diff options
author | Ard Biesheuvel <ardb@kernel.org> | 2022-01-27 14:35:43 +0300 |
---|---|---|
committer | Herbert Xu <herbert@gondor.apana.org.au> | 2022-02-05 07:10:51 +0300 |
commit | c8bf850e991a1838964ed8d8426f96cf8d3fbab5 (patch) | |
tree | 4845e70ee700096733b810eaab6281dc15441e3b /arch/arm/crypto/aes-neonbs-glue.c | |
parent | 8daa399edeed4cfa792ccea12beda50d445ab6a0 (diff) | |
download | linux-c8bf850e991a1838964ed8d8426f96cf8d3fbab5.tar.xz |
crypto: arm/aes-neonbs-ctr - deal with non-multiples of AES block size
Instead of falling back to C code to deal with the final bit of input
that is not a round multiple of the block size, handle this in the asm
code, permitting us to use overlapping loads and stores for performance,
and implement the 16-byte wide XOR using a single NEON instruction.
Since NEON loads and stores have a natural width of 16 bytes, we need to
handle inputs of less than 16 bytes in a special way, but this rarely
occurs in practice so it does not impact performance. All other input
sizes can be consumed directly by the NEON asm code, although it should
be noted that the core AES transform can still only process 128 bytes (8
AES blocks) at a time.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Diffstat (limited to 'arch/arm/crypto/aes-neonbs-glue.c')
-rw-r--r-- | arch/arm/crypto/aes-neonbs-glue.c | 35 |
1 files changed, 14 insertions, 21 deletions
diff --git a/arch/arm/crypto/aes-neonbs-glue.c b/arch/arm/crypto/aes-neonbs-glue.c index 5c6cd3c63cbc..f00f042ef357 100644 --- a/arch/arm/crypto/aes-neonbs-glue.c +++ b/arch/arm/crypto/aes-neonbs-glue.c @@ -37,7 +37,7 @@ asmlinkage void aesbs_cbc_decrypt(u8 out[], u8 const in[], u8 const rk[], int rounds, int blocks, u8 iv[]); asmlinkage void aesbs_ctr_encrypt(u8 out[], u8 const in[], u8 const rk[], - int rounds, int blocks, u8 ctr[], u8 final[]); + int rounds, int blocks, u8 ctr[]); asmlinkage void aesbs_xts_encrypt(u8 out[], u8 const in[], u8 const rk[], int rounds, int blocks, u8 iv[], int); @@ -243,32 +243,25 @@ static int ctr_encrypt(struct skcipher_request *req) err = skcipher_walk_virt(&walk, req, false); while (walk.nbytes > 0) { - unsigned int blocks = walk.nbytes / AES_BLOCK_SIZE; - u8 *final = (walk.total % AES_BLOCK_SIZE) ? buf : NULL; + const u8 *src = walk.src.virt.addr; + u8 *dst = walk.dst.virt.addr; + int bytes = walk.nbytes; - if (walk.nbytes < walk.total) { - blocks = round_down(blocks, - walk.stride / AES_BLOCK_SIZE); - final = NULL; - } + if (unlikely(bytes < AES_BLOCK_SIZE)) + src = dst = memcpy(buf + sizeof(buf) - bytes, + src, bytes); + else if (walk.nbytes < walk.total) + bytes &= ~(8 * AES_BLOCK_SIZE - 1); kernel_neon_begin(); - aesbs_ctr_encrypt(walk.dst.virt.addr, walk.src.virt.addr, - ctx->rk, ctx->rounds, blocks, walk.iv, final); + aesbs_ctr_encrypt(dst, src, ctx->rk, ctx->rounds, bytes, walk.iv); kernel_neon_end(); - if (final) { - u8 *dst = walk.dst.virt.addr + blocks * AES_BLOCK_SIZE; - u8 *src = walk.src.virt.addr + blocks * AES_BLOCK_SIZE; + if (unlikely(bytes < AES_BLOCK_SIZE)) + memcpy(walk.dst.virt.addr, + buf + sizeof(buf) - bytes, bytes); - crypto_xor_cpy(dst, src, final, - walk.total % AES_BLOCK_SIZE); - - err = skcipher_walk_done(&walk, 0); - break; - } - err = skcipher_walk_done(&walk, - walk.nbytes - blocks * AES_BLOCK_SIZE); + err = skcipher_walk_done(&walk, walk.nbytes - bytes); } return err; |