From: Tim Chen Subject: [PATCH v2 10/10] Create module providing optimized SHA512 routines using SSSE3, AVX or AVX2 instructions. Date: Tue, 26 Mar 2013 14:00:02 -0700 Message-ID: <1364331602.27102.20.camel@schen9-DESK> References: Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: Jussi Kivilinna , Jim Kukunas , Kirk Yap , David Cote , James Guilford , Wajdi Feghali , linux-kernel , linux-crypto@vger.kernel.org To: Herbert Xu , "H. Peter Anvin" , "David S.Miller" Return-path: Received: from mga02.intel.com ([134.134.136.20]:46440 "EHLO mga02.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754911Ab3CZVAB (ORCPT ); Tue, 26 Mar 2013 17:00:01 -0400 In-Reply-To: Sender: linux-crypto-owner@vger.kernel.org List-ID: We added glue code and config options to create crypto module that uses SSE/AVX/AVX2 optimized SHA512 x86_64 assembly routines. Signed-off-by: Tim Chen --- arch/x86/crypto/Makefile | 2 + arch/x86/crypto/sha512_ssse3_glue.c | 282 ++++++++++++++++++++++++++++++++++++ crypto/Kconfig | 11 ++ 3 files changed, 295 insertions(+) create mode 100644 arch/x86/crypto/sha512_ssse3_glue.c diff --git a/arch/x86/crypto/Makefile b/arch/x86/crypto/Makefile index b45f599..3b3d1aa 100644 --- a/arch/x86/crypto/Makefile +++ b/arch/x86/crypto/Makefile @@ -29,6 +29,7 @@ obj-$(CONFIG_CRYPTO_CRC32C_INTEL) += crc32c-intel.o obj-$(CONFIG_CRYPTO_SHA1_SSSE3) += sha1-ssse3.o obj-$(CONFIG_CRYPTO_CRC32_PCLMUL) += crc32-pclmul.o obj-$(CONFIG_CRYPTO_SHA256_SSSE3) += sha256-ssse3.o +obj-$(CONFIG_CRYPTO_SHA512_SSSE3) += sha512-ssse3.o aes-i586-y := aes-i586-asm_32.o aes_glue.o twofish-i586-y := twofish-i586-asm_32.o twofish_glue.o @@ -56,3 +57,4 @@ crc32c-intel-y := crc32c-intel_glue.o crc32c-intel-$(CONFIG_64BIT) += crc32c-pcl-intel-asm_64.o crc32-pclmul-y := crc32-pclmul_asm.o crc32-pclmul_glue.o sha256-ssse3-y := sha256-ssse3-asm.o sha256-avx-asm.o sha256-avx2-asm.o sha256_ssse3_glue.o +sha512-ssse3-y := sha512-ssse3-asm.o sha512-avx-asm.o sha512-avx2-asm.o sha512_ssse3_glue.o diff --git a/arch/x86/crypto/sha512_ssse3_glue.c b/arch/x86/crypto/sha512_ssse3_glue.c new file mode 100644 index 0000000..295f790 --- /dev/null +++ b/arch/x86/crypto/sha512_ssse3_glue.c @@ -0,0 +1,282 @@ +/* + * Cryptographic API. + * + * Glue code for the SHA512 Secure Hash Algorithm assembler + * implementation using supplemental SSE3 / AVX / AVX2 instructions. + * + * This file is based on sha512_generic.c + * + * Copyright (C) 2013 Intel Corporation + * Author: Tim Chen + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the Free + * Software Foundation; either version 2 of the License, or (at your option) + * any later version. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include + +asmlinkage void sha512_transform_ssse3(const char *data, u64 *digest, + u64 rounds); +#ifdef CONFIG_AS_AVX +asmlinkage void sha512_transform_avx(const char *data, u64 *digest, + u64 rounds); +#endif +#ifdef CONFIG_AS_AVX2 +asmlinkage void sha512_transform_rorx(const char *data, u64 *digest, + u64 rounds); +#endif + +asmlinkage void (*sha512_transform_asm)(const char *, u64 *, u64); + + +static int sha512_ssse3_init(struct shash_desc *desc) +{ + struct sha512_state *sctx = shash_desc_ctx(desc); + + sctx->state[0] = SHA512_H0; + sctx->state[1] = SHA512_H1; + sctx->state[2] = SHA512_H2; + sctx->state[3] = SHA512_H3; + sctx->state[4] = SHA512_H4; + sctx->state[5] = SHA512_H5; + sctx->state[6] = SHA512_H6; + sctx->state[7] = SHA512_H7; + sctx->count[0] = sctx->count[1] = 0; + + return 0; +} + +static int __sha512_ssse3_update(struct shash_desc *desc, const u8 *data, + unsigned int len, unsigned int partial) +{ + struct sha512_state *sctx = shash_desc_ctx(desc); + unsigned int done = 0; + + sctx->count[0] += len; + if (sctx->count[0] < len) + sctx->count[1]++; + + if (partial) { + done = SHA512_BLOCK_SIZE - partial; + memcpy(sctx->buf + partial, data, done); + sha512_transform_asm(sctx->buf, sctx->state, 1); + } + + if (len - done >= SHA512_BLOCK_SIZE) { + const unsigned int rounds = (len - done) / SHA512_BLOCK_SIZE; + + sha512_transform_asm(data + done, sctx->state, (u64) rounds); + + done += rounds * SHA512_BLOCK_SIZE; + } + + memcpy(sctx->buf, data + done, len - done); + + return 0; +} + +static int sha512_ssse3_update(struct shash_desc *desc, const u8 *data, + unsigned int len) +{ + struct sha512_state *sctx = shash_desc_ctx(desc); + unsigned int partial = sctx->count[0] % SHA512_BLOCK_SIZE; + int res; + + /* Handle the fast case right here */ + if (partial + len < SHA512_BLOCK_SIZE) { + sctx->count[0] += len; + if (sctx->count[0] < len) + sctx->count[1]++; + memcpy(sctx->buf + partial, data, len); + + return 0; + } + + if (!irq_fpu_usable()) { + res = crypto_sha512_update(desc, data, len); + } else { + kernel_fpu_begin(); + res = __sha512_ssse3_update(desc, data, len, partial); + kernel_fpu_end(); + } + + return res; +} + + +/* Add padding and return the message digest. */ +static int sha512_ssse3_final(struct shash_desc *desc, u8 *out) +{ + struct sha512_state *sctx = shash_desc_ctx(desc); + unsigned int i, index, padlen; + __be64 *dst = (__be64 *)out; + __be64 bits[2]; + static const u8 padding[SHA512_BLOCK_SIZE] = { 0x80, }; + + /* save number of bits */ + bits[1] = cpu_to_be64(sctx->count[0] << 3); + bits[0] = cpu_to_be64(sctx->count[1] << 3) | sctx->count[0] >> 61; + + /* Pad out to 112 mod 128 and append length */ + index = sctx->count[0] & 0x7f; + padlen = (index < 112) ? (112 - index) : ((128+112) - index); + + if (!irq_fpu_usable()) { + crypto_sha512_update(desc, padding, padlen); + crypto_sha512_update(desc, (const u8 *)&bits, sizeof(bits)); + } else { + kernel_fpu_begin(); + /* We need to fill a whole block for __sha512_ssse3_update() */ + if (padlen <= 112) { + sctx->count[0] += padlen; + if (sctx->count[0] < padlen) + sctx->count[1]++; + memcpy(sctx->buf + index, padding, padlen); + } else { + __sha512_ssse3_update(desc, padding, padlen, index); + } + __sha512_ssse3_update(desc, (const u8 *)&bits, + sizeof(bits), 112); + kernel_fpu_end(); + } + + /* Store state in digest */ + for (i = 0; i < 8; i++) + dst[i] = cpu_to_be64(sctx->state[i]); + + /* Wipe context */ + memset(sctx, 0, sizeof(*sctx)); + + return 0; +} + +static int sha512_ssse3_export(struct shash_desc *desc, void *out) +{ + struct sha512_state *sctx = shash_desc_ctx(desc); + + memcpy(out, sctx, sizeof(*sctx)); + + return 0; +} + +static int sha512_ssse3_import(struct shash_desc *desc, const void *in) +{ + struct sha512_state *sctx = shash_desc_ctx(desc); + + memcpy(sctx, in, sizeof(*sctx)); + + return 0; +} + +static struct shash_alg alg = { + .digestsize = SHA512_DIGEST_SIZE, + .init = sha512_ssse3_init, + .update = sha512_ssse3_update, + .final = sha512_ssse3_final, + .export = sha512_ssse3_export, + .import = sha512_ssse3_import, + .descsize = sizeof(struct sha512_state), + .statesize = sizeof(struct sha512_state), + .base = { + .cra_name = "sha512", + .cra_driver_name = "sha512-ssse3", + .cra_priority = 150, + .cra_flags = CRYPTO_ALG_TYPE_SHASH, + .cra_blocksize = SHA512_BLOCK_SIZE, + .cra_module = THIS_MODULE, + } +}; + +#ifdef CONFIG_AS_AVX +static bool __init avx_usable(void) +{ + u64 xcr0; + + if (!cpu_has_avx || !cpu_has_osxsave) + return false; + + xcr0 = xgetbv(XCR_XFEATURE_ENABLED_MASK); + if ((xcr0 & (XSTATE_SSE | XSTATE_YMM)) != (XSTATE_SSE | XSTATE_YMM)) { + pr_info("AVX detected but unusable.\n"); + + return false; + } + + return true; +} +#endif + +static int __init sha512_ssse3_mod_init(void) +{ + /* test for SSE3 first */ + if (cpu_has_ssse3) + sha512_transform_asm = sha512_transform_ssse3; + +#ifdef CONFIG_AS_AVX + /* allow AVX to override SSSE3, it's a little faster */ + if (avx_usable()) { +#ifdef CONFIG_AS_AVX2 + if (boot_cpu_has(X86_FEATURE_AVX2)) + sha512_transform_asm = sha512_transform_rorx; + else +#endif + sha512_transform_asm = sha512_transform_avx; + } +#endif + + if (sha512_transform_asm) { +#ifdef CONFIG_AS_AVX + if (sha512_transform_asm == sha512_transform_avx) + pr_info("Using AVX optimized SHA-512 implementation\n"); +#ifdef CONFIG_AS_AVX2 + else if (sha512_transform_asm == sha512_transform_rorx) + pr_info("Using AVX2 optimized SHA-512 implementation\n"); +#endif + else +#endif + pr_info("Using SSSE3 optimized SHA-512 implementation\n"); + return crypto_register_shash(&alg); + } + pr_info("Neither AVX nor SSSE3 is available/usable.\n"); + + return -ENODEV; +} + +static void __exit sha512_ssse3_mod_fini(void) +{ + crypto_unregister_shash(&alg); +} + +module_init(sha512_ssse3_mod_init); +module_exit(sha512_ssse3_mod_fini); + +MODULE_LICENSE("GPL"); +MODULE_DESCRIPTION("SHA512 Secure Hash Algorithm, Supplemental SSE3 accelerated"); + +MODULE_ALIAS("sha512"); diff --git a/crypto/Kconfig b/crypto/Kconfig index 8064ef1..a654b13 100644 --- a/crypto/Kconfig +++ b/crypto/Kconfig @@ -481,6 +481,17 @@ config CRYPTO_SHA256_SSSE3 Extensions version 1 (AVX1), or Advanced Vector Extensions version 2 (AVX2) instructions, when available. +config CRYPTO_SHA512_SSSE3 + tristate "SHA512 digest algorithm (SSSE3/AVX/AVX2)" + depends on X86 && 64BIT + select CRYPTO_SHA512 + select CRYPTO_HASH + help + SHA-512 secure hash standard (DFIPS 180-2) implemented + using Supplemental SSE3 (SSSE3) instructions, or Advanced Vector + Extensions version 1 (AVX1), or Advanced Vector Extensions + version 2 (AVX2) instructions, when available. + config CRYPTO_SHA1_SPARC64 tristate "SHA1 digest algorithm (SPARC64)" depends on SPARC64 -- 1.7.11.7