Support for Signal calls.

Merge in RedPhone

// FREEBIE

jni/openssl/crypto/aes/README

@@ -0,0 +1,3 @@
+This is an OpenSSL-compatible version of AES (also called Rijndael).
+aes_core.c is basically the same as rijndael-alg-fst.c but with an
+API that looks like the rest of the OpenSSL symmetric cipher suite.

jni/openssl/crypto/aes/aes.h

@@ -0,0 +1,147 @@
+/* crypto/aes/aes.h -*- mode:C; c-file-style: "eay" -*- */
+/* ====================================================================
+ * Copyright (c) 1998-2002 The OpenSSL Project.  All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer. 
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *
+ * 3. All advertising materials mentioning features or use of this
+ *    software must display the following acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit. (http://www.openssl.org/)"
+ *
+ * 4. The names "OpenSSL Toolkit" and "OpenSSL Project" must not be used to
+ *    endorse or promote products derived from this software without
+ *    prior written permission. For written permission, please contact
+ *    openssl-core@openssl.org.
+ *
+ * 5. Products derived from this software may not be called "OpenSSL"
+ *    nor may "OpenSSL" appear in their names without prior written
+ *    permission of the OpenSSL Project.
+ *
+ * 6. Redistributions of any form whatsoever must retain the following
+ *    acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit (http://www.openssl.org/)"
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE OpenSSL PROJECT ``AS IS'' AND ANY
+ * EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE OpenSSL PROJECT OR
+ * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ * ====================================================================
+ *
+ */
+
+#ifndef HEADER_AES_H
+#define HEADER_AES_H
+
+#include <openssl/opensslconf.h>
+
+#ifdef OPENSSL_NO_AES
+#error AES is disabled.
+#endif
+
+#include <stddef.h>
+
+#define AES_ENCRYPT	1
+#define AES_DECRYPT	0
+
+/* Because array size can't be a const in C, the following two are macros.
+   Both sizes are in bytes. */
+#define AES_MAXNR 14
+#define AES_BLOCK_SIZE 16
+
+#ifdef  __cplusplus
+extern "C" {
+#endif
+
+/* This should be a hidden type, but EVP requires that the size be known */
+struct aes_key_st {
+#ifdef AES_LONG
+    unsigned long rd_key[4 *(AES_MAXNR + 1)];
+#else
+    unsigned int rd_key[4 *(AES_MAXNR + 1)];
+#endif
+    int rounds;
+};
+typedef struct aes_key_st AES_KEY;
+
+const char *AES_options(void);
+
+int AES_set_encrypt_key(const unsigned char *userKey, const int bits,
+	AES_KEY *key);
+int AES_set_decrypt_key(const unsigned char *userKey, const int bits,
+	AES_KEY *key);
+
+int private_AES_set_encrypt_key(const unsigned char *userKey, const int bits,
+	AES_KEY *key);
+int private_AES_set_decrypt_key(const unsigned char *userKey, const int bits,
+	AES_KEY *key);
+
+void AES_encrypt(const unsigned char *in, unsigned char *out,
+	const AES_KEY *key);
+void AES_decrypt(const unsigned char *in, unsigned char *out,
+	const AES_KEY *key);
+
+void AES_ecb_encrypt(const unsigned char *in, unsigned char *out,
+	const AES_KEY *key, const int enc);
+void AES_cbc_encrypt(const unsigned char *in, unsigned char *out,
+	size_t length, const AES_KEY *key,
+	unsigned char *ivec, const int enc);
+void AES_cfb128_encrypt(const unsigned char *in, unsigned char *out,
+	size_t length, const AES_KEY *key,
+	unsigned char *ivec, int *num, const int enc);
+void AES_cfb1_encrypt(const unsigned char *in, unsigned char *out,
+	size_t length, const AES_KEY *key,
+	unsigned char *ivec, int *num, const int enc);
+void AES_cfb8_encrypt(const unsigned char *in, unsigned char *out,
+	size_t length, const AES_KEY *key,
+	unsigned char *ivec, int *num, const int enc);
+void AES_ofb128_encrypt(const unsigned char *in, unsigned char *out,
+	size_t length, const AES_KEY *key,
+	unsigned char *ivec, int *num);
+void AES_ctr128_encrypt(const unsigned char *in, unsigned char *out,
+	size_t length, const AES_KEY *key,
+	unsigned char ivec[AES_BLOCK_SIZE],
+	unsigned char ecount_buf[AES_BLOCK_SIZE],
+	unsigned int *num);
+/* NB: the IV is _two_ blocks long */
+void AES_ige_encrypt(const unsigned char *in, unsigned char *out,
+		     size_t length, const AES_KEY *key,
+		     unsigned char *ivec, const int enc);
+/* NB: the IV is _four_ blocks long */
+void AES_bi_ige_encrypt(const unsigned char *in, unsigned char *out,
+			size_t length, const AES_KEY *key,
+			const AES_KEY *key2, const unsigned char *ivec,
+			const int enc);
+
+int AES_wrap_key(AES_KEY *key, const unsigned char *iv,
+		unsigned char *out,
+		const unsigned char *in, unsigned int inlen);
+int AES_unwrap_key(AES_KEY *key, const unsigned char *iv,
+		unsigned char *out,
+		const unsigned char *in, unsigned int inlen);
+
+
+#ifdef  __cplusplus
+}
+#endif
+
+#endif /* !HEADER_AES_H */

jni/openssl/crypto/aes/aes_cbc.c

@@ -0,0 +1,63 @@
+/* crypto/aes/aes_cbc.c -*- mode:C; c-file-style: "eay" -*- */
+/* ====================================================================
+ * Copyright (c) 1998-2002 The OpenSSL Project.  All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer. 
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *
+ * 3. All advertising materials mentioning features or use of this
+ *    software must display the following acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit. (http://www.openssl.org/)"
+ *
+ * 4. The names "OpenSSL Toolkit" and "OpenSSL Project" must not be used to
+ *    endorse or promote products derived from this software without
+ *    prior written permission. For written permission, please contact
+ *    openssl-core@openssl.org.
+ *
+ * 5. Products derived from this software may not be called "OpenSSL"
+ *    nor may "OpenSSL" appear in their names without prior written
+ *    permission of the OpenSSL Project.
+ *
+ * 6. Redistributions of any form whatsoever must retain the following
+ *    acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit (http://www.openssl.org/)"
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE OpenSSL PROJECT ``AS IS'' AND ANY
+ * EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE OpenSSL PROJECT OR
+ * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ * ====================================================================
+ *
+ */
+
+#include <openssl/aes.h>
+#include <openssl/modes.h>
+
+void AES_cbc_encrypt(const unsigned char *in, unsigned char *out,
+		     size_t len, const AES_KEY *key,
+		     unsigned char *ivec, const int enc) {
+
+	if (enc)
+		CRYPTO_cbc128_encrypt(in,out,len,key,ivec,(block128_f)AES_encrypt);
+	else
+		CRYPTO_cbc128_decrypt(in,out,len,key,ivec,(block128_f)AES_decrypt);
+}

jni/openssl/crypto/aes/aes_cfb.c

@@ -0,0 +1,81 @@
+/* crypto/aes/aes_cfb.c -*- mode:C; c-file-style: "eay" -*- */
+/* ====================================================================
+ * Copyright (c) 2002-2006 The OpenSSL Project.  All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer. 
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *
+ * 3. All advertising materials mentioning features or use of this
+ *    software must display the following acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit. (http://www.openssl.org/)"
+ *
+ * 4. The names "OpenSSL Toolkit" and "OpenSSL Project" must not be used to
+ *    endorse or promote products derived from this software without
+ *    prior written permission. For written permission, please contact
+ *    openssl-core@openssl.org.
+ *
+ * 5. Products derived from this software may not be called "OpenSSL"
+ *    nor may "OpenSSL" appear in their names without prior written
+ *    permission of the OpenSSL Project.
+ *
+ * 6. Redistributions of any form whatsoever must retain the following
+ *    acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit (http://www.openssl.org/)"
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE OpenSSL PROJECT ``AS IS'' AND ANY
+ * EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE OpenSSL PROJECT OR
+ * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ * ====================================================================
+ *
+ */
+
+#include <openssl/aes.h>
+#include <openssl/modes.h>
+
+/* The input and output encrypted as though 128bit cfb mode is being
+ * used.  The extra state information to record how much of the
+ * 128bit block we have used is contained in *num;
+ */
+
+void AES_cfb128_encrypt(const unsigned char *in, unsigned char *out,
+	size_t length, const AES_KEY *key,
+	unsigned char *ivec, int *num, const int enc) {
+
+	CRYPTO_cfb128_encrypt(in,out,length,key,ivec,num,enc,(block128_f)AES_encrypt);
+}
+
+/* N.B. This expects the input to be packed, MS bit first */
+void AES_cfb1_encrypt(const unsigned char *in, unsigned char *out,
+		      size_t length, const AES_KEY *key,
+		      unsigned char *ivec, int *num, const int enc)
+    {
+    CRYPTO_cfb128_1_encrypt(in,out,length,key,ivec,num,enc,(block128_f)AES_encrypt);
+    }
+
+void AES_cfb8_encrypt(const unsigned char *in, unsigned char *out,
+		      size_t length, const AES_KEY *key,
+		      unsigned char *ivec, int *num, const int enc)
+    {
+    CRYPTO_cfb128_8_encrypt(in,out,length,key,ivec,num,enc,(block128_f)AES_encrypt);
+    }
+

