rijndael: split AES-NI functions to separate file

* cipher/Makefile.in: Add 'rijndael-aesni.c'.
* cipher/rijndael-aesni.c: New.
* cipher/rijndael-internal.h: New.
* cipher/rijndael.c (MAXKC, MAXROUNDS, BLOCKSIZE, ATTR_ALIGNED_16)
(USE_AMD64_ASM, USE_ARM_ASM, USE_PADLOCK, USE_AESNI, RIJNDAEL_context)
(keyschenc, keyschdec, padlockkey): Move to 'rijndael-internal.h'.
(u128_s, aesni_prepare, aesni_cleanup, aesni_cleanup_2_6)
(aesni_do_setkey, do_aesni_enc, do_aesni_dec, do_aesni_enc_vec4)
(do_aesni_dec_vec4, do_aesni_cfb, do_aesni_ctr, do_aesni_ctr_4): Move
to 'rijndael-aesni.c'.
(prepare_decryption, rijndael_encrypt, _gcry_aes_cfb_enc)
(_gcry_aes_cbc_enc, _gcry_aes_ctr_enc, rijndael_decrypt)
(_gcry_aes_cfb_dec, _gcry_aes_cbc_dec) [USE_AESNI]: Move to functions
in 'rijdael-aesni.c'.
* configure.ac [mpi_cpu_arch=x86]: Add 'rijndael-aesni.lo'.
--

Clean-up rijndael.c before new new hardware acceleration support gets added.

Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>

1 files changed, 1288 insertions, 0 deletions

diff --git a/cipher/rijndael-aesni.c b/cipher/rijndael-aesni.c
new file mode 100644
index 00000000..15ed4ad9
--- /dev/null
+++ b/cipher/rijndael-aesni.c
@@ -0,0 +1,1288 @@
+/* AES-NI accelerated AES for Libgcrypt
+ * Copyright (C) 2000, 2001, 2002, 2003, 2007,
+ *               2008, 2011, 2012 Free Software Foundation, Inc.
+ *
+ * This file is part of Libgcrypt.
+ *
+ * Libgcrypt is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU Lesser General Public License as
+ * published by the Free Software Foundation; either version 2.1 of
+ * the License, or (at your option) any later version.
+ *
+ * Libgcrypt 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 Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this program; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <config.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h> /* for memcmp() */
+
+#include "types.h"  /* for byte and u32 typedefs */
+#include "g10lib.h"
+#include "cipher.h"
+#include "bufhelp.h"
+#include "cipher-selftest.h"
+#include "rijndael-internal.h"
+
+
+#ifdef USE_AESNI
+
+
+typedef struct u128_s { u32 a, b, c, d; } u128_t;
+
+
+/* Two macros to be called prior and after the use of AESNI
+   instructions.  There should be no external function calls between
+   the use of these macros.  There purpose is to make sure that the
+   SSE regsiters are cleared and won't reveal any information about
+   the key or the data.  */
+#define aesni_prepare() do { } while (0)
+#define aesni_cleanup()                                                \
+  do { asm volatile ("pxor %%xmm0, %%xmm0\n\t"                         \
+                     "pxor %%xmm1, %%xmm1\n" :: );                     \
+  } while (0)
+#define aesni_cleanup_2_6()                                            \
+  do { asm volatile ("pxor %%xmm2, %%xmm2\n\t"                         \
+                     "pxor %%xmm3, %%xmm3\n"                           \
+                     "pxor %%xmm4, %%xmm4\n"                           \
+                     "pxor %%xmm5, %%xmm5\n"                           \
+                     "pxor %%xmm6, %%xmm6\n":: );                      \
+  } while (0)
+
+
+void
+_gcry_aes_aesni_do_setkey (RIJNDAEL_context *ctx, const byte *key)
+{
+  aesni_prepare();
+
+  if (ctx->rounds < 12)
+    {
+      /* 128-bit key */
+#define AESKEYGENASSIST_xmm1_xmm2(imm8) \
+	".byte 0x66, 0x0f, 0x3a, 0xdf, 0xd1, " #imm8 " \n\t"
+#define AESKEY_EXPAND128 \
+	"pshufd $0xff, %%xmm2, %%xmm2\n\t" \
+	"movdqa %%xmm1, %%xmm3\n\t" \
+	"pslldq $4, %%xmm3\n\t" \
+	"pxor   %%xmm3, %%xmm1\n\t" \
+	"pslldq $4, %%xmm3\n\t" \
+	"pxor   %%xmm3, %%xmm1\n\t" \
+	"pslldq $4, %%xmm3\n\t" \
+	"pxor   %%xmm3, %%xmm2\n\t" \
+	"pxor   %%xmm2, %%xmm1\n\t"
+
+      asm volatile ("movdqu (%[key]), %%xmm1\n\t"     /* xmm1 := key   */
+                    "movdqa %%xmm1, (%[ksch])\n\t"     /* ksch[0] := xmm1  */
+                    AESKEYGENASSIST_xmm1_xmm2(0x01)
+                    AESKEY_EXPAND128
+                    "movdqa %%xmm1, 0x10(%[ksch])\n\t" /* ksch[1] := xmm1  */
+                    AESKEYGENASSIST_xmm1_xmm2(0x02)
+                    AESKEY_EXPAND128
+                    "movdqa %%xmm1, 0x20(%[ksch])\n\t" /* ksch[2] := xmm1  */
+                    AESKEYGENASSIST_xmm1_xmm2(0x04)
+                    AESKEY_EXPAND128
+                    "movdqa %%xmm1, 0x30(%[ksch])\n\t" /* ksch[3] := xmm1  */
+                    AESKEYGENASSIST_xmm1_xmm2(0x08)
+                    AESKEY_EXPAND128
+                    "movdqa %%xmm1, 0x40(%[ksch])\n\t" /* ksch[4] := xmm1  */
+                    AESKEYGENASSIST_xmm1_xmm2(0x10)
+                    AESKEY_EXPAND128
+                    "movdqa %%xmm1, 0x50(%[ksch])\n\t" /* ksch[5] := xmm1  */
+                    AESKEYGENASSIST_xmm1_xmm2(0x20)
+                    AESKEY_EXPAND128
+                    "movdqa %%xmm1, 0x60(%[ksch])\n\t" /* ksch[6] := xmm1  */
+                    AESKEYGENASSIST_xmm1_xmm2(0x40)
+                    AESKEY_EXPAND128
+                    "movdqa %%xmm1, 0x70(%[ksch])\n\t" /* ksch[7] := xmm1  */
+                    AESKEYGENASSIST_xmm1_xmm2(0x80)
+                    AESKEY_EXPAND128
+                    "movdqa %%xmm1, 0x80(%[ksch])\n\t" /* ksch[8] := xmm1  */
+                    AESKEYGENASSIST_xmm1_xmm2(0x1b)
+                    AESKEY_EXPAND128
+                    "movdqa %%xmm1, 0x90(%[ksch])\n\t" /* ksch[9] := xmm1  */
+                    AESKEYGENASSIST_xmm1_xmm2(0x36)
+                    AESKEY_EXPAND128
+                    "movdqa %%xmm1, 0xa0(%[ksch])\n\t" /* ksch[10] := xmm1  */
+                    :
+                    : [key] "r" (key), [ksch] "r" (ctx->keyschenc)
+                    : "cc", "memory" );
+#undef AESKEYGENASSIST_xmm1_xmm2
+#undef AESKEY_EXPAND128
+    }
+  else if (ctx->rounds == 12)
+    {
+      /* 192-bit key */
+#define AESKEYGENASSIST_xmm3_xmm2(imm8) \
+	".byte 0x66, 0x0f, 0x3a, 0xdf, 0xd3, " #imm8 " \n\t"
+#define AESKEY_EXPAND192 \
+	"pshufd $0x55, %%xmm2, %%xmm2\n\t" \
+	"movdqu %%xmm1, %%xmm4\n\t" \
+	"pslldq $4, %%xmm4\n\t" \
+	"pxor %%xmm4, %%xmm1\n\t" \
+	"pslldq $4, %%xmm4\n\t" \
+	"pxor %%xmm4, %%xmm1\n\t" \
+	"pslldq $4, %%xmm4\n\t" \
+	"pxor %%xmm4, %%xmm1\n\t" \
