diff options
Diffstat (limited to 'lib/crypto/aes.c')
| -rw-r--r-- | lib/crypto/aes.c | 697 |
1 files changed, 550 insertions, 147 deletions
diff --git a/lib/crypto/aes.c b/lib/crypto/aes.c index eafe14d021f5..ca733f15b2a8 100644 --- a/lib/crypto/aes.c +++ b/lib/crypto/aes.c @@ -1,18 +1,20 @@ // SPDX-License-Identifier: GPL-2.0 /* * Copyright (C) 2017-2019 Linaro Ltd <ard.biesheuvel@linaro.org> + * Copyright 2026 Google LLC */ +#include <crypto/aes-cbc-macs.h> #include <crypto/aes.h> +#include <crypto/utils.h> +#include <linux/cache.h> #include <linux/crypto.h> +#include <linux/export.h> #include <linux/module.h> #include <linux/unaligned.h> +#include "fips.h" -/* - * Emit the sbox as volatile const to prevent the compiler from doing - * constant folding on sbox references involving fixed indexes. - */ -static volatile const u8 __cacheline_aligned aes_sbox[] = { +static const u8 ____cacheline_aligned aes_sbox[] = { 0x63, 0x7c, 0x77, 0x7b, 0xf2, 0x6b, 0x6f, 0xc5, 0x30, 0x01, 0x67, 0x2b, 0xfe, 0xd7, 0xab, 0x76, 0xca, 0x82, 0xc9, 0x7d, 0xfa, 0x59, 0x47, 0xf0, @@ -47,7 +49,7 @@ static volatile const u8 __cacheline_aligned aes_sbox[] = { 0x41, 0x99, 0x2d, 0x0f, 0xb0, 0x54, 0xbb, 0x16, }; -static volatile const u8 __cacheline_aligned aes_inv_sbox[] = { +static const u8 ____cacheline_aligned aes_inv_sbox[] = { 0x52, 0x09, 0x6a, 0xd5, 0x30, 0x36, 0xa5, 0x38, 0xbf, 0x40, 0xa3, 0x9e, 0x81, 0xf3, 0xd7, 0xfb, 0x7c, 0xe3, 0x39, 0x82, 0x9b, 0x2f, 0xff, 0x87, @@ -88,6 +90,110 @@ extern const u8 crypto_aes_inv_sbox[256] __alias(aes_inv_sbox); EXPORT_SYMBOL(crypto_aes_sbox); EXPORT_SYMBOL(crypto_aes_inv_sbox); +/* aes_enc_tab[i] contains MixColumn([SubByte(i), 0, 0, 0]). */ +const u32 ____cacheline_aligned aes_enc_tab[256] = { + 0xa56363c6, 0x847c7cf8, 0x997777ee, 0x8d7b7bf6, 0x0df2f2ff, 0xbd6b6bd6, + 0xb16f6fde, 0x54c5c591, 0x50303060, 0x03010102, 0xa96767ce, 0x7d2b2b56, + 0x19fefee7, 0x62d7d7b5, 0xe6abab4d, 0x9a7676ec, 0x45caca8f, 0x9d82821f, + 0x40c9c989, 0x877d7dfa, 0x15fafaef, 0xeb5959b2, 0xc947478e, 0x0bf0f0fb, + 0xecadad41, 0x67d4d4b3, 0xfda2a25f, 0xeaafaf45, 0xbf9c9c23, 0xf7a4a453, + 0x967272e4, 0x5bc0c09b, 0xc2b7b775, 0x1cfdfde1, 0xae93933d, 0x6a26264c, + 0x5a36366c, 0x413f3f7e, 0x02f7f7f5, 0x4fcccc83, 0x5c343468, 0xf4a5a551, + 0x34e5e5d1, 0x08f1f1f9, 0x937171e2, 0x73d8d8ab, 0x53313162, 0x3f15152a, + 0x0c040408, 0x52c7c795, 0x65232346, 0x5ec3c39d, 0x28181830, 0xa1969637, + 0x0f05050a, 0xb59a9a2f, 0x0907070e, 0x36121224, 0x9b80801b, 0x3de2e2df, + 0x26ebebcd, 0x6927274e, 0xcdb2b27f, 0x9f7575ea, 0x1b090912, 0x9e83831d, + 0x742c2c58, 0x2e1a1a34, 0x2d1b1b36, 0xb26e6edc, 0xee5a5ab4, 0xfba0a05b, + 0xf65252a4, 0x4d3b3b76, 0x61d6d6b7, 0xceb3b37d, 0x7b292952, 0x3ee3e3dd, + 0x712f2f5e, 0x97848413, 0xf55353a6, 0x68d1d1b9, 0x00000000, 0x2cededc1, + 0x60202040, 0x1ffcfce3, 0xc8b1b179, 0xed5b5bb6, 0xbe6a6ad4, 0x46cbcb8d, + 0xd9bebe67, 0x4b393972, 0xde4a4a94, 0xd44c4c98, 0xe85858b0, 0x4acfcf85, + 0x6bd0d0bb, 0x2aefefc5, 0xe5aaaa4f, 0x16fbfbed, 0xc5434386, 0xd74d4d9a, + 0x55333366, 0x94858511, 0xcf45458a, 0x10f9f9e9, 0x06020204, 0x817f7ffe, + 0xf05050a0, 0x443c3c78, 0xba9f9f25, 0xe3a8a84b, 0xf35151a2, 0xfea3a35d, + 0xc0404080, 0x8a8f8f05, 0xad92923f, 0xbc9d9d21, 0x48383870, 0x04f5f5f1, + 0xdfbcbc63, 0xc1b6b677, 0x75dadaaf, 0x63212142, 0x30101020, 0x1affffe5, + 0x0ef3f3fd, 0x6dd2d2bf, 0x4ccdcd81, 0x140c0c18, 0x35131326, 0x2fececc3, + 0xe15f5fbe, 0xa2979735, 0xcc444488, 0x3917172e, 0x57c4c493, 0xf2a7a755, + 0x827e7efc, 0x473d3d7a, 0xac6464c8, 0xe75d5dba, 0x2b191932, 0x957373e6, + 0xa06060c0, 0x98818119, 0xd14f4f9e, 0x7fdcdca3, 0x66222244, 0x7e2a2a54, + 0xab90903b, 0x8388880b, 0xca46468c, 0x29eeeec7, 0xd3b8b86b, 0x3c141428, + 0x79dedea7, 0xe25e5ebc, 0x1d0b0b16, 