304]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 312]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 576], rax",
            "mov qword ptr [{s} + 584], rdx",
            "mov qword ptr [{s} + 592], {c}",
            "mov qword ptr [{s} + 600], {t3}",
            // R = sub(s2, s1)
            "mov {t0}, qword ptr [{s} + 512]",
            "mov {t1}, qword ptr [{s} + 520]",
            "mov {t2}, qword ptr [{s} + 528]",
            "mov {t3}, qword ptr [{s} + 536]",
            "sub {t0}, qword ptr [{s} + 480]",
            "sbb {t1}, qword ptr [{s} + 488]",
            "sbb {t2}, qword ptr [{s} + 496]",
            "sbb {t3}, qword ptr [{s} + 504]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 320], {t0}",
            "mov qword ptr [{s} + 328], {t1}",
            "mov qword ptr [{s} + 336], {t2}",
            "mov qword ptr [{s} + 344], {t3}",
            // R = add(R, R)
            "mov {t0}, qword ptr [{s} + 320]",
            "mov {t1}, qword ptr [{s} + 328]",
            "mov {t2}, qword ptr [{s} + 336]",
            "mov {t3}, qword ptr [{s} + 344]",
            "xor {c:e}, {c:e}",
            "add {t0}, qword ptr [{s} + 320]",
            "adc {t1}, qword ptr [{s} + 328]",
            "adc {t2}, qword ptr [{s} + 336]",
            "adc {t3}, qword ptr [{s} + 344]",
            "adc {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 320], {t4}",
            "mov qword ptr [{s} + 328], {t5}",
            "mov qword ptr [{s} + 336], {t6}",
            "mov qword ptr [{s} + 344], {t7}",
            // X3 = sqr(R)
            "mov rdx, qword ptr [{s} + 320]",
            "mulx {t2}, {t1}, qword ptr [{s} + 328]",
            "mulx {t3}, rax, qword ptr [{s} + 336]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 344]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 328]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 336]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 344]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 336]",
            "mulx {t6}, rax, qword ptr [{s} + 344]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 320]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 328]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 336]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 344]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 192], rax",
            "mov qword ptr [{s} + 200], rdx",
            "mov qword ptr [{s} + 208], {t2}",
            "mov qword ptr [{s} + 216], {t3}",
            // Z3 = add(Z1, Z2)
            "mov {t0}, qword ptr [{s} + 64]",
            "mov {t1}, qword ptr [{s} + 72]",
            "mov {t2}, qword ptr [{s} + 80]",
            "mov {t3}, qword ptr [{s} + 88]",
            "xor {c:e}, {c:e}",
            "add {t0}, qword ptr [{s} + 160]",
            "adc {t1}, qword ptr [{s} + 168]",
            "adc {t2}, qword ptr [{s} + 176]",
            "adc {t3}, qword ptr [{s} + 184]",
            "adc {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 256], {t4}",
            "mov qword ptr [{s} + 264], {t5}",
            "mov qword ptr [{s} + 272], {t6}",
            "mov qword ptr [{s} + 280], {t7}",
            // Z3 = sqr(Z3)
            "mov rdx, qword ptr [{s} + 256]",
            "mulx {t2}, {t1}, qword ptr [{s} + 264]",
            "mulx {t3}, rax, qword ptr [{s} + 272]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 280]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 264]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 272]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 280]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 272]",
            "mulx {t6}, rax, qword ptr [{s} + 280]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 256]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 264]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 272]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 280]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 256], rax",
            "mov qword ptr [{s} + 264], rdx",
            "mov qword ptr [{s} + 272], {t2}",
            "mov qword ptr [{s} + 280], {t3}",
            // t3 = mul(s1, j)
            "mov rdx, qword ptr [{s} + 576]",
            "mulx {t1}, {t0}, qword ptr [{s} + 480]",
            "mulx {t2}, rax, qword ptr [{s} + 488]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 496]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 504]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 584]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 480]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 488]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 496]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 504]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 592]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 480]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 488]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 496]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 504]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 600]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 480]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 488]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 496]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 504]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 704], rax",
            "mov qword ptr [{s} + 712], rdx",
            "mov qword ptr [{s} + 720], {c}",
            "mov qword ptr [{s} + 728], {t3}",
            // t1 = add(v, v)
            "mov {t0}, qword ptr [{s} + 608]",
            "mov {t1}, qword ptr [{s} + 616]",
            "mov {t2}, qword ptr [{s} + 624]",
            "mov {t3}, qword ptr [{s} + 632]",
            "xor {c:e}, {c:e}",
            "add {t0}, qword ptr [{s} + 608]",
            "adc {t1}, qword ptr [{s} + 616]",
            "adc {t2}, qword ptr [{s} + 624]",
            "adc {t3}, qword ptr [{s} + 632]",
            "adc {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 640], {t4}",
            "mov qword ptr [{s} + 648], {t5}",
            "mov qword ptr [{s} + 656], {t6}",
            "mov qword ptr [{s} + 664], {t7}",
            // X3 = sub(X3, j)
            "mov {t0}, qword ptr [{s} + 192]",
            "mov {t1}, qword ptr [{s} + 200]",
            "mov {t2}, qword ptr [{s} + 208]",
            "mov {t3}, qword ptr [{s} + 216]",
            "sub {t0}, qword ptr [{s} + 576]",
            "sbb {t1}, qword ptr [{s} + 584]",
            "sbb {t2}, qword ptr [{s} + 592]",
            "sbb {t3}, qword ptr [{s} + 600]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 192], {t0}",
            "mov qword ptr [{s} + 200], {t1}",
            "mov qword ptr [{s} + 208], {t2}",
            "mov qword ptr [{s} + 216], {t3}",
            // X3 = sub(X3, t1)
            "mov {t0}, qword ptr [{s} + 192]",
            "mov {t1}, qword ptr [{s} + 200]",
            "mov {t2}, qword ptr [{s} + 208]",
            "mov {t3}, qword ptr [{s} + 216]",
            "sub {t0}, qword ptr [{s} + 640]",
            "sbb {t1}, qword ptr [{s} + 648]",
            "sbb {t2}, qword ptr [{s} + 656]",
            "sbb {t3}, qword ptr [{s} + 664]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 192], {t0}",
            "mov qword ptr [{s} + 200], {t1}",
            "mov qword ptr [{s} + 208], {t2}",
