linux/arch/x86/crypto/serpent_avx_glue.c

615 lines
17 KiB
C
Raw Normal View History

crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
/*
* Glue Code for AVX assembler versions of Serpent Cipher
*
* Copyright (C) 2012 Johannes Goetzfried
* <Johannes.Goetzfried@informatik.stud.uni-erlangen.de>
*
* Copyright © 2011-2013 Jussi Kivilinna <jussi.kivilinna@iki.fi>
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
*
* 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.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
* USA
*
*/
#include <linux/module.h>
#include <linux/hardirq.h>
#include <linux/types.h>
#include <linux/crypto.h>
#include <linux/err.h>
#include <crypto/ablk_helper.h>
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
#include <crypto/algapi.h>
#include <crypto/serpent.h>
#include <crypto/cryptd.h>
#include <crypto/b128ops.h>
#include <crypto/ctr.h>
#include <crypto/lrw.h>
#include <crypto/xts.h>
#include <asm/fpu/api.h>
#include <asm/crypto/serpent-avx.h>
#include <asm/crypto/glue_helper.h>
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
/* 8-way parallel cipher functions */
asmlinkage void serpent_ecb_enc_8way_avx(struct serpent_ctx *ctx, u8 *dst,
const u8 *src);
EXPORT_SYMBOL_GPL(serpent_ecb_enc_8way_avx);
asmlinkage void serpent_ecb_dec_8way_avx(struct serpent_ctx *ctx, u8 *dst,
const u8 *src);
EXPORT_SYMBOL_GPL(serpent_ecb_dec_8way_avx);
asmlinkage void serpent_cbc_dec_8way_avx(struct serpent_ctx *ctx, u8 *dst,
const u8 *src);
EXPORT_SYMBOL_GPL(serpent_cbc_dec_8way_avx);
asmlinkage void serpent_ctr_8way_avx(struct serpent_ctx *ctx, u8 *dst,
const u8 *src, le128 *iv);
EXPORT_SYMBOL_GPL(serpent_ctr_8way_avx);
asmlinkage void serpent_xts_enc_8way_avx(struct serpent_ctx *ctx, u8 *dst,
const u8 *src, le128 *iv);
EXPORT_SYMBOL_GPL(serpent_xts_enc_8way_avx);
asmlinkage void serpent_xts_dec_8way_avx(struct serpent_ctx *ctx, u8 *dst,
const u8 *src, le128 *iv);
EXPORT_SYMBOL_GPL(serpent_xts_dec_8way_avx);
void __serpent_crypt_ctr(void *ctx, u128 *dst, const u128 *src, le128 *iv)
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
{
be128 ctrblk;
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
le128_to_be128(&ctrblk, iv);
le128_inc(iv);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
__serpent_encrypt(ctx, (u8 *)&ctrblk, (u8 *)&ctrblk);
u128_xor(dst, src, (u128 *)&ctrblk);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
}
EXPORT_SYMBOL_GPL(__serpent_crypt_ctr);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
void serpent_xts_enc(void *ctx, u128 *dst, const u128 *src, le128 *iv)
{
glue_xts_crypt_128bit_one(ctx, dst, src, iv,
GLUE_FUNC_CAST(__serpent_encrypt));
}
EXPORT_SYMBOL_GPL(serpent_xts_enc);
void serpent_xts_dec(void *ctx, u128 *dst, const u128 *src, le128 *iv)
{
glue_xts_crypt_128bit_one(ctx, dst, src, iv,
GLUE_FUNC_CAST(__serpent_decrypt));
}
EXPORT_SYMBOL_GPL(serpent_xts_dec);
static const struct common_glue_ctx serpent_enc = {
.num_funcs = 2,
.fpu_blocks_limit = SERPENT_PARALLEL_BLOCKS,
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
.funcs = { {
.num_blocks = SERPENT_PARALLEL_BLOCKS,