+	"pxor %%xmm2, %%xmm1\n\t" \
+	"pshufd $0xff, %%xmm1, %%xmm2\n\t" \
+	"movdqu %%xmm3, %%xmm4\n\t" \
+	"pslldq $4, %%xmm4\n\t" \
+	"pxor %%xmm4, %%xmm3\n\t" \
+	"pxor %%xmm2, %%xmm3\n\t"
+
+      asm volatile ("movdqu (%[key]), %%xmm1\n\t"     /* xmm1 := key[0..15]   */
+                    "movq 16(%[key]), %%xmm3\n\t"     /* xmm3 := key[16..23]  */
+                    "movdqa %%xmm1, (%[ksch])\n\t"    /* ksch[0] := xmm1  */
+                    "movdqa %%xmm3, %%xmm5\n\t"
+
+                    AESKEYGENASSIST_xmm3_xmm2(0x01)
+                    AESKEY_EXPAND192
+                    "shufpd $0, %%xmm1, %%xmm5\n\t"
+                    "movdqa %%xmm5, 0x10(%[ksch])\n\t" /* ksch[1] := xmm5  */
+                    "movdqa %%xmm1, %%xmm6\n\t"
+                    "shufpd $1, %%xmm3, %%xmm6\n\t"
+                    "movdqa %%xmm6, 0x20(%[ksch])\n\t" /* ksch[2] := xmm6  */
+                    AESKEYGENASSIST_xmm3_xmm2(0x02)
+                    AESKEY_EXPAND192
+                    "movdqa %%xmm1, 0x30(%[ksch])\n\t" /* ksch[3] := xmm1  */
+                    "movdqa %%xmm3, %%xmm5\n\t"
+
+                    AESKEYGENASSIST_xmm3_xmm2(0x04)
+                    AESKEY_EXPAND192
+                    "shufpd $0, %%xmm1, %%xmm5\n\t"
+                    "movdqa %%xmm5, 0x40(%[ksch])\n\t" /* ksch[4] := xmm5  */
+                    "movdqa %%xmm1, %%xmm6\n\t"
+                    "shufpd $1, %%xmm3, %%xmm6\n\t"
+                    "movdqa %%xmm6, 0x50(%[ksch])\n\t" /* ksch[5] := xmm6  */
+                    AESKEYGENASSIST_xmm3_xmm2(0x08)
+                    AESKEY_EXPAND192
+                    "movdqa %%xmm1, 0x60(%[ksch])\n\t" /* ksch[6] := xmm1  */
+                    "movdqa %%xmm3, %%xmm5\n\t"
+
+                    AESKEYGENASSIST_xmm3_xmm2(0x10)
+                    AESKEY_EXPAND192
+                    "shufpd $0, %%xmm1, %%xmm5\n\t"
+                    "movdqa %%xmm5, 0x70(%[ksch])\n\t" /* ksch[7] := xmm5  */
+                    "movdqa %%xmm1, %%xmm6\n\t"
+                    "shufpd $1, %%xmm3, %%xmm6\n\t"
+                    "movdqa %%xmm6, 0x80(%[ksch])\n\t" /* ksch[8] := xmm6  */
+                    AESKEYGENASSIST_xmm3_xmm2(0x20)
+                    AESKEY_EXPAND192
+                    "movdqa %%xmm1, 0x90(%[ksch])\n\t" /* ksch[9] := xmm1  */
+                    "movdqa %%xmm3, %%xmm5\n\t"
+
+                    AESKEYGENASSIST_xmm3_xmm2(0x40)
+                    AESKEY_EXPAND192
+                    "shufpd $0, %%xmm1, %%xmm5\n\t"
+                    "movdqa %%xmm5, 0xa0(%[ksch])\n\t" /* ksch[10] := xmm5  */
+                    "movdqa %%xmm1, %%xmm6\n\t"
+                    "shufpd $1, %%xmm3, %%xmm6\n\t"
+                    "movdqa %%xmm6, 0xb0(%[ksch])\n\t" /* ksch[11] := xmm6  */
+                    AESKEYGENASSIST_xmm3_xmm2(0x80)
+                    AESKEY_EXPAND192
+                    "movdqa %%xmm1, 0xc0(%[ksch])\n\t" /* ksch[12] := xmm1  */
+                    :
+                    : [key] "r" (key), [ksch] "r" (ctx->keyschenc)
+                    : "cc", "memory" );
+#undef AESKEYGENASSIST_xmm3_xmm2
+#undef AESKEY_EXPAND192
+    }
+  else if (ctx->rounds > 12)
+    {
+      /* 256-bit key */
+#define AESKEYGENASSIST_xmm1_xmm2(imm8) \
+	".byte 0x66, 0x0f, 0x3a, 0xdf, 0xd1, " #imm8 " \n\t"
+#define AESKEYGENASSIST_xmm3_xmm2(imm8) \
+	".byte 0x66, 0x0f, 0x3a, 0xdf, 0xd3, " #imm8 " \n\t"
+#define AESKEY_EXPAND256_A \
+	"pshufd $0xff, %%xmm2, %%xmm2\n\t" \
+	"movdqa %%xmm1, %%xmm4\n\t" \
+	"pslldq $4, %%xmm4\n\t" \
+	"pxor %%xmm4, %%xmm1\n\t" \
+	"pslldq $4, %%xmm4\n\t" \
+	"pxor %%xmm4, %%xmm1\n\t" \
+	"pslldq $4, %%xmm4\n\t" \
+	"pxor %%xmm4, %%xmm1\n\t" \
+	"pxor %%xmm2, %%xmm1\n\t"
+#define AESKEY_EXPAND256_B \
+	"pshufd $0xaa, %%xmm2, %%xmm2\n\t" \
+	"movdqa %%xmm3, %%xmm4\n\t" \
+	"pslldq $4, %%xmm4\n\t" \
+	"pxor %%xmm4, %%xmm3\n\t" \
+	"pslldq $4, %%xmm4\n\t" \
+	"pxor %%xmm4, %%xmm3\n\t" \
+	"pslldq $4, %%xmm4\n\t" \
+	"pxor %%xmm4, %%xmm3\n\t" \
+	"pxor %%xmm2, %%xmm3\n\t"
+
+      asm volatile ("movdqu (%[key]), %%xmm1\n\t"     /* xmm1 := key[0..15]   */
+                    "movdqu 16(%[key]), %%xmm3\n\t"   /* xmm3 := key[16..31]  */
+                    "movdqa %%xmm1, (%[ksch])\n\t"     /* ksch[0] := xmm1  */
+                    "movdqa %%xmm3, 0x10(%[ksch])\n\t" /* ksch[1] := xmm3  */
+
+                    AESKEYGENASSIST_xmm3_xmm2(0x01)
+                    AESKEY_EXPAND256_A
+                    "movdqa %%xmm1, 0x20(%[ksch])\n\t" /* ksch[2] := xmm1  */
+                    AESKEYGENASSIST_xmm1_xmm2(0x00)
+                    AESKEY_EXPAND256_B
+                    "movdqa %%xmm3, 0x30(%[ksch])\n\t" /* ksch[3] := xmm3  */
+
+                    AESKEYGENASSIST_xmm3_xmm2(0x02)
+                    AESKEY_EXPAND256_A
+                    "movdqa %%xmm1, 0x40(%[ksch])\n\t" /* ksch[4] := xmm1  */
+                    AESKEYGENASSIST_xmm1_xmm2(0x00)
+                    AESKEY_EXPAND256_B
+                    "movdqa %%xmm3, 0x50(%[ksch])\n\t" /* ksch[5] := xmm3  */
+
+                    AESKEYGENASSIST_xmm3_xmm2(0x04)
+                    AESKEY_EXPAND256_A
+                    "movdqa %%xmm1, 0x60(%[ksch])\n\t" /* ksch[6] := xmm1  */
+                    AESKEYGENASSIST_xmm1_xmm2(0x00)
+                    AESKEY_EXPAND256_B
+                    "movdqa %%xmm3, 0x70(%[ksch])\n\t" /* ksch[7] := xmm3  */
+
+                    AESKEYGENASSIST_xmm3_xmm2(0x08)
+                    AESKEY_EXPAND256_A
+                    "movdqa %%xmm1, 0x80(%[ksch])\n\t" /* ksch[8] := xmm1  */
+                    AESKEYGENASSIST_xmm1_xmm2(0x00)
+                    AESKEY_EXPAND256_B
+                    "movdqa %%xmm3, 0x90(%[ksch])\n\t" /* ksch[9] := xmm3  */
+
+                    AESKEYGENASSIST_xmm3_xmm2(0x10)
+                    AESKEY_EXPAND256_A
+                    "movdqa %%xmm1, 0xa0(%[ksch])\n\t" /* ksch[10] := xmm1  */
+                    AESKEYGENASSIST_xmm1_xmm2(0x00)
+                    AESKEY_EXPAND256_B
+                    "movdqa %%xmm3, 0xb0(%[ksch])\n\t" /* ksch[11] := xmm3  */
+
+                    AESKEYGENASSIST_xmm3_xmm2(0x20)
+                    AESKEY_EXPAND256_A
+                    "movdqa %%xmm1, 0xc0(%[ksch])\n\t" /* ksch[12] := xmm1  */
+                    AESKEYGENASSIST_xmm1_xmm2(0x00)
