/* * (C) 2026 Jack Lloyd * * Botan is released under the Simplified BSD License (see license.txt) */ #include #include #include #include #include namespace Botan { namespace Camellia_AVX512 { namespace { constexpr uint64_t pre123_a = gfni_matrix(R"( 1 1 1 0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 1 0 0 0 0 1 0 1 1 0 0 1 1 0 0 0 0 1 1 0 0 1 0 1 0 0 1 0 0 0 0 1 0 1 1 0 0 1 0 0 0 0 1 1 0)"); constexpr uint64_t pre4_a = gfni_matrix(R"( 1 1 0 1 1 0 1 1 0 1 1 0 0 1 0 0 1 0 1 0 0 0 0 1 0 1 1 0 0 1 1 1 0 0 0 1 1 0 0 0 0 1 0 0 1 0 0 1 0 1 0 1 1 0 0 0 0 0 0 0 1 1 0 1)"); constexpr uint8_t pre_c = 0b01000101; constexpr uint64_t post2_a = gfni_matrix(R"( 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 1 0 1 1 0 0 1 1 0 1 0 1 1 1 1 1 0 0 0 0 1 1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 1 1 1 0 0 1 1 1 1 1 1 1)"); constexpr uint64_t post3_a = gfni_matrix(R"( 0 1 1 0 0 1 1 0 1 0 1 1 1 1 1 0 0 0 0 1 1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 1 1 1 0 0 1 1 1 1 1 1 1 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 1)"); constexpr uint64_t post14_a = gfni_matrix(R"( 0 0 0 0 0 0 0 1 0 1 1 0 0 1 1 0 1 0 1 1 1 1 1 0 0 0 0 1 1 0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 1 1 1 0 0 1 1 1 1 1 1 1 0 0 0 1 1 1 0 0)"); // NOLINTBEGIN(portability-simd-intrinsics) BOTAN_FORCE_INLINE BOTAN_FN_ISA_AVX512_GFNI SIMD_8x64 camellia_f(SIMD_8x64 x) { const __m512i xr = x.raw(); /* * Camellia sends different bytes of each word through different sboxes; we * resolve this like cavemen by computing all 4 S-box variants over the full * vector in parallel, then blending the results. */ // Compute S1(x), S2(x), S3(x), S4(x) for all bytes const __m512i y123 = _mm512_gf2p8affine_epi64_epi8(xr, _mm512_set1_epi64(pre123_a), pre_c); const __m512i y4 = _mm512_gf2p8affine_epi64_epi8(xr, _mm512_set1_epi64(pre4_a), pre_c); const __m512i s1 = _mm512_gf2p8affineinv_epi64_epi8(y123, _mm512_set1_epi64(post14_a), 0x6E); const __m512i s2 = _mm512_gf2p8affineinv_epi64_epi8(y123, _mm512_set1_epi64(post2_a), 0xDC); const __m512i s3 = _mm512_gf2p8affineinv_epi64_epi8(y123, _mm512_set1_epi64(post3_a), 0x37); const __m512i s4 = _mm512_gf2p8affineinv_epi64_epi8(y4, _mm512_set1_epi64(post14_a), 0x6E); // Blend to find correct S(x) for each byte position auto sx = s1; sx = _mm512_mask_blend_epi8(__mmask64(0x4848484848484848), sx, s2); // s2 at bytes {3,6} sx = _mm512_mask_blend_epi8(__mmask64(0x2424242424242424), sx, s3); // s3 at bytes {2,5} sx = _mm512_mask_blend_epi8(__mmask64(0x1212121212121212), sx, s4); // s4 at bytes {1,4} // Linear mixing layer const