jni/openssl/crypto/aes/aes_ctr.c

@@ -0,0 +1,61 @@
+/* crypto/aes/aes_ctr.c -*- mode:C; c-file-style: "eay" -*- */
+/* ====================================================================
+ * Copyright (c) 1998-2002 The OpenSSL Project.  All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer. 
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *
+ * 3. All advertising materials mentioning features or use of this
+ *    software must display the following acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit. (http://www.openssl.org/)"
+ *
+ * 4. The names "OpenSSL Toolkit" and "OpenSSL Project" must not be used to
+ *    endorse or promote products derived from this software without
+ *    prior written permission. For written permission, please contact
+ *    openssl-core@openssl.org.
+ *
+ * 5. Products derived from this software may not be called "OpenSSL"
+ *    nor may "OpenSSL" appear in their names without prior written
+ *    permission of the OpenSSL Project.
+ *
+ * 6. Redistributions of any form whatsoever must retain the following
+ *    acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit (http://www.openssl.org/)"
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE OpenSSL PROJECT ``AS IS'' AND ANY
+ * EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE OpenSSL PROJECT OR
+ * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ * ====================================================================
+ *
+ */
+
+#include <openssl/aes.h>
+#include <openssl/modes.h>
+
+void AES_ctr128_encrypt(const unsigned char *in, unsigned char *out,
+			size_t length, const AES_KEY *key,
+			unsigned char ivec[AES_BLOCK_SIZE],
+			unsigned char ecount_buf[AES_BLOCK_SIZE],
+			unsigned int *num) {
+	CRYPTO_ctr128_encrypt(in,out,length,key,ivec,ecount_buf,num,(block128_f)AES_encrypt);
+}

jni/openssl/crypto/aes/aes_ecb.c

@@ -0,0 +1,73 @@
+/* crypto/aes/aes_ecb.c -*- mode:C; c-file-style: "eay" -*- */
+/* ====================================================================
+ * Copyright (c) 1998-2002 The OpenSSL Project.  All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer. 
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *
+ * 3. All advertising materials mentioning features or use of this
+ *    software must display the following acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit. (http://www.openssl.org/)"
+ *
+ * 4. The names "OpenSSL Toolkit" and "OpenSSL Project" must not be used to
+ *    endorse or promote products derived from this software without
+ *    prior written permission. For written permission, please contact
+ *    openssl-core@openssl.org.
+ *
+ * 5. Products derived from this software may not be called "OpenSSL"
+ *    nor may "OpenSSL" appear in their names without prior written
+ *    permission of the OpenSSL Project.
+ *
+ * 6. Redistributions of any form whatsoever must retain the following
+ *    acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit (http://www.openssl.org/)"
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE OpenSSL PROJECT ``AS IS'' AND ANY
+ * EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE OpenSSL PROJECT OR
+ * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ * ====================================================================
+ *
+ */
+
+#ifndef AES_DEBUG
+# ifndef NDEBUG
+#  define NDEBUG
+# endif
+#endif
+#include <assert.h>
+
+#include <openssl/aes.h>
+#include "aes_locl.h"
+
+void AES_ecb_encrypt(const unsigned char *in, unsigned char *out,
+		     const AES_KEY *key, const int enc) {
+
+        assert(in && out && key);
+	assert((AES_ENCRYPT == enc)||(AES_DECRYPT == enc));
+
+	if (AES_ENCRYPT == enc)
+		AES_encrypt(in, out, key);
+	else
+		AES_decrypt(in, out, key);
+}
+