.fn_u = { .ecb = GLUE_FUNC_CAST(serpent_ecb_enc_8way_avx) }
}, {
.num_blocks = 1,
.fn_u = { .ecb = GLUE_FUNC_CAST(__serpent_encrypt) }
} }
};
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
static const struct common_glue_ctx serpent_ctr = {
.num_funcs = 2,
.fpu_blocks_limit = SERPENT_PARALLEL_BLOCKS,
.funcs = { {
.num_blocks = SERPENT_PARALLEL_BLOCKS,
.fn_u = { .ctr = GLUE_CTR_FUNC_CAST(serpent_ctr_8way_avx) }
}, {
.num_blocks = 1,
.fn_u = { .ctr = GLUE_CTR_FUNC_CAST(__serpent_crypt_ctr) }
} }
};
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
static const struct common_glue_ctx serpent_enc_xts = {
.num_funcs = 2,
.fpu_blocks_limit = SERPENT_PARALLEL_BLOCKS,
.funcs = { {
.num_blocks = SERPENT_PARALLEL_BLOCKS,
.fn_u = { .xts = GLUE_XTS_FUNC_CAST(serpent_xts_enc_8way_avx) }
}, {
.num_blocks = 1,
.fn_u = { .xts = GLUE_XTS_FUNC_CAST(serpent_xts_enc) }
} }
};
static const struct common_glue_ctx serpent_dec = {
.num_funcs = 2,
.fpu_blocks_limit = SERPENT_PARALLEL_BLOCKS,
.funcs = { {
.num_blocks = SERPENT_PARALLEL_BLOCKS,
.fn_u = { .ecb = GLUE_FUNC_CAST(serpent_ecb_dec_8way_avx) }
}, {
.num_blocks = 1,
.fn_u = { .ecb = GLUE_FUNC_CAST(__serpent_decrypt) }
} }
};
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
static const struct common_glue_ctx serpent_dec_cbc = {
.num_funcs = 2,
.fpu_blocks_limit = SERPENT_PARALLEL_BLOCKS,
.funcs = { {
.num_blocks = SERPENT_PARALLEL_BLOCKS,
.fn_u = { .cbc = GLUE_CBC_FUNC_CAST(serpent_cbc_dec_8way_avx) }
}, {
.num_blocks = 1,
.fn_u = { .cbc = GLUE_CBC_FUNC_CAST(__serpent_decrypt) }
} }
};
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
static const struct common_glue_ctx serpent_dec_xts = {
.num_funcs = 2,
.fpu_blocks_limit = SERPENT_PARALLEL_BLOCKS,
.funcs = { {
.num_blocks = SERPENT_PARALLEL_BLOCKS,
.fn_u = { .xts = GLUE_XTS_FUNC_CAST(serpent_xts_dec_8way_avx) }
}, {
.num_blocks = 1,
.fn_u = { .xts = GLUE_XTS_FUNC_CAST(serpent_xts_dec) }
} }
};
static int ecb_encrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
struct scatterlist *src, unsigned int nbytes)
{
return glue_ecb_crypt_128bit(&serpent_enc, desc, dst, src, nbytes);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
}
static int ecb_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
struct scatterlist *src, unsigned int nbytes)
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
{
return glue_ecb_crypt_128bit(&serpent_dec, desc, dst, src, nbytes);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
}
static int cbc_encrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
struct scatterlist *src, unsigned int nbytes)
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
{
return glue_cbc_encrypt_128bit(GLUE_FUNC_CAST(__serpent_encrypt), desc,
dst, src, nbytes);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
}
static int cbc_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
struct scatterlist *src, unsigned int nbytes)
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
{
return glue_cbc_decrypt_128bit(&serpent_dec_cbc, desc, dst, src,
nbytes);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
}