auto P1 = _mm512_set_epi64(0x0808080908080809, 0x0000000100000001, 0x0808080908080809, 0x0000000100000001, 0x0808080908080809, 0x0000000100000001, 0x0808080908080809, 0x0000000100000001); const auto P2 = _mm512_set_epi64(0x09090A0A09090A0A, 0x0101020201010202, 0x09090A0A09090A0A, 0x0101020201010202, 0x09090A0A09090A0A, 0x0101020201010202, 0x09090A0A09090A0A, 0x0101020201010202); const auto P3 = _mm512_set_epi64(0x0A0B0B0B0A0B0B0B, 0x0203030302030303, 0x0A0B0B0B0A0B0B0B, 0x0203030302030303, 0x0A0B0B0B0A0B0B0B, 0x0203030302030303, 0x0A0B0B0B0A0B0B0B, 0x0203030302030303); const auto P4 = _mm512_set_epi64(0x0C0C0D0C0E0D0C0C, 0x0404050406050404, 0x0C0C0D0C0E0D0C0C, 0x0404050406050404, 0x0C0C0D0C0E0D0C0C, 0x0404050406050404, 0x0C0C0D0C0E0D0C0C, 0x0404050406050404); const auto P5 = _mm512_set_epi64(0x0D0E0E0D0F0E0D0F, 0x0506060507060507, 0x0D0E0E0D0F0E0D0F, 0x0506060507060507, 0x0D0E0E0D0F0E0D0F, 0x0506060507060507, 0x0D0E0E0D0F0E0D0F, 0x0506060507060507); const auto P6 = _mm512_set_epi64(0x0F0F0F0EFFFFFFFF, 0x07070706FFFFFFFF, 0x0F0F0F0EFFFFFFFF, 0x07070706FFFFFFFF, 0x0F0F0F0EFFFFFFFF, 0x07070706FFFFFFFF, 0x0F0F0F0EFFFFFFFF, 0x07070706FFFFFFFF); const auto t1 = SIMD_8x64(_mm512_shuffle_epi8(sx, P1)); const auto t2 = SIMD_8x64(_mm512_shuffle_epi8(sx, P2)); const auto t3 = SIMD_8x64(_mm512_shuffle_epi8(sx, P3)); const auto t4 = SIMD_8x64(_mm512_shuffle_epi8(sx, P4)); const auto t5 = SIMD_8x64(_mm512_shuffle_epi8(sx, P5)); const auto t6 = SIMD_8x64(_mm512_shuffle_epi8(sx, P6)); return (t1 ^ t2 ^ t3 ^ t4 ^ t5 ^ t6); } BOTAN_FORCE_INLINE BOTAN_FN_ISA_AVX512 SIMD_8x64 FL_8(SIMD_8x64 v, uint64_t K) { const auto Kv = _mm512_set1_epi64(K); auto vr = v.raw(); // x2 ^= rotl<1>(x1 & k1): AND, rotate 32-bit elements, shift high->low, XOR vr = _mm512_xor_si512(vr, _mm512_srli_epi64(_mm512_rol_epi32(_mm512_and_si512(vr, Kv), 1), 32)); // x1 ^= (x2 | k2): OR, shift low->high, XOR vr = _mm512_xor_si512(vr, _mm512_slli_epi64(_mm512_or_si512(vr, Kv), 32)); return SIMD_8x64(vr); } BOTAN_FORCE_INLINE BOTAN_FN_ISA_AVX512 SIMD_8x64 FLINV_8(SIMD_8x64 v, uint64_t K) { const auto Kv = _mm512_set1_epi64(K); auto vr = v.raw(); // x1 ^= (x2 | k2): OR, shift low->high, XOR vr = _mm512_xor_si512(vr, _mm512_slli_epi64(_mm512_or_si512(vr, Kv), 32)); // x2 ^= rotl<1>(x1 & k1): AND, rotate 32-bit elements, shift high->low, XOR vr = _mm512_xor_si512(vr, _mm512_srli_epi64(_mm512_rol_epi32(_mm512_and_si512(vr, Kv), 1), 32)); return SIMD_8x64(vr); } /* * Load 8 blocks, byte-swap, and deinterleave into L (even) and R (odd) halves */ BOTAN_FORCE_INLINE BOTAN_FN_ISA_AVX512 void load_and_deinterleave(const