+                    AESKEY_EXPAND256_B
+                    "movdqa %%xmm3, 0xd0(%[ksch])\n\t" /* ksch[13] := xmm3  */
+
+                    AESKEYGENASSIST_xmm3_xmm2(0x40)
+                    AESKEY_EXPAND256_A
+                    "movdqa %%xmm1, 0xe0(%[ksch])\n\t" /* ksch[14] := xmm1  */
+
+                    :
+                    : [key] "r" (key), [ksch] "r" (ctx->keyschenc)
+                    : "cc", "memory" );
+#undef AESKEYGENASSIST_xmm1_xmm2
+#undef AESKEYGENASSIST_xmm3_xmm2
+#undef AESKEY_EXPAND256_A
+#undef AESKEY_EXPAND256_B
+    }
+
+  aesni_cleanup();
+  aesni_cleanup_2_6();
+}
+
+
+/* Make a decryption key from an encryption key. */
+void
+_gcry_aes_aesni_prepare_decryption (RIJNDAEL_context *ctx)
+{
+  /* The AES-NI decrypt instructions use the Equivalent Inverse
+     Cipher, thus we can't use the the standard decrypt key
+     preparation.  */
+  u128_t *ekey = (u128_t *)ctx->keyschenc;
+  u128_t *dkey = (u128_t *)ctx->keyschdec;
+  int rr;
+  int r;
+
+  aesni_prepare();
+
+#define DO_AESNI_AESIMC() \
+  asm volatile ("movdqa %[ekey], %%xmm1\n\t" \
+                /*"aesimc %%xmm1, %%xmm1\n\t"*/ \
+                ".byte 0x66, 0x0f, 0x38, 0xdb, 0xc9\n\t" \
+                "movdqa %%xmm1, %[dkey]" \
+                : [dkey] "=m" (dkey[r]) \
+                : [ekey] "m" (ekey[rr]) \
+                : "memory")
+
+  dkey[0] = ekey[ctx->rounds];
+  r=1;
+  rr=ctx->rounds-1;
+  DO_AESNI_AESIMC(); r++; rr--; /* round 1 */
+  DO_AESNI_AESIMC(); r++; rr--; /* round 2 */
+  DO_AESNI_AESIMC(); r++; rr--; /* round 3 */
+  DO_AESNI_AESIMC(); r++; rr--; /* round 4 */
+  DO_AESNI_AESIMC(); r++; rr--; /* round 5 */
+  DO_AESNI_AESIMC(); r++; rr--; /* round 6 */
+  DO_AESNI_AESIMC(); r++; rr--; /* round 7 */
+  DO_AESNI_AESIMC(); r++; rr--; /* round 8 */
+  DO_AESNI_AESIMC(); r++; rr--; /* round 9 */
+  if (ctx->rounds > 10)
+    {
+      DO_AESNI_AESIMC(); r++; rr--; /* round 10 */
+      DO_AESNI_AESIMC(); r++; rr--; /* round 11 */
+      if (ctx->rounds > 12)
+        {
+          DO_AESNI_AESIMC(); r++; rr--; /* round 12 */
+          DO_AESNI_AESIMC(); r++; rr--; /* round 13 */
+        }
+    }
+
+  dkey[r] = ekey[0];
+
+#undef DO_AESNI_AESIMC
+
+  aesni_cleanup();
+}
+
+
+/* Encrypt one block using the Intel AES-NI instructions.  A and B may
+   be the same.
+
+   Our problem here is that gcc does not allow the "x" constraint for
+   SSE registers in asm unless you compile with -msse.  The common
+   wisdom is to use a separate file for SSE instructions and build it
+   separately.  This would require a lot of extra build system stuff,
+   similar to what we do in mpi/ for the asm stuff.  What we do
+   instead is to use standard registers and a bit more of plain asm
+   which copies the data and key stuff to the SSE registers and later
+   back.  If we decide to implement some block modes with parallelized
+   AES instructions, it might indeed be better to use plain asm ala
+   mpi/.  */
+static inline void
+do_aesni_enc (const RIJNDAEL_context *ctx, unsigned char *b,
+              const unsigned char *a)
+{
+#define aesenc_xmm1_xmm0      ".byte 0x66, 0x0f, 0x38, 0xdc, 0xc1\n\t"
+#define aesenclast_xmm1_xmm0  ".byte 0x66, 0x0f, 0x38, 0xdd, 0xc1\n\t"
+  /* Note: For now we relax the alignment requirement for A and B: It
+     does not make much difference because in many case we would need
+     to memcpy them to an extra buffer; using the movdqu is much faster
+     that memcpy and movdqa.  For CFB we know that the IV is properly
+     aligned but that is a special case.  We should better implement
+     CFB direct in asm.  */
+  asm volatile ("movdqu %[src], %%xmm0\n\t"     /* xmm0 := *a     */
+                "movdqa (%[key]), %%xmm1\n\t"    /* xmm1 := key[0] */
+                "pxor   %%xmm1, %%xmm0\n\t"     /* xmm0 ^= key[0] */
+                "movdqa 0x10(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x20(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x30(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x40(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x50(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x60(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x70(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x80(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x90(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0xa0(%[key]), %%xmm1\n\t"
+                "cmpl $10, %[rounds]\n\t"
+                "jz .Lenclast%=\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0xb0(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0xc0(%[key]), %%xmm1\n\t"
+                "cmpl $12, %[rounds]\n\t"
+                "jz .Lenclast%=\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0xd0(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0xe0(%[key]), %%xmm1\n"
+
+                ".Lenclast%=:\n\t"
+                aesenclast_xmm1_xmm0
+                "movdqu %%xmm0, %[dst]\n"
+                : [dst] "=m" (*b)
+                : [src] "m" (*a),
+                  [key] "r" (ctx->keyschenc),
+                  [rounds] "r" (ctx->rounds)
+                : "cc", "memory");
+#undef aesenc_xmm1_xmm0
+#undef aesenclast_xmm1_xmm0
+}
+
+
+static inline void
+do_aesni_dec (const RIJNDAEL_context *ctx, unsigned char *b,
+              const unsigned char *a)
+{
+#define aesdec_xmm1_xmm0      ".byte 0x66, 0x0f, 0x38, 0xde, 0xc1\n\t"
+#define aesdeclast_xmm1_xmm0  ".byte 0x66, 0x0f, 0x38, 0xdf, 0xc1\n\t"
+  asm volatile ("movdqu %[src], %%xmm0\n\t"     /* xmm0 := *a     */
+                "movdqa (%[key]), %%xmm1\n\t"
+                "pxor   %%xmm1, %%xmm0\n\t"     /* xmm0 ^= key[0] */
+                "movdqa 0x10(%[key]), %%xmm1\n\t"
+                aesdec_xmm1_xmm0
+                "movdqa 0x20(%[key]), %%xmm1\n\t"
+                aesdec_xmm1_xmm0
+                "movdqa 0x30(%[key]), %%xmm1\n\t"
+                aesdec_xmm1_xmm0
+                "movdqa 0x40(%[key]), %%xmm1\n\t"
+                aesdec_xmm1_xmm0
+                "movdqa 0x50(%[key]), %%xmm1\n\t"
+                aesdec_xmm1_xmm0
+                "movdqa 0x60(%[key]), %%xmm1\n\t"
+                aesdec_xmm1_xmm0
+                "movdqa 0x70(%[key]), %%xmm1\n\t"
+                aesdec_xmm1_xmm0