            "mov qword ptr [{s} + 216], {t3}",
            // t4 = sub(v, X3)
            "mov {t0}, qword ptr [{s} + 608]",
            "mov {t1}, qword ptr [{s} + 616]",
            "mov {t2}, qword ptr [{s} + 624]",
            "mov {t3}, qword ptr [{s} + 632]",
            "sub {t0}, qword ptr [{s} + 192]",
            "sbb {t1}, qword ptr [{s} + 200]",
            "sbb {t2}, qword ptr [{s} + 208]",
            "sbb {t3}, qword ptr [{s} + 216]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 736], {t0}",
            "mov qword ptr [{s} + 744], {t1}",
            "mov qword ptr [{s} + 752], {t2}",
            "mov qword ptr [{s} + 760], {t3}",
            // Z3 = sub(Z3, z1z1)
            "mov {t0}, qword ptr [{s} + 256]",
            "mov {t1}, qword ptr [{s} + 264]",
            "mov {t2}, qword ptr [{s} + 272]",
            "mov {t3}, qword ptr [{s} + 280]",
            "sub {t0}, qword ptr [{s} + 352]",
            "sbb {t1}, qword ptr [{s} + 360]",
            "sbb {t2}, qword ptr [{s} + 368]",
            "sbb {t3}, qword ptr [{s} + 376]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 256], {t0}",
            "mov qword ptr [{s} + 264], {t1}",
            "mov qword ptr [{s} + 272], {t2}",
            "mov qword ptr [{s} + 280], {t3}",
            // Y3 = mul(R, t4)
            "mov rdx, qword ptr [{s} + 736]",
            "mulx {t1}, {t0}, qword ptr [{s} + 320]",
            "mulx {t2}, rax, qword ptr [{s} + 328]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 336]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 344]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 744]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 320]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 328]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 336]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 344]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 752]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 320]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 328]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 336]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 344]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 760]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 320]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 328]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 336]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 344]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 224], rax",
            "mov qword ptr [{s} + 232], rdx",
            "mov qword ptr [{s} + 240], {c}",
            "mov qword ptr [{s} + 248], {t3}",
            // Z3 = sub(Z3, z2z2)
            "mov {t0}, qword ptr [{s} + 256]",
            "mov {t1}, qword ptr [{s} + 264]",
            "mov {t2}, qword ptr [{s} + 272]",
            "mov {t3}, qword ptr [{s} + 280]",
            "sub {t0}, qword ptr [{s} + 384]",
            "sbb {t1}, qword ptr [{s} + 392]",
            "sbb {t2}, qword ptr [{s} + 400]",
            "sbb {t3}, qword ptr [{s} + 408]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 256], {t0}",
            "mov qword ptr [{s} + 264], {t1}",
            "mov qword ptr [{s} + 272], {t2}",
            "mov qword ptr [{s} + 280], {t3}",
            // Z3 = mul(Z3, H)
            "mov rdx, qword ptr [{s} + 288]",
            "mulx {t1}, {t0}, qword ptr [{s} + 256]",
            "mulx {t2}, rax, qword ptr [{s} + 264]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 272]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 280]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 296]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 256]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 264]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 272]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 280]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 304]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 256]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 264]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 272]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 280]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 312]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 256]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 264]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 272]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 280]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 256], rax",
            "mov qword ptr [{s} + 264], rdx",
            "mov qword ptr [{s} + 272], {c}",
            "mov qword ptr [{s} + 280], {t3}",
            // t3 = add(t3, t3)
            "mov {t0}, qword ptr [{s} + 704]",
            "mov {t1}, qword ptr [{s} + 712]",
            "mov {t2}, qword ptr [{s} + 720]",
            "mov {t3}, qword ptr [{s} + 728]",
            "xor {c:e}, {c:e}",
            "add {t0}, qword ptr [{s} + 704]",
            "adc {t1}, qword ptr [{s} + 712]",
            "adc {t2}, qword ptr [{s} + 720]",
            "adc {t3}, qword ptr [{s} + 728]",
            "adc {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 704], {t4}",
            "mov qword ptr [{s} + 712], {t5}",
            "mov qword ptr [{s} + 720], {t6}",
            "mov qword ptr [{s} + 728], {t7}",
            // Y3 = sub(Y3, t3)
            "mov {t0}, qword ptr [{s} + 224]",
            "mov {t1}, qword ptr [{s} + 232]",
            "mov {t2}, qword ptr [{s} + 240]",
            "mov {t3}, qword ptr [{s} + 248]",
            "sub {t0}, qword ptr [{s} + 704]",
            "sbb {t1}, qword ptr [{s} + 712]",
            "sbb {t2}, qword ptr [{s} + 720]",
            "sbb {t3}, qword ptr [{s} + 728]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 224], {t0}",
            "mov qword ptr [{s} + 232], {t1}",
            "mov qword ptr [{s} + 240], {t2}",
            "mov qword ptr [{s} + 248], {t3}",
            s = in(reg) buf.as_mut_ptr(),
            t0 = out(reg) _,
            t1 = out(reg) _,
            t2 = out(reg) _,
            t3 = out(reg) _,
            t4 = out(reg) _,
            t5 = out(reg) _,
            t6 = out(reg) _,
            t7 = out(reg) _,
            c = out(reg) _,
            out("rax") _,
            out("rdx") _,
            options(nostack),
        );
    }
}
#[allow(dead_code)]
pub const ADD_X3: usize = 6;
#[allow(dead_code)]
pub const ADD_Y3: usize = 7;
#[allow(dead_code)]
pub const ADD_Z3: usize = 8;
#[allow(dead_code)]
pub const ADD_H: usize = 9;
#[allow(dead_code)]
pub const ADD_R: usize = 10;

/// Mixed addition, `madd-2007-bl`, as `Point::add_affine_formula`; also leaves `H` and `R`.
///
/// Slots, 32 bytes each, in order: `X1`, `Y1`, `Z1`, `x2`, `y2`, `X3`, `Y3`, `Z3`, `H`, `R`, `z1z1`, `u2`, `s2`, `hh`, `i`, `j`, `v`, `t`, `t2`, `t3`.
///
/// # Safety
///
/// The CPU must support BMI2 and ADX; see `super::p256_x86::has_adx`.
#[inline(never)]
pub unsafe fn add_affine_adx(buf: &mut [MaybeUninit<u64>; 80]) {
    // SAFETY: the caller guarantees the instructions exist. The block reads
    // and writes only the buffer behind `{s}`, a live, exclusive reference
    // of 640 bytes, at offsets below that; it writes only the
    // registers it declares, and does not touch the stack. It reads no slot
    // before writing it but the inputs, which the caller has written.