0x76dbdbad, 0x3be0e0db, 0x56323264, + 0x4e3a3a74, 0x1e0a0a14, 0xdb494992, 0x0a06060c, 0x6c242448, 0xe45c5cb8, + 0x5dc2c29f, 0x6ed3d3bd, 0xefacac43, 0xa66262c4, 0xa8919139, 0xa4959531, + 0x37e4e4d3, 0x8b7979f2, 0x32e7e7d5, 0x43c8c88b, 0x5937376e, 0xb76d6dda, + 0x8c8d8d01, 0x64d5d5b1, 0xd24e4e9c, 0xe0a9a949, 0xb46c6cd8, 0xfa5656ac, + 0x07f4f4f3, 0x25eaeacf, 0xaf6565ca, 0x8e7a7af4, 0xe9aeae47, 0x18080810, + 0xd5baba6f, 0x887878f0, 0x6f25254a, 0x722e2e5c, 0x241c1c38, 0xf1a6a657, + 0xc7b4b473, 0x51c6c697, 0x23e8e8cb, 0x7cdddda1, 0x9c7474e8, 0x211f1f3e, + 0xdd4b4b96, 0xdcbdbd61, 0x868b8b0d, 0x858a8a0f, 0x907070e0, 0x423e3e7c, + 0xc4b5b571, 0xaa6666cc, 0xd8484890, 0x05030306, 0x01f6f6f7, 0x120e0e1c, + 0xa36161c2, 0x5f35356a, 0xf95757ae, 0xd0b9b969, 0x91868617, 0x58c1c199, + 0x271d1d3a, 0xb99e9e27, 0x38e1e1d9, 0x13f8f8eb, 0xb398982b, 0x33111122, + 0xbb6969d2, 0x70d9d9a9, 0x898e8e07, 0xa7949433, 0xb69b9b2d, 0x221e1e3c, + 0x92878715, 0x20e9e9c9, 0x49cece87, 0xff5555aa, 0x78282850, 0x7adfdfa5, + 0x8f8c8c03, 0xf8a1a159, 0x80898909, 0x170d0d1a, 0xdabfbf65, 0x31e6e6d7, + 0xc6424284, 0xb86868d0, 0xc3414182, 0xb0999929, 0x772d2d5a, 0x110f0f1e, + 0xcbb0b07b, 0xfc5454a8, 0xd6bbbb6d, 0x3a16162c, +}; +EXPORT_SYMBOL(aes_enc_tab); + +/* aes_dec_tab[i] contains InvMixColumn([InvSubByte(i), 0, 0, 0]). */ +const u32 ____cacheline_aligned aes_dec_tab[256] = { + 0x50a7f451, 0x5365417e, 0xc3a4171a, 0x965e273a, 0xcb6bab3b, 0xf1459d1f, + 0xab58faac, 0x9303e34b, 0x55fa3020, 0xf66d76ad, 0x9176cc88, 0x254c02f5, + 0xfcd7e54f, 0xd7cb2ac5, 0x80443526, 0x8fa362b5, 0x495ab1de, 0x671bba25, + 0x980eea45, 0xe1c0fe5d, 0x02752fc3, 0x12f04c81, 0xa397468d, 0xc6f9d36b, + 0xe75f8f03, 0x959c9215, 0xeb7a6dbf, 0xda595295, 0x2d83bed4, 0xd3217458, + 0x2969e049, 0x44c8c98e, 0x6a89c275, 0x78798ef4, 0x6b3e5899, 0xdd71b927, + 0xb64fe1be, 0x17ad88f0, 0x66ac20c9, 0xb43ace7d, 0x184adf63, 0x82311ae5, + 0x60335197, 0x457f5362, 0xe07764b1, 0x84ae6bbb, 0x1ca081fe, 0x942b08f9, + 0x58684870, 0x19fd458f, 0x876cde94, 0xb7f87b52, 0x23d373ab, 0xe2024b72, + 0x578f1fe3, 0x2aab5566, 0x0728ebb2, 0x03c2b52f, 0x9a7bc586, 0xa50837d3, + 0xf2872830, 0xb2a5bf23, 0xba6a0302, 0x5c8216ed, 0x2b1ccf8a, 0x92b479a7, + 0xf0f207f3, 0xa1e2694e, 0xcdf4da65, 0xd5be0506, 0x1f6234d1, 0x8afea6c4, + 0x9d532e34, 0xa055f3a2, 0x32e18a05, 0x75ebf6a4, 0x39ec830b, 0xaaef6040, + 0x069f715e, 0x51106ebd, 0xf98a213e, 0x3d06dd96, 0xae053edd, 0x46bde64d, + 0xb58d5491, 0x055dc471, 0x6fd40604, 0xff155060, 0x24fb9819, 0x97e9bdd6, + 0xcc434089, 0x779ed967, 0xbd42e8b0, 0x888b8907, 0x385b19e7, 0xdbeec879, + 0x470a7ca1, 0xe90f427c, 0xc91e84f8, 0x00000000, 0x83868009, 0x48ed2b32, + 0xac70111e, 0x4e725a6c, 0xfbff0efd, 0x5638850f, 0x1ed5ae3d, 0x27392d36, + 0x64d90f0a, 0x21a65c68, 0xd1545b9b, 0x3a2e3624, 0xb1670a0c, 0x0fe75793, + 0xd296eeb4, 0x9e919b1b, 0x4fc5c080, 0xa220dc61, 0x694b775a, 0x161a121c, + 0x0aba93e2, 0xe52aa0c0, 0x43e0223c, 0x1d171b12, 0x0b0d090e, 0xadc78bf2, + 0xb9a8b62d, 0xc8a91e14, 0x8519f157, 0x4c0775af, 0xbbdd99ee, 0xfd607fa3, + 0x9f2601f7, 0xbcf5725c, 0xc53b6644, 0x347efb5b, 0x7629438b, 0xdcc623cb, + 0x68fcedb6, 0x63f1e4b8, 0xcadc31d7, 0x10856342, 0x40229713, 0x2011c684, + 0x7d244a85, 0xf83dbbd2, 0x1132f9ae, 0x6da129c7, 0x4b2f9e1d, 0xf330b2dc, + 0xec52860d, 0xd0e3c177, 0x6c16b32b, 0x99b970a9, 0xfa489411, 0x2264e947, + 0xc48cfca8, 0x1a3ff0a0, 0xd82c7d56, 0xef903322, 0xc74e4987, 0xc1d138d9, + 0xfea2ca8c, 0x360bd498, 0xcf81f5a6, 0x28de7aa5, 0x268eb7da, 0xa4bfad3f, + 0xe49d3a2c, 0x0d927850, 0x9bcc5f6a, 0x62467e54, 0xc2138df6, 0xe8b8d890, + 0x5ef7392e, 0xf5afc382, 0xbe805d9f, 0x7c93d069, 0xa92dd56f, 0xb31225cf, + 0x3b99acc8, 0xa77d1810, 0x6e639ce8, 0x7bbb3bdb, 0x097826cd, 0xf418596e, + 0x01b79aec, 0xa89a4f83, 0x656e95e6, 0x7ee6ffaa, 0x08cfbc21, 