+                "movdqa 0x80(%[key]), %%xmm1\n\t"
+                aesdec_xmm1_xmm0
+                "movdqa 0x90(%[key]), %%xmm1\n\t"
+                aesdec_xmm1_xmm0
+                "movdqa 0xa0(%[key]), %%xmm1\n\t"
+                "cmpl $10, %[rounds]\n\t"
+                "jz .Ldeclast%=\n\t"
+                aesdec_xmm1_xmm0
+                "movdqa 0xb0(%[key]), %%xmm1\n\t"
+                aesdec_xmm1_xmm0
+                "movdqa 0xc0(%[key]), %%xmm1\n\t"
+                "cmpl $12, %[rounds]\n\t"
+                "jz .Ldeclast%=\n\t"
+                aesdec_xmm1_xmm0
+                "movdqa 0xd0(%[key]), %%xmm1\n\t"
+                aesdec_xmm1_xmm0
+                "movdqa 0xe0(%[key]), %%xmm1\n"
+
+                ".Ldeclast%=:\n\t"
+                aesdeclast_xmm1_xmm0
+                "movdqu %%xmm0, %[dst]\n"
+                : [dst] "=m" (*b)
+                : [src] "m" (*a),
+                  [key] "r" (ctx->keyschdec),
+                  [rounds] "r" (ctx->rounds)
+                : "cc", "memory");
+#undef aesdec_xmm1_xmm0
+#undef aesdeclast_xmm1_xmm0
+}
+
+
+/* Encrypt four blocks using the Intel AES-NI instructions.  Blocks are input
+ * and output through SSE registers xmm1 to xmm4.  */
+static inline void
+do_aesni_enc_vec4 (const RIJNDAEL_context *ctx)
+{
+#define aesenc_xmm0_xmm1      ".byte 0x66, 0x0f, 0x38, 0xdc, 0xc8\n\t"
+#define aesenc_xmm0_xmm2      ".byte 0x66, 0x0f, 0x38, 0xdc, 0xd0\n\t"
+#define aesenc_xmm0_xmm3      ".byte 0x66, 0x0f, 0x38, 0xdc, 0xd8\n\t"
+#define aesenc_xmm0_xmm4      ".byte 0x66, 0x0f, 0x38, 0xdc, 0xe0\n\t"
+#define aesenclast_xmm0_xmm1  ".byte 0x66, 0x0f, 0x38, 0xdd, 0xc8\n\t"
+#define aesenclast_xmm0_xmm2  ".byte 0x66, 0x0f, 0x38, 0xdd, 0xd0\n\t"
+#define aesenclast_xmm0_xmm3  ".byte 0x66, 0x0f, 0x38, 0xdd, 0xd8\n\t"
+#define aesenclast_xmm0_xmm4  ".byte 0x66, 0x0f, 0x38, 0xdd, 0xe0\n\t"
+  asm volatile ("movdqa (%[key]), %%xmm0\n\t"
+                "pxor   %%xmm0, %%xmm1\n\t"     /* xmm1 ^= key[0] */
+                "pxor   %%xmm0, %%xmm2\n\t"     /* xmm2 ^= key[0] */
+                "pxor   %%xmm0, %%xmm3\n\t"     /* xmm3 ^= key[0] */
+                "pxor   %%xmm0, %%xmm4\n\t"     /* xmm4 ^= key[0] */
+                "movdqa 0x10(%[key]), %%xmm0\n\t"
+                aesenc_xmm0_xmm1
+                aesenc_xmm0_xmm2
+                aesenc_xmm0_xmm3
+                aesenc_xmm0_xmm4
+                "movdqa 0x20(%[key]), %%xmm0\n\t"
+                aesenc_xmm0_xmm1
+                aesenc_xmm0_xmm2
+                aesenc_xmm0_xmm3
+                aesenc_xmm0_xmm4
+                "movdqa 0x30(%[key]), %%xmm0\n\t"
+                aesenc_xmm0_xmm1
+                aesenc_xmm0_xmm2
+                aesenc_xmm0_xmm3
+                aesenc_xmm0_xmm4
+                "movdqa 0x40(%[key]), %%xmm0\n\t"
+                aesenc_xmm0_xmm1
+                aesenc_xmm0_xmm2
+                aesenc_xmm0_xmm3
+                aesenc_xmm0_xmm4
+                "movdqa 0x50(%[key]), %%xmm0\n\t"
+                aesenc_xmm0_xmm1
+                aesenc_xmm0_xmm2
+                aesenc_xmm0_xmm3
+                aesenc_xmm0_xmm4
+                "movdqa 0x60(%[key]), %%xmm0\n\t"
+                aesenc_xmm0_xmm1
+                aesenc_xmm0_xmm2
+                aesenc_xmm0_xmm3
+                aesenc_xmm0_xmm4
+                "movdqa 0x70(%[key]), %%xmm0\n\t"
+                aesenc_xmm0_xmm1
+                aesenc_xmm0_xmm2
+                aesenc_xmm0_xmm3
+                aesenc_xmm0_xmm4
+                "movdqa 0x80(%[key]), %%xmm0\n\t"
+                aesenc_xmm0_xmm1
+                aesenc_xmm0_xmm2
+                aesenc_xmm0_xmm3
+                aesenc_xmm0_xmm4
+                "movdqa 0x90(%[key]), %%xmm0\n\t"
+                aesenc_xmm0_xmm1
+                aesenc_xmm0_xmm2
+                aesenc_xmm0_xmm3
+                aesenc_xmm0_xmm4
+                "movdqa 0xa0(%[key]), %%xmm0\n\t"
+                "cmpl $10, %[rounds]\n\t"
+                "jz .Ldeclast%=\n\t"
+                aesenc_xmm0_xmm1
+                aesenc_xmm0_xmm2
+                aesenc_xmm0_xmm3
+                aesenc_xmm0_xmm4
+                "movdqa 0xb0(%[key]), %%xmm0\n\t"
+                aesenc_xmm0_xmm1
+                aesenc_xmm0_xmm2
+                aesenc_xmm0_xmm3
+                aesenc_xmm0_xmm4
+                "movdqa 0xc0(%[key]), %%xmm0\n\t"
+                "cmpl $12, %[rounds]\n\t"
+                "jz .Ldeclast%=\n\t"
+                aesenc_xmm0_xmm1
+                aesenc_xmm0_xmm2
+                aesenc_xmm0_xmm3
+                aesenc_xmm0_xmm4
+                "movdqa 0xd0(%[key]), %%xmm0\n\t"
+                aesenc_xmm0_xmm1
+                aesenc_xmm0_xmm2
+                aesenc_xmm0_xmm3
+                aesenc_xmm0_xmm4
+                "movdqa 0xe0(%[key]), %%xmm0\n"
+
+                ".Ldeclast%=:\n\t"
+                aesenclast_xmm0_xmm1
+                aesenclast_xmm0_xmm2
+                aesenclast_xmm0_xmm3
+                aesenclast_xmm0_xmm4
+                : /* no output */
+                : [key] "r" (ctx->keyschenc),
+                  [rounds] "r" (ctx->rounds)
+                : "cc", "memory");
+#undef aesenc_xmm0_xmm1
+#undef aesenc_xmm0_xmm2
+#undef aesenc_xmm0_xmm3
+#undef aesenc_xmm0_xmm4
+#undef aesenclast_xmm0_xmm1
+#undef aesenclast_xmm0_xmm2
+#undef aesenclast_xmm0_xmm3
+#undef aesenclast_xmm0_xmm4
+}
+
+
+/* Decrypt four blocks using the Intel AES-NI instructions.  Blocks are input
+ * and output through SSE registers xmm1 to xmm4.  */
+static inline void
+do_aesni_dec_vec4 (const RIJNDAEL_context *ctx)
+{
+#define aesdec_xmm0_xmm1 ".byte 0x66, 0x0f, 0x38, 0xde, 0xc8\n\t"
+#define aesdec_xmm0_xmm2 ".byte 0x66, 0x0f, 0x38, 0xde, 0xd0\n\t"
+#define aesdec_xmm0_xmm3 ".byte 0x66, 0x0f, 0x38, 0xde, 0xd8\n\t"
+#define aesdec_xmm0_xmm4 ".byte 0x66, 0x0f, 0x38, 0xde, 0xe0\n\t"
+#define aesdeclast_xmm0_xmm1 ".byte 0x66, 0x0f, 0x38, 0xdf, 0xc8\n\t"
+#define aesdeclast_xmm0_xmm2 ".byte 0x66, 0x0f, 0x38, 0xdf, 0xd0\n\t"
+#define aesdeclast_xmm0_xmm3 ".byte 0x66, 0x0f, 0x38, 0xdf, 0xd8\n\t"
+#define aesdeclast_xmm0_xmm4 ".byte 0x66, 0x0f, 0x38, 0xdf, 0xe0\n\t"
+  asm volatile ("movdqa (%[key]), %%xmm0\n\t"
+                "pxor   %%xmm0, %%xmm1\n\t"     /* xmm1 ^= key[0] */