static int ctr_crypt(struct blkcipher_desc *desc, struct scatterlist *dst,
struct scatterlist *src, unsigned int nbytes)
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
{
return glue_ctr_crypt_128bit(&serpent_ctr, desc, dst, src, nbytes);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
}
static inline bool serpent_fpu_begin(bool fpu_enabled, unsigned int nbytes)
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
{
return glue_fpu_begin(SERPENT_BLOCK_SIZE, SERPENT_PARALLEL_BLOCKS,
NULL, fpu_enabled, nbytes);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
}
static inline void serpent_fpu_end(bool fpu_enabled)
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
{
glue_fpu_end(fpu_enabled);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
}
struct crypt_priv {
struct serpent_ctx *ctx;
bool fpu_enabled;
};
static void encrypt_callback(void *priv, u8 *srcdst, unsigned int nbytes)
{
const unsigned int bsize = SERPENT_BLOCK_SIZE;
struct crypt_priv *ctx = priv;
int i;
ctx->fpu_enabled = serpent_fpu_begin(ctx->fpu_enabled, nbytes);
if (nbytes == bsize * SERPENT_PARALLEL_BLOCKS) {
serpent_ecb_enc_8way_avx(ctx->ctx, srcdst, srcdst);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
return;
}
for (i = 0; i < nbytes / bsize; i++, srcdst += bsize)
__serpent_encrypt(ctx->ctx, srcdst, srcdst);
}
static void decrypt_callback(void *priv, u8 *srcdst, unsigned int nbytes)
{
const unsigned int bsize = SERPENT_BLOCK_SIZE;
struct crypt_priv *ctx = priv;
int i;
ctx->fpu_enabled = serpent_fpu_begin(ctx->fpu_enabled, nbytes);
if (nbytes == bsize * SERPENT_PARALLEL_BLOCKS) {
serpent_ecb_dec_8way_avx(ctx->ctx, srcdst, srcdst);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
return;
}
for (i = 0; i < nbytes / bsize; i++, srcdst += bsize)
__serpent_decrypt(ctx->ctx, srcdst, srcdst);
}
int lrw_serpent_setkey(struct crypto_tfm *tfm, const u8 *key,
unsigned int keylen)
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
{
struct serpent_lrw_ctx *ctx = crypto_tfm_ctx(tfm);
int err;
err = __serpent_setkey(&ctx->serpent_ctx, key, keylen -
SERPENT_BLOCK_SIZE);
if (err)
return err;
return lrw_init_table(&ctx->lrw_table, key + keylen -
SERPENT_BLOCK_SIZE);
}
EXPORT_SYMBOL_GPL(lrw_serpent_setkey);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
static int lrw_encrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
struct scatterlist *src, unsigned int nbytes)
{
struct serpent_lrw_ctx *ctx = crypto_blkcipher_ctx(desc->tfm);
be128 buf[SERPENT_PARALLEL_BLOCKS];
struct crypt_priv crypt_ctx = {
.ctx = &ctx->serpent_ctx,
.fpu_enabled = false,
};
struct lrw_crypt_req req = {
.tbuf = buf,
.tbuflen = sizeof(buf),
.table_ctx = &ctx->lrw_table,
.crypt_ctx = &crypt_ctx,
.crypt_fn = encrypt_callback,
};
int ret;
desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP;
ret = lrw_crypt(desc, dst, src, nbytes, &req);
serpent_fpu_end(crypt_ctx.fpu_enabled);
return ret;
}
static int lrw_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
struct scatterlist *src, unsigned int nbytes)
{
struct serpent_lrw_ctx *ctx = crypto_blkcipher_ctx(desc->tfm);