    unsafe {
        asm!(
            ".p2align 6",
            // z1z1 = sqr(Z1)
            "mov rdx, qword ptr [{s} + 64]",
            "mulx {t2}, {t1}, qword ptr [{s} + 72]",
            "mulx {t3}, rax, qword ptr [{s} + 80]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 88]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 72]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 80]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 88]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 80]",
            "mulx {t6}, rax, qword ptr [{s} + 88]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 64]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 72]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 80]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 88]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 320], rax",
            "mov qword ptr [{s} + 328], rdx",
            "mov qword ptr [{s} + 336], {t2}",
            "mov qword ptr [{s} + 344], {t3}",
            // t = mul(y2, Z1)
            "mov rdx, qword ptr [{s} + 64]",
            "mulx {t1}, {t0}, qword ptr [{s} + 128]",
            "mulx {t2}, rax, qword ptr [{s} + 136]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 144]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 152]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 72]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 128]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 136]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 144]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 152]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 80]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 128]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 136]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 144]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 152]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 88]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 128]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 136]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 144]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 152]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 544], rax",
            "mov qword ptr [{s} + 552], rdx",
            "mov qword ptr [{s} + 560], {c}",
            "mov qword ptr [{s} + 568], {t3}",
            // u2 = mul(x2, z1z1)
            "mov rdx, qword ptr [{s} + 320]",
            "mulx {t1}, {t0}, qword ptr [{s} + 96]",
            "mulx {t2}, rax, qword ptr [{s} + 104]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 112]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 120]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 328]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 96]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 104]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 112]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 120]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 336]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 96]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 104]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 112]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 120]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 344]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 96]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 104]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 112]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 120]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 352], rax",
            "mov qword ptr [{s} + 360], rdx",
            "mov qword ptr [{s} + 368], {c}",
            "mov qword ptr [{s} + 376], {t3}",
            // s2 = mul(t, z1z1)
            "mov rdx, qword ptr [{s} + 320]",
            "mulx {t1}, {t0}, qword ptr [{s} + 544]",
            "mulx {t2}, rax, qword ptr [{s} + 552]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 560]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 568]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 328]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 544]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 552]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 560]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 568]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 336]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 544]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 552]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 560]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 568]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 344]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 544]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 552]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 560]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 568]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 384], rax",
            "mov qword ptr [{s} + 392], rdx",
            "mov qword ptr [{s} + 400], {c}",
            "mov qword ptr [{s} + 408], {t3}",
            // H = sub(u2, X1)
            "mov {t0}, qword ptr [{s} + 352]",
            "mov {t1}, qword ptr [{s} + 360]",
            "mov {t2}, qword ptr [{s} + 368]",
            "mov {t3}, qword ptr [{s} + 376]",
            "sub {t0}, qword ptr [{s}]",
            "sbb {t1}, qword ptr [{s} + 8]",
            "sbb {t2}, qword ptr [{s} + 16]",
            "sbb {t3}, qword ptr [{s} + 24]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 256], {t0}",
            "mov qword ptr [{s} + 264], {t1}",
            "mov qword ptr [{s} + 272], {t2}",
            "mov qword ptr [{s} + 280], {t3}",
            // hh = sqr(H)
            "mov rdx, qword ptr [{s} + 256]",
            "mulx {t2}, {t1}, qword ptr [{s} + 264]",
            "mulx {t3}, rax, qword ptr [{s} + 272]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 280]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 264]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 272]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 280]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 272]",
            "mulx {t6}, rax, qword ptr [{s} + 280]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 256]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 264]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 272]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 280]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 416], rax",
            "mov qword ptr [{s} + 424], rdx",
            "mov qword ptr [{s} + 432], {t2}",
            "mov qword ptr [{s} + 440], {t3}",
            // R = sub(s2, Y1)
            "mov {t0}, qword ptr [{s} + 384]",
            "mov {t1}, qword ptr [{s} + 392]",
            "mov {t2}, qword ptr [{s} + 400]",
            "mov {t3}, qword ptr [{s} + 408]",
            "sub {t0}, qword ptr [{s} + 32]",
            "sbb {t1}, qword ptr [{s} + 40]",
            "sbb {t2}, qword ptr [{s} + 48]",
            "sbb {t3}, qword ptr [{s} + 56]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 288], {t0}",
            "mov qword ptr [{s} + 296], {t1}",
            "mov qword ptr [{s} + 304], {t2}",
            "mov qword ptr [{s} + 312], {t3}",
            // R = add(R, R)
            "mov {t0}, qword ptr [{s} + 288]",
            "mov {t1}, qword ptr [{s} + 296]",
            "mov {t2}, qword ptr [{s} + 304]",
            "mov {t3}, qword ptr [{s} + 312]",
            "xor {c:e}, {c:e}",
            "add {t0}, qword ptr [{s} + 288]",
            "adc {t1}, qword ptr [{s} + 296]",
            "adc {t2}, qword ptr [{s} + 304]",
            "adc {t3}, qword ptr [{s} + 312]",
            "adc {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 288], {t4}",
