From 7dc094b2917544123a3f9582a14c581dec51b809 Mon Sep 17 00:00:00 2001 From: "Pavel V. Shatov (Meister)" Date: Mon, 24 Sep 2018 21:36:19 +0300 Subject: Shared piece of microcode (modular inversion) --- curve25519/curve25519_fpga_microcode.cpp | 243 +++++++++++++++++++++++++++++++ curve25519/curve25519_fpga_microcode.h | 126 ++++++++++++++++ 2 files changed, 369 insertions(+) create mode 100644 curve25519/curve25519_fpga_microcode.cpp create mode 100644 curve25519/curve25519_fpga_microcode.h diff --git a/curve25519/curve25519_fpga_microcode.cpp b/curve25519/curve25519_fpga_microcode.cpp new file mode 100644 index 0000000..37d2f8e --- /dev/null +++ b/curve25519/curve25519_fpga_microcode.cpp @@ -0,0 +1,243 @@ +//------------------------------------------------------------------------------ +// +// curve25519_fpga_microcode.cpp +// ------------------------------------- +// Microcode Architecture for Curve25519 +// +// Authors: Pavel Shatov +// +// Copyright (c) 2018 NORDUnet A/S +// +// Redistribution and use in source and binary forms, with or without +// modification, are permitted provided that the following conditions are met: +// +// - Redistributions of source code must retain the above copyright notice, +// this list of conditions and the following disclaimer. +// +// - Redistributions in binary form must reproduce the above copyright notice, +// this list of conditions and the following disclaimer in the documentation +// and/or other materials provided with the distribution. +// +// - Neither the name of the NORDUnet nor the names of its contributors may be +// used to endorse or promote products derived from this software without +// specific prior written permission. +// +// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" +// AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +// IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +// ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE +// LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR +// CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF +// SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS +// INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN +// CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) +// ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE +// POSSIBILITY OF SUCH DAMAGE. +// +//------------------------------------------------------------------------------ + + +//------------------------------------------------------------------------------ +// Headers +//------------------------------------------------------------------------------ +#include "curve25519_fpga_model.h" + + +//------------------------------------------------------------------------------ +void uop_move (UOP_BANK src, int s_op, + UOP_BANK dst, int d_op, + FPGA_BUFFER *buf_lo, FPGA_BUFFER *buf_hi) +//------------------------------------------------------------------------------ +{ + FPGA_BUFFER *s_ptr = NULL; + FPGA_BUFFER *d_ptr = NULL; + + if (src == BANK_LO) s_ptr = &buf_lo[s_op]; + if (src == BANK_HI) s_ptr = &buf_hi[s_op]; + if (dst == BANK_LO) d_ptr = &buf_lo[d_op]; + if (dst == BANK_HI) d_ptr = &buf_hi[d_op]; + + fpga_multiword_copy(s_ptr, d_ptr); +} + + +//------------------------------------------------------------------------------ +void uop_move2 (UOP_BANK src, int s_op1, int s_op2, + UOP_BANK dst, int d_op1, int d_op2, + FPGA_BUFFER *buf_lo, FPGA_BUFFER *buf_hi) +//------------------------------------------------------------------------------ +{ + uop_move(src, s_op1, dst, d_op1, buf_lo, buf_hi); + uop_move(src, s_op2, dst, d_op2, buf_lo, buf_hi); +} + + +//------------------------------------------------------------------------------ +void uop_calc (UOP_MATH math, + UOP_BANK src, int s_op1, int s_op2, + UOP_BANK dst, int d_op, + FPGA_BUFFER *buf_lo, FPGA_BUFFER *buf_hi, + UOP_MODULUS mod) +//------------------------------------------------------------------------------ +{ + FPGA_BUFFER *s_ptr1 = NULL; + FPGA_BUFFER *s_ptr2 = NULL; + FPGA_BUFFER *d_ptr = NULL; + FPGA_BUFFER *n_ptr = NULL; + + if (src == BANK_LO) + { s_ptr1 = &buf_lo[s_op1]; + s_ptr2 = &buf_lo[s_op2]; + } + if (src == BANK_HI) + { s_ptr1 = &buf_hi[s_op1]; + s_ptr2 = &buf_hi[s_op2]; + } + if (dst == BANK_LO) + { d_ptr = &buf_lo[d_op]; + } + if (dst == BANK_HI) + { d_ptr = &buf_hi[d_op]; + } + + if (mod == MOD_1P) n_ptr = &CURVE25519_1P; + if (mod == MOD_2P) n_ptr = &CURVE25519_2P; + + if (math == ADD) fpga_modular_add(s_ptr1, s_ptr2, d_ptr, n_ptr); + if (math == SUB) fpga_modular_sub(s_ptr1, s_ptr2, d_ptr, n_ptr); + if (math == MUL) fpga_modular_mul(s_ptr1, s_ptr2, d_ptr, n_ptr); +} + + +//------------------------------------------------------------------------------ +void uop_load(const FPGA_BUFFER *mem, UOP_BANK dst, int d_op, FPGA_BUFFER *buf_lo, FPGA_BUFFER *buf_hi) +//------------------------------------------------------------------------------ +{ + FPGA_BUFFER *d_ptr = NULL; + if (dst == BANK_LO) d_ptr = &buf_lo[d_op]; + if (dst == BANK_HI) d_ptr = &buf_hi[d_op]; + + fpga_multiword_copy(mem, d_ptr); +} + + +//------------------------------------------------------------------------------ +void uop_stor(const FPGA_BUFFER *buf_lo, const FPGA_BUFFER *buf_hi, UOP_BANK src, int s_op, FPGA_BUFFER *mem) +//------------------------------------------------------------------------------ +{ + FPGA_BUFFER *s_ptr = NULL; + if (src == BANK_LO) + { s_ptr = (FPGA_BUFFER *)&buf_lo[s_op]; + } + if (src == BANK_HI) + { s_ptr = (FPGA_BUFFER *)&buf_hi[s_op]; + } + + fpga_multiword_copy(s_ptr, mem); +} + + +//------------------------------------------------------------------------------ +void fpga_modular_inv_microcode(FPGA_BUFFER *buf_lo, FPGA_BUFFER *buf_hi) +//------------------------------------------------------------------------------ +{ + int cyc_count; // counters + + // T_1 + uop_move2(BANK_LO, INVERT_T_1, INVERT_T_1, BANK_HI, INVERT_T_1, INVERT_T_1, buf_lo, buf_hi); + + // T_10 + uop_calc(MUL, BANK_LO, INVERT_T_1, INVERT_T_1, BANK_HI, INVERT_T_10, buf_lo, buf_hi, MOD_2P); + + // T_1001 + uop_calc(MUL, BANK_HI, INVERT_T_10, INVERT_T_10, BANK_LO, INVERT_R1, buf_lo, buf_hi, MOD_2P); + uop_calc(MUL, BANK_LO, INVERT_R1, INVERT_R1, BANK_HI, INVERT_R2, buf_lo, buf_hi, MOD_2P); + uop_calc(MUL, BANK_HI, INVERT_R2, INVERT_T_1, BANK_LO, INVERT_T_1001, buf_lo, buf_hi, MOD_2P); + + // T_1011 + uop_move2(BANK_HI, INVERT_T_10, INVERT_T_10, BANK_LO, INVERT_T_10, INVERT_T_10, buf_lo, buf_hi); + uop_calc(MUL, BANK_LO, INVERT_T_1001, INVERT_T_10, BANK_HI, INVERT_T_1011, buf_lo, buf_hi, MOD_2P); + + // T_X5 + uop_calc(MUL, BANK_HI, INVERT_T_1011, INVERT_T_1011, BANK_LO, INVERT_R1, buf_lo, buf_hi, MOD_2P); + uop_calc(MUL, BANK_LO, INVERT_R1, INVERT_T_1001, BANK_HI, INVERT_T_X5, buf_lo, buf_hi, MOD_2P); + + // T_X10 + uop_move2(BANK_HI, INVERT_T_X5, INVERT_T_X5, BANK_LO, INVERT_R1, INVERT_R1, buf_lo, buf_hi); + + for (cyc_count=0; cyc_count<4; cyc_count++) + if (!(cyc_count % 2)) uop_calc(MUL, BANK_LO, INVERT_R1, INVERT_R1, BANK_HI, INVERT_R2, buf_lo, buf_hi, MOD_2P); + else uop_calc(MUL, BANK_HI, INVERT_R2, INVERT_R2, BANK_LO, INVERT_R1, buf_lo, buf_hi, MOD_2P); + + uop_calc(MUL, BANK_LO, INVERT_R1, INVERT_R1, BANK_HI, INVERT_R2, buf_lo, buf_hi, MOD_2P); + uop_calc(MUL, BANK_HI, INVERT_R2, INVERT_T_X5, BANK_LO, INVERT_T_X10, buf_lo, buf_hi, MOD_2P); + + // T_X20 + uop_move2(BANK_LO, INVERT_T_X10, INVERT_T_X10, BANK_HI, INVERT_R1, INVERT_R1, buf_lo, buf_hi); + uop_move2(BANK_LO, INVERT_T_X10, INVERT_T_X10, BANK_HI, INVERT_T_X10, INVERT_T_X10, buf_lo, buf_hi); + + for (cyc_count=0; cyc_count<10; cyc_count++) + if (!(cyc_count % 2)) uop_calc(MUL, BANK_HI, INVERT_R1, INVERT_R1, BANK_LO, INVERT_R2, buf_lo, buf_hi, MOD_2P); + else uop_calc(MUL, BANK_LO, INVERT_R2, INVERT_R2, BANK_HI, INVERT_R1, buf_lo, buf_hi, MOD_2P); + + uop_calc(MUL, BANK_HI, INVERT_R1, INVERT_T_X10, BANK_LO, INVERT_T_X20, buf_lo, buf_hi, MOD_2P); + + // T_X40 + uop_move2(BANK_LO, INVERT_T_X20, INVERT_T_X20, BANK_HI, INVERT_R1, INVERT_R1, buf_lo, buf_hi); + uop_move2(BANK_LO, INVERT_T_X20, INVERT_T_X20, BANK_HI, INVERT_T_X20, INVERT_T_X20, buf_lo, buf_hi); + + for (cyc_count=0; cyc_count<20; cyc_count++) + if (!(cyc_count % 2)) uop_calc(MUL, BANK_HI, INVERT_R1, INVERT_R1, BANK_LO, INVERT_R2, buf_lo, buf_hi, MOD_2P); + else uop_calc(MUL, BANK_LO, INVERT_R2, INVERT_R2, BANK_HI, INVERT_R1, buf_lo, buf_hi, MOD_2P); + + uop_calc(MUL, BANK_HI, INVERT_R1, INVERT_T_X20, BANK_LO, INVERT_T_X40, buf_lo, buf_hi, MOD_2P); + + // T_X50 + uop_move2(BANK_LO, INVERT_T_X40, INVERT_T_X40, BANK_HI, INVERT_R1, INVERT_R1, buf_lo, buf_hi); + + for (cyc_count=0; cyc_count<10; cyc_count++) + if (!(cyc_count % 2)) uop_calc(MUL, BANK_HI, INVERT_R1, INVERT_R1, BANK_LO, INVERT_R2, buf_lo, buf_hi, MOD_2P); + else uop_calc(MUL, BANK_LO, INVERT_R2, INVERT_R2, BANK_HI, INVERT_R1, buf_lo, buf_hi, MOD_2P); + + uop_calc(MUL, BANK_HI, INVERT_R1, INVERT_T_X10, BANK_LO, INVERT_T_X50, buf_lo, buf_hi, MOD_2P); + + // T_X100 + uop_move2(BANK_LO, INVERT_T_X50, INVERT_T_X50, BANK_HI, INVERT_R1, INVERT_R1, buf_lo, buf_hi); + uop_move2(BANK_LO, INVERT_T_X50, INVERT_T_X50, BANK_HI, INVERT_T_X50, INVERT_T_X50, buf_lo, buf_hi); + + for (cyc_count=0; cyc_count<50; cyc_count++) + if (!