be128 buf[SERPENT_PARALLEL_BLOCKS];
struct crypt_priv crypt_ctx = {
.ctx = &ctx->serpent_ctx,
.fpu_enabled = false,
};
struct lrw_crypt_req req = {
.tbuf = buf,
.tbuflen = sizeof(buf),
.table_ctx = &ctx->lrw_table,
.crypt_ctx = &crypt_ctx,
.crypt_fn = decrypt_callback,
};
int ret;
desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP;
ret = lrw_crypt(desc, dst, src, nbytes, &req);
serpent_fpu_end(crypt_ctx.fpu_enabled);
return ret;
}
void lrw_serpent_exit_tfm(struct crypto_tfm *tfm)
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
{
struct serpent_lrw_ctx *ctx = crypto_tfm_ctx(tfm);
lrw_free_table(&ctx->lrw_table);
}
EXPORT_SYMBOL_GPL(lrw_serpent_exit_tfm);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
int xts_serpent_setkey(struct crypto_tfm *tfm, const u8 *key,
unsigned int keylen)
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
{
struct serpent_xts_ctx *ctx = crypto_tfm_ctx(tfm);
int err;
err = xts_check_key(tfm, key, keylen);
if (err)
return err;
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
/* first half of xts-key is for crypt */
err = __serpent_setkey(&ctx->crypt_ctx, key, keylen / 2);
if (err)
return err;
/* second half of xts-key is for tweak */
return __serpent_setkey(&ctx->tweak_ctx, key + keylen / 2, keylen / 2);
}
EXPORT_SYMBOL_GPL(xts_serpent_setkey);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
static int xts_encrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
struct scatterlist *src, unsigned int nbytes)
{
struct serpent_xts_ctx *ctx = crypto_blkcipher_ctx(desc->tfm);
return glue_xts_crypt_128bit(&serpent_enc_xts, desc, dst, src, nbytes,
XTS_TWEAK_CAST(__serpent_encrypt),
&ctx->tweak_ctx, &ctx->crypt_ctx);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
}
static int xts_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
struct scatterlist *src, unsigned int nbytes)
{
struct serpent_xts_ctx *ctx = crypto_blkcipher_ctx(desc->tfm);
return glue_xts_crypt_128bit(&serpent_dec_xts, desc, dst, src, nbytes,
XTS_TWEAK_CAST(__serpent_encrypt),
&ctx->tweak_ctx, &ctx->crypt_ctx);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
}
static struct crypto_alg serpent_algs[10] = { {
.cra_name = "__ecb-serpent-avx",
.cra_driver_name = "__driver-ecb-serpent-avx",
.cra_priority = 0,
.cra_flags = CRYPTO_ALG_TYPE_BLKCIPHER |
CRYPTO_ALG_INTERNAL,
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
.cra_blocksize = SERPENT_BLOCK_SIZE,
.cra_ctxsize = sizeof(struct serpent_ctx),
.cra_alignmask = 0,
.cra_type = &crypto_blkcipher_type,
.cra_module = THIS_MODULE,
.cra_u = {
.blkcipher = {
.min_keysize = SERPENT_MIN_KEY_SIZE,
.max_keysize = SERPENT_MAX_KEY_SIZE,
.setkey = serpent_setkey,
.encrypt = ecb_encrypt,
.decrypt = ecb_decrypt,
},
},
}, {
.cra_name = "__cbc-serpent-avx",
.cra_driver_name = "__driver-cbc-serpent-avx",
.cra_priority = 0,
.cra_flags = CRYPTO_ALG_TYPE_BLKCIPHER |
CRYPTO_ALG_INTERNAL,
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
.cra_blocksize = SERPENT_BLOCK_SIZE,
.cra_ctxsize = sizeof(struct serpent_ctx),
.cra_alignmask = 0,
.cra_type = &crypto_blkcipher_type,
.cra_module = THIS_MODULE,