jni/openssl/crypto/aes/aes_ige.c

@@ -0,0 +1,323 @@
+/* crypto/aes/aes_ige.c -*- mode:C; c-file-style: "eay" -*- */
+/* ====================================================================
+ * Copyright (c) 2006 The OpenSSL Project.  All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer. 
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *
+ * 3. All advertising materials mentioning features or use of this
+ *    software must display the following acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit. (http://www.openssl.org/)"
+ *
+ * 4. The names "OpenSSL Toolkit" and "OpenSSL Project" must not be used to
+ *    endorse or promote products derived from this software without
+ *    prior written permission. For written permission, please contact
+ *    openssl-core@openssl.org.
+ *
+ * 5. Products derived from this software may not be called "OpenSSL"
+ *    nor may "OpenSSL" appear in their names without prior written
+ *    permission of the OpenSSL Project.
+ *
+ * 6. Redistributions of any form whatsoever must retain the following
+ *    acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit (http://www.openssl.org/)"
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE OpenSSL PROJECT ``AS IS'' AND ANY
+ * EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE OpenSSL PROJECT OR
+ * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ * ====================================================================
+ *
+ */
+
+#include "cryptlib.h"
+
+#include <openssl/aes.h>
+#include "aes_locl.h"
+
+#define N_WORDS (AES_BLOCK_SIZE / sizeof(unsigned long))
+typedef struct {
+        unsigned long data[N_WORDS];
+} aes_block_t;
+
+/* XXX: probably some better way to do this */
+#if defined(__i386__) || defined(__x86_64__)
+#define UNALIGNED_MEMOPS_ARE_FAST 1
+#else
+#define UNALIGNED_MEMOPS_ARE_FAST 0
+#endif
+
+#if UNALIGNED_MEMOPS_ARE_FAST
+#define load_block(d, s)        (d) = *(const aes_block_t *)(s)
+#define store_block(d, s)       *(aes_block_t *)(d) = (s)
+#else
+#define load_block(d, s)        memcpy((d).data, (s), AES_BLOCK_SIZE)
+#define store_block(d, s)       memcpy((d), (s).data, AES_BLOCK_SIZE)
+#endif
+
+/* N.B. The IV for this mode is _twice_ the block size */
+
+void AES_ige_encrypt(const unsigned char *in, unsigned char *out,
+					 size_t length, const AES_KEY *key,
+					 unsigned char *ivec, const int enc)
+	{
+	size_t n;
+	size_t len = length;
+
+	OPENSSL_assert(in && out && key && ivec);
+	OPENSSL_assert((AES_ENCRYPT == enc)||(AES_DECRYPT == enc));
+	OPENSSL_assert((length%AES_BLOCK_SIZE) == 0);
+
+	len = length / AES_BLOCK_SIZE;
+
+	if (AES_ENCRYPT == enc)
+		{
+		if (in != out &&
+		    (UNALIGNED_MEMOPS_ARE_FAST || ((size_t)in|(size_t)out|(size_t)ivec)%sizeof(long)==0))
+			{
+			aes_block_t *ivp = (aes_block_t *)ivec;
+			aes_block_t *iv2p = (aes_block_t *)(ivec + AES_BLOCK_SIZE);
+
+			while (len)
+				{
+				aes_block_t *inp = (aes_block_t *)in;
+				aes_block_t *outp = (aes_block_t *)out;
+
+				for(n=0 ; n < N_WORDS; ++n)
+					outp->data[n] = inp->data[n] ^ ivp->data[n];
+				AES_encrypt((unsigned char *)outp->data, (unsigned char *)outp->data, key);
+				for(n=0 ; n < N_WORDS; ++n)
+					outp->data[n] ^= iv2p->data[n];
+				ivp = outp;
+				iv2p = inp;
+				--len;
+				in += AES_BLOCK_SIZE;
+				out += AES_BLOCK_SIZE;
+				}
+			memcpy(ivec, ivp->data, AES_BLOCK_SIZE);
+			memcpy(ivec + AES_BLOCK_SIZE, iv2p->data, AES_BLOCK_SIZE);
+			}
+		else
+			{
+			aes_block_t tmp, tmp2;
+			aes_block_t iv;
+			aes_block_t iv2;
+
+			load_block(iv, ivec);
+			load_block(iv2, ivec + AES_BLOCK_SIZE);
+
+			while (len)
+				{
+				load_block(tmp, in);
+				for(n=0 ; n < N_WORDS; ++n)
+					tmp2.data[n] = tmp.data[n] ^ iv.data[n];
+				AES_encrypt((unsigned char *)tmp2.data, (unsigned char *)tmp2.data, key);
+				for(n=0 ; n < N_WORDS; ++n)
+					tmp2.data[n] ^= iv2.data[n];
+				store_block(out, tmp2);
+				iv = tmp2;
+				iv2 = tmp;
+				--len;
+				in += AES_BLOCK_SIZE;
+				out += AES_BLOCK_SIZE;
+				}
+			memcpy(ivec, iv.data, AES_BLOCK_SIZE);
+			memcpy(ivec + AES_BLOCK_SIZE, iv2.data, AES_BLOCK_SIZE);
+			}
+		}
+	else
+		{
+		if (in != out &&
+		    (UNALIGNED_MEMOPS_ARE_FAST || ((size_t)in|(size_t)out|(size_t)ivec)%sizeof(long)==0))
+			{
+			aes_block_t *ivp = (aes_block_t *)ivec;
+			aes_block_t *iv2p = (aes_block_t *)(ivec + AES_BLOCK_SIZE);
+
+			while (len)
+				{
+				aes_block_t tmp;
+				aes_block_t *inp = (aes_block_t *)in;
+				aes_block_t *outp = (aes_block_t *)out;
+
+				for(n=0 ; n < N_WORDS; ++n)
+					tmp.data[n] = inp->data[n] ^ iv2p->data[n];
+				AES_decrypt((unsigned char *)tmp.data, (unsigned char *)outp->data, key);
+				for(n=0 ; n < N_WORDS; ++n)
+					outp->data[n] ^= ivp->data[n];
+				ivp = inp;
+				iv2p = outp;
+				--len;
+				in += AES_BLOCK_SIZE;
+				out += AES_BLOCK_SIZE;
+				}
+			memcpy(ivec, ivp->data, AES_BLOCK_SIZE);
+			memcpy(ivec + AES_BLOCK_SIZE, iv2p->data, AES_BLOCK_SIZE);
+			}
+		else
+			{
+			aes_block_t tmp, tmp2;
+			aes_block_t iv;
+			aes_block_t iv2;
+
+			load_block(iv, ivec);
+			load_block(iv2, ivec + AES_BLOCK_SIZE);
+
+			while (len)
+				{
+				load_block(tmp, in);
+				tmp2 = tmp;
+				for(n=0 ; n < N_WORDS; ++n)
+					tmp.data[n] ^= iv2.data[n];
+				AES_decrypt((unsigned char *)tmp.data, (unsigned char *)tmp.data, key);
+				for(n=0 ; n < N_WORDS; ++n)
+					tmp.data[n] ^= iv.data[n];
+				store_block(out, tmp);
+				iv = tmp2;
+				iv2 = tmp;
+				--len;
+				in += AES_BLOCK_SIZE;
+				out += AES_BLOCK_SIZE;
+				}
+			memcpy(ivec, iv.data, AES_BLOCK_SIZE);
+			memcpy(ivec + AES_BLOCK_SIZE, iv2.data, AES_BLOCK_SIZE);
+			}
+		}
+	}
+
+/*
+ * Note that its effectively impossible to do biIGE in anything other
+ * than a single pass, so no provision is made for chaining.
+ */
+
+/* N.B. The IV for this mode is _four times_ the block size */
+
+void AES_bi_ige_encrypt(const unsigned char *in, unsigned char *out,
+						size_t length, const AES_KEY *key,
+						const AES_KEY *key2, const unsigned char *ivec,
+						const int enc)
+	{
+	size_t n;
+	size_t len = length;
+	unsigned char tmp[AES_BLOCK_SIZE];
+	unsigned char tmp2[AES_BLOCK_SIZE];
+	unsigned char tmp3[AES_BLOCK_SIZE];
+	unsigned char prev[AES_BLOCK_SIZE];
+	const unsigned char *iv;
+	const unsigned char *iv2;
+
+	OPENSSL_assert(in && out && key && ivec);
+	OPENSSL_assert((AES_ENCRYPT == enc)||(AES_DECRYPT == enc));
+	OPENSSL_assert((length%AES_BLOCK_SIZE) == 0);
+
+	if (AES_ENCRYPT == enc)
+		{
+		/* XXX: Do a separate case for when in != out (strictly should
+		   check for overlap, too) */
+
+		/* First the forward pass */ 
+		iv = ivec;
+		iv2 = ivec + AES_BLOCK_SIZE;
+		while (len >= AES_BLOCK_SIZE)
+			{
+			for(n=0 ; n < AES_BLOCK_SIZE ; ++n)
+				out[n] = in[n] ^ iv[n];
+			AES_encrypt(out, out, key);
+			for(n=0 ; n < AES_BLOCK_SIZE ; ++n)
+				out[n] ^= iv2[n];
+			iv = out;
+			memcpy(prev, in, AES_BLOCK_SIZE);
+			iv2 = prev;
+			len -= AES_BLOCK_SIZE;
+			in += AES_BLOCK_SIZE;
+			out += AES_BLOCK_SIZE;
+			}
+
+		/* And now backwards */
+		iv = ivec + AES_BLOCK_SIZE*2;
+		iv2 = ivec + AES_BLOCK_SIZE*3;
+		len = length;
+		while(len >= AES_BLOCK_SIZE)
+			{
+			out -= AES_BLOCK_SIZE;
+			/* XXX: reduce copies by alternating between buffers */
+			memcpy(tmp, out, AES_BLOCK_SIZE);
+			for(n=0 ; n < AES_BLOCK_SIZE ; ++n)
+				out[n] ^= iv[n];
+			/*			hexdump(stdout, "out ^ iv", out, AES_BLOCK_SIZE); */
+			AES_encrypt(out, out, key);
+			/*			hexdump(stdout,"enc", out, AES_BLOCK_SIZE); */
+			/*			hexdump(stdout,"iv2", iv2, AES_BLOCK_SIZE); */
+			for(n=0 ; n < AES_BLOCK_SIZE ; ++n)
+				out[n] ^= iv2[n];
+			/*			hexdump(stdout,"out", out, AES_BLOCK_SIZE); */
+			iv = out;
+			memcpy(prev, tmp, AES_BLOCK_SIZE);
+			iv2 = prev;
+			len -= AES_BLOCK_SIZE;
+			}
+		}
+	else
+		{
+		/* First backwards */
+		iv = ivec + AES_BLOCK_SIZE*2;
+		iv2 = ivec + AES_BLOCK_SIZE*3;
+		in += length;
+		out += length;
+		while (len >= AES_BLOCK_SIZE)
+			{
+			in -= AES_BLOCK_SIZE;
+			out -= AES_BLOCK_SIZE;
+			memcpy(tmp, in, AES_BLOCK_SIZE);
+			memcpy(tmp2, in, AES_BLOCK_SIZE);
+			for(n=0 ; n < AES_BLOCK_SIZE ; ++n)
+				tmp[n] ^= iv2[n];
+			AES_decrypt(tmp, out, key);
+			for(n=0 ; n < AES_BLOCK_SIZE ; ++n)
+				out[n] ^= iv[n];
+			memcpy(tmp3, tmp2, AES_BLOCK_SIZE);
+			iv = tmp3;
+			iv2 = out;
+			len -= AES_BLOCK_SIZE;
+			}
+
+		/* And now forwards */
+		iv = ivec;
+		iv2 = ivec + AES_BLOCK_SIZE;
+		len = length;
+		while (len >= AES_BLOCK_SIZE)
+			{
+			memcpy(tmp, out, AES_BLOCK_SIZE);
+			memcpy(tmp2, out, AES_BLOCK_SIZE);
+			for(n=0 ; n < AES_BLOCK_SIZE ; ++n)
+				tmp[n] ^= iv2[n];
+			AES_decrypt(tmp, out, key);
+			for(n=0 ; n < AES_BLOCK_SIZE ; ++n)
+				out[n] ^= iv[n];
+			memcpy(tmp3, tmp2, AES_BLOCK_SIZE);
+			iv = tmp3;
+			iv2 = out;
+			len -= AES_BLOCK_SIZE;
+			in += AES_BLOCK_SIZE;
+			out += AES_BLOCK_SIZE;
+			}
+		}
+	}