uint8_t in[], SIMD_8x64& L, SIMD_8x64& R) { const auto idx_l = _mm512_set_epi64(0x0E, 0x0C, 0x0A, 0x08, 0x06, 0x04, 0x02, 0x00); const auto idx_r = _mm512_set_epi64(0x0F, 0x0D, 0x0B, 0x09, 0x07, 0x05, 0x03, 0x01); auto A = SIMD_8x64::load_be(in); auto B = SIMD_8x64::load_be(in + 64); L = SIMD_8x64(_mm512_permutex2var_epi64(A.raw(), idx_l, B.raw())); R = SIMD_8x64(_mm512_permutex2var_epi64(A.raw(), idx_r, B.raw())); } /* * Interleave R/L halves (note swap), byte-swap, and store 8 blocks */ BOTAN_FORCE_INLINE BOTAN_FN_ISA_AVX512 void interleave_and_store(uint8_t out[], SIMD_8x64 L, SIMD_8x64 R) { const auto idx_lo = _mm512_set_epi64(0x0B, 0x03, 0x0A, 0x02, 0x09, 0x01, 0x08, 0x00); const auto idx_hi = _mm512_set_epi64(0x0F, 0x07, 0x0E, 0x06, 0x0D, 0x05, 0x0C, 0x04); auto A = SIMD_8x64(_mm512_permutex2var_epi64(R.raw(), idx_lo, L.raw())); auto B = SIMD_8x64(_mm512_permutex2var_epi64(R.raw(), idx_hi, L.raw())); A.store_be(out); B.store_be(out + 64); } // NOLINTEND(portability-simd-intrinsics) // Helpers for 6 round iterations BOTAN_FORCE_INLINE BOTAN_FN_ISA_AVX512_GFNI void six_e_rounds(SIMD_8x64& L, SIMD_8x64& R, std::span SK) { R ^= camellia_f(L ^ SIMD_8x64::splat(SK[0])); L ^= camellia_f(R ^ SIMD_8x64::splat(SK[1])); R ^= camellia_f(L ^ SIMD_8x64::splat(SK[2])); L ^= camellia_f(R ^ SIMD_8x64::splat(SK[3])); R ^= camellia_f(L ^ SIMD_8x64::splat(SK[4])); L ^= camellia_f(R ^ SIMD_8x64::splat(SK[5])); } BOTAN_FORCE_INLINE BOTAN_FN_ISA_AVX512_GFNI void six_d_rounds(SIMD_8x64& L, SIMD_8x64& R, std::span SK) { R ^= camellia_f(L ^ SIMD_8x64::splat(SK[5])); L ^= camellia_f(R ^ SIMD_8x64::splat(SK[4])); R ^= camellia_f(L ^ SIMD_8x64::splat(SK[3])); L ^= camellia_f(R ^ SIMD_8x64::splat(SK[2])); R ^= camellia_f(L ^ SIMD_8x64::splat(SK[1])); L ^= camellia_f(R ^ SIMD_8x64::splat(SK[0])); } BOTAN_FORCE_INLINE BOTAN_FN_ISA_AVX512_GFNI void six_e_rounds_x2( SIMD_8x64& L1, SIMD_8x64& R1, SIMD_8x64& L2, SIMD_8x64& R2, std::span SK) { const auto K0 = SIMD_8x64::splat(SK[0]); const auto K1 = SIMD_8x64::splat(SK[1]); const auto K2 = SIMD_8x64::splat(SK[2]); const auto K3 = SIMD_8x64::splat(SK[3]); const auto K4 = SIMD_8x64::splat(SK[4]); const auto K5 = SIMD_8x64::splat(SK[5]); R1 ^= camellia_f(L1 ^ K0); R2 ^= camellia_f(L2 ^ K0); L1 ^= camellia_f(R1 ^ K1); L2 ^= camellia_f(R2 ^ K1); R1 ^= camellia_f(L1 ^ K2); R2 ^= camellia_f(L2 ^ K2); L1 ^= camellia_f(R1 ^ K3); L2 ^= camellia_f(R2 ^ K3); R1 ^= camellia_f(L1 ^ K4); R2 ^= camellia_f(L2 ^ K4); L1 ^= camellia_f(R1 ^ K5); L2 ^= camellia_f(R2 ^ K5); } BOTAN_FORCE_INLINE BOTAN_FN_ISA_AVX512_GFNI void