            "mov qword ptr [{s} + 296], {t5}",
            "mov qword ptr [{s} + 304], {t6}",
            "mov qword ptr [{s} + 312], {t7}",
            // X3 = sqr(R)
            "mov rdx, qword ptr [{s} + 288]",
            "mulx {t2}, {t1}, qword ptr [{s} + 296]",
            "mulx {t3}, rax, qword ptr [{s} + 304]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 312]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 296]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 304]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 312]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 304]",
            "mulx {t6}, rax, qword ptr [{s} + 312]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 288]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 296]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 304]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 312]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 160], rax",
            "mov qword ptr [{s} + 168], rdx",
            "mov qword ptr [{s} + 176], {t2}",
            "mov qword ptr [{s} + 184], {t3}",
            // i = mulk(hh, 4)
            "mov {t0}, qword ptr [{s} + 416]",
            "mov {t1}, qword ptr [{s} + 424]",
            "mov {t2}, qword ptr [{s} + 432]",
            "mov {t3}, qword ptr [{s} + 440]",
            "mov {t4}, {t3}",
            "shr {t4}, 62",
            "shld {t3}, {t2}, 2",
            "shld {t2}, {t1}, 2",
            "shld {t1}, {t0}, 2",
            "shl {t0}, 2",
            "mov rax, {t4}",
            "shl rax, 32",
            "xor {c:e}, {c:e}",
            "add {t0}, {t4}",
            "adc {t1}, 0",
            "adc {t2}, 0",
            "adc {t3}, rax",
            "adc {c}, 0",
            "sub {t1}, rax",
            "sbb {t2}, 0",
            "sbb {t3}, {t4}",
            "sbb {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 448], {t4}",
            "mov qword ptr [{s} + 456], {t5}",
            "mov qword ptr [{s} + 464], {t6}",
            "mov qword ptr [{s} + 472], {t7}",
            // v = mul(X1, i)
            "mov rdx, qword ptr [{s} + 448]",
            "mulx {t1}, {t0}, qword ptr [{s}]",
            "mulx {t2}, rax, qword ptr [{s} + 8]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 16]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 24]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 456]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s}]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 8]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 16]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 24]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 464]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s}]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 8]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 16]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 24]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 472]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s}]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 8]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 16]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 24]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 512], rax",
            "mov qword ptr [{s} + 520], rdx",
            "mov qword ptr [{s} + 528], {c}",
            "mov qword ptr [{s} + 536], {t3}",
            // j = mul(H, i)
            "mov rdx, qword ptr [{s} + 448]",
            "mulx {t1}, {t0}, qword ptr [{s} + 256]",
            "mulx {t2}, rax, qword ptr [{s} + 264]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 272]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 280]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 456]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 256]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 264]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 272]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 280]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 464]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 256]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 264]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 272]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 280]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 472]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 256]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 264]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 272]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 280]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 480], rax",
            "mov qword ptr [{s} + 488], rdx",
            "mov qword ptr [{s} + 496], {c}",
            "mov qword ptr [{s} + 504], {t3}",
            // Z3 = add(Z1, H)
            "mov {t0}, qword ptr [{s} + 64]",
            "mov {t1}, qword ptr [{s} + 72]",
            "mov {t2}, qword ptr [{s} + 80]",
            "mov {t3}, qword ptr [{s} + 88]",
            "xor {c:e}, {c:e}",
            "add {t0}, qword ptr [{s} + 256]",
            "adc {t1}, qword ptr [{s} + 264]",
            "adc {t2}, qword ptr [{s} + 272]",
            "adc {t3}, qword ptr [{s} + 280]",
            "adc {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 224], {t4}",
            "mov qword ptr [{s} + 232], {t5}",
            "mov qword ptr [{s} + 240], {t6}",
            "mov qword ptr [{s} + 248], {t7}",
            // Z3 = sqr(Z3)
            "mov rdx, qword ptr [{s} + 224]",
            "mulx {t2}, {t1}, qword ptr [{s} + 232]",
            "mulx {t3}, rax, qword ptr [{s} + 240]",
            "add {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 248]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "mov rdx, qword ptr [{s} + 232]",
            "xor {t5:e}, {t5:e}",
            "mulx {c}, rax, qword ptr [{s} + 240]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 248]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "mov rdx, qword ptr [{s} + 240]",
            "mulx {t6}, rax, qword ptr [{s} + 248]",
            "add {t5}, rax",
            "adc {t6}, 0",
            "xor {t7:e}, {t7:e}",
            "add {t1}, {t1}",
            "adc {t2}, {t2}",
            "adc {t3}, {t3}",
            "adc {t4}, {t4}",
            "adc {t5}, {t5}",
            "adc {t6}, {t6}",
            "adc {t7}, 0",
            "mov rdx, qword ptr [{s} + 224]",
            "mulx {c}, {t0}, rdx",
            "add {t1}, {c}",
            "mov rdx, qword ptr [{s} + 232]",
            "mulx {c}, rax, rdx",
            "adc {t2}, rax",
            "adc {t3}, {c}",
            "mov rdx, qword ptr [{s} + 240]",
            "mulx {c}, rax, rdx",
            "adc {t4}, rax",
            "adc {t5}, {c}",
            "mov rdx, qword ptr [{s} + 248]",
            "mulx {c}, rax, rdx",
            "adc {t6}, rax",
            "adc {t7}, {c}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov rdx, {t0}",
            "shr rdx, 32",
            "mov {c}, {t0}",
            "sub {c}, rax",
            "sbb {t0}, rdx",
            "add {t1}, rax",
            "adc {t2}, rdx",
            "adc {t3}, {c}",
            "adc {t0}, 0",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov rdx, {t1}",
            "shr rdx, 32",
            "mov {c}, {t1}",
            "sub {c}, rax",
            "sbb {t1}, rdx",
            "add {t2}, rax",
            "adc {t3}, rdx",
            "adc {t0}, {c}",
            "adc {t1}, 0",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov rdx, {t2}",
            "shr rdx, 32",
            "mov {c}, {t2}",
            "sub {c}, rax",
            "sbb {t2}, rdx",
            "add {t3}, rax",
            "adc {t0}, rdx",
            "adc {t1}, {c}",
            "adc {t2}, 0",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov rdx, {t3}",