jni/openssl/crypto/aes/aes_locl.h

@@ -0,0 +1,89 @@
+/* crypto/aes/aes.h -*- mode:C; c-file-style: "eay" -*- */
+/* ====================================================================
+ * Copyright (c) 1998-2002 The OpenSSL Project.  All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer. 
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *
+ * 3. All advertising materials mentioning features or use of this
+ *    software must display the following acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit. (http://www.openssl.org/)"
+ *
+ * 4. The names "OpenSSL Toolkit" and "OpenSSL Project" must not be used to
+ *    endorse or promote products derived from this software without
+ *    prior written permission. For written permission, please contact
+ *    openssl-core@openssl.org.
+ *
+ * 5. Products derived from this software may not be called "OpenSSL"
+ *    nor may "OpenSSL" appear in their names without prior written
+ *    permission of the OpenSSL Project.
+ *
+ * 6. Redistributions of any form whatsoever must retain the following
+ *    acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit (http://www.openssl.org/)"
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE OpenSSL PROJECT ``AS IS'' AND ANY
+ * EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE OpenSSL PROJECT OR
+ * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ * ====================================================================
+ *
+ */
+
+#ifndef HEADER_AES_LOCL_H
+#define HEADER_AES_LOCL_H
+
+#include <openssl/e_os2.h>
+
+#ifdef OPENSSL_NO_AES
+#error AES is disabled.
+#endif
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+#if defined(_MSC_VER) && (defined(_M_IX86) || defined(_M_AMD64) || defined(_M_X64))
+# define SWAP(x) (_lrotl(x, 8) & 0x00ff00ff | _lrotr(x, 8) & 0xff00ff00)
+# define GETU32(p) SWAP(*((u32 *)(p)))
+# define PUTU32(ct, st) { *((u32 *)(ct)) = SWAP((st)); }
+#else
+# define GETU32(pt) (((u32)(pt)[0] << 24) ^ ((u32)(pt)[1] << 16) ^ ((u32)(pt)[2] <<  8) ^ ((u32)(pt)[3]))
+# define PUTU32(ct, st) { (ct)[0] = (u8)((st) >> 24); (ct)[1] = (u8)((st) >> 16); (ct)[2] = (u8)((st) >>  8); (ct)[3] = (u8)(st); }
+#endif
+
+#ifdef AES_LONG
+typedef unsigned long u32;
+#else
+typedef unsigned int u32;
+#endif
+typedef unsigned short u16;
+typedef unsigned char u8;
+
+#define MAXKC   (256/32)
+#define MAXKB   (256/8)
+#define MAXNR   14
+
+/* This controls loop-unrolling in aes_core.c */
+#undef FULL_UNROLL
+
+#endif /* !HEADER_AES_LOCL_H */

jni/openssl/crypto/aes/aes_misc.c

@@ -0,0 +1,85 @@
+/* crypto/aes/aes_misc.c -*- mode:C; c-file-style: "eay" -*- */
+/* ====================================================================
+ * Copyright (c) 1998-2002 The OpenSSL Project.  All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer. 
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *
+ * 3. All advertising materials mentioning features or use of this
+ *    software must display the following acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit. (http://www.openssl.org/)"
+ *
+ * 4. The names "OpenSSL Toolkit" and "OpenSSL Project" must not be used to
+ *    endorse or promote products derived from this software without
+ *    prior written permission. For written permission, please contact
+ *    openssl-core@openssl.org.
+ *
+ * 5. Products derived from this software may not be called "OpenSSL"
+ *    nor may "OpenSSL" appear in their names without prior written
+ *    permission of the OpenSSL Project.
+ *
+ * 6. Redistributions of any form whatsoever must retain the following
+ *    acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit (http://www.openssl.org/)"
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE OpenSSL PROJECT ``AS IS'' AND ANY
+ * EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE OpenSSL PROJECT OR
+ * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ * ====================================================================
+ *
+ */
+
+#include <openssl/opensslv.h>
+#include <openssl/crypto.h>
+#include <openssl/aes.h>
+#include "aes_locl.h"
+
+const char AES_version[]="AES" OPENSSL_VERSION_PTEXT;
+
+const char *AES_options(void) {
+#ifdef FULL_UNROLL
+        return "aes(full)";
+#else   
+        return "aes(partial)";
+#endif
+}
+
+/* FIPS wrapper functions to block low level AES calls in FIPS mode */
+
+int AES_set_encrypt_key(const unsigned char *userKey, const int bits,
+			AES_KEY *key)
+	{
+#ifdef OPENSSL_FIPS
+	fips_cipher_abort(AES);
+#endif
+	return private_AES_set_encrypt_key(userKey, bits, key);
+	}
+
+int AES_set_decrypt_key(const unsigned char *userKey, const int bits,
+			AES_KEY *key)
+	{
+#ifdef OPENSSL_FIPS
+	fips_cipher_abort(AES);
+#endif
+	return private_AES_set_decrypt_key(userKey, bits, key);
+	}

jni/openssl/crypto/aes/aes_ofb.c

@@ -0,0 +1,60 @@
+/* crypto/aes/aes_ofb.c -*- mode:C; c-file-style: "eay" -*- */
+/* ====================================================================
+ * Copyright (c) 2002-2006 The OpenSSL Project.  All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer. 
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *
+ * 3. All advertising materials mentioning features or use of this
+ *    software must display the following acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit. (http://www.openssl.org/)"
+ *
+ * 4. The names "OpenSSL Toolkit" and "OpenSSL Project" must not be used to
+ *    endorse or promote products derived from this software without
+ *    prior written permission. For written permission, please contact
+ *    openssl-core@openssl.org.
+ *
+ * 5. Products derived from this software may not be called "OpenSSL"
+ *    nor may "OpenSSL" appear in their names without prior written
+ *    permission of the OpenSSL Project.
+ *
+ * 6. Redistributions of any form whatsoever must retain the following
+ *    acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit (http://www.openssl.org/)"
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE OpenSSL PROJECT ``AS IS'' AND ANY
+ * EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE OpenSSL PROJECT OR
+ * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ * ====================================================================
+ *
+ */
+
+#include <openssl/aes.h>
+#include <openssl/modes.h>
+
+void AES_ofb128_encrypt(const unsigned char *in, unsigned char *out,
+	size_t length, const AES_KEY *key,
+	unsigned char *ivec, int *num)
+{
+	CRYPTO_ofb128_encrypt(in,out,length,key,ivec,num,(block128_f)AES_encrypt);
+}