six_d_rounds_x2( SIMD_8x64& L1, SIMD_8x64& R1, SIMD_8x64& L2, SIMD_8x64& R2, std::span SK) { const auto K0 = SIMD_8x64::splat(SK[0]); const auto K1 = SIMD_8x64::splat(SK[1]); const auto K2 = SIMD_8x64::splat(SK[2]); const auto K3 = SIMD_8x64::splat(SK[3]); const auto K4 = SIMD_8x64::splat(SK[4]); const auto K5 = SIMD_8x64::splat(SK[5]); R1 ^= camellia_f(L1 ^ K5); R2 ^= camellia_f(L2 ^ K5); L1 ^= camellia_f(R1 ^ K4); L2 ^= camellia_f(R2 ^ K4); R1 ^= camellia_f(L1 ^ K3); R2 ^= camellia_f(L2 ^ K3); L1 ^= camellia_f(R1 ^ K2); L2 ^= camellia_f(R2 ^ K2); R1 ^= camellia_f(L1 ^ K1); R2 ^= camellia_f(L2 ^ K1); L1 ^= camellia_f(R1 ^ K0); L2 ^= camellia_f(R2 ^ K0); } BOTAN_FN_ISA_AVX512_GFNI void camellia_encrypt_x16_18r(const uint8_t in[], uint8_t out[], std::span SK) { SIMD_8x64 L1; SIMD_8x64 R1; SIMD_8x64 L2; SIMD_8x64 R2; load_and_deinterleave(in, L1, R1); load_and_deinterleave(in + 128, L2, R2); const auto K0 = SIMD_8x64::splat(SK[0]); const auto K1 = SIMD_8x64::splat(SK[1]); L1 ^= K0; L2 ^= K0; R1 ^= K1; R2 ^= K1; six_e_rounds_x2(L1, R1, L2, R2, SK.subspan(2)); L1 = FL_8(L1, SK[8]); L2 = FL_8(L2, SK[8]); R1 = FLINV_8(R1, SK[9]); R2 = FLINV_8(R2, SK[9]); six_e_rounds_x2(L1, R1, L2, R2, SK.subspan(10)); L1 = FL_8(L1, SK[16]); L2 = FL_8(L2, SK[16]); R1 = FLINV_8(R1, SK[17]); R2 = FLINV_8(R2, SK[17]); six_e_rounds_x2(L1, R1, L2, R2, SK.subspan(18)); const auto K24 = SIMD_8x64::splat(SK[24]); const auto K25 = SIMD_8x64::splat(SK[25]); R1 ^= K24; R2 ^= K24; L1 ^= K25; L2 ^= K25; interleave_and_store(out, L1, R1); interleave_and_store(out + 128, L2, R2); } BOTAN_FN_ISA_AVX512_GFNI void camellia_decrypt_x16_18r(const uint8_t in[], uint8_t out[], std::span SK) { SIMD_8x64 L1; SIMD_8x64 R1; SIMD_8x64 L2; SIMD_8x64 R2; load_and_deinterleave(in, L1, R1); load_and_deinterleave(in + 128, L2, R2); const auto K25 = SIMD_8x64::splat(SK[25]); const auto K24 = SIMD_8x64::splat(SK[24]); R1 ^= K25; R2 ^= K25; L1 ^= K24; L2 ^= K24; six_d_rounds_x2(L1, R1, L2, R2, SK.subspan(18)); L1 = FL_8(L1, SK[17]); L2 = FL_8(L2, SK[17]); R1 = FLINV_8(R1, SK[16]); R2 = FLINV_8(R2, SK[16]); six_d_rounds_x2(L1, R1, L2, R2, SK.subspan(10)); L1 = FL_8(L1, SK[9]); L2 = FL_8(L2, SK[9]); R1 = FLINV_8(R1, SK[8]); R2 = FLINV_8(R2, SK[8]); six_d_rounds_x2(L1, R1, L2, R2, SK.subspan(2)); const auto K1 = SIMD_8x64::splat(SK[1]); const auto K0 = SIMD_8x64::splat(SK[0]); L1 ^= K1; L2 ^= K1; R1 ^= K0; R2 ^= K0; interleave_and_store(out, L1, R1); interleave_and_store(out + 128, L2, R2); } BOTAN_FN_ISA_AVX512_GFNI void camellia_encrypt_x16_24r(const uint8_t in[], uint8_t out[], std::span SK) { SIMD_8x64 L1; SIMD_8x64 