            "shr rdx, 32",
            "mov {c}, {t3}",
            "sub {c}, rax",
            "sbb {t3}, rdx",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, {c}",
            "adc {t3}, 0",
            "xor {c:e}, {c:e}",
            "add {t4}, {t0}",
            "adc {t5}, {t1}",
            "adc {t6}, {t2}",
            "adc {t7}, {t3}",
            "adc {c}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {t2}, {t6}",
            "mov {t3}, {t7}",
            "mov {t0:e}, 0xffffffff",
            "mov {t1}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t0}",
            "sbb {t2}, 0",
            "sbb {t3}, {t1}",
            "sbb {c}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {t2}, {t6}",
            "cmovc {t3}, {t7}",
            "mov qword ptr [{s} + 224], rax",
            "mov qword ptr [{s} + 232], rdx",
            "mov qword ptr [{s} + 240], {t2}",
            "mov qword ptr [{s} + 248], {t3}",
            // t2 = mul(Y1, j)
            "mov rdx, qword ptr [{s} + 480]",
            "mulx {t1}, {t0}, qword ptr [{s} + 32]",
            "mulx {t2}, rax, qword ptr [{s} + 40]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 48]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 56]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 488]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 32]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 40]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 48]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 56]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 496]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 32]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 40]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 48]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 56]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 504]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 32]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 40]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 48]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 56]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 576], rax",
            "mov qword ptr [{s} + 584], rdx",
            "mov qword ptr [{s} + 592], {c}",
            "mov qword ptr [{s} + 600], {t3}",
            // t = add(v, v)
            "mov {t0}, qword ptr [{s} + 512]",
            "mov {t1}, qword ptr [{s} + 520]",
            "mov {t2}, qword ptr [{s} + 528]",
            "mov {t3}, qword ptr [{s} + 536]",
            "xor {c:e}, {c:e}",
            "add {t0}, qword ptr [{s} + 512]",
            "adc {t1}, qword ptr [{s} + 520]",
            "adc {t2}, qword ptr [{s} + 528]",
            "adc {t3}, qword ptr [{s} + 536]",
            "adc {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 544], {t4}",
            "mov qword ptr [{s} + 552], {t5}",
            "mov qword ptr [{s} + 560], {t6}",
            "mov qword ptr [{s} + 568], {t7}",
            // X3 = sub(X3, j)
            "mov {t0}, qword ptr [{s} + 160]",
            "mov {t1}, qword ptr [{s} + 168]",
            "mov {t2}, qword ptr [{s} + 176]",
            "mov {t3}, qword ptr [{s} + 184]",
            "sub {t0}, qword ptr [{s} + 480]",
            "sbb {t1}, qword ptr [{s} + 488]",
            "sbb {t2}, qword ptr [{s} + 496]",
            "sbb {t3}, qword ptr [{s} + 504]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 160], {t0}",
            "mov qword ptr [{s} + 168], {t1}",
            "mov qword ptr [{s} + 176], {t2}",
            "mov qword ptr [{s} + 184], {t3}",
            // X3 = sub(X3, t)
            "mov {t0}, qword ptr [{s} + 160]",
            "mov {t1}, qword ptr [{s} + 168]",
            "mov {t2}, qword ptr [{s} + 176]",
            "mov {t3}, qword ptr [{s} + 184]",
            "sub {t0}, qword ptr [{s} + 544]",
            "sbb {t1}, qword ptr [{s} + 552]",
            "sbb {t2}, qword ptr [{s} + 560]",
            "sbb {t3}, qword ptr [{s} + 568]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 160], {t0}",
            "mov qword ptr [{s} + 168], {t1}",
            "mov qword ptr [{s} + 176], {t2}",
            "mov qword ptr [{s} + 184], {t3}",
            // t3 = sub(v, X3)
            "mov {t0}, qword ptr [{s} + 512]",
            "mov {t1}, qword ptr [{s} + 520]",
            "mov {t2}, qword ptr [{s} + 528]",
            "mov {t3}, qword ptr [{s} + 536]",
            "sub {t0}, qword ptr [{s} + 160]",
            "sbb {t1}, qword ptr [{s} + 168]",
            "sbb {t2}, qword ptr [{s} + 176]",
            "sbb {t3}, qword ptr [{s} + 184]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 608], {t0}",
            "mov qword ptr [{s} + 616], {t1}",
            "mov qword ptr [{s} + 624], {t2}",
            "mov qword ptr [{s} + 632], {t3}",
            // Z3 = sub(Z3, z1z1)
            "mov {t0}, qword ptr [{s} + 224]",
            "mov {t1}, qword ptr [{s} + 232]",
            "mov {t2}, qword ptr [{s} + 240]",
            "mov {t3}, qword ptr [{s} + 248]",
            "sub {t0}, qword ptr [{s} + 320]",
            "sbb {t1}, qword ptr [{s} + 328]",
            "sbb {t2}, qword ptr [{s} + 336]",
            "sbb {t3}, qword ptr [{s} + 344]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 224], {t0}",
            "mov qword ptr [{s} + 232], {t1}",
            "mov qword ptr [{s} + 240], {t2}",
            "mov qword ptr [{s} + 248], {t3}",
            // Y3 = mul(R, t3)
            "mov rdx, qword ptr [{s} + 608]",
            "mulx {t1}, {t0}, qword ptr [{s} + 288]",
            "mulx {t2}, rax, qword ptr [{s} + 296]",
            "add {t1}, rax",
            "mulx {t3}, rax, qword ptr [{s} + 304]",
            "adc {t2}, rax",
            "mulx {t4}, rax, qword ptr [{s} + 312]",
            "adc {t3}, rax",
            "adc {t4}, 0",
            "xor {t5:e}, {t5:e}",
            "mov rax, {t0}",
            "shl rax, 32",
            "mov {c}, {t0}",
            "shr {c}, 32",
            "mov rdx, {t0}",
            "sub rdx, rax",
            "sbb {t0}, {c}",
            "add {t1}, rax",
            "adc {t2}, {c}",
            "adc {t3}, rdx",
            "adc {t4}, {t0}",
            "adc {t5}, 0",
            "mov rdx, qword ptr [{s} + 616]",
            "xor {t0:e}, {t0:e}",
            "mulx {c}, rax, qword ptr [{s} + 288]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 296]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 304]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 312]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mov eax, 0",
            "adcx {t5}, rax",
            "adox {t0}, rax",
            "adcx {t0}, rax",
            "mov rax, {t1}",
            "shl rax, 32",
            "mov {c}, {t1}",
            "shr {c}, 32",
            "mov rdx, {t1}",
            "sub rdx, rax",
            "sbb {t1}, {c}",
            "add {t2}, rax",
            "adc {t3}, {c}",
            "adc {t4}, rdx",
            "adc {t5}, {t1}",
            "adc {t0}, 0",
            "mov rdx, qword ptr [{s} + 624]",
            "xor {t1:e}, {t1:e}",
            "mulx {c}, rax, qword ptr [{s} + 288]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 296]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 304]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 312]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mov eax, 0",
            "adcx {t0}, rax",
            "adox {t1}, rax",
            "adcx {t1}, rax",
            "mov rax, {t2}",
            "shl rax, 32",
            "mov {c}, {t2}",
            "shr {c}, 32",
            "mov rdx, {t2}",
            "sub rdx, rax",
            "sbb {t2}, {c}",
            "add {t3}, rax",
            "adc {t4}, {c}",
            "adc {t5}, rdx",
            "adc {t0}, {t2}",