jni/openssl/crypto/aes/aes_wrap.c

@@ -0,0 +1,259 @@
+/* crypto/aes/aes_wrap.c */
+/* Written by Dr Stephen N Henson (steve@openssl.org) for the OpenSSL
+ * project.
+ */
+/* ====================================================================
+ * Copyright (c) 2008 The OpenSSL Project.  All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer. 
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *
+ * 3. All advertising materials mentioning features or use of this
+ *    software must display the following acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit. (http://www.OpenSSL.org/)"
+ *
+ * 4. The names "OpenSSL Toolkit" and "OpenSSL Project" must not be used to
+ *    endorse or promote products derived from this software without
+ *    prior written permission. For written permission, please contact
+ *    licensing@OpenSSL.org.
+ *
+ * 5. Products derived from this software may not be called "OpenSSL"
+ *    nor may "OpenSSL" appear in their names without prior written
+ *    permission of the OpenSSL Project.
+ *
+ * 6. Redistributions of any form whatsoever must retain the following
+ *    acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit (http://www.OpenSSL.org/)"
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE OpenSSL PROJECT ``AS IS'' AND ANY
+ * EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE OpenSSL PROJECT OR
+ * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ * ====================================================================
+ */
+
+#include "cryptlib.h"
+#include <openssl/aes.h>
+#include <openssl/bio.h>
+
+static const unsigned char default_iv[] = {
+  0xA6, 0xA6, 0xA6, 0xA6, 0xA6, 0xA6, 0xA6, 0xA6,
+};
+
+int AES_wrap_key(AES_KEY *key, const unsigned char *iv,
+		unsigned char *out,
+		const unsigned char *in, unsigned int inlen)
+	{
+	unsigned char *A, B[16], *R;
+	unsigned int i, j, t;
+	if ((inlen & 0x7) || (inlen < 8))
+		return -1;
+	A = B;
+	t = 1;
+	memcpy(out + 8, in, inlen);
+	if (!iv)
+		iv = default_iv;
+
+	memcpy(A, iv, 8);
+
+	for (j = 0; j < 6; j++)
+		{
+		R = out + 8;
+		for (i = 0; i < inlen; i += 8, t++, R += 8)
+			{
+			memcpy(B + 8, R, 8);
+			AES_encrypt(B, B, key);
+			A[7] ^= (unsigned char)(t & 0xff);
+			if (t > 0xff)	
+				{
+				A[6] ^= (unsigned char)((t >> 8) & 0xff);
+				A[5] ^= (unsigned char)((t >> 16) & 0xff);
+				A[4] ^= (unsigned char)((t >> 24) & 0xff);
+				}
+			memcpy(R, B + 8, 8);
+			}
+		}
+	memcpy(out, A, 8);
+	return inlen + 8;
+	}
+
+int AES_unwrap_key(AES_KEY *key, const unsigned char *iv,
+		unsigned char *out,
+		const unsigned char *in, unsigned int inlen)
+	{
+	unsigned char *A, B[16], *R;
+	unsigned int i, j, t;
+	inlen -= 8;
+	if (inlen & 0x7)
+		return -1;
+	if (inlen < 8)
+		return -1;
+	A = B;
+	t =  6 * (inlen >> 3);
+	memcpy(A, in, 8);
+	memcpy(out, in + 8, inlen);
+	for (j = 0; j < 6; j++)
+		{
+		R = out + inlen - 8;
+		for (i = 0; i < inlen; i += 8, t--, R -= 8)
+			{
+			A[7] ^= (unsigned char)(t & 0xff);
+			if (t > 0xff)	
+				{
+				A[6] ^= (unsigned char)((t >> 8) & 0xff);
+				A[5] ^= (unsigned char)((t >> 16) & 0xff);
+				A[4] ^= (unsigned char)((t >> 24) & 0xff);
+				}
+			memcpy(B + 8, R, 8);
+			AES_decrypt(B, B, key);
+			memcpy(R, B + 8, 8);
+			}
+		}
+	if (!iv)
+		iv = default_iv;
+	if (memcmp(A, iv, 8))
+		{
+		OPENSSL_cleanse(out, inlen);
+		return 0;
+		}
+	return inlen;
+	}
+
+#ifdef AES_WRAP_TEST
+
+int AES_wrap_unwrap_test(const unsigned char *kek, int keybits,
+			 const unsigned char *iv,
+			 const unsigned char *eout,
+			 const unsigned char *key, int keylen)
+	{
+	unsigned char *otmp = NULL, *ptmp = NULL;
+	int r, ret = 0;
+	AES_KEY wctx;
+	otmp = OPENSSL_malloc(keylen + 8);
+	ptmp = OPENSSL_malloc(keylen);
+	if (!otmp || !ptmp)
+		return 0;
+	if (AES_set_encrypt_key(kek, keybits, &wctx))
+		goto err;
+	r = AES_wrap_key(&wctx, iv, otmp, key, keylen);
+	if (r <= 0)
+		goto err;
+
+	if (eout && memcmp(eout, otmp, keylen))
+		goto err;
+		
+	if (AES_set_decrypt_key(kek, keybits, &wctx))
+		goto err;
+	r = AES_unwrap_key(&wctx, iv, ptmp, otmp, r);
+
+	if (memcmp(key, ptmp, keylen))
+		goto err;
+
+	ret = 1;
+
+	err:
+	if (otmp)
+		OPENSSL_free(otmp);
+	if (ptmp)
+		OPENSSL_free(ptmp);
+
+	return ret;
+
+	}
+
+
+
+int main(int argc, char **argv)
+{
+
+static const unsigned char kek[] = {
+  0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
+  0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
+  0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17,
+  0x18, 0x19, 0x1a, 0x1b, 0x1c, 0x1d, 0x1e, 0x1f
+};
+
+static const unsigned char key[] = {
+  0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77,
+  0x88, 0x99, 0xaa, 0xbb, 0xcc, 0xdd, 0xee, 0xff,
+  0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
+  0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f
+};
+
+static const unsigned char e1[] = {
+  0x1f, 0xa6, 0x8b, 0x0a, 0x81, 0x12, 0xb4, 0x47,
+  0xae, 0xf3, 0x4b, 0xd8, 0xfb, 0x5a, 0x7b, 0x82,
+  0x9d, 0x3e, 0x86, 0x23, 0x71, 0xd2, 0xcf, 0xe5
+};
+
+static const unsigned char e2[] = {
+  0x96, 0x77, 0x8b, 0x25, 0xae, 0x6c, 0xa4, 0x35,
+  0xf9, 0x2b, 0x5b, 0x97, 0xc0, 0x50, 0xae, 0xd2,
+  0x46, 0x8a, 0xb8, 0xa1, 0x7a, 0xd8, 0x4e, 0x5d
+};
+
+static const unsigned char e3[] = {
+  0x64, 0xe8, 0xc3, 0xf9, 0xce, 0x0f, 0x5b, 0xa2,
+  0x63, 0xe9, 0x77, 0x79, 0x05, 0x81, 0x8a, 0x2a,
+  0x93, 0xc8, 0x19, 0x1e, 0x7d, 0x6e, 0x8a, 0xe7
+};
+
+static const unsigned char e4[] = {
+  0x03, 0x1d, 0x33, 0x26, 0x4e, 0x15, 0xd3, 0x32,
+  0x68, 0xf2, 0x4e, 0xc2, 0x60, 0x74, 0x3e, 0xdc,
+  0xe1, 0xc6, 0xc7, 0xdd, 0xee, 0x72, 0x5a, 0x93,
+  0x6b, 0xa8, 0x14, 0x91, 0x5c, 0x67, 0x62, 0xd2
+};
+
+static const unsigned char e5[] = {
+  0xa8, 0xf9, 0xbc, 0x16, 0x12, 0xc6, 0x8b, 0x3f,
+  0xf6, 0xe6, 0xf4, 0xfb, 0xe3, 0x0e, 0x71, 0xe4,
+  0x76, 0x9c, 0x8b, 0x80, 0xa3, 0x2c, 0xb8, 0x95,
+  0x8c, 0xd5, 0xd1, 0x7d, 0x6b, 0x25, 0x4d, 0xa1
+};
+
+static const unsigned char e6[] = {
+  0x28, 0xc9, 0xf4, 0x04, 0xc4, 0xb8, 0x10, 0xf4,
+  0xcb, 0xcc, 0xb3, 0x5c, 0xfb, 0x87, 0xf8, 0x26,
+  0x3f, 0x57, 0x86, 0xe2, 0xd8, 0x0e, 0xd3, 0x26,
+  0xcb, 0xc7, 0xf0, 0xe7, 0x1a, 0x99, 0xf4, 0x3b,
+  0xfb, 0x98, 0x8b, 0x9b, 0x7a, 0x02, 0xdd, 0x21
+};
+
+	AES_KEY wctx, xctx;
+	int ret;
+	ret = AES_wrap_unwrap_test(kek, 128, NULL, e1, key, 16);
+	fprintf(stderr, "Key test result %d\n", ret);
+	ret = AES_wrap_unwrap_test(kek, 192, NULL, e2, key, 16);
+	fprintf(stderr, "Key test result %d\n", ret);
+	ret = AES_wrap_unwrap_test(kek, 256, NULL, e3, key, 16);
+	fprintf(stderr, "Key test result %d\n", ret);
+	ret = AES_wrap_unwrap_test(kek, 192, NULL, e4, key, 24);
+	fprintf(stderr, "Key test result %d\n", ret);
+	ret = AES_wrap_unwrap_test(kek, 256, NULL, e5, key, 24);
+	fprintf(stderr, "Key test result %d\n", ret);
+	ret = AES_wrap_unwrap_test(kek, 256, NULL, e6, key, 32);
+	fprintf(stderr, "Key test result %d\n", ret);
+}
+	
+	
+#endif