0xe6e815ef, + 0xd99be7ba, 0xce366f4a, 0xd4099fea, 0xd67cb029, 0xafb2a431, 0x31233f2a, + 0x3094a5c6, 0xc066a235, 0x37bc4e74, 0xa6ca82fc, 0xb0d090e0, 0x15d8a733, + 0x4a9804f1, 0xf7daec41, 0x0e50cd7f, 0x2ff69117, 0x8dd64d76, 0x4db0ef43, + 0x544daacc, 0xdf0496e4, 0xe3b5d19e, 0x1b886a4c, 0xb81f2cc1, 0x7f516546, + 0x04ea5e9d, 0x5d358c01, 0x737487fa, 0x2e410bfb, 0x5a1d67b3, 0x52d2db92, + 0x335610e9, 0x1347d66d, 0x8c61d79a, 0x7a0ca137, 0x8e14f859, 0x893c13eb, + 0xee27a9ce, 0x35c961b7, 0xede51ce1, 0x3cb1477a, 0x59dfd29c, 0x3f73f255, + 0x79ce1418, 0xbf37c773, 0xeacdf753, 0x5baafd5f, 0x146f3ddf, 0x86db4478, + 0x81f3afca, 0x3ec468b9, 0x2c342438, 0x5f40a3c2, 0x72c31d16, 0x0c25e2bc, + 0x8b493c28, 0x41950dff, 0x7101a839, 0xdeb30c08, 0x9ce4b4d8, 0x90c15664, + 0x6184cb7b, 0x70b632d5, 0x745c6c48, 0x4257b8d0, +}; +EXPORT_SYMBOL(aes_dec_tab); + +/* Prefetch data into L1 cache. @mem should be cacheline-aligned. */ +static __always_inline void aes_prefetch(const void *mem, size_t len) +{ + for (size_t i = 0; i < len; i += L1_CACHE_BYTES) + *(volatile const u8 *)(mem + i); + barrier(); +} + static u32 mul_by_x(u32 w) { u32 x = w & 0x7f7f7f7f; @@ -144,22 +250,6 @@ static u32 inv_mix_columns(u32 x) return mix_columns(x ^ y ^ ror32(y, 16)); } -static __always_inline u32 subshift(u32 in[], int pos) -{ - return (aes_sbox[in[pos] & 0xff]) ^ - (aes_sbox[(in[(pos + 1) % 4] >> 8) & 0xff] << 8) ^ - (aes_sbox[(in[(pos + 2) % 4] >> 16) & 0xff] << 16) ^ - (aes_sbox[(in[(pos + 3) % 4] >> 24) & 0xff] << 24); -} - -static __always_inline u32 inv_subshift(u32 in[], int pos) -{ - return (aes_inv_sbox[in[pos] & 0xff]) ^ - (aes_inv_sbox[(in[(pos + 3) % 4] >> 8) & 0xff] << 8) ^ - (aes_inv_sbox[(in[(pos + 2) % 4] >> 16) & 0xff] << 16) ^ - (aes_inv_sbox[(in[(pos + 1) % 4] >> 24) & 0xff] << 24); -} - static u32 subw(u32 in) { return (aes_sbox[in & 0xff]) ^ @@ -168,38 +258,17 @@ static u32 subw(u32 in) (aes_sbox[(in >> 24) & 0xff] << 24); } -/** - * aes_expandkey - Expands the AES key as described in FIPS-197 - * @ctx: The location where the computed key will be stored. - * @in_key: The supplied key. - * @key_len: The length of the supplied key. - * - * Returns 0 on success. The function fails only if an invalid key size (or - * pointer) is supplied. - * The expanded key size is 240 bytes (max of 14 rounds with a unique 16 bytes - * key schedule plus a 16 bytes key which is used before the first round). - * The decryption key is prepared for the "Equivalent Inverse Cipher" as - * described in FIPS-197. The first slot (16 bytes) of each key (enc or dec) is - * for the initial combination, the second slot for the first round and so on. - */ -int aes_expandkey(struct crypto_aes_ctx *ctx, const u8 *in_key, - unsigned int key_len) +static void aes_expandkey_generic(u32 rndkeys[], u32 *inv_rndkeys, + const u8 *in_key, int key_len) { u32 kwords = key_len / sizeof(u32); u32 rc, i, j; - int err; - - err = aes_check_keylen(key_len); - if (err) - return err; - - ctx->key_length = key_len; for (i = 0; i < kwords; i++) - ctx->key_enc[i] = get_unaligned_le32(in_key + i * sizeof(u32)); + rndkeys[i] = get_unaligned_le32(&in_key[i * sizeof(u32)]); for (i = 0, rc = 1; i < 10; i++, rc = mul_by_x(rc)) { - u32 *rki = ctx->key_enc + (i * kwords); + u32 *rki = &rndkeys[i * kwords]; u32 *rko = rki + kwords; rko[0] = ror32(subw(rki[kwords - 1]), 8) ^ rc ^ rki[0]; @@ -228,129 +297,463 @@ int aes_expandkey(struct crypto_aes_ctx *ctx, const u8 *in_key, * the Inverse Mix Columns transformation to all but the first and * the last one. */ - ctx->key_dec[0] = ctx->key_enc[key_len + 24]; - ctx->key_dec[1] = ctx->key_enc[key_len + 25]; - ctx->key_dec[2] = ctx->key_enc[key_len + 26]; - ctx->key_dec[3] = ctx->key_enc[key_len + 27]; - - for (i = 4, j = key_len + 20; j > 0; i += 4, j -= 4) { - ctx->key_dec[i] = inv_mix_columns(ctx->key_enc[j]); - ctx->key_dec[i + 1] = inv_mix_columns(ctx->key_enc[j + 1]); - ctx->key_dec[i + 2] = inv_mix_columns(ctx->key_enc[j + 2]); - ctx->key_dec[i + 3] = inv_mix_columns(ctx->key_enc[j + 3]); - } + if (inv_rndkeys) { + inv_rndkeys[0] = rndkeys[key_len + 24]; + inv_rndkeys[1] = rndkeys[key_len + 25]; + inv_rndkeys[2] = rndkeys[key_len + 26]; + inv_rndkeys[3] = rndkeys[key_len + 27]; + + for (i = 4, j = key_len + 20; j > 0; i += 4, j -= 4) { + inv_rndkeys[i] = inv_mix_columns(rndkeys[j]); + inv_rndkeys[i + 1] = inv_mix_columns(rndkeys[j + 1]); + inv_rndkeys[i + 2] = inv_mix_columns(rndkeys[j + 2]); + inv_rndkeys[i + 3] = inv_mix_columns(rndkeys[j + 3]); + } - ctx->key_dec[i] = ctx->key_enc[0]; - ctx->key_dec[i + 1] = ctx->key_enc[1]; - ctx->key_dec[i + 2] = ctx->key_enc[2]; - ctx->key_dec[i + 3] = ctx->key_enc[3]; + inv_rndkeys[i] = rndkeys[0]; + inv_rndkeys[i + 1] = rndkeys[1]; + inv_rndkeys[i + 2] = rndkeys[2]; + inv_rndkeys[i + 3] = rndkeys[3]; + } +} +int aes_expandkey(struct crypto_aes_ctx *ctx, const u8 *in_key, + unsigned int key_len) +{ + if (aes_check_keylen(key_len) != 0) + return -EINVAL; + ctx->key_length = key_len; + aes_expandkey_generic(ctx->key_enc, ctx->key_dec, in_key, key_len); return 0; } EXPORT_SYMBOL(aes_expandkey); -/** - * aes_encrypt - Encrypt a single AES block - * @ctx: Context struct containing the key schedule - * @out: Buffer to store the ciphertext - * @in: Buffer containing the plaintext - */ -void aes_encrypt(const struct crypto_aes_ctx *ctx, u8 *out, const u8 *in) +static __always_inline u32 enc_quarterround(const u32 w[4], int i, u32 rk) { - const u32 *rkp = ctx->key_enc + 4; - int rounds = 6 + ctx->key_length / 4; - u32 st0[4], st1[4]; - int round; + return rk ^ aes_enc_tab[(u8)w[i]] ^ + rol32(aes_enc_tab[(u8)(w[(i + 1) % 4] >> 8)], 8) ^ + rol32(aes_enc_tab[(u8)(w[(i + 2) % 4] >> 16)], 16) ^ + rol32(aes_enc_tab[(u8)(w[(i + 3) % 4] >> 24)], 24); +} + +static __always_inline u32 enclast_quarterround(const u32 w[4], int i, u32 rk) +{ + return rk ^ ((aes_enc_tab[(u8)w[i]] & 0x0000ff00) >> 8) ^ + (aes_enc_tab[(u8)(w[(i + 1) % 4] >> 8)] & 0x0000ff00) ^ + ((aes_enc_tab[(u8)(w[(i + 2) % 4] >> 16)] & 0x0000ff00) << 8) ^ + ((aes_enc_tab[(u8)(w[(i + 3) % 4] >> 24)] & 0x0000ff00) << 16); +} - st0[0] = ctx->key_enc[0] ^ get_unaligned_le32(in); - st0[1] = ctx->key_enc[1] ^ get_unaligned_le32(in + 4); - st0[2] = ctx->key_enc[2] ^ get_unaligned_le32(in + 8); - st0[3] = ctx->key_enc[3] ^ get_unaligned_le32(in + 12); +static void __maybe_unused aes_encrypt_generic(const u32 rndkeys[], int nrounds, + u8 out[AES_BLOCK_SIZE], + const u8 in[AES_BLOCK_SIZE]) +{ + const u32 *rkp = rndkeys; + int n = nrounds - 1; + u32 w[4]; + + w[0] = get_unaligned_le32(&in[0]) ^ *rkp++; + w[1] = get_unaligned_le32(&in[4]) ^ *rkp++; + w[2] = get_unaligned_le32(&in[8]) ^ *rkp++; + w[3] = get_unaligned_le32(&in[12]) ^ *rkp++; /* - * Force the compiler to emit data independent Sbox references, - * by xoring the input with Sbox values that are known to add up - * to zero. This pulls the entire Sbox into the D-cache before any - * data dependent lookups are done. + * Prefetch the table before doing data and key-dependent loads from it. + * + * This is intended only as a basic constant-time hardening measure that + * avoids interfering with performance too much. Its effectiveness is + * not guaranteed. For proper constant-time AES, a CPU that supports + * AES instructions should be used instead. */ - st0[0] ^= aes_sbox[ 0] ^ aes_sbox[ 64] ^ aes_sbox[134] ^ aes_sbox[195]; - st0[1] ^= aes_sbox[16] ^ aes_sbox[ 82] ^ aes_sbox[158] ^ aes_sbox[221]; - st0[2] ^= aes_sbox[32] ^ aes_sbox[ 96] ^ aes_sbox[160] ^ aes_sbox[234]; - st0[3] ^= aes_sbox[48] ^ aes_sbox[112] ^ aes_sbox[186] ^ aes_sbox[241]; - - for (round = 0;; round += 2, rkp += 8) { - st1[0] = mix_columns(subshift(st0, 0)) ^ rkp[0]; - st1[1] = mix_columns(subshift(st0, 1)) ^ rkp[1]; - st1[2] = mix_columns(subshift(st0, 2)) ^ rkp[2]; - st1[3] = mix_columns(subshift(st0, 3)) ^ rkp[3]; - - if (round == rounds - 2) - break; - - st0[0] = mix_columns(subshift(st1, 0)) ^ rkp[4]; - st0[1] = mix_columns(subshift(st1, 1)) ^ rkp[5]; - st0[2] = mix_columns(subshift(st1, 2)) ^ rkp[6]; - st0[3] = mix_columns(subshift(st1, 3)) ^ rkp[7]; - } + aes_prefetch(aes_enc_tab, sizeof(aes_enc_tab)); + + do { + u32 w0 = enc_quarterround(w, 0, *rkp++); + u32 w1 = enc_quarterround(w, 1, *rkp++); + u32 w2 = enc_quarterround(w, 2, *rkp++); + u32 w3 = enc_quarterround(w, 3, *rkp++); + + w[0] = w0; + w[1] = w1; + w[2] = w2; + w[3] = w3; + } while (--n); + + put_unaligned_le32(enclast_quarterround(w, 0, *rkp++), &out[0]); + put_unaligned_le32(enclast_quarterround(w, 1, *rkp++), &out[4]); + put_unaligned_le32(enclast_quarterround(w, 2, *rkp++), &out[8]); + put_unaligned_le32(enclast_quarterround(w, 3, *rkp++), &out[12]); +} - put_unaligned_le32(subshift(st1, 0) ^ rkp[4], out); - put_unaligned_le32(subshift(st1, 1) ^ rkp[5], out + 4); - put_unaligned_le32(subshift(st1, 2) ^ rkp[6], out + 8); - put_unaligned_le32(subshift(st1, 3) ^ rkp[7], out + 12); +static __always_inline u32 dec_quarterround(const u32 w[4], int i, u32 rk) +{ + return rk ^ aes_dec_tab[(u8)w[i]] ^ + rol32(aes_dec_tab[(u8)(w[(i + 3) % 4] >> 8)], 8) ^ + rol32(aes_dec_tab[(u8)(w[(i + 2) % 4] >> 16)], 16) ^ + rol32(aes_dec_tab[(u8)(w[(i + 1) % 4] >> 24)], 24); +} + +static __always_inline u32 declast_quarterround(const u32 w[4], int i, u32 rk) +{ + return rk ^ aes_inv_sbox[(u8)w[i]] ^ + ((u32)aes_inv_sbox[(u8)(w[(i + 3) % 4] >> 8)] << 8) ^ + ((u32)aes_inv_sbox[(u8)(w[(i + 2) % 4] >> 16)] << 16) ^ + ((u32)aes_inv_sbox[(u8)(w[(i + 1) % 4] >> 24)] << 24); +} + +static void __maybe_unused aes_decrypt_generic(const u32 inv_rndkeys[], + int nrounds, + u8 out[AES_BLOCK_SIZE], + const u8 in[AES_BLOCK_SIZE]) +{ + const u32 *rkp = inv_rndkeys; + int n = nrounds - 1; + u32 w[4]; + + w[0] = get_unaligned_le32(&in[0]) ^ *rkp++; + w[1] = get_unaligned_le32(&in[4]) ^ *rkp++; + w[2] = get_unaligned_le32(&in[8]) ^ *rkp++; + w[3] = get_unaligned_le32(&in[12]) ^ *rkp++; + + aes_prefetch(aes_dec_tab, sizeof(aes_dec_tab)); + + do { + u32 w0 = dec_quarterround(w, 0, *rkp++); + u32 w1 = dec_quarterround(w, 1, *rkp++); + u32 w2 = dec_quarterround(w, 2, *rkp++); + u32 w3 = dec_quarterround(w, 3, *rkp++); + + w[0] = w0; + w[1] = w1; + w[2] = w2; + w[3] = w3; + } while (--n); + + aes_prefetch(aes_inv_sbox, sizeof(aes_inv_sbox)); + put_unaligned_le32(declast_quarterround(w, 0, *rkp++), &out[0]); + put_unaligned_le32(declast_quarterround(w, 1, *rkp++), &out[4]); + put_unaligned_le32(declast_quarterround(w, 2, *rkp++), &out[8]); + put_unaligned_le32(declast_quarterround(w, 3, *rkp++), &out[12]); } -EXPORT_SYMBOL(aes_encrypt); -/** - * aes_decrypt - Decrypt a single AES block - * @ctx: Context struct containing the key schedule - * @out: Buffer to store the plaintext - * @in: Buffer containing the ciphertext +/* + * Note: the aes_prepare*key_* names reflect the fact that the implementation + * might not actually expand the key. (The s390 code for example doesn't.) + * Where the key is expanded we use the more specific names aes_expandkey_*. + * + * aes_preparekey_arch() is passed an optional pointer 'inv_k' which points to + * the area to store the prepared decryption key. It will be NULL if the user + * is requesting