R1; SIMD_8x64 L2; SIMD_8x64 R2; load_and_deinterleave(in, L1, R1); load_and_deinterleave(in + 128, L2, R2); const auto K0 = SIMD_8x64::splat(SK[0]); const auto K1 = SIMD_8x64::splat(SK[1]); L1 ^= K0; L2 ^= K0; R1 ^= K1; R2 ^= K1; six_e_rounds_x2(L1, R1, L2, R2, SK.subspan(2)); L1 = FL_8(L1, SK[8]); L2 = FL_8(L2, SK[8]); R1 = FLINV_8(R1, SK[9]); R2 = FLINV_8(R2, SK[9]); six_e_rounds_x2(L1, R1, L2, R2, SK.subspan(10)); L1 = FL_8(L1, SK[16]); L2 = FL_8(L2, SK[16]); R1 = FLINV_8(R1, SK[17]); R2 = FLINV_8(R2, SK[17]); six_e_rounds_x2(L1, R1, L2, R2, SK.subspan(18)); L1 = FL_8(L1, SK[24]); L2 = FL_8(L2, SK[24]); R1 = FLINV_8(R1, SK[25]); R2 = FLINV_8(R2, SK[25]); six_e_rounds_x2(L1, R1, L2, R2, SK.subspan(26)); const auto K32 = SIMD_8x64::splat(SK[32]); const auto K33 = SIMD_8x64::splat(SK[33]); R1 ^= K32; R2 ^= K32; L1 ^= K33; L2 ^= K33; interleave_and_store(out, L1, R1); interleave_and_store(out + 128, L2, R2); } BOTAN_FN_ISA_AVX512_GFNI void camellia_decrypt_x16_24r(const uint8_t in[], uint8_t out[], std::span SK) { SIMD_8x64 L1; SIMD_8x64 R1; SIMD_8x64 L2; SIMD_8x64 R2; load_and_deinterleave(in, L1, R1); load_and_deinterleave(in + 128, L2, R2); const auto K33 = SIMD_8x64::splat(SK[33]); const auto K32 = SIMD_8x64::splat(SK[32]); R1 ^= K33; R2 ^= K33; L1 ^= K32; L2 ^= K32; six_d_rounds_x2(L1, R1, L2, R2, SK.subspan(26)); L1 = FL_8(L1, SK[25]); L2 = FL_8(L2, SK[25]); R1 = FLINV_8(R1, SK[24]); R2 = FLINV_8(R2, SK[24]); six_d_rounds_x2(L1, R1, L2, R2, SK.subspan(18)); L1 = FL_8(L1, SK[17]); L2 = FL_8(L2, SK[17]); R1 = FLINV_8(R1, SK[16]); R2 = FLINV_8(R2, SK[16]); six_d_rounds_x2(L1, R1, L2, R2, SK.subspan(10)); L1 = FL_8(L1, SK[9]); L2 = FL_8(L2, SK[9]); R1 = FLINV_8(R1, SK[8]); R2 = FLINV_8(R2, SK[8]); six_d_rounds_x2(L1, R1, L2, R2, SK.subspan(2)); const auto K1 = SIMD_8x64::splat(SK[1]); const auto K0 = SIMD_8x64::splat(SK[0]); L1 ^= K1; L2 ^= K1; R1 ^= K0; R2 ^= K0; interleave_and_store(out, L1, R1); interleave_and_store(out + 128, L2, R2); } BOTAN_FN_ISA_AVX512_GFNI void camellia_encrypt_x8_18r(const uint8_t in[], uint8_t out[], std::span SK) { SIMD_8x64 L; SIMD_8x64 R; load_and_deinterleave(in, L, R); L ^= SIMD_8x64::splat(SK[0]); R ^= SIMD_8x64::splat(SK[1]); six_e_rounds(L, R, SK.subspan(2)); L = FL_8(L, SK[8]); R = FLINV_8(R, SK[9]); six_e_rounds(L, R, SK.subspan(10)); L = FL_8(L, SK[16]); R = FLINV_8(R, SK[17]); six_e_rounds(L, R, SK.subspan(18)); R ^= SIMD_8x64::splat(SK[24]); L ^= SIMD_8x64::splat(SK[25]); interleave_and_store(out, L, R); } BOTAN_FN_ISA_AVX512_GFNI void camellia_decrypt_x8_18r(const uint8_t in[], uint8_t out[], std::span SK) { SIMD_8x64 