.cra_u = {
.blkcipher = {
.min_keysize = SERPENT_MIN_KEY_SIZE,
.max_keysize = SERPENT_MAX_KEY_SIZE,
.setkey = serpent_setkey,
.encrypt = cbc_encrypt,
.decrypt = cbc_decrypt,
},
},
}, {
.cra_name = "__ctr-serpent-avx",
.cra_driver_name = "__driver-ctr-serpent-avx",
.cra_priority = 0,
.cra_flags = CRYPTO_ALG_TYPE_BLKCIPHER |
CRYPTO_ALG_INTERNAL,
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
.cra_blocksize = 1,
.cra_ctxsize = sizeof(struct serpent_ctx),
.cra_alignmask = 0,
.cra_type = &crypto_blkcipher_type,
.cra_module = THIS_MODULE,
.cra_u = {
.blkcipher = {
.min_keysize = SERPENT_MIN_KEY_SIZE,
.max_keysize = SERPENT_MAX_KEY_SIZE,
.ivsize = SERPENT_BLOCK_SIZE,
.setkey = serpent_setkey,
.encrypt = ctr_crypt,
.decrypt = ctr_crypt,
},
},
}, {
.cra_name = "__lrw-serpent-avx",
.cra_driver_name = "__driver-lrw-serpent-avx",
.cra_priority = 0,
.cra_flags = CRYPTO_ALG_TYPE_BLKCIPHER |
CRYPTO_ALG_INTERNAL,
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
.cra_blocksize = SERPENT_BLOCK_SIZE,
.cra_ctxsize = sizeof(struct serpent_lrw_ctx),
.cra_alignmask = 0,
.cra_type = &crypto_blkcipher_type,
.cra_module = THIS_MODULE,
.cra_exit = lrw_serpent_exit_tfm,
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
.cra_u = {
.blkcipher = {
.min_keysize = SERPENT_MIN_KEY_SIZE +
SERPENT_BLOCK_SIZE,
.max_keysize = SERPENT_MAX_KEY_SIZE +
SERPENT_BLOCK_SIZE,
.ivsize = SERPENT_BLOCK_SIZE,
.setkey = lrw_serpent_setkey,
.encrypt = lrw_encrypt,
.decrypt = lrw_decrypt,
},
},
}, {
.cra_name = "__xts-serpent-avx",
.cra_driver_name = "__driver-xts-serpent-avx",
.cra_priority = 0,
.cra_flags = CRYPTO_ALG_TYPE_BLKCIPHER |
CRYPTO_ALG_INTERNAL,
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
.cra_blocksize = SERPENT_BLOCK_SIZE,
.cra_ctxsize = sizeof(struct serpent_xts_ctx),
.cra_alignmask = 0,
.cra_type = &crypto_blkcipher_type,
.cra_module = THIS_MODULE,
.cra_u = {
.blkcipher = {
.min_keysize = SERPENT_MIN_KEY_SIZE * 2,
.max_keysize = SERPENT_MAX_KEY_SIZE * 2,
.ivsize = SERPENT_BLOCK_SIZE,
.setkey = xts_serpent_setkey,
.encrypt = xts_encrypt,
.decrypt = xts_decrypt,
},
},
}, {
.cra_name = "ecb(serpent)",
.cra_driver_name = "ecb-serpent-avx",
.cra_priority = 500,
.cra_flags = CRYPTO_ALG_TYPE_ABLKCIPHER | CRYPTO_ALG_ASYNC,
.cra_blocksize = SERPENT_BLOCK_SIZE,
.cra_ctxsize = sizeof(struct async_helper_ctx),
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
.cra_alignmask = 0,
.cra_type = &crypto_ablkcipher_type,
.cra_module = THIS_MODULE,
.cra_init = ablk_init,
.cra_exit = ablk_exit,
.cra_u = {
.ablkcipher = {
.min_keysize = SERPENT_MIN_KEY_SIZE,
.max_keysize = SERPENT_MAX_KEY_SIZE,
.setkey = ablk_set_key,
.encrypt = ablk_encrypt,
.decrypt = ablk_decrypt,
},
},
}, {
.cra_name = "cbc(serpent)",
.cra_driver_name = "cbc-serpent-avx",
.cra_priority = 500,
.cra_flags = CRYPTO_ALG_TYPE_ABLKCIPHER | CRYPTO_ALG_ASYNC,
.cra_blocksize = SERPENT_BLOCK_SIZE,
.cra_ctxsize = sizeof(struct async_helper_ctx),
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
.cra_alignmask = 0,
.cra_type = &crypto_ablkcipher_type,
.cra_module = THIS_MODULE,
.cra_init = ablk_init,