(cyc_count % 2)) uop_calc(MUL, BANK_HI, INVERT_R1, INVERT_R1, BANK_LO, INVERT_R2, buf_lo, buf_hi, MOD_2P); + else uop_calc(MUL, BANK_LO, INVERT_R2, INVERT_R2, BANK_HI, INVERT_R1, buf_lo, buf_hi, MOD_2P); + + uop_calc(MUL, BANK_HI, INVERT_R1, INVERT_T_X50, BANK_LO, INVERT_T_X100, buf_lo, buf_hi, MOD_2P); + + uop_move2(BANK_LO, INVERT_T_X100, INVERT_T_X100, BANK_HI, INVERT_R1, INVERT_R1, buf_lo, buf_hi); + uop_move2(BANK_LO, INVERT_T_X100, INVERT_T_X100, BANK_HI, INVERT_T_X100, INVERT_T_X100, buf_lo, buf_hi); + + for (cyc_count=0; cyc_count<100; cyc_count++) + if (!(cyc_count % 2)) uop_calc(MUL, BANK_HI, INVERT_R1, INVERT_R1, BANK_LO, INVERT_R2, buf_lo, buf_hi, MOD_2P); + else uop_calc(MUL, BANK_LO, INVERT_R2, INVERT_R2, BANK_HI, INVERT_R1, buf_lo, buf_hi, MOD_2P); + + uop_calc(MUL, BANK_HI, INVERT_R1, INVERT_T_X100, BANK_LO, INVERT_R2, buf_lo, buf_hi, MOD_2P); + + for (cyc_count=0; cyc_count<50; cyc_count++) + if (!(cyc_count % 2)) uop_calc(MUL, BANK_LO, INVERT_R2, INVERT_R2, BANK_HI, INVERT_R1, buf_lo, buf_hi, MOD_2P); + else uop_calc(MUL, BANK_HI, INVERT_R1, INVERT_R1, BANK_LO, INVERT_R2, buf_lo, buf_hi, MOD_2P); + + uop_calc(MUL, BANK_LO, INVERT_R2, INVERT_T_X50, BANK_HI, INVERT_R1, buf_lo, buf_hi, MOD_2P); + + for (cyc_count=0; cyc_count<4; cyc_count++) + if (!(cyc_count % 2)) uop_calc(MUL, BANK_HI, INVERT_R1, INVERT_R1, BANK_LO, INVERT_R2, buf_lo, buf_hi, MOD_2P); + else uop_calc(MUL, BANK_LO, INVERT_R2, INVERT_R2, BANK_HI, INVERT_R1, buf_lo, buf_hi, MOD_2P); + + uop_calc(MUL, BANK_HI, INVERT_R1, INVERT_R1, BANK_LO, INVERT_R2, buf_lo, buf_hi, MOD_2P); + + uop_move2(BANK_HI, INVERT_T_1011, INVERT_T_1011, BANK_LO, INVERT_T_1011, INVERT_T_1011, buf_lo, buf_hi); + + uop_calc(MUL, BANK_LO, INVERT_R2, INVERT_T_1011, BANK_HI, INVERT_R1, buf_lo, buf_hi, MOD_2P); +} + + +//------------------------------------------------------------------------------ +// End-of-File +//------------------------------------------------------------------------------ diff --git a/curve25519/curve25519_fpga_microcode.h b/curve25519/curve25519_fpga_microcode.h new file mode 100644 index 0000000..10d29ca --- /dev/null +++ b/curve25519/curve25519_fpga_microcode.h @@ -0,0 +1,126 @@ +//------------------------------------------------------------------------------ +// +// curve25519_fpga_microcode.h +// ------------------------------------- +// Microcode Architecture for Curve25519 +// +// Authors: Pavel Shatov +// +// Copyright (c) 2018 NORDUnet A/S +// +// Redistribution and use in source and binary forms, with or without +// modification, are permitted provided that the following conditions are met: +// +// - Redistributions of source code must retain the above copyright notice, +// this list of conditions and the following disclaimer. +// +// - Redistributions in binary form must reproduce the above copyright notice, +// this list of conditions and the following disclaimer in the documentation +// and/or other materials provided with the distribution. +// +// - Neither the name of the NORDUnet nor the names of its contributors may be +// used to endorse or promote products derived from this software without +// specific prior written