+                "pxor   %%xmm0, %%xmm2\n\t"     /* xmm2 ^= key[0] */
+                "pxor   %%xmm0, %%xmm3\n\t"     /* xmm3 ^= key[0] */
+                "pxor   %%xmm0, %%xmm4\n\t"     /* xmm4 ^= key[0] */
+                "movdqa 0x10(%[key]), %%xmm0\n\t"
+                aesdec_xmm0_xmm1
+                aesdec_xmm0_xmm2
+                aesdec_xmm0_xmm3
+                aesdec_xmm0_xmm4
+                "movdqa 0x20(%[key]), %%xmm0\n\t"
+                aesdec_xmm0_xmm1
+                aesdec_xmm0_xmm2
+                aesdec_xmm0_xmm3
+                aesdec_xmm0_xmm4
+                "movdqa 0x30(%[key]), %%xmm0\n\t"
+                aesdec_xmm0_xmm1
+                aesdec_xmm0_xmm2
+                aesdec_xmm0_xmm3
+                aesdec_xmm0_xmm4
+                "movdqa 0x40(%[key]), %%xmm0\n\t"
+                aesdec_xmm0_xmm1
+                aesdec_xmm0_xmm2
+                aesdec_xmm0_xmm3
+                aesdec_xmm0_xmm4
+                "movdqa 0x50(%[key]), %%xmm0\n\t"
+                aesdec_xmm0_xmm1
+                aesdec_xmm0_xmm2
+                aesdec_xmm0_xmm3
+                aesdec_xmm0_xmm4
+                "movdqa 0x60(%[key]), %%xmm0\n\t"
+                aesdec_xmm0_xmm1
+                aesdec_xmm0_xmm2
+                aesdec_xmm0_xmm3
+                aesdec_xmm0_xmm4
+                "movdqa 0x70(%[key]), %%xmm0\n\t"
+                aesdec_xmm0_xmm1
+                aesdec_xmm0_xmm2
+                aesdec_xmm0_xmm3
+                aesdec_xmm0_xmm4
+                "movdqa 0x80(%[key]), %%xmm0\n\t"
+                aesdec_xmm0_xmm1
+                aesdec_xmm0_xmm2
+                aesdec_xmm0_xmm3
+                aesdec_xmm0_xmm4
+                "movdqa 0x90(%[key]), %%xmm0\n\t"
+                aesdec_xmm0_xmm1
+                aesdec_xmm0_xmm2
+                aesdec_xmm0_xmm3
+                aesdec_xmm0_xmm4
+                "movdqa 0xa0(%[key]), %%xmm0\n\t"
+                "cmpl $10, %[rounds]\n\t"
+                "jz .Ldeclast%=\n\t"
+                aesdec_xmm0_xmm1
+                aesdec_xmm0_xmm2
+                aesdec_xmm0_xmm3
+                aesdec_xmm0_xmm4
+                "movdqa 0xb0(%[key]), %%xmm0\n\t"
+                aesdec_xmm0_xmm1
+                aesdec_xmm0_xmm2
+                aesdec_xmm0_xmm3
+                aesdec_xmm0_xmm4
+                "movdqa 0xc0(%[key]), %%xmm0\n\t"
+                "cmpl $12, %[rounds]\n\t"
+                "jz .Ldeclast%=\n\t"
+                aesdec_xmm0_xmm1
+                aesdec_xmm0_xmm2
+                aesdec_xmm0_xmm3
+                aesdec_xmm0_xmm4
+                "movdqa 0xd0(%[key]), %%xmm0\n\t"
+                aesdec_xmm0_xmm1
+                aesdec_xmm0_xmm2
+                aesdec_xmm0_xmm3
+                aesdec_xmm0_xmm4
+                "movdqa 0xe0(%[key]), %%xmm0\n"
+
+                ".Ldeclast%=:\n\t"
+                aesdeclast_xmm0_xmm1
+                aesdeclast_xmm0_xmm2
+                aesdeclast_xmm0_xmm3
+                aesdeclast_xmm0_xmm4
+                : /* no output */
+                : [key] "r" (ctx->keyschdec),
+                  [rounds] "r" (ctx->rounds)
+                : "cc", "memory");
+#undef aesdec_xmm0_xmm1
+#undef aesdec_xmm0_xmm2
+#undef aesdec_xmm0_xmm3
+#undef aesdec_xmm0_xmm4
+#undef aesdeclast_xmm0_xmm1
+#undef aesdeclast_xmm0_xmm2
+#undef aesdeclast_xmm0_xmm3
+#undef aesdeclast_xmm0_xmm4
+}
+
+
+/* Perform a CFB encryption or decryption round using the
+   initialization vector IV and the input block A.  Write the result
+   to the output block B and update IV.  IV needs to be 16 byte
+   aligned.  */
+static inline void
+do_aesni_cfb (const RIJNDAEL_context *ctx, int decrypt_flag,
+              unsigned char *iv, unsigned char *b, const unsigned char *a)
+{
+#define aesenc_xmm1_xmm0      ".byte 0x66, 0x0f, 0x38, 0xdc, 0xc1\n\t"
+#define aesenclast_xmm1_xmm0  ".byte 0x66, 0x0f, 0x38, 0xdd, 0xc1\n\t"
+  asm volatile ("movdqa %[iv], %%xmm0\n\t"      /* xmm0 := IV     */
+                "movdqa (%[key]), %%xmm1\n\t"    /* xmm1 := key[0] */
+                "pxor   %%xmm1, %%xmm0\n\t"     /* xmm0 ^= key[0] */
+                "movdqa 0x10(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x20(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x30(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x40(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x50(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x60(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x70(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x80(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x90(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0xa0(%[key]), %%xmm1\n\t"
+                "cmpl $10, %[rounds]\n\t"
+                "jz .Lenclast%=\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0xb0(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0xc0(%[key]), %%xmm1\n\t"
+                "cmpl $12, %[rounds]\n\t"
+                "jz .Lenclast%=\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0xd0(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0xe0(%[key]), %%xmm1\n"
+
+                ".Lenclast%=:\n\t"
+                aesenclast_xmm1_xmm0
+                "movdqu %[src], %%xmm1\n\t"      /* Save input.  */
+                "pxor %%xmm1, %%xmm0\n\t"        /* xmm0 = input ^ IV  */
+
+                "cmpl $1, %[decrypt]\n\t"
+                "jz .Ldecrypt_%=\n\t"
+                "movdqa %%xmm0, %[iv]\n\t"       /* [encrypt] Store IV.  */
+                "jmp .Lleave_%=\n"
+                ".Ldecrypt_%=:\n\t"
+                "movdqa %%xmm1, %[iv]\n"         /* [decrypt] Store IV.  */
+                ".Lleave_%=:\n\t"
+                "movdqu %%xmm0, %[dst]\n"        /* Store output.   */
+                : [iv] "+m" (*iv), [dst] "=m" (*b)
+                : [src] "m" (*a),
+                  [key] "r" (ctx->keyschenc),
+                  [rounds] "g" (ctx->rounds),
+                  [decrypt] "m" (decrypt_flag)
+                : "cc", "memory");
+#undef aesenc_xmm1_xmm0
+#undef aesenclast_xmm1_xmm0
+}
+
+/* Perform a CTR encryption round using the counter CTR and the input
+   block A.  Write the result to the output block B and update CTR.