            "adc {t1}, 0",
            "mov rdx, qword ptr [{s} + 632]",
            "xor {t2:e}, {t2:e}",
            "mulx {c}, rax, qword ptr [{s} + 288]",
            "adcx {t3}, rax",
            "adox {t4}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 296]",
            "adcx {t4}, rax",
            "adox {t5}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 304]",
            "adcx {t5}, rax",
            "adox {t0}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 312]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mov eax, 0",
            "adcx {t1}, rax",
            "adox {t2}, rax",
            "adcx {t2}, rax",
            "mov rax, {t3}",
            "shl rax, 32",
            "mov {c}, {t3}",
            "shr {c}, 32",
            "mov rdx, {t3}",
            "sub rdx, rax",
            "sbb {t3}, {c}",
            "add {t4}, rax",
            "adc {t5}, {c}",
            "adc {t0}, rdx",
            "adc {t1}, {t3}",
            "adc {t2}, 0",
            "mov rax, {t4}",
            "mov rdx, {t5}",
            "mov {c}, {t0}",
            "mov {t3}, {t1}",
            "mov {t6:e}, 0xffffffff",
            "mov {t7}, 0xffffffff00000001",
            "sub rax, -1",
            "sbb rdx, {t6}",
            "sbb {c}, 0",
            "sbb {t3}, {t7}",
            "sbb {t2}, 0",
            "cmovc rax, {t4}",
            "cmovc rdx, {t5}",
            "cmovc {c}, {t0}",
            "cmovc {t3}, {t1}",
            "mov qword ptr [{s} + 192], rax",
            "mov qword ptr [{s} + 200], rdx",
            "mov qword ptr [{s} + 208], {c}",
            "mov qword ptr [{s} + 216], {t3}",
            // Z3 = sub(Z3, hh)
            "mov {t0}, qword ptr [{s} + 224]",
            "mov {t1}, qword ptr [{s} + 232]",
            "mov {t2}, qword ptr [{s} + 240]",
            "mov {t3}, qword ptr [{s} + 248]",
            "sub {t0}, qword ptr [{s} + 416]",
            "sbb {t1}, qword ptr [{s} + 424]",
            "sbb {t2}, qword ptr [{s} + 432]",
            "sbb {t3}, qword ptr [{s} + 440]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 224], {t0}",
            "mov qword ptr [{s} + 232], {t1}",
            "mov qword ptr [{s} + 240], {t2}",
            "mov qword ptr [{s} + 248], {t3}",
            // t2 = add(t2, t2)
            "mov {t0}, qword ptr [{s} + 576]",
            "mov {t1}, qword ptr [{s} + 584]",
            "mov {t2}, qword ptr [{s} + 592]",
            "mov {t3}, qword ptr [{s} + 600]",
            "xor {c:e}, {c:e}",
            "add {t0}, qword ptr [{s} + 576]",
            "adc {t1}, qword ptr [{s} + 584]",
            "adc {t2}, qword ptr [{s} + 592]",
            "adc {t3}, qword ptr [{s} + 600]",
            "adc {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 576], {t4}",
            "mov qword ptr [{s} + 584], {t5}",
            "mov qword ptr [{s} + 592], {t6}",
            "mov qword ptr [{s} + 600], {t7}",
            // Y3 = sub(Y3, t2)
            "mov {t0}, qword ptr [{s} + 192]",
            "mov {t1}, qword ptr [{s} + 200]",
            "mov {t2}, qword ptr [{s} + 208]",
            "mov {t3}, qword ptr [{s} + 216]",
            "sub {t0}, qword ptr [{s} + 576]",
            "sbb {t1}, qword ptr [{s} + 584]",
            "sbb {t2}, qword ptr [{s} + 592]",
            "sbb {t3}, qword ptr [{s} + 600]",
            "sbb {c}, {c}",
            "mov rax, {c}",
            "shr rax, 32",
            "mov rdx, 0xffffffff00000001",
            "and rdx, {c}",
            "add {t0}, {c}",
            "adc {t1}, rax",
            "adc {t2}, 0",
            "adc {t3}, rdx",
            "mov qword ptr [{s} + 192], {t0}",
            "mov qword ptr [{s} + 200], {t1}",
            "mov qword ptr [{s} + 208], {t2}",
            "mov qword ptr [{s} + 216], {t3}",
            s = in(reg) buf.as_mut_ptr(),
            t0 = out(reg) _,
            t1 = out(reg) _,
            t2 = out(reg) _,
            t3 = out(reg) _,
            t4 = out(reg) _,
            t5 = out(reg) _,
            t6 = out(reg) _,
            t7 = out(reg) _,
            c = out(reg) _,
            out("rax") _,
            out("rdx") _,
            options(nostack),
        );
    }
}
#[allow(dead_code)]
pub const ADD_AFFINE_X1: usize = 0;
#[allow(dead_code)]
pub const ADD_AFFINE_Y1: usize = 1;
#[allow(dead_code)]
pub const ADD_AFFINE_Z1: usize = 2;
#[allow(dead_code)]
pub const ADD_AFFINE_X2: usize = 3;
#[allow(dead_code)]
pub const ADD_AFFINE_Y2: usize = 4;
#[allow(dead_code)]
pub const ADD_AFFINE_X3: usize = 5;
#[allow(dead_code)]
pub const ADD_AFFINE_Y3: usize = 6;
#[allow(dead_code)]
pub const ADD_AFFINE_Z3: usize = 7;
#[allow(dead_code)]
pub const ADD_AFFINE_H: usize = 8;
#[allow(dead_code)]
pub const ADD_AFFINE_R: usize = 9;

/// `3 A - 8 B`, alone, for the tests: a fused operation the doubling uses.
///
/// Slots, 32 bytes each, in order: `A`, `B`, `OUT`.
///
/// # Safety
///
/// The CPU must support BMI2 and ADX; see `super::p256_x86::has_adx`.
#[cfg(test)]
#[inline(never)]
pub unsafe fn cms_3_8_adx(buf: &mut [MaybeUninit<u64>; 12]) {
    // SAFETY: the caller guarantees the instructions exist. The block reads
    // and writes only the buffer behind `{s}`, a live, exclusive reference
    // of 96 bytes, at offsets below that; it writes only the
    // registers it declares, and does not touch the stack. It reads no slot
    // before writing it but the inputs, which the caller has written.
    unsafe {
        asm!(
            ".p2align 6",
            // OUT = cms(A, B, 3, 8)
            "mov {t0}, -1",
            "sub {t0}, qword ptr [{s} + 32]",
            "mov {t1:e}, 0xffffffff",
            "sbb {t1}, qword ptr [{s} + 40]",
            "mov {t2:e}, 0",
            "sbb {t2}, qword ptr [{s} + 48]",
            "mov {t3}, 0xffffffff00000001",
            "sbb {t3}, qword ptr [{s} + 56]",
            "mov {t4}, {t3}",
            "shr {t4}, 61",
            "shld {t3}, {t2}, 3",
            "shld {t2}, {t1}, 3",
            "shld {t1}, {t0}, 3",
            "shl {t0}, 3",
            "mov edx, 3",
            "xor {t6:e}, {t6:e}",
            "mulx {c}, rax, qword ptr [{s}]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 8]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 16]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx rdx, rax, qword ptr [{s} + 24]",
            "adcx {t3}, rax",
            "adox rdx, {t4}",
            "adcx rdx, {t6}",
            "add rdx, 1",
            "mov rax, rdx",
            "shl rax, 32",
            "mov {c}, rax",
            "neg rax",
            "sub {c}, rdx",
            "sub {c}, 1",
            "add {t0}, rdx",
            "adc {t1}, rax",
            "adc {t2}, -1",
            "adc {t3}, {c}",
            "cmc",
            "sbb rax, rax",
            "mov rdx, rax",
            "shr rdx, 32",
            "mov {c}, 0xffffffff00000001",
            "and {c}, rax",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, 0",
            "adc {t3}, {c}",
            "mov qword ptr [{s} + 64], {t0}",
            "mov qword ptr [{s} + 72], {t1}",
            "mov qword ptr [{s} + 80], {t2}",
            "mov qword ptr [{s} + 88], {t3}",
            s = in(reg) buf.as_mut_ptr(),
            t0 = out(reg) _,
            t1 = out(reg) _,
            t2 = out(reg) _,
            t3 = out(reg) _,
            t4 = out(reg) _,
            t6 = out(reg) _,
            c = out(reg) _,
            out("rax") _,
            out("rdx") _,
            options(nostack),
        );
    }
}
#[cfg(test)]
#[allow(dead_code)]
pub const CMS_3_8_OUT: usize = 2;

/// `4 A - 1 B`, alone, for the tests: a fused operation the doubling uses.
///
/// Slots, 32 bytes each, in order: `A`, `B`, `OUT`.
///
/// # Safety
///
/// The CPU must support BMI2 and ADX; see `super::p256_x86::has_adx`.