jni/openssl/crypto/aes/asm/aesv8-armx-64.S

@@ -0,0 +1,761 @@
+#include "arm_arch.h"
+
+#if __ARM_ARCH__>=7
+.text
+.arch	armv8-a+crypto
+.align	5
+rcon:
+.long	0x01,0x01,0x01,0x01
+.long	0x0c0f0e0d,0x0c0f0e0d,0x0c0f0e0d,0x0c0f0e0d	// rotate-n-splat
+.long	0x1b,0x1b,0x1b,0x1b
+
+.globl	aes_v8_set_encrypt_key
+.type	aes_v8_set_encrypt_key,%function
+.align	5
+aes_v8_set_encrypt_key:
+.Lenc_key:
+	stp	x29,x30,[sp,#-16]!
+	add	x29,sp,#0
+	adr	x3,rcon
+	cmp	w1,#192
+
+	eor	v0.16b,v0.16b,v0.16b
+	ld1	{v3.16b},[x0],#16
+	mov	w1,#8		// reuse w1
+	ld1	{v1.4s,v2.4s},[x3],#32
+
+	b.lt	.Loop128
+	b.eq	.L192
+	b	.L256
+
+.align	4
+.Loop128:
+	tbl	v6.16b,{v3.16b},v2.16b
+	ext	v5.16b,v0.16b,v3.16b,#12
+	st1	{v3.4s},[x2],#16
+	aese	v6.16b,v0.16b
+	subs	w1,w1,#1
+
+	eor	v3.16b,v3.16b,v5.16b
+	ext	v5.16b,v0.16b,v5.16b,#12
+	eor	v3.16b,v3.16b,v5.16b
+	ext	v5.16b,v0.16b,v5.16b,#12
+	 eor	v6.16b,v6.16b,v1.16b
+	eor	v3.16b,v3.16b,v5.16b
+	shl	v1.16b,v1.16b,#1
+	eor	v3.16b,v3.16b,v6.16b
+	b.ne	.Loop128
+
+	ld1	{v1.4s},[x3]
+
+	tbl	v6.16b,{v3.16b},v2.16b
+	ext	v5.16b,v0.16b,v3.16b,#12
+	st1	{v3.4s},[x2],#16
+	aese	v6.16b,v0.16b
+
+	eor	v3.16b,v3.16b,v5.16b
+	ext	v5.16b,v0.16b,v5.16b,#12
+	eor	v3.16b,v3.16b,v5.16b
+	ext	v5.16b,v0.16b,v5.16b,#12
+	 eor	v6.16b,v6.16b,v1.16b
+	eor	v3.16b,v3.16b,v5.16b
+	shl	v1.16b,v1.16b,#1
+	eor	v3.16b,v3.16b,v6.16b
+
+	tbl	v6.16b,{v3.16b},v2.16b
+	ext	v5.16b,v0.16b,v3.16b,#12
+	st1	{v3.4s},[x2],#16
+	aese	v6.16b,v0.16b
+
+	eor	v3.16b,v3.16b,v5.16b
+	ext	v5.16b,v0.16b,v5.16b,#12
+	eor	v3.16b,v3.16b,v5.16b
+	ext	v5.16b,v0.16b,v5.16b,#12
+	 eor	v6.16b,v6.16b,v1.16b
+	eor	v3.16b,v3.16b,v5.16b
+	eor	v3.16b,v3.16b,v6.16b
+	st1	{v3.4s},[x2]
+	add	x2,x2,#0x50
+
+	mov	w12,#10
+	b	.Ldone
+
+.align	4
+.L192:
+	ld1	{v4.8b},[x0],#8
+	movi	v6.16b,#8			// borrow v6.16b
+	st1	{v3.4s},[x2],#16
+	sub	v2.16b,v2.16b,v6.16b	// adjust the mask
+
+.Loop192:
+	tbl	v6.16b,{v4.16b},v2.16b
+	ext	v5.16b,v0.16b,v3.16b,#12
+	st1	{v4.8b},[x2],#8
+	aese	v6.16b,v0.16b
+	subs	w1,w1,#1
+
+	eor	v3.16b,v3.16b,v5.16b
+	ext	v5.16b,v0.16b,v5.16b,#12
+	eor	v3.16b,v3.16b,v5.16b
+	ext	v5.16b,v0.16b,v5.16b,#12
+	eor	v3.16b,v3.16b,v5.16b
+
+	dup	v5.4s,v3.s[3]
+	eor	v5.16b,v5.16b,v4.16b
+	 eor	v6.16b,v6.16b,v1.16b
+	ext	v4.16b,v0.16b,v4.16b,#12
+	shl	v1.16b,v1.16b,#1
+	eor	v4.16b,v4.16b,v5.16b
+	eor	v3.16b,v3.16b,v6.16b
+	eor	v4.16b,v4.16b,v6.16b
+	st1	{v3.4s},[x2],#16
+	b.ne	.Loop192
+
+	mov	w12,#12
+	add	x2,x2,#0x20
+	b	.Ldone
+
+.align	4
+.L256:
+	ld1	{v4.16b},[x0]
+	mov	w1,#7
+	mov	w12,#14
+	st1	{v3.4s},[x2],#16
+
+.Loop256:
+	tbl	v6.16b,{v4.16b},v2.16b
+	ext	v5.16b,v0.16b,v3.16b,#12
+	st1	{v4.4s},[x2],#16
+	aese	v6.16b,v0.16b
+	subs	w1,w1,#1
+
+	eor	v3.16b,v3.16b,v5.16b
+	ext	v5.16b,v0.16b,v5.16b,#12
+	eor	v3.16b,v3.16b,v5.16b
+	ext	v5.16b,v0.16b,v5.16b,#12
+	 eor	v6.16b,v6.16b,v1.16b
+	eor	v3.16b,v3.16b,v5.16b
+	shl	v1.16b,v1.16b,#1
+	eor	v3.16b,v3.16b,v6.16b
+	st1	{v3.4s},[x2],#16
+	b.eq	.Ldone
+
+	dup	v6.4s,v3.s[3]		// just splat
+	ext	v5.16b,v0.16b,v4.16b,#12
+	aese	v6.16b,v0.16b
+
+	eor	v4.16b,v4.16b,v5.16b
+	ext	v5.16b,v0.16b,v5.16b,#12
+	eor	v4.16b,v4.16b,v5.16b
+	ext	v5.16b,v0.16b,v5.16b,#12
+	eor	v4.16b,v4.16b,v5.16b
+
+	eor	v4.16b,v4.16b,v6.16b
+	b	.Loop256
+
+.Ldone:
+	str	w12,[x2]
+
+	eor	x0,x0,x0		// return value
+	ldr	x29,[sp],#16
+	ret
+.size	aes_v8_set_encrypt_key,.-aes_v8_set_encrypt_key
+
+.globl	aes_v8_set_decrypt_key
+.type	aes_v8_set_decrypt_key,%function
+.align	5
+aes_v8_set_decrypt_key:
+	stp	x29,x30,[sp,#-16]!
+	add	x29,sp,#0
+	bl	.Lenc_key
+
+	sub	x2,x2,#240		// restore original x2
+	mov	x4,#-16
+	add	x0,x2,x12,lsl#4	// end of key schedule
+
+	ld1	{v0.4s},[x2]
+	ld1	{v1.4s},[x0]
+	st1	{v0.4s},[x0],x4
+	st1	{v1.4s},[x2],#16
+
+.Loop_imc:
+	ld1	{v0.4s},[x2]
+	ld1	{v1.4s},[x0]
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	st1	{v0.4s},[x0],x4
+	st1	{v1.4s},[x2],#16
+	cmp	x0,x2
+	b.hi	.Loop_imc
+
+	ld1	{v0.4s},[x2]
+	aesimc	v0.16b,v0.16b
+	st1	{v0.4s},[x0]
+
+	eor	x0,x0,x0		// return value
+	ldp	x29,x30,[sp],#16
+	ret
+.size	aes_v8_set_decrypt_key,.-aes_v8_set_decrypt_key
+.globl	aes_v8_encrypt
+.type	aes_v8_encrypt,%function
+.align	5
+aes_v8_encrypt:
+	ldr	w3,[x2,#240]
+	ld1	{v0.4s},[x2],#16
+	ld1	{v2.16b},[x0]
+	sub	w3,w3,#2
+	ld1	{v1.4s},[x2],#16
+
+.Loop_enc:
+	aese	v2.16b,v0.16b
+	ld1	{v0.4s},[x2],#16
+	aesmc	v2.16b,v2.16b
+	subs	w3,w3,#2
+	aese	v2.16b,v1.16b
+	ld1	{v1.4s},[x2],#16
+	aesmc	v2.16b,v2.16b
+	b.gt	.Loop_enc
+
+	aese	v2.16b,v0.16b
+	ld1	{v0.4s},[x2]
+	aesmc	v2.16b,v2.16b
+	aese	v2.16b,v1.16b
+	eor	v2.16b,v2.16b,v0.16b
+
+	st1	{v2.16b},[x1]
+	ret
+.size	aes_v8_encrypt,.-aes_v8_encrypt
+.globl	aes_v8_decrypt
+.type	aes_v8_decrypt,%function
+.align	5
+aes_v8_decrypt:
+	ldr	w3,[x2,#240]
+	ld1	{v0.4s},[x2],#16
+	ld1	{v2.16b},[x0]
+	sub	w3,w3,#2
+	ld1	{v1.4s},[x2],#16