+   CTR needs to be a 16 byte aligned little-endian value.  */
+static void
+do_aesni_ctr (const RIJNDAEL_context *ctx,
+              unsigned char *ctr, unsigned char *b, const unsigned char *a)
+{
+#define aesenc_xmm1_xmm0      ".byte 0x66, 0x0f, 0x38, 0xdc, 0xc1\n\t"
+#define aesenclast_xmm1_xmm0  ".byte 0x66, 0x0f, 0x38, 0xdd, 0xc1\n\t"
+
+  asm volatile ("movdqa %%xmm5, %%xmm0\n\t"     /* xmm0 := CTR (xmm5)  */
+                "pcmpeqd %%xmm1, %%xmm1\n\t"
+                "psrldq $8, %%xmm1\n\t"         /* xmm1 = -1 */
+
+                "pshufb %%xmm6, %%xmm5\n\t"
+                "psubq  %%xmm1, %%xmm5\n\t"     /* xmm5++ (big endian) */
+
+                /* detect if 64-bit carry handling is needed */
+                "cmpl   $0xffffffff, 8(%[ctr])\n\t"
+                "jne    .Lno_carry%=\n\t"
+                "cmpl   $0xffffffff, 12(%[ctr])\n\t"
+                "jne    .Lno_carry%=\n\t"
+
+                "pslldq $8, %%xmm1\n\t"         /* move lower 64-bit to high */
+                "psubq   %%xmm1, %%xmm5\n\t"    /* add carry to upper 64bits */
+
+                ".Lno_carry%=:\n\t"
+
+                "pshufb %%xmm6, %%xmm5\n\t"
+                "movdqa %%xmm5, (%[ctr])\n\t"   /* Update CTR (mem).       */
+
+                "pxor (%[key]), %%xmm0\n\t"     /* xmm1 ^= key[0]    */
+                "movdqa 0x10(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x20(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x30(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x40(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x50(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x60(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x70(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x80(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0x90(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0xa0(%[key]), %%xmm1\n\t"
+                "cmpl $10, %[rounds]\n\t"
+                "jz .Lenclast%=\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0xb0(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0xc0(%[key]), %%xmm1\n\t"
+                "cmpl $12, %[rounds]\n\t"
+                "jz .Lenclast%=\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0xd0(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                "movdqa 0xe0(%[key]), %%xmm1\n"
+
+                ".Lenclast%=:\n\t"
+                aesenclast_xmm1_xmm0
+                "movdqu %[src], %%xmm1\n\t"      /* xmm1 := input   */
+                "pxor %%xmm1, %%xmm0\n\t"        /* EncCTR ^= input  */
+                "movdqu %%xmm0, %[dst]"          /* Store EncCTR.    */
+
+                : [dst] "=m" (*b)
+                : [src] "m" (*a),
+                  [ctr] "r" (ctr),
+                  [key] "r" (ctx->keyschenc),
+                  [rounds] "g" (ctx->rounds)
+                : "cc", "memory");
+#undef aesenc_xmm1_xmm0
+#undef aesenclast_xmm1_xmm0
+}
+
+
+/* Four blocks at a time variant of do_aesni_ctr.  */
+static void
+do_aesni_ctr_4 (const RIJNDAEL_context *ctx,
+                unsigned char *ctr, unsigned char *b, const unsigned char *a)
+{
+#define aesenc_xmm1_xmm0      ".byte 0x66, 0x0f, 0x38, 0xdc, 0xc1\n\t"
+#define aesenc_xmm1_xmm2      ".byte 0x66, 0x0f, 0x38, 0xdc, 0xd1\n\t"
+#define aesenc_xmm1_xmm3      ".byte 0x66, 0x0f, 0x38, 0xdc, 0xd9\n\t"
+#define aesenc_xmm1_xmm4      ".byte 0x66, 0x0f, 0x38, 0xdc, 0xe1\n\t"
+#define aesenclast_xmm1_xmm0  ".byte 0x66, 0x0f, 0x38, 0xdd, 0xc1\n\t"
+#define aesenclast_xmm1_xmm2  ".byte 0x66, 0x0f, 0x38, 0xdd, 0xd1\n\t"
+#define aesenclast_xmm1_xmm3  ".byte 0x66, 0x0f, 0x38, 0xdd, 0xd9\n\t"
+#define aesenclast_xmm1_xmm4  ".byte 0x66, 0x0f, 0x38, 0xdd, 0xe1\n\t"
+
+  /* Register usage:
+      esi   keyschedule
+      xmm0  CTR-0
+      xmm1  temp / round key
+      xmm2  CTR-1
+      xmm3  CTR-2
+      xmm4  CTR-3
+      xmm5  copy of *ctr
+      xmm6  endian swapping mask
+   */
+
+  asm volatile ("movdqa %%xmm5, %%xmm0\n\t"     /* xmm0, xmm2 := CTR (xmm5) */
+                "movdqa %%xmm0, %%xmm2\n\t"
+                "pcmpeqd %%xmm1, %%xmm1\n\t"
+                "psrldq $8, %%xmm1\n\t"         /* xmm1 = -1 */
+
+                "pshufb %%xmm6, %%xmm2\n\t"     /* xmm2 := le(xmm2) */
+                "psubq  %%xmm1, %%xmm2\n\t"     /* xmm2++           */
+                "movdqa %%xmm2, %%xmm3\n\t"     /* xmm3 := xmm2     */
+                "psubq  %%xmm1, %%xmm3\n\t"     /* xmm3++           */
+                "movdqa %%xmm3, %%xmm4\n\t"     /* xmm4 := xmm3     */
+                "psubq  %%xmm1, %%xmm4\n\t"     /* xmm4++           */
+                "movdqa %%xmm4, %%xmm5\n\t"     /* xmm5 := xmm4     */
+                "psubq  %%xmm1, %%xmm5\n\t"     /* xmm5++           */
+
+                /* detect if 64-bit carry handling is needed */
+                "cmpl   $0xffffffff, 8(%[ctr])\n\t"
+                "jne    .Lno_carry%=\n\t"
+                "movl   12(%[ctr]), %%esi\n\t"
+                "bswapl %%esi\n\t"
+                "cmpl   $0xfffffffc, %%esi\n\t"
+                "jb     .Lno_carry%=\n\t"       /* no carry */
+
+                "pslldq $8, %%xmm1\n\t"         /* move lower 64-bit to high */
+                "je     .Lcarry_xmm5%=\n\t"     /* esi == 0xfffffffc */
+                "cmpl   $0xfffffffe, %%esi\n\t"
+                "jb     .Lcarry_xmm4%=\n\t"     /* esi == 0xfffffffd */
+                "je     .Lcarry_xmm3%=\n\t"     /* esi == 0xfffffffe */
+                /* esi == 0xffffffff */
+
+                "psubq   %%xmm1, %%xmm2\n\t"
+                ".Lcarry_xmm3%=:\n\t"
+                "psubq   %%xmm1, %%xmm3\n\t"
+                ".Lcarry_xmm4%=:\n\t"
+                "psubq   %%xmm1, %%xmm4\n\t"
+                ".Lcarry_xmm5%=:\n\t"
+                "psubq   %%xmm1, %%xmm5\n\t"
+
+                ".Lno_carry%=:\n\t"
+                "movdqa (%[key]), %%xmm1\n\t"   /* xmm1 := key[0]    */
+                "movl %[rounds], %%esi\n\t"
+
+                "pshufb %%xmm6, %%xmm2\n\t"     /* xmm2 := be(xmm2) */