L; SIMD_8x64 R; load_and_deinterleave(in, L, R); R ^= SIMD_8x64::splat(SK[25]); L ^= SIMD_8x64::splat(SK[24]); six_d_rounds(L, R, SK.subspan(18)); L = FL_8(L, SK[17]); R = FLINV_8(R, SK[16]); six_d_rounds(L, R, SK.subspan(10)); L = FL_8(L, SK[9]); R = FLINV_8(R, SK[8]); six_d_rounds(L, R, SK.subspan(2)); L ^= SIMD_8x64::splat(SK[1]); R ^= SIMD_8x64::splat(SK[0]); interleave_and_store(out, L, R); } BOTAN_FN_ISA_AVX512_GFNI void camellia_encrypt_x8_24r(const uint8_t in[], uint8_t out[], std::span SK) { SIMD_8x64 L; SIMD_8x64 R; load_and_deinterleave(in, L, R); L ^= SIMD_8x64::splat(SK[0]); R ^= SIMD_8x64::splat(SK[1]); six_e_rounds(L, R, SK.subspan(2)); L = FL_8(L, SK[8]); R = FLINV_8(R, SK[9]); six_e_rounds(L, R, SK.subspan(10)); L = FL_8(L, SK[16]); R = FLINV_8(R, SK[17]); six_e_rounds(L, R, SK.subspan(18)); L = FL_8(L, SK[24]); R = FLINV_8(R, SK[25]); six_e_rounds(L, R, SK.subspan(26)); R ^= SIMD_8x64::splat(SK[32]); L ^= SIMD_8x64::splat(SK[33]); interleave_and_store(out, L, R); } BOTAN_FN_ISA_AVX512_GFNI void camellia_decrypt_x8_24r(const uint8_t in[], uint8_t out[], std::span SK) { SIMD_8x64 L; SIMD_8x64 R; load_and_deinterleave(in, L, R); R ^= SIMD_8x64::splat(SK[33]); L ^= SIMD_8x64::splat(SK[32]); six_d_rounds(L, R, SK.subspan(26)); L = FL_8(L, SK[25]); R = FLINV_8(R, SK[24]); six_d_rounds(L, R, SK.subspan(18)); L = FL_8(L, SK[17]); R = FLINV_8(R, SK[16]); six_d_rounds(L, R, SK.subspan(10)); L = FL_8(L, SK[9]); R = FLINV_8(R, SK[8]); six_d_rounds(L, R, SK.subspan(2)); L ^= SIMD_8x64::splat(SK[1]); R ^= SIMD_8x64::splat(SK[0]); interleave_and_store(out, L, R); } } // namespace } // namespace Camellia_AVX512 // static void BOTAN_FN_ISA_AVX512_GFNI Camellia_128::avx512_gfni_encrypt(const uint8_t in[], uint8_t out[], size_t blocks, std::span SK) { while(blocks >= 16) { Camellia_AVX512::camellia_encrypt_x16_18r(in, out, SK); in += 16 * 16; out += 16 * 16; blocks -= 16; } while(blocks >= 8) { Camellia_AVX512::camellia_encrypt_x8_18r(in, out, SK); in += 8 * 16; out += 8 * 16; blocks -= 8; } if(blocks > 0) { uint8_t ibuf[8 * 16] = {0}; uint8_t obuf[8 * 16] = {0}; copy_mem(ibuf, in, blocks * 16); Camellia_AVX512::camellia_encrypt_x8_18r(ibuf, obuf, SK); copy_mem(out, obuf, blocks * 16); } } // static void BOTAN_FN_ISA_AVX512_GFNI Camellia_128::avx512_gfni_decrypt(const uint8_t in[], uint8_t out[], size_t blocks, std::span SK) { while(blocks >= 16) { Camellia_AVX512::camellia_decrypt_x16_18r(in, out, SK); in += 16 * 16; out += 16 * 16; blocks -= 16; } while(blocks >= 8) { Camellia_AVX512::camellia_decrypt_x8_18r(in, out, SK); in += 8 * 16; out += 8 * 16; blocks -= 8; } if(blocks > 0) { uint8_t ibuf[8 * 16] = {0}; uint8_t obuf[8 * 16] = {0}; copy_mem(ibuf, in, blocks * 16); Camellia_AVX512::camellia_decrypt_x8_18r(ibuf, obuf, SK); copy_mem(out, obuf, blocks * 16); } } // static void BOTAN_FN_ISA_AVX512_GFNI Camellia_192::avx512_gfni_encrypt(const uint8_t in[], uint8_t out[], size_t blocks, std::span SK) { while(blocks >= 16) { Camellia_AVX512::camellia_encrypt_x16_24r(in, out, SK); in += 16 * 16; out += 16 * 16; blocks -= 16; } while(blocks >= 8) { Camellia_AVX512::camellia_encrypt_x8_24r(in, out, SK); in += 8 * 16; out += 8 * 16; blocks -= 8; } if(blocks > 0) { uint8_t ibuf[8 * 16] = {0}; uint8_t obuf[8 * 16] = {0}; copy_mem(ibuf, in, blocks * 16); Camellia_AVX512::camellia_encrypt_x8_24r(ibuf, obuf, SK); copy_mem(out, obuf, blocks * 16); } } // static void BOTAN_FN_ISA_AVX512_GFNI Camellia_192::avx512_gfni_decrypt(const uint8_t in[], uint8_t out[], size_t blocks, std::span SK) { while(blocks >= 16) { Camellia_AVX512::camellia_decrypt_x16_24r(in, out, SK); in += 16 * 16; out += 16 * 16; blocks -= 16; } while(blocks >= 8) { Camellia_AVX512::camellia_decrypt_x8_24r(in, out, SK); in += 8 * 16; out += 8 * 16; blocks -= 8; } if(blocks > 0) { uint8_t ibuf[8 * 16] = {0}; uint8_t obuf[8 * 16] = {0}; copy_mem(ibuf, in, blocks * 16); Camellia_AVX512::camellia_decrypt_x8_24r(ibuf, obuf, SK); copy_mem(out, obuf, blocks * 16); } } // static void BOTAN_FN_ISA_AVX512_GFNI Camellia_256::avx512_gfni_encrypt(const uint8_t in[], uint8_t out[], size_t blocks, std::span SK) { while(blocks >= 16) { Camellia_AVX512::camellia_encrypt_x16_24r(in, out, SK); in += 16 * 16; out += 16 * 16; blocks -= 16; } while(blocks >= 8) { Camellia_AVX512::camellia_encrypt_x8_24r(in, out, SK); in += 8 * 16; out += 8 * 16; blocks -= 8; } if(blocks > 0) { uint8_t ibuf[8 * 16] = {0}; uint8_t obuf[8 * 16] = {0}; copy_mem(ibuf, in, blocks * 16); Camellia_AVX512::camellia_encrypt_x8_24r(ibuf, obuf, SK); copy_mem(out, obuf, blocks * 16); } } // static void BOTAN_FN_ISA_AVX512_GFNI Camellia_256::avx512_gfni_decrypt(const uint8_t in[], uint8_t out[], size_t blocks, std::span SK) { while(blocks >= 16) { Camellia_AVX512::camellia_decrypt_x16_24r(in, out, SK); in += 16 * 16; out += 16 * 16; blocks -= 16; } while(blocks >= 8) { Camellia_AVX512::camellia_decrypt_x8_24r(in, out, SK); in += 8 * 16; out += 8 * 16; blocks -= 8; } if(blocks > 0) { uint8_t ibuf[8 * 16] = {0}; uint8_t obuf[8 * 16] = {0}; copy_mem(ibuf, in, blocks * 16); Camellia_AVX512::camellia_decrypt_x8_24r(ibuf, obuf, SK); copy_mem(out, obuf, blocks * 16); } } } // namespace Botan