+
+.Loop_dec:
+	aesd	v2.16b,v0.16b
+	ld1	{v0.4s},[x2],#16
+	aesimc	v2.16b,v2.16b
+	subs	w3,w3,#2
+	aesd	v2.16b,v1.16b
+	ld1	{v1.4s},[x2],#16
+	aesimc	v2.16b,v2.16b
+	b.gt	.Loop_dec
+
+	aesd	v2.16b,v0.16b
+	ld1	{v0.4s},[x2]
+	aesimc	v2.16b,v2.16b
+	aesd	v2.16b,v1.16b
+	eor	v2.16b,v2.16b,v0.16b
+
+	st1	{v2.16b},[x1]
+	ret
+.size	aes_v8_decrypt,.-aes_v8_decrypt
+.globl	aes_v8_cbc_encrypt
+.type	aes_v8_cbc_encrypt,%function
+.align	5
+aes_v8_cbc_encrypt:
+	stp	x29,x30,[sp,#-16]!
+	add	x29,sp,#0
+	subs	x2,x2,#16
+	mov	x8,#16
+	b.lo	.Lcbc_abort
+	csel	x8,xzr,x8,eq
+
+	cmp	w5,#0			// en- or decrypting?
+	ldr	w5,[x3,#240]
+	and	x2,x2,#-16
+	ld1	{v6.16b},[x4]
+	ld1	{v0.16b},[x0],x8
+
+	ld1	{v16.4s-v17.4s},[x3]		// load key schedule...
+	sub	w5,w5,#6
+	add	x7,x3,x5,lsl#4	// pointer to last 7 round keys
+	sub	w5,w5,#2
+	ld1	{v18.4s-v19.4s},[x7],#32
+	ld1	{v20.4s-v21.4s},[x7],#32
+	ld1	{v22.4s-v23.4s},[x7],#32
+	ld1	{v7.4s},[x7]
+
+	add	x7,x3,#32
+	mov	w6,w5
+	b.eq	.Lcbc_dec
+
+	cmp	w5,#2
+	eor	v0.16b,v0.16b,v6.16b
+	eor	v5.16b,v16.16b,v7.16b
+	b.eq	.Lcbc_enc128
+
+.Loop_cbc_enc:
+	aese	v0.16b,v16.16b
+	ld1	{v16.4s},[x7],#16
+	aesmc	v0.16b,v0.16b
+	subs	w6,w6,#2
+	aese	v0.16b,v17.16b
+	ld1	{v17.4s},[x7],#16
+	aesmc	v0.16b,v0.16b
+	b.gt	.Loop_cbc_enc
+
+	aese	v0.16b,v16.16b
+	aesmc	v0.16b,v0.16b
+	 subs	x2,x2,#16
+	aese	v0.16b,v17.16b
+	aesmc	v0.16b,v0.16b
+	 csel	x8,xzr,x8,eq
+	aese	v0.16b,v18.16b
+	aesmc	v0.16b,v0.16b
+	 add	x7,x3,#16
+	aese	v0.16b,v19.16b
+	aesmc	v0.16b,v0.16b
+	 ld1	{v16.16b},[x0],x8
+	aese	v0.16b,v20.16b
+	aesmc	v0.16b,v0.16b
+	 eor	v16.16b,v16.16b,v5.16b
+	aese	v0.16b,v21.16b
+	aesmc	v0.16b,v0.16b
+	 ld1 {v17.4s},[x7],#16	// re-pre-load rndkey[1]
+	aese	v0.16b,v22.16b
+	aesmc	v0.16b,v0.16b
+	aese	v0.16b,v23.16b
+
+	 mov	w6,w5
+	eor	v6.16b,v0.16b,v7.16b
+	st1	{v6.16b},[x1],#16
+	b.hs	.Loop_cbc_enc
+
+	b	.Lcbc_done
+
+.align	5
+.Lcbc_enc128:
+	ld1	{v2.4s-v3.4s},[x7]
+	aese	v0.16b,v16.16b
+	aesmc	v0.16b,v0.16b
+	b	.Lenter_cbc_enc128
+.Loop_cbc_enc128:
+	aese	v0.16b,v16.16b
+	aesmc	v0.16b,v0.16b
+	 st1	{v6.16b},[x1],#16
+.Lenter_cbc_enc128:
+	aese	v0.16b,v17.16b
+	aesmc	v0.16b,v0.16b
+	 subs	x2,x2,#16
+	aese	v0.16b,v2.16b
+	aesmc	v0.16b,v0.16b
+	 csel	x8,xzr,x8,eq
+	aese	v0.16b,v3.16b
+	aesmc	v0.16b,v0.16b
+	aese	v0.16b,v18.16b
+	aesmc	v0.16b,v0.16b
+	aese	v0.16b,v19.16b
+	aesmc	v0.16b,v0.16b
+	 ld1	{v16.16b},[x0],x8
+	aese	v0.16b,v20.16b
+	aesmc	v0.16b,v0.16b
+	aese	v0.16b,v21.16b
+	aesmc	v0.16b,v0.16b
+	aese	v0.16b,v22.16b
+	aesmc	v0.16b,v0.16b
+	 eor	v16.16b,v16.16b,v5.16b
+	aese	v0.16b,v23.16b
+	eor	v6.16b,v0.16b,v7.16b
+	b.hs	.Loop_cbc_enc128
+
+	st1	{v6.16b},[x1],#16
+	b	.Lcbc_done
+
+.align	5
+.Lcbc_dec128:
+	ld1	{v4.4s-v5.4s},[x7]
+	eor	v6.16b,v6.16b,v7.16b
+	eor	v2.16b,v0.16b,v7.16b
+	mov	x12,x8
+
+.Loop2x_cbc_dec128:
+	aesd	v0.16b,v16.16b
+	aesd	v1.16b,v16.16b
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	 subs	x2,x2,#32
+	aesd	v0.16b,v17.16b
+	aesd	v1.16b,v17.16b
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	 csel	x8,xzr,x8,lo
+	aesd	v0.16b,v4.16b
+	aesd	v1.16b,v4.16b
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	 csel	x12,xzr,x12,ls
+	aesd	v0.16b,v5.16b
+	aesd	v1.16b,v5.16b
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	aesd	v0.16b,v18.16b
+	aesd	v1.16b,v18.16b
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	aesd	v0.16b,v19.16b
+	aesd	v1.16b,v19.16b
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	aesd	v0.16b,v20.16b
+	aesd	v1.16b,v20.16b
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	aesd	v0.16b,v21.16b
+	aesd	v1.16b,v21.16b
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	aesd	v0.16b,v22.16b
+	aesd	v1.16b,v22.16b
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	aesd	v0.16b,v23.16b
+	aesd	v1.16b,v23.16b
+
+	eor	v6.16b,v6.16b,v0.16b
+	ld1	{v0.16b},[x0],x8
+	eor	v2.16b,v2.16b,v1.16b
+	ld1	{v1.16b},[x0],x12
+	st1	{v6.16b},[x1],#16
+	eor	v6.16b,v3.16b,v7.16b
+	st1	{v2.16b},[x1],#16
+	eor	v2.16b,v0.16b,v7.16b
+	orr	v3.16b,v1.16b,v1.16b
+	b.hs	.Loop2x_cbc_dec128
+
+	adds	x2,x2,#32
+	eor	v6.16b,v6.16b,v7.16b
+	b.eq	.Lcbc_done
+	eor	v2.16b,v2.16b,v7.16b
+	b	.Lcbc_dec_tail
+
+.align	5
+.Lcbc_dec:
+	subs	x2,x2,#16
+	orr	v2.16b,v0.16b,v0.16b
+	b.lo	.Lcbc_dec_tail
+
+	csel	x8,xzr,x8,eq
+	cmp	w5,#2
+	ld1	{v1.16b},[x0],x8
+	orr	v3.16b,v1.16b,v1.16b
+	b.eq	.Lcbc_dec128
+
+.Loop2x_cbc_dec:
+	aesd	v0.16b,v16.16b
+	aesd	v1.16b,v16.16b
+	ld1	{v16.4s},[x7],#16
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	subs	w6,w6,#2
+	aesd	v0.16b,v17.16b
+	aesd	v1.16b,v17.16b
+	ld1	{v17.4s},[x7],#16
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	b.gt	.Loop2x_cbc_dec
+
+	aesd	v0.16b,v16.16b
+	aesd	v1.16b,v16.16b
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	 eor	v4.16b,v6.16b,v7.16b
+	 eor	v5.16b,v2.16b,v7.16b