+                "pshufb %%xmm6, %%xmm3\n\t"     /* xmm3 := be(xmm3) */
+                "pshufb %%xmm6, %%xmm4\n\t"     /* xmm4 := be(xmm4) */
+                "pshufb %%xmm6, %%xmm5\n\t"     /* xmm5 := be(xmm5) */
+                "movdqa %%xmm5, (%[ctr])\n\t"   /* Update CTR (mem).  */
+
+                "pxor   %%xmm1, %%xmm0\n\t"     /* xmm0 ^= key[0]    */
+                "pxor   %%xmm1, %%xmm2\n\t"     /* xmm2 ^= key[0]    */
+                "pxor   %%xmm1, %%xmm3\n\t"     /* xmm3 ^= key[0]    */
+                "pxor   %%xmm1, %%xmm4\n\t"     /* xmm4 ^= key[0]    */
+                "movdqa 0x10(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                aesenc_xmm1_xmm2
+                aesenc_xmm1_xmm3
+                aesenc_xmm1_xmm4
+                "movdqa 0x20(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                aesenc_xmm1_xmm2
+                aesenc_xmm1_xmm3
+                aesenc_xmm1_xmm4
+                "movdqa 0x30(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                aesenc_xmm1_xmm2
+                aesenc_xmm1_xmm3
+                aesenc_xmm1_xmm4
+                "movdqa 0x40(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                aesenc_xmm1_xmm2
+                aesenc_xmm1_xmm3
+                aesenc_xmm1_xmm4
+                "movdqa 0x50(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                aesenc_xmm1_xmm2
+                aesenc_xmm1_xmm3
+                aesenc_xmm1_xmm4
+                "movdqa 0x60(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                aesenc_xmm1_xmm2
+                aesenc_xmm1_xmm3
+                aesenc_xmm1_xmm4
+                "movdqa 0x70(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                aesenc_xmm1_xmm2
+                aesenc_xmm1_xmm3
+                aesenc_xmm1_xmm4
+                "movdqa 0x80(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                aesenc_xmm1_xmm2
+                aesenc_xmm1_xmm3
+                aesenc_xmm1_xmm4
+                "movdqa 0x90(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                aesenc_xmm1_xmm2
+                aesenc_xmm1_xmm3
+                aesenc_xmm1_xmm4
+                "movdqa 0xa0(%[key]), %%xmm1\n\t"
+                "cmpl $10, %%esi\n\t"
+                "jz .Lenclast%=\n\t"
+                aesenc_xmm1_xmm0
+                aesenc_xmm1_xmm2
+                aesenc_xmm1_xmm3
+                aesenc_xmm1_xmm4
+                "movdqa 0xb0(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                aesenc_xmm1_xmm2
+                aesenc_xmm1_xmm3
+                aesenc_xmm1_xmm4
+                "movdqa 0xc0(%[key]), %%xmm1\n\t"
+                "cmpl $12, %%esi\n\t"
+                "jz .Lenclast%=\n\t"
+                aesenc_xmm1_xmm0
+                aesenc_xmm1_xmm2
+                aesenc_xmm1_xmm3
+                aesenc_xmm1_xmm4
+                "movdqa 0xd0(%[key]), %%xmm1\n\t"
+                aesenc_xmm1_xmm0
+                aesenc_xmm1_xmm2
+                aesenc_xmm1_xmm3
+                aesenc_xmm1_xmm4
+                "movdqa 0xe0(%[key]), %%xmm1\n"
+
+                ".Lenclast%=:\n\t"
+                aesenclast_xmm1_xmm0
+                aesenclast_xmm1_xmm2
+                aesenclast_xmm1_xmm3
+                aesenclast_xmm1_xmm4
+
+                "movdqu (%[src]), %%xmm1\n\t"    /* Get block 1.      */
+                "pxor %%xmm1, %%xmm0\n\t"        /* EncCTR-1 ^= input */
+                "movdqu %%xmm0, (%[dst])\n\t"    /* Store block 1     */
+
+                "movdqu 16(%[src]), %%xmm1\n\t"  /* Get block 2.      */
+                "pxor %%xmm1, %%xmm2\n\t"        /* EncCTR-2 ^= input */
+                "movdqu %%xmm2, 16(%[dst])\n\t"  /* Store block 2.    */
+
+                "movdqu 32(%[src]), %%xmm1\n\t"  /* Get block 3.      */
+                "pxor %%xmm1, %%xmm3\n\t"        /* EncCTR-3 ^= input */
+                "movdqu %%xmm3, 32(%[dst])\n\t"  /* Store block 3.    */
+
+                "movdqu 48(%[src]), %%xmm1\n\t"  /* Get block 4.      */
+                "pxor %%xmm1, %%xmm4\n\t"        /* EncCTR-4 ^= input */
+                "movdqu %%xmm4, 48(%[dst])"      /* Store block 4.   */
+
+                :
+                : [ctr] "r" (ctr),
+                  [src] "r" (a),
+                  [dst] "r" (b),
+                  [key] "r" (ctx->keyschenc),
+                  [rounds] "g" (ctx->rounds)
+                : "%esi", "cc", "memory");
+#undef aesenc_xmm1_xmm0
+#undef aesenc_xmm1_xmm2
+#undef aesenc_xmm1_xmm3
+#undef aesenc_xmm1_xmm4
+#undef aesenclast_xmm1_xmm0
+#undef aesenclast_xmm1_xmm2
+#undef aesenclast_xmm1_xmm3
+#undef aesenclast_xmm1_xmm4
+}
+
+
+void
+_gcry_aes_aesni_encrypt (RIJNDAEL_context *ctx, unsigned char *dst,
+                         const unsigned char *src)
+{
+  aesni_prepare ();
+  do_aesni_enc (ctx, dst, src);
+  aesni_cleanup ();
+}
+
+
+void
+_gcry_aes_aesni_cfb_enc (RIJNDAEL_context *ctx, unsigned char *outbuf,
+                         const unsigned char *inbuf, unsigned char *iv,
+                         size_t nblocks)
+{
+  aesni_prepare ();
+  for ( ;nblocks; nblocks-- )
+    {
+      do_aesni_cfb (ctx, 0, iv, outbuf, inbuf);
+      outbuf += BLOCKSIZE;
+      inbuf  += BLOCKSIZE;
+    }
+  aesni_cleanup ();
+}
+
+
+void
+_gcry_aes_aesni_cbc_enc (RIJNDAEL_context *ctx, unsigned char *outbuf,
+                         const unsigned char *inbuf, unsigned char *iv,
+                         size_t nblocks, int cbc_mac)
+{
+  unsigned char *last_iv;
+
+  aesni_prepare ();
+
+  last_iv = iv;
+
+  for ( ;nblocks; nblocks-- )
+    {
+      /* ~35% speed up on Sandy-Bridge when doing xoring and copying with
+          SSE registers.  */
+      asm volatile ("movdqu %[iv], %%xmm0\n\t"
+                    "movdqu %[inbuf], %%xmm1\n\t"
+                    "pxor %%xmm0, %%xmm1\n\t"
+                    "movdqu %%xmm1, %[outbuf]\n\t"
+                    : /* No output */
+                    : [iv] "m" (*last_iv),