permission. +// +// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" +// AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +// IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +// ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE +// LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR +// CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF +// SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS +// INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN +// CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) +// ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE +// POSSIBILITY OF SUCH DAMAGE. +// +//------------------------------------------------------------------------------ + + +//------------------------------------------------------------------------------ +// Headers +//------------------------------------------------------------------------------ +#include + + +//------------------------------------------------------------------------------ +enum UOP_BANK +//------------------------------------------------------------------------------ +{ + BANK_LO, BANK_HI +}; + + +//-------------------------- +enum CURVE25519_UOP_OPERAND +//-------------------------- +{ + CONST_ZERO, + CONST_ONE, + + INVERT_R1, + INVERT_R2, + + INVERT_T_1, + INVERT_T_10, + INVERT_T_1001, + INVERT_T_1011, + + INVERT_T_X5, + INVERT_T_X10, + INVERT_T_X20, + INVERT_T_X40, + INVERT_T_X50, + INVERT_T_X100, + + CURVE25519_UOP_OPERAND_COUNT +}; + + +//------------------------------------------------------------------------------ +enum UOP_MODULUS +//------------------------------------------------------------------------------ +{ + MOD_1P, + MOD_2P +}; + + +//------------------------------------------------------------------------------ +enum UOP_MATH +//------------------------------------------------------------------------------ +{ + ADD, SUB, MUL +}; + + +//------------------------------------------------------------------------------ +// Prototypes (Micro-Operations) +//------------------------------------------------------------------------------ +void uop_move (UOP_BANK src, int s_op1, + UOP_BANK dst, int d_op1, + FPGA_BUFFER *buf_lo, FPGA_BUFFER *buf_hi); + +void uop_move2 (UOP_BANK src, int s_op1, int s_op2, + UOP_BANK dst, int d_op1, int d_op2, + FPGA_BUFFER *buf_lo, FPGA_BUFFER *buf_hi); + +void uop_calc (UOP_MATH math, + UOP_BANK src, int s_op1, int s_op2, + UOP_BANK dst, int d_op, + FPGA_BUFFER *buf_lo, FPGA_BUFFER *buf_hi, + UOP_MODULUS mod); + +void uop_load (const FPGA_BUFFER *mem, UOP_BANK dst, int d_op, FPGA_BUFFER *buf_lo, FPGA_BUFFER *buf_hi); +void uop_stor (const FPGA_BUFFER *buf_lo, const FPGA_BUFFER *buf_hi, UOP_BANK src, int s_op, FPGA_BUFFER *mem); + + +//------------------------------------------------------------------------------ +// Prototype (Macro-Operation) +//------------------------------------------------------------------------------ +void fpga_modular_inv_microcode (FPGA_BUFFER *buf_lo, FPGA_BUFFER *buf_hi); + + +//------------------------------------------------------------------------------ +// End-of-File +//------------------------------------------------------------------------------ -- cgit v1.2.3