#[cfg(test)]
#[inline(never)]
pub unsafe fn cms_4_1_adx(buf: &mut [MaybeUninit<u64>; 12]) {
    // SAFETY: the caller guarantees the instructions exist. The block reads
    // and writes only the buffer behind `{s}`, a live, exclusive reference
    // of 96 bytes, at offsets below that; it writes only the
    // registers it declares, and does not touch the stack. It reads no slot
    // before writing it but the inputs, which the caller has written.
    unsafe {
        asm!(
            ".p2align 6",
            // OUT = cms(A, B, 4, 1)
            "xor {t4:e}, {t4:e}",
            "mov {t0}, -1",
            "sub {t0}, qword ptr [{s} + 32]",
            "mov {t1:e}, 0xffffffff",
            "sbb {t1}, qword ptr [{s} + 40]",
            "mov {t2:e}, 0",
            "sbb {t2}, qword ptr [{s} + 48]",
            "mov {t3}, 0xffffffff00000001",
            "sbb {t3}, qword ptr [{s} + 56]",
            "mov edx, 4",
            "xor {t6:e}, {t6:e}",
            "mulx {c}, rax, qword ptr [{s}]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 8]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 16]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx rdx, rax, qword ptr [{s} + 24]",
            "adcx {t3}, rax",
            "adox rdx, {t4}",
            "adcx rdx, {t6}",
            "add rdx, 1",
            "mov rax, rdx",
            "shl rax, 32",
            "mov {c}, rax",
            "neg rax",
            "sub {c}, rdx",
            "sub {c}, 1",
            "add {t0}, rdx",
            "adc {t1}, rax",
            "adc {t2}, -1",
            "adc {t3}, {c}",
            "cmc",
            "sbb rax, rax",
            "mov rdx, rax",
            "shr rdx, 32",
            "mov {c}, 0xffffffff00000001",
            "and {c}, rax",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, 0",
            "adc {t3}, {c}",
            "mov qword ptr [{s} + 64], {t0}",
            "mov qword ptr [{s} + 72], {t1}",
            "mov qword ptr [{s} + 80], {t2}",
            "mov qword ptr [{s} + 88], {t3}",
            s = in(reg) buf.as_mut_ptr(),
            t0 = out(reg) _,
            t1 = out(reg) _,
            t2 = out(reg) _,
            t3 = out(reg) _,
            t4 = out(reg) _,
            t6 = out(reg) _,
            c = out(reg) _,
            out("rax") _,
            out("rdx") _,
            options(nostack),
        );
    }
}
#[cfg(test)]
#[allow(dead_code)]
pub const CMS_4_1_OUT: usize = 2;

/// `12 A - 9 B`, alone, for the tests: a fused operation the doubling uses.
///
/// Slots, 32 bytes each, in order: `A`, `B`, `OUT`.
///
/// # Safety
///
/// The CPU must support BMI2 and ADX; see `super::p256_x86::has_adx`.
#[cfg(test)]
#[inline(never)]
pub unsafe fn cms_12_9_adx(buf: &mut [MaybeUninit<u64>; 12]) {
    // SAFETY: the caller guarantees the instructions exist. The block reads
    // and writes only the buffer behind `{s}`, a live, exclusive reference
    // of 96 bytes, at offsets below that; it writes only the
    // registers it declares, and does not touch the stack. It reads no slot
    // before writing it but the inputs, which the caller has written.
    unsafe {
        asm!(
            ".p2align 6",
            // OUT = cms(A, B, 12, 9)
            "mov {t0}, -1",
            "sub {t0}, qword ptr [{s} + 32]",
            "mov {t1:e}, 0xffffffff",
            "sbb {t1}, qword ptr [{s} + 40]",
            "mov {t2:e}, 0",
            "sbb {t2}, qword ptr [{s} + 48]",
            "mov {t3}, 0xffffffff00000001",
            "sbb {t3}, qword ptr [{s} + 56]",
            "mov edx, 9",
            "mulx {c}, {t0}, {t0}",
            "mulx {t5}, {t1}, {t1}",
            "add {t1}, {c}",
            "mulx {c}, {t2}, {t2}",
            "adc {t2}, {t5}",
            "mulx {t4}, {t3}, {t3}",
            "adc {t3}, {c}",
            "adc {t4}, 0",
            "mov edx, 12",
            "xor {t6:e}, {t6:e}",
            "mulx {c}, rax, qword ptr [{s}]",
            "adcx {t0}, rax",
            "adox {t1}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 8]",
            "adcx {t1}, rax",
            "adox {t2}, {c}",
            "mulx {c}, rax, qword ptr [{s} + 16]",
            "adcx {t2}, rax",
            "adox {t3}, {c}",
            "mulx rdx, rax, qword ptr [{s} + 24]",
            "adcx {t3}, rax",
            "adox rdx, {t4}",
            "adcx rdx, {t6}",
            "add rdx, 1",
            "mov rax, rdx",
            "shl rax, 32",
            "mov {c}, rax",
            "neg rax",
            "sub {c}, rdx",
            "sub {c}, 1",
            "add {t0}, rdx",
            "adc {t1}, rax",
            "adc {t2}, -1",
            "adc {t3}, {c}",
            "cmc",
            "sbb rax, rax",
            "mov rdx, rax",
            "shr rdx, 32",
            "mov {c}, 0xffffffff00000001",
            "and {c}, rax",
            "add {t0}, rax",
            "adc {t1}, rdx",
            "adc {t2}, 0",
            "adc {t3}, {c}",
            "mov qword ptr [{s} + 64], {t0}",
            "mov qword ptr [{s} + 72], {t1}",
            "mov qword ptr [{s} + 80], {t2}",
            "mov qword ptr [{s} + 88], {t3}",
            s = in(reg) buf.as_mut_ptr(),
            t0 = out(reg) _,
            t1 = out(reg) _,
            t2 = out(reg) _,
            t3 = out(reg) _,
            t4 = out(reg) _,
            t5 = out(reg) _,
            t6 = out(reg) _,
            c = out(reg) _,
            out("rax") _,
            out("rdx") _,
            options(nostack),
        );
    }
}
#[cfg(test)]
#[allow(dead_code)]
pub const CMS_12_9_OUT: usize = 2;

/// `3 * A`, alone, for the tests: the small-multiple operation the formulas use.
///
/// Slots, 32 bytes each, in order: `A`, `OUT`.
///
/// # Safety
///
/// The CPU must support BMI2 and ADX; see `super::p256_x86::has_adx`.