.cra_exit = ablk_exit,
.cra_u = {
.ablkcipher = {
.min_keysize = SERPENT_MIN_KEY_SIZE,
.max_keysize = SERPENT_MAX_KEY_SIZE,
.ivsize = SERPENT_BLOCK_SIZE,
.setkey = ablk_set_key,
.encrypt = __ablk_encrypt,
.decrypt = ablk_decrypt,
},
},
}, {
.cra_name = "ctr(serpent)",
.cra_driver_name = "ctr-serpent-avx",
.cra_priority = 500,
.cra_flags = CRYPTO_ALG_TYPE_ABLKCIPHER | CRYPTO_ALG_ASYNC,
.cra_blocksize = 1,
.cra_ctxsize = sizeof(struct async_helper_ctx),
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
.cra_alignmask = 0,
.cra_type = &crypto_ablkcipher_type,
.cra_module = THIS_MODULE,
.cra_init = ablk_init,
.cra_exit = ablk_exit,
.cra_u = {
.ablkcipher = {
.min_keysize = SERPENT_MIN_KEY_SIZE,
.max_keysize = SERPENT_MAX_KEY_SIZE,
.ivsize = SERPENT_BLOCK_SIZE,
.setkey = ablk_set_key,
.encrypt = ablk_encrypt,
.decrypt = ablk_encrypt,
.geniv = "chainiv",
},
},
}, {
.cra_name = "lrw(serpent)",
.cra_driver_name = "lrw-serpent-avx",
.cra_priority = 500,
.cra_flags = CRYPTO_ALG_TYPE_ABLKCIPHER | CRYPTO_ALG_ASYNC,
.cra_blocksize = SERPENT_BLOCK_SIZE,
.cra_ctxsize = sizeof(struct async_helper_ctx),
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
.cra_alignmask = 0,
.cra_type = &crypto_ablkcipher_type,
.cra_module = THIS_MODULE,
.cra_init = ablk_init,
.cra_exit = ablk_exit,
.cra_u = {
.ablkcipher = {
.min_keysize = SERPENT_MIN_KEY_SIZE +
SERPENT_BLOCK_SIZE,
.max_keysize = SERPENT_MAX_KEY_SIZE +
SERPENT_BLOCK_SIZE,
.ivsize = SERPENT_BLOCK_SIZE,
.setkey = ablk_set_key,
.encrypt = ablk_encrypt,
.decrypt = ablk_decrypt,
},
},
}, {
.cra_name = "xts(serpent)",
.cra_driver_name = "xts-serpent-avx",
.cra_priority = 500,
.cra_flags = CRYPTO_ALG_TYPE_ABLKCIPHER | CRYPTO_ALG_ASYNC,
.cra_blocksize = SERPENT_BLOCK_SIZE,
.cra_ctxsize = sizeof(struct async_helper_ctx),
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
.cra_alignmask = 0,
.cra_type = &crypto_ablkcipher_type,
.cra_module = THIS_MODULE,
.cra_init = ablk_init,
.cra_exit = ablk_exit,
.cra_u = {
.ablkcipher = {
.min_keysize = SERPENT_MIN_KEY_SIZE * 2,
.max_keysize = SERPENT_MAX_KEY_SIZE * 2,
.ivsize = SERPENT_BLOCK_SIZE,
.setkey = ablk_set_key,
.encrypt = ablk_encrypt,
.decrypt = ablk_decrypt,
},
},
} };
static int __init serpent_init(void)
{
const char *feature_name;
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
x86/fpu: Rename XSAVE macros There are two concepts that have some confusing naming: 1. Extended State Component numbers (currently called XFEATURE_BIT_*) 2. Extended State Component masks (currently called XSTATE_*) The numbers are (currently) from 0-9. State component 3 is the bounds registers for MPX, for instance. But when we want to enable "state component 3", we go set a bit in XCR0. The bit we set is 1<<3. We can check to see if a state component feature is enabled by looking at its bit. The current 'xfeature_bit's are at best xfeature bit _numbers_. Calling them bits is at best inconsistent with ending the enum list with 'XFEATURES_NR_MAX'. This patch renames the enum to be 'xfeature'. These also happen to be what the Intel documentation calls a "state component". We also want to differentiate these from the "XSTATE_*" macros. The "XSTATE_*" macros are a mask, and we rename them to match. These macros are reasonably widely used so this patch is a wee bit big, but this really is just a rename. The only non-mechanical part of this is the s/XSTATE_EXTEND_MASK/XFEATURE_MASK_EXTEND/ We need a better name for it, but that's another patch. Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Brian Gerst <brgerst@gmail.com> Cc: Denys Vlasenko <dvlasenk@redhat.com> Cc: Fenghua Yu <fenghua.yu@intel.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Tim Chen <tim.c.chen@linux.intel.com> Cc: dave@sr71.net Cc: linux-kernel@vger.kernel.org Link: http://lkml.kernel.org/r/20150902233126.38653250@viggo.jf.intel.com [ Ported to v4.3-rc1. ] Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-09-03 07:31:26 +08:00
if (!cpu_has_xfeatures(XFEATURE_MASK_SSE | XFEATURE_MASK_YMM,
&feature_name)) {
pr_info("CPU feature '%s' is not supported.\n", feature_name);
crypto: serpent - add x86_64/avx assembler implementation This patch adds a x86_64/avx assembler implementation of the Serpent block cipher. The implementation is very similar to the sse2 implementation and processes eight blocks in parallel. Because of the new non-destructive three operand syntax all move-instructions can be removed and therefore a little performance increase is provided. Patch has been tested with tcrypt and automated filesystem tests. Tcrypt benchmark results: Intel Core i5-2500 CPU (fam:6, model:42, step:7) serpent-avx-x86_64 vs. serpent-sse2-x86_64 128bit key: (lrw:256bit) (xts:256bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.03x 1.01x 1.01x 1.01x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 0.99x 1.00x 1.01x 1.00x 1.00x 256B 1.05x 1.03x 1.00x 1.02x 1.05x 1.06x 1.05x 1.02x 1.05x 1.02x 1024B 1.05x 1.02x 1.00x 1.02x 1.05x 1.06x 1.05x 1.03x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.03x 1.04x 1.02x 256bit key: (lrw:384bit) (xts:512bit) size ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec 16B 1.01x 1.00x 1.01x 1.01x 1.00x 1.00x 0.99x 1.03x 1.01x 1.01x 64B 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.00x 1.01x 1.00x 1.02x 256B 1.05x 1.02x 1.00x 1.02x 1.05x 1.02x 1.04x 1.05x 1.05x 1.02x 1024B 1.06x 1.02x 1.00x 1.02x 1.07x 1.06x 1.05x 1.04x 1.05x 1.02x 8192B 1.05x 1.02x 1.00x 1.02x 1.06x 1.06x 1.04x 1.05x 1.05x 1.02x serpent-avx-x86_64 vs aes-asm (8kB block): 128bit 256bit ecb-enc 1.26x 1.73x ecb-dec 1.20x 1.64x cbc-enc 0.33x 0.45x cbc-dec 1.24x 1.67x ctr-enc 1.32x 1.76x ctr-dec 1.32x 1.76x lrw-enc 1.20x 1.60x lrw-dec 1.15x 1.54x xts-enc 1.22x 1.64x xts-dec 1.17x 1.57x Signed-off-by: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2012-06-12 16:47:43 +08:00
return -ENODEV;
}
return crypto_register_algs(serpent_algs, ARRAY_SIZE(serpent_algs));
}
static void __exit serpent_exit(void)
{
crypto_unregister_algs(serpent_algs, ARRAY_SIZE(serpent_algs));
}
module_init(serpent_init);
module_exit(serpent_exit);
MODULE_DESCRIPTION("Serpent Cipher Algorithm, AVX optimized");
MODULE_LICENSE("GPL");
MODULE_ALIAS_CRYPTO("serpent");