encryption-only. aes_preparekey_arch() is also passed a valid + * 'key_len' and 'nrounds', corresponding to AES-128, AES-192, or AES-256. */ -void aes_decrypt(const struct crypto_aes_ctx *ctx, u8 *out, const u8 *in) +#ifdef CONFIG_CRYPTO_LIB_AES_ARCH +/* An arch-specific implementation of AES is available. Include it. */ +#include "aes.h" /* $(SRCARCH)/aes.h */ +#else +/* No arch-specific implementation of AES is available. Use generic code. */ + +static void aes_preparekey_arch(union aes_enckey_arch *k, + union aes_invkey_arch *inv_k, + const u8 *in_key, int key_len, int nrounds) +{ + aes_expandkey_generic(k->rndkeys, inv_k ? inv_k->inv_rndkeys : NULL, + in_key, key_len); +} + +static void aes_encrypt_arch(const struct aes_enckey *key, + u8 out[AES_BLOCK_SIZE], + const u8 in[AES_BLOCK_SIZE]) +{ + aes_encrypt_generic(key->k.rndkeys, key->nrounds, out, in); +} + +static void aes_decrypt_arch(const struct aes_key *key, + u8 out[AES_BLOCK_SIZE], + const u8 in[AES_BLOCK_SIZE]) +{ + aes_decrypt_generic(key->inv_k.inv_rndkeys, key->nrounds, out, in); +} +#endif + +static int __aes_preparekey(struct aes_enckey *enc_key, + union aes_invkey_arch *inv_k, + const u8 *in_key, size_t key_len) { - const u32 *rkp = ctx->key_dec + 4; - int rounds = 6 + ctx->key_length / 4; - u32 st0[4], st1[4]; - int round; + if (aes_check_keylen(key_len) != 0) + return -EINVAL; + enc_key->len = key_len; + enc_key->nrounds = 6 + key_len / 4; + aes_preparekey_arch(&enc_key->k, inv_k, in_key, key_len, + enc_key->nrounds); + return 0; +} - st0[0] = ctx->key_dec[0] ^ get_unaligned_le32(in); - st0[1] = ctx->key_dec[1] ^ get_unaligned_le32(in + 4); - st0[2] = ctx->key_dec[2] ^ get_unaligned_le32(in + 8); - st0[3] = ctx->key_dec[3] ^ get_unaligned_le32(in + 12); +int aes_preparekey(struct aes_key *key, const u8 *in_key, size_t key_len) +{ + return __aes_preparekey((struct aes_enckey *)key, &key->inv_k, + in_key, key_len); +} +EXPORT_SYMBOL(aes_preparekey); + +int aes_prepareenckey(struct aes_enckey *key, const u8 *in_key, size_t key_len) +{ + return __aes_preparekey(key, NULL, in_key, key_len); +} +EXPORT_SYMBOL(aes_prepareenckey); + +void aes_encrypt(aes_encrypt_arg key, u8 out[AES_BLOCK_SIZE], + const u8 in[AES_BLOCK_SIZE]) +{ + aes_encrypt_arch(key.enc_key, out, in); +} +EXPORT_SYMBOL(aes_encrypt); + +void aes_decrypt(const struct aes_key *key, u8 out[AES_BLOCK_SIZE], + const u8 in[AES_BLOCK_SIZE]) +{ + aes_decrypt_arch(key, out, in); +} +EXPORT_SYMBOL(aes_decrypt); + +#if IS_ENABLED(CONFIG_CRYPTO_LIB_AES_CBC_MACS) + +#ifndef aes_cbcmac_blocks_arch +static bool aes_cbcmac_blocks_arch(u8 h[AES_BLOCK_SIZE], + const struct aes_enckey *key, const u8 *data, + size_t nblocks, bool enc_before, + bool enc_after) +{ + return false; +} +#endif + +/* This assumes nblocks >= 1. */ +static void aes_cbcmac_blocks(u8 h[AES_BLOCK_SIZE], + const struct aes_enckey *key, const u8 *data, + size_t nblocks, bool enc_before, bool enc_after) +{ + if (aes_cbcmac_blocks_arch(h, key, data, nblocks, enc_before, + enc_after)) + return; + + if (enc_before) + aes_encrypt(key, h, h); + for (; nblocks > 1; nblocks--) { + crypto_xor(h, data, AES_BLOCK_SIZE); + data += AES_BLOCK_SIZE; + aes_encrypt(key, h, h); + } + crypto_xor(h, data, AES_BLOCK_SIZE); + if (enc_after) + aes_encrypt(key, h, h); +} + +int aes_cmac_preparekey(struct aes_cmac_key *key, const u8 *in_key, + size_t key_len) +{ + u64 hi, lo, mask; + int err; + + /* Prepare the AES key. */ + err = aes_prepareenckey(&key->aes, in_key, key_len); + if (err) + return err; /* - * Force the compiler to emit data independent Sbox references, - * by xoring the input with Sbox values that are known to add up - * to zero. This pulls the entire Sbox into