#[cfg(test)]
#[inline(never)]
pub unsafe fn times3_adx(buf: &mut [MaybeUninit<u64>; 8]) {
    // SAFETY: the caller guarantees the instructions exist. The block reads
    // and writes only the buffer behind `{s}`, a live, exclusive reference
    // of 64 bytes, at offsets below that; it writes only the
    // registers it declares, and does not touch the stack. It reads no slot
    // before writing it but the inputs, which the caller has written.
    unsafe {
        asm!(
            ".p2align 6",
            // OUT = mulk(A, 3)
            "mov {t0}, qword ptr [{s}]",
            "mov {t1}, qword ptr [{s} + 8]",
            "mov {t2}, qword ptr [{s} + 16]",
            "mov {t3}, qword ptr [{s} + 24]",
            "mov {t4}, {t3}",
            "shr {t4}, 63",
            "shld {t3}, {t2}, 1",
            "shld {t2}, {t1}, 1",
            "shld {t1}, {t0}, 1",
            "shl {t0}, 1",
            "add {t0}, qword ptr [{s}]",
            "adc {t1}, qword ptr [{s} + 8]",
            "adc {t2}, qword ptr [{s} + 16]",
            "adc {t3}, qword ptr [{s} + 24]",
            "adc {t4}, 0",
            "mov rax, {t4}",
            "shl rax, 32",
            "xor {c:e}, {c:e}",
            "add {t0}, {t4}",
            "adc {t1}, 0",
            "adc {t2}, 0",
            "adc {t3}, rax",
            "adc {c}, 0",
            "sub {t1}, rax",
            "sbb {t2}, 0",
            "sbb {t3}, {t4}",
            "sbb {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 32], {t4}",
            "mov qword ptr [{s} + 40], {t5}",
            "mov qword ptr [{s} + 48], {t6}",
            "mov qword ptr [{s} + 56], {t7}",
            s = in(reg) buf.as_mut_ptr(),
            t0 = out(reg) _,
            t1 = out(reg) _,
            t2 = out(reg) _,
            t3 = out(reg) _,
            t4 = out(reg) _,
            t5 = out(reg) _,
            t6 = out(reg) _,
            t7 = out(reg) _,
            c = out(reg) _,
            out("rax") _,
            out("rdx") _,
            options(nostack),
        );
    }
}
#[cfg(test)]
#[allow(dead_code)]
pub const TIMES3_OUT: usize = 1;

/// `4 * A`, alone, for the tests: the small-multiple operation the formulas use.
///
/// Slots, 32 bytes each, in order: `A`, `OUT`.
///
/// # Safety
///
/// The CPU must support BMI2 and ADX; see `super::p256_x86::has_adx`.
#[cfg(test)]
#[inline(never)]
pub unsafe fn times4_adx(buf: &mut [MaybeUninit<u64>; 8]) {
    // SAFETY: the caller guarantees the instructions exist. The block reads
    // and writes only the buffer behind `{s}`, a live, exclusive reference
    // of 64 bytes, at offsets below that; it writes only the
    // registers it declares, and does not touch the stack. It reads no slot
    // before writing it but the inputs, which the caller has written.
    unsafe {
        asm!(
            ".p2align 6",
            // OUT = mulk(A, 4)
            "mov {t0}, qword ptr [{s}]",
            "mov {t1}, qword ptr [{s} + 8]",
            "mov {t2}, qword ptr [{s} + 16]",
            "mov {t3}, qword ptr [{s} + 24]",
            "mov {t4}, {t3}",
            "shr {t4}, 62",
            "shld {t3}, {t2}, 2",
            "shld {t2}, {t1}, 2",
            "shld {t1}, {t0}, 2",
            "shl {t0}, 2",
            "mov rax, {t4}",
            "shl rax, 32",
            "xor {c:e}, {c:e}",
            "add {t0}, {t4}",
            "adc {t1}, 0",
            "adc {t2}, 0",
            "adc {t3}, rax",
            "adc {c}, 0",
            "sub {t1}, rax",
            "sbb {t2}, 0",
            "sbb {t3}, {t4}",
            "sbb {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 32], {t4}",
            "mov qword ptr [{s} + 40], {t5}",
            "mov qword ptr [{s} + 48], {t6}",
            "mov qword ptr [{s} + 56], {t7}",
            s = in(reg) buf.as_mut_ptr(),
            t0 = out(reg) _,
            t1 = out(reg) _,
            t2 = out(reg) _,
            t3 = out(reg) _,
            t4 = out(reg) _,
            t5 = out(reg) _,
            t6 = out(reg) _,
            t7 = out(reg) _,
            c = out(reg) _,
            out("rax") _,
            out("rdx") _,
            options(nostack),
        );
    }
}
#[cfg(test)]
#[allow(dead_code)]
pub const TIMES4_OUT: usize = 1;

/// `8 * A`, alone, for the tests: the small-multiple operation the formulas use.
///
/// Slots, 32 bytes each, in order: `A`, `OUT`.
///
/// # Safety
///
/// The CPU must support BMI2 and ADX; see `super::p256_x86::has_adx`.
#[cfg(test)]
#[inline(never)]
pub unsafe fn times8_adx(buf: &mut [MaybeUninit<u64>; 8]) {
    // SAFETY: the caller guarantees the instructions exist. The block reads
    // and writes only the buffer behind `{s}`, a live, exclusive reference
    // of 64 bytes, at offsets below that; it writes only the
    // registers it declares, and does not touch the stack. It reads no slot
    // before writing it but the inputs, which the caller has written.
    unsafe {
        asm!(
            ".p2align 6",
            // OUT = mulk(A, 8)
            "mov {t0}, qword ptr [{s}]",
            "mov {t1}, qword ptr [{s} + 8]",
            "mov {t2}, qword ptr [{s} + 16]",
            "mov {t3}, qword ptr [{s} + 24]",
            "mov {t4}, {t3}",
            "shr {t4}, 61",
            "shld {t3}, {t2}, 3",
            "shld {t2}, {t1}, 3",
            "shld {t1}, {t0}, 3",
            "shl {t0}, 3",
            "mov rax, {t4}",
            "shl rax, 32",
            "xor {c:e}, {c:e}",
            "add {t0}, {t4}",
            "adc {t1}, 0",
            "adc {t2}, 0",
            "adc {t3}, rax",
            "adc {c}, 0",
            "sub {t1}, rax",
            "sbb {t2}, 0",
            "sbb {t3}, {t4}",
            "sbb {c}, 0",
            "mov {t4}, {t0}",
            "mov {t5}, {t1}",
            "mov {t6}, {t2}",
            "mov {t7}, {t3}",
            "mov eax, 0xffffffff",
            "mov rdx, 0xffffffff00000001",
            "sub {t4}, -1",
            "sbb {t5}, rax",
            "sbb {t6}, 0",
            "sbb {t7}, rdx",
            "sbb {c}, 0",
            "cmovc {t4}, {t0}",
            "cmovc {t5}, {t1}",
            "cmovc {t6}, {t2}",
            "cmovc {t7}, {t3}",
            "mov qword ptr [{s} + 32], {t4}",
            "mov qword ptr [{s} + 40], {t5}",
            "mov qword ptr [{s} + 48], {t6}",
            "mov qword ptr [{s} + 56], {t7}",
            s = in(reg) buf.as_mut_ptr(),
            t0 = out(reg) _,
            t1 = out(reg) _,
            t2 = out(reg) _,
            t3 = out(reg) _,
            t4 = out(reg) _,
            t5 = out(reg) _,
            t6 = out(reg) _,
            t7 = out(reg) _,
            c = out(reg) _,
            out("rax") _,
            out("rdx") _,
            options(nostack),
        );
    }
}
#[cfg(test)]
#[allow(dead_code)]
pub const TIMES8_OUT: usize = 1;