+                      [inbuf] "m" (*inbuf),
+                      [outbuf] "m" (*outbuf)
+                    : "memory" );
+
+      do_aesni_enc (ctx, outbuf, outbuf);
+
+      last_iv = outbuf;
+      inbuf += BLOCKSIZE;
+      if (!cbc_mac)
+        outbuf += BLOCKSIZE;
+    }
+
+  if (last_iv != iv)
+    {
+      asm volatile ("movdqu %[last], %%xmm0\n\t"
+                    "movdqu %%xmm0, %[iv]\n\t"
+                    : /* No output */
+                    : [last] "m" (*last_iv),
+                      [iv] "m" (*iv)
+                    : "memory" );
+    }
+
+  aesni_cleanup ();
+}
+
+
+void
+_gcry_aes_aesni_ctr_enc (RIJNDAEL_context *ctx, unsigned char *outbuf,
+                         const unsigned char *inbuf, unsigned char *ctr,
+                         size_t nblocks)
+{
+  static const unsigned char be_mask[16] __attribute__ ((aligned (16))) =
+    { 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0 };
+
+  aesni_prepare ();
+
+  asm volatile ("movdqa %[mask], %%xmm6\n\t" /* Preload mask */
+                "movdqa %[ctr], %%xmm5\n\t"  /* Preload CTR */
+                : /* No output */
+                : [mask] "m" (*be_mask),
+                  [ctr] "m" (*ctr)
+                : "memory");
+
+  for ( ;nblocks > 3 ; nblocks -= 4 )
+    {
+      do_aesni_ctr_4 (ctx, ctr, outbuf, inbuf);
+      outbuf += 4*BLOCKSIZE;
+      inbuf  += 4*BLOCKSIZE;
+    }
+  for ( ;nblocks; nblocks-- )
+    {
+      do_aesni_ctr (ctx, ctr, outbuf, inbuf);
+      outbuf += BLOCKSIZE;
+      inbuf  += BLOCKSIZE;
+    }
+  aesni_cleanup ();
+  aesni_cleanup_2_6 ();
+}
+
+
+void
+_gcry_aes_aesni_decrypt (RIJNDAEL_context *ctx, unsigned char *dst,
+                         const unsigned char *src)
+{
+  aesni_prepare ();
+  do_aesni_dec (ctx, dst, src);
+  aesni_cleanup ();
+}
+
+
+void
+_gcry_aes_aesni_cfb_dec (RIJNDAEL_context *ctx, unsigned char *outbuf,
+                         const unsigned char *inbuf, unsigned char *iv,
+                         size_t nblocks)
+{
+  aesni_prepare ();
+
+  /* CFB decryption can be parallelized */
+  for ( ;nblocks >= 4; nblocks -= 4)
+    {
+      asm volatile
+        ("movdqu (%[iv]),        %%xmm1\n\t" /* load input blocks */
+         "movdqu 0*16(%[inbuf]), %%xmm2\n\t"
+         "movdqu 1*16(%[inbuf]), %%xmm3\n\t"
+         "movdqu 2*16(%[inbuf]), %%xmm4\n\t"
+
+         "movdqu 3*16(%[inbuf]), %%xmm0\n\t" /* update IV */
+         "movdqu %%xmm0,         (%[iv])\n\t"
+         : /* No output */
+         : [inbuf] "r" (inbuf), [iv] "r" (iv)
+         : "memory");
+
+      do_aesni_enc_vec4 (ctx);
+
+      asm volatile
+        ("movdqu 0*16(%[inbuf]), %%xmm5\n\t"
+         "pxor %%xmm5, %%xmm1\n\t"
+         "movdqu %%xmm1, 0*16(%[outbuf])\n\t"
+
+         "movdqu 1*16(%[inbuf]), %%xmm5\n\t"
+         "pxor %%xmm5, %%xmm2\n\t"
+         "movdqu %%xmm2, 1*16(%[outbuf])\n\t"
+
+         "movdqu 2*16(%[inbuf]), %%xmm5\n\t"
+         "pxor %%xmm5, %%xmm3\n\t"
+         "movdqu %%xmm3, 2*16(%[outbuf])\n\t"
+
+         "movdqu 3*16(%[inbuf]), %%xmm5\n\t"
+         "pxor %%xmm5, %%xmm4\n\t"
+         "movdqu %%xmm4, 3*16(%[outbuf])\n\t"
+
+         : /* No output */
+         : [inbuf] "r" (inbuf),
+           [outbuf] "r" (outbuf)
+         : "memory");
+
+      outbuf += 4*BLOCKSIZE;
+      inbuf  += 4*BLOCKSIZE;
+    }
+
+  for ( ;nblocks; nblocks-- )
+    {
+      do_aesni_cfb (ctx, 1, iv, outbuf, inbuf);
+      outbuf += BLOCKSIZE;
+      inbuf  += BLOCKSIZE;
+    }
+  aesni_cleanup ();
+  aesni_cleanup_2_6 ();
+}
+
+
+void
+_gcry_aes_aesni_cbc_dec (RIJNDAEL_context *ctx, unsigned char *outbuf,
+			 const unsigned char *inbuf, unsigned char *iv,
+			 size_t nblocks)
+{
+  aesni_prepare ();
+
+  asm volatile
+    ("movdqu %[iv], %%xmm5\n\t"	/* use xmm5 as fast IV storage */
+     : /* No output */
+     : [iv] "m" (*iv)
+     : "memory");
+
+  for ( ;nblocks > 3 ; nblocks -= 4 )
+    {
+      asm volatile
+        ("movdqu 0*16(%[inbuf]), %%xmm1\n\t"	/* load input blocks */
+         "movdqu 1*16(%[inbuf]), %%xmm2\n\t"
+         "movdqu 2*16(%[inbuf]), %%xmm3\n\t"
+         "movdqu 3*16(%[inbuf]), %%xmm4\n\t"
+         : /* No output */
+         : [inbuf] "r" (inbuf)
+         : "memory");
+
+      do_aesni_dec_vec4 (ctx);
+
+      asm volatile
+        ("pxor %%xmm5, %%xmm1\n\t"		/* xor IV with output */
+         "movdqu 0*16(%[inbuf]), %%xmm5\n\t"	/* load new IV */
+         "movdqu %%xmm1, 0*16(%[outbuf])\n\t"
+
+         "pxor %%xmm5, %%xmm2\n\t"		/* xor IV with output */
+         "movdqu 1*16(%[inbuf]), %%xmm5\n\t"	/* load new IV */
+         "movdqu %%xmm2, 1*16(%[outbuf])\n\t"
+
+         "pxor %%xmm5, %%xmm3\n\t"		/* xor IV with output */
+         "movdqu 2*16(%[inbuf]), %%xmm5\n\t"	/* load new IV */
+         "movdqu %%xmm3, 2*16(%[outbuf])\n\t"
+
+         "pxor %%xmm5, %%xmm4\n\t"		/* xor IV with output */
+         "movdqu 3*16(%[inbuf]), %%xmm5\n\t"	/* load new IV */
+         "movdqu %%xmm4, 3*16(%[outbuf])\n\t"
+
+         : /* No output */
+         : [inbuf] "r" (inbuf),
+           [outbuf] "r" (outbuf)
+         : "memory");
+
+      outbuf += 4*BLOCKSIZE;
+      inbuf  += 4*BLOCKSIZE;
+    }
+
+  for ( ;nblocks; nblocks-- )
+    {
+      asm volatile
+        ("movdqu %[inbuf], %%xmm2\n\t"	/* use xmm2 as savebuf */
+         : /* No output */
+         : [inbuf] "m" (*inbuf)
+         : "memory");
+
+      /* uses only xmm0 and xmm1 */
+      do_aesni_dec (ctx, outbuf, inbuf);
+
+      asm volatile
+        ("movdqu %[outbuf], %%xmm0\n\t"
+         "pxor %%xmm5, %%xmm0\n\t"	/* xor IV with output */
+         "movdqu %%xmm0, %[outbuf]\n\t"
+         "movdqu %%xmm2, %%xmm5\n\t"	/* store savebuf as new IV */
+         : /* No output */
+         : [outbuf] "m" (*outbuf)
+         : "memory");
+
+      outbuf += BLOCKSIZE;
+      inbuf  += BLOCKSIZE;
+    }
+
+  asm volatile
+    ("movdqu %%xmm5, %[iv]\n\t"	/* store IV */
+     : /* No output */
+     : [iv] "m" (*iv)
+     : "memory");
+
+  aesni_cleanup ();
+  aesni_cleanup_2_6 ();
+}
+
+#endif /* USE_AESNI */