the D-cache before any - * data dependent lookups are done. + * Prepare the subkeys K1 and K2 by encrypting the all-zeroes block, + * then multiplying by 'x' and 'x^2' (respectively) in GF(2^128). + * Reference: NIST SP 800-38B, Section 6.1 "Subkey Generation". */ - st0[0] ^= aes_inv_sbox[ 0] ^ aes_inv_sbox[ 64] ^ aes_inv_sbox[129] ^ aes_inv_sbox[200]; - st0[1] ^= aes_inv_sbox[16] ^ aes_inv_sbox[ 83] ^ aes_inv_sbox[150] ^ aes_inv_sbox[212]; - st0[2] ^= aes_inv_sbox[32] ^ aes_inv_sbox[ 96] ^ aes_inv_sbox[160] ^ aes_inv_sbox[236]; - st0[3] ^= aes_inv_sbox[48] ^ aes_inv_sbox[112] ^ aes_inv_sbox[187] ^ aes_inv_sbox[247]; - - for (round = 0;; round += 2, rkp += 8) { - st1[0] = inv_mix_columns(inv_subshift(st0, 0)) ^ rkp[0]; - st1[1] = inv_mix_columns(inv_subshift(st0, 1)) ^ rkp[1]; - st1[2] = inv_mix_columns(inv_subshift(st0, 2)) ^ rkp[2]; - st1[3] = inv_mix_columns(inv_subshift(st0, 3)) ^ rkp[3]; - - if (round == rounds - 2) - break; - - st0[0] = inv_mix_columns(inv_subshift(st1, 0)) ^ rkp[4]; - st0[1] = inv_mix_columns(inv_subshift(st1, 1)) ^ rkp[5]; - st0[2] = inv_mix_columns(inv_subshift(st1, 2)) ^ rkp[6]; - st0[3] = inv_mix_columns(inv_subshift(st1, 3)) ^ rkp[7]; + memset(key->k_final[0].b, 0, AES_BLOCK_SIZE); + aes_encrypt(&key->aes, key->k_final[0].b, key->k_final[0].b); + hi = be64_to_cpu(key->k_final[0].w[0]); + lo = be64_to_cpu(key->k_final[0].w[1]); + for (int i = 0; i < 2; i++) { + mask = ((s64)hi >> 63) & 0x87; + hi = (hi << 1) ^ (lo >> 63); + lo = (lo << 1) ^ mask; + key->k_final[i].w[0] = cpu_to_be64(hi); + key->k_final[i].w[1] = cpu_to_be64(lo); } + return 0; +} +EXPORT_SYMBOL_GPL(aes_cmac_preparekey); - put_unaligned_le32(inv_subshift(st1, 0) ^ rkp[4], out); - put_unaligned_le32(inv_subshift(st1, 1) ^ rkp[5], out + 4); - put_unaligned_le32(inv_subshift(st1, 2) ^ rkp[6], out + 8); - put_unaligned_le32(inv_subshift(st1, 3) ^ rkp[7], out + 12); +void aes_xcbcmac_preparekey(struct aes_cmac_key *key, + const u8 in_key[AES_KEYSIZE_128]) +{ + static const u8 constants[3][AES_BLOCK_SIZE] = { + { [0 ... AES_BLOCK_SIZE - 1] = 0x1 }, + { [0 ... AES_BLOCK_SIZE - 1] = 0x2 }, + { [0 ... AES_BLOCK_SIZE - 1] = 0x3 }, + }; + u8 new_aes_key[AES_BLOCK_SIZE]; + + static_assert(AES_BLOCK_SIZE == AES_KEYSIZE_128); + aes_prepareenckey(&key->aes, in_key, AES_BLOCK_SIZE); + aes_encrypt(&key->aes, new_aes_key, constants[0]); + aes_encrypt(&key->aes, key->k_final[0].b, constants[1]); + aes_encrypt(&key->aes, key->k_final[1].b, constants[2]); + aes_prepareenckey(&key->aes, new_aes_key, AES_BLOCK_SIZE); + memzero_explicit(new_aes_key, AES_BLOCK_SIZE); } -EXPORT_SYMBOL(aes_decrypt); +EXPORT_SYMBOL_GPL(aes_xcbcmac_preparekey); + +void aes_cmac_update(struct aes_cmac_ctx *ctx, const u8 *data, size_t data_len) +{ + bool enc_before = false; + size_t nblocks; + + if (ctx->partial_len) { + /* XOR data into a pending block. */ + size_t l = min(data_len, AES_BLOCK_SIZE - ctx->partial_len); + + crypto_xor(&ctx->h[ctx->partial_len], data, l); + data += l; + data_len -= l; + ctx->partial_len += l; + if (data_len == 0) { + /* + * Either the pending block hasn't been filled yet, or + * no more data was given so it's not yet known whether + * the block is the final block. + */ + return; + } + /* Pending block has been filled and isn't the final block. */ + enc_before = true; + } + + nblocks = data_len / AES_BLOCK_SIZE; + data_len %= AES_BLOCK_SIZE; + if (nblocks == 0) { + /* 0 additional full blocks, then optionally a partial block */ + if (enc_before) + aes_encrypt(&ctx->key->aes, ctx->h, ctx->h); + crypto_xor(ctx->h, data, data_len); + ctx->partial_len = data_len; + } else if (data_len != 0) { + /* 1 or more additional full blocks, then a partial block */ + aes_cbcmac_blocks(ctx->h, &ctx->key->aes, data, nblocks, + enc_before, /* enc_after= */ true); + data += nblocks * AES_BLOCK_SIZE; + crypto_xor(ctx->h, data, data_len); + ctx->partial_len = data_len; + } else { + /* + * 1 or more additional full blocks only. Encryption of the + * last block is delayed until it's known whether it's the final + * block in the message or not. + */ + aes_cbcmac_blocks(ctx->h, &ctx->key->aes, data, nblocks, + enc_before, /* enc_after= */ false); + ctx->partial_len = AES_BLOCK_SIZE; + } +} +EXPORT_SYMBOL_GPL(aes_cmac_update); + +void aes_cmac_final(struct aes_cmac_ctx *ctx, u8 out[AES_BLOCK_SIZE]) +{ + if (ctx->partial_len == AES_BLOCK_SIZE) { + /* Final block is a full block. Use k_final[0]. */ + crypto_xor(ctx->h, ctx->key->k_final[0].b, AES_BLOCK_SIZE); + } else { + /* Final block is a partial block. Pad, and use k_final[1]. */ + ctx->h[ctx->partial_len] ^= 0x80; + crypto_xor(ctx->h, ctx->key->k_final[1].b, AES_BLOCK_SIZE); + } + aes_encrypt(&ctx->key->aes, out, ctx->h); + memzero_explicit(ctx, sizeof(*ctx)); +} +EXPORT_SYMBOL_GPL(aes_cmac_final); + +void aes_cbcmac_update(struct aes_cbcmac_ctx *ctx, const u8 *data, + size_t data_len) +{ + bool enc_before = false; + size_t nblocks; + + if (ctx->partial_len) { + size_t l = min(data_len, AES_BLOCK_SIZE - ctx->partial_len); + + crypto_xor(&ctx->h[ctx->partial_len], data, l); + data += l; + data_len -= l; + ctx->partial_len += l; + if (ctx->partial_len < AES_BLOCK_SIZE) + return; + enc_before = true; + } + + nblocks = data_len / AES_BLOCK_SIZE; + data_len %= AES_BLOCK_SIZE; + if (nblocks == 0) { + if (enc_before) + aes_encrypt(ctx->key, ctx->h, ctx->h); + } else { + aes_cbcmac_blocks(ctx->h, ctx->key, data, nblocks, enc_before, + /* enc_after= */ true); + data += nblocks * AES_BLOCK_SIZE; + } + crypto_xor(ctx->h, data, data_len); + ctx->partial_len = data_len; +} +EXPORT_SYMBOL_NS_GPL(aes_cbcmac_update, "CRYPTO_INTERNAL"); + +void aes_cbcmac_final(struct aes_cbcmac_ctx *ctx, u8 out[AES_BLOCK_SIZE]) +{ + if (ctx->partial_len) + aes_encrypt(ctx->key, out, ctx->h); + else + memcpy(out, ctx->h, AES_BLOCK_SIZE); + memzero_explicit(ctx, sizeof(*ctx)); +} +EXPORT_SYMBOL_NS_GPL(aes_cbcmac_final, "CRYPTO_INTERNAL"); + +/* + * FIPS cryptographic algorithm self-test for AES-CMAC. As per the FIPS 140-3 + * Implementation Guidance, a cryptographic algorithm self-test for at least one + * of AES-GCM, AES-CCM, AES-CMAC, or AES-GMAC is required if any of those modes + * is implemented. This fulfills that requirement via AES-CMAC. + * + * This is just for FIPS. The full tests are in the KUnit test suite. + */ +static void __init aes_cmac_fips_test(void) +{ + struct aes_cmac_key key; + u8 mac[AES_BLOCK_SIZE]; + + if (aes_cmac_preparekey(&key, fips_test_key, sizeof(fips_test_key)) != + 0) + panic("aes: CMAC FIPS self-test failed (preparekey)\n"); + aes_cmac(&key, fips_test_data, sizeof(fips_test_data), mac); + if (memcmp(fips_test_aes_cmac_value, mac, sizeof(mac)) != 0) + panic("aes: CMAC FIPS self-test failed (wrong MAC)\n"); + memzero_explicit(&key, sizeof(key)); +} +#else /* CONFIG_CRYPTO_LIB_AES_CBC_MACS */ +static inline void aes_cmac_fips_test(void) +{ +} +#endif /* !CONFIG_CRYPTO_LIB_AES_CBC_MACS */ + +static int __init aes_mod_init(void) +{ +#ifdef aes_mod_init_arch + aes_mod_init_arch(); +#endif + if (fips_enabled) + aes_cmac_fips_test(); + return 0; +} +subsys_initcall(aes_mod_init); + +static void __exit aes_mod_exit(void) +{ +} +module_exit(aes_mod_exit); -MODULE_DESCRIPTION("Generic AES library"); +MODULE_DESCRIPTION("AES block cipher"); MODULE_AUTHOR("Ard Biesheuvel <ard.biesheuvel@linaro.org>"); +MODULE_AUTHOR("Eric Biggers <ebiggers@kernel.org>"); MODULE_LICENSE("GPL v2"); |