+	aesd	v0.16b,v17.16b
+	aesd	v1.16b,v17.16b
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	 orr	v6.16b,v3.16b,v3.16b
+	 subs	x2,x2,#32
+	aesd	v0.16b,v18.16b
+	aesd	v1.16b,v18.16b
+	aesimc	v0.16b,v0.16b
+	 csel	x8,xzr,x8,lo
+	aesimc	v1.16b,v1.16b
+	 mov	x7,x3
+	aesd	v0.16b,v19.16b
+	aesd	v1.16b,v19.16b
+	aesimc	v0.16b,v0.16b
+	 ld1	{v2.16b},[x0],x8
+	aesimc	v1.16b,v1.16b
+	 csel	x8,xzr,x8,ls
+	aesd	v0.16b,v20.16b
+	aesd	v1.16b,v20.16b
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	 ld1	{v3.16b},[x0],x8
+	aesd	v0.16b,v21.16b
+	aesd	v1.16b,v21.16b
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	 ld1 {v16.4s},[x7],#16	// re-pre-load rndkey[0]
+	aesd	v0.16b,v22.16b
+	aesd	v1.16b,v22.16b
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	 ld1 {v17.4s},[x7],#16	// re-pre-load rndkey[1]
+	aesd	v0.16b,v23.16b
+	aesd	v1.16b,v23.16b
+
+	 mov	w6,w5
+	eor	v4.16b,v4.16b,v0.16b
+	eor	v5.16b,v5.16b,v1.16b
+	 orr	v0.16b,v2.16b,v2.16b
+	st1	{v4.16b},[x1],#16
+	 orr	v1.16b,v3.16b,v3.16b
+	st1	{v5.16b},[x1],#16
+	b.hs	.Loop2x_cbc_dec
+
+	adds	x2,x2,#32
+	b.eq	.Lcbc_done
+
+.Lcbc_dec_tail:
+	aesd	v0.16b,v16.16b
+	ld1	{v16.4s},[x7],#16
+	aesimc	v0.16b,v0.16b
+	subs	w6,w6,#2
+	aesd	v0.16b,v17.16b
+	ld1	{v17.4s},[x7],#16
+	aesimc	v0.16b,v0.16b
+	b.gt	.Lcbc_dec_tail
+
+	aesd	v0.16b,v16.16b
+	aesimc	v0.16b,v0.16b
+	aesd	v0.16b,v17.16b
+	aesimc	v0.16b,v0.16b
+	 eor	v4.16b,v6.16b,v7.16b
+	aesd	v0.16b,v18.16b
+	aesimc	v0.16b,v0.16b
+	 orr	v6.16b,v2.16b,v2.16b
+	aesd	v0.16b,v19.16b
+	aesimc	v0.16b,v0.16b
+	aesd	v0.16b,v20.16b
+	aesimc	v0.16b,v0.16b
+	aesd	v0.16b,v21.16b
+	aesimc	v0.16b,v0.16b
+	aesd	v0.16b,v22.16b
+	aesimc	v0.16b,v0.16b
+	aesd	v0.16b,v23.16b
+
+	eor	v4.16b,v4.16b,v0.16b
+	st1	{v4.16b},[x1],#16
+
+.Lcbc_done:
+	st1	{v6.16b},[x4]
+.Lcbc_abort:
+	ldr	x29,[sp],#16
+	ret
+.size	aes_v8_cbc_encrypt,.-aes_v8_cbc_encrypt
+.globl	aes_v8_ctr32_encrypt_blocks
+.type	aes_v8_ctr32_encrypt_blocks,%function
+.align	5
+aes_v8_ctr32_encrypt_blocks:
+	stp		x29,x30,[sp,#-16]!
+	add		x29,sp,#0
+	ldr		w5,[x3,#240]
+
+	ldr		w8, [x4, #12]
+	ld1		{v0.4s},[x4]
+
+	ld1		{v16.4s-v17.4s},[x3]		// load key schedule...
+	sub		w5,w5,#6
+	add		x7,x3,x5,lsl#4	// pointer to last 7 round keys
+	sub		w5,w5,#2
+	ld1		{v18.4s-v19.4s},[x7],#32
+	ld1		{v20.4s-v21.4s},[x7],#32
+	ld1		{v22.4s-v23.4s},[x7],#32
+	ld1		{v7.4s},[x7]
+
+	add		x7,x3,#32
+	mov		w6,w5
+
+	subs		x2,x2,#2
+	b.lo		.Lctr32_tail
+
+#ifndef __ARMEB__
+	rev		w8, w8
+#endif
+	orr		v1.16b,v0.16b,v0.16b
+	add		w8, w8, #1
+	orr		v6.16b,v0.16b,v0.16b
+	rev		w10, w8
+	cmp		w5,#2
+	mov		v1.s[3],w10
+	b.eq		.Lctr32_128
+
+.Loop2x_ctr32:
+	aese		v0.16b,v16.16b
+	aese		v1.16b,v16.16b
+	ld1		{v16.4s},[x7],#16
+	aesmc		v0.16b,v0.16b
+	aesmc		v1.16b,v1.16b
+	subs		w6,w6,#2
+	aese		v0.16b,v17.16b
+	aese		v1.16b,v17.16b
+	ld1		{v17.4s},[x7],#16
+	aesmc		v0.16b,v0.16b
+	aesmc		v1.16b,v1.16b
+	b.gt		.Loop2x_ctr32
+
+	aese		v0.16b,v16.16b
+	aese		v1.16b,v16.16b
+	aesmc		v4.16b,v0.16b
+	 orr		v0.16b,v6.16b,v6.16b
+	aesmc		v5.16b,v1.16b
+	 orr		v1.16b,v6.16b,v6.16b
+	aese		v4.16b,v17.16b
+	aese		v5.16b,v17.16b
+	 ld1		{v2.16b},[x0],#16
+	aesmc		v4.16b,v4.16b
+	 ld1		{v3.16b},[x0],#16
+	aesmc		v5.16b,v5.16b
+	 add		w8,w8,#1
+	aese		v4.16b,v18.16b
+	aese		v5.16b,v18.16b
+	 rev		w9,w8
+	aesmc		v4.16b,v4.16b
+	aesmc		v5.16b,v5.16b
+	 add		w8,w8,#1
+	aese		v4.16b,v19.16b
+	aese		v5.16b,v19.16b
+	 eor		v2.16b,v2.16b,v7.16b
+	 rev		w10,w8
+	aesmc		v4.16b,v4.16b
+	aesmc		v5.16b,v5.16b
+	 eor		v3.16b,v3.16b,v7.16b
+	 mov		x7,x3
+	aese		v4.16b,v20.16b
+	aese		v5.16b,v20.16b
+	 subs		x2,x2,#2
+	aesmc		v4.16b,v4.16b
+	aesmc		v5.16b,v5.16b
+	 ld1	 {v16.4s-v17.4s},[x7],#32	// re-pre-load rndkey[0-1]
+	aese		v4.16b,v21.16b
+	aese		v5.16b,v21.16b
+	aesmc		v4.16b,v4.16b
+	aesmc		v5.16b,v5.16b
+	aese		v4.16b,v22.16b
+	aese		v5.16b,v22.16b
+	 mov	v0.s[3], w9
+	aesmc		v4.16b,v4.16b
+	 mov	v1.s[3], w10
+	aesmc		v5.16b,v5.16b
+	aese		v4.16b,v23.16b
+	aese		v5.16b,v23.16b
+
+	 mov		w6,w5
+	eor		v2.16b,v2.16b,v4.16b
+	eor		v3.16b,v3.16b,v5.16b
+	st1		{v2.16b},[x1],#16
+	st1		{v3.16b},[x1],#16
+	b.hs		.Loop2x_ctr32
+
+	adds		x2,x2,#2
+	b.eq		.Lctr32_done
+	b		.Lctr32_tail
+
+.Lctr32_128:
+	ld1		{v4.4s-v5.4s},[x7]
+
+.Loop2x_ctr32_128:
+	aese		v0.16b,v16.16b
+	aese		v1.16b,v16.16b
+	aesmc		v0.16b,v0.16b
+	 ld1		{v2.16b},[x0],#16
+	aesmc		v1.16b,v1.16b
+	 ld1		{v3.16b},[x0],#16
+	aese		v0.16b,v17.16b
+	aese		v1.16b,v17.16b
+	 add		w8,w8,#1
+	aesmc		v0.16b,v0.16b
+	aesmc		v1.16b,v1.16b
+	 rev		w9,w8
+	aese		v0.16b,v4.16b
+	aese		v1.16b,v4.16b
+	 add		w8,w8,#1
+	aesmc		v0.16b,v0.16b
+	aesmc		v1.16b,v1.16b
+	 rev		w10,w8
+	aese		v0.16b,v5.16b
+	aese		v1.16b,v5.16b
+	 subs		x2,x2,#2
+	aesmc		v0.16b,v0.16b
+	aesmc		v1.16b,v1.16b
+	aese		v0.16b,v18.16b
+	aese		v1.16b,v18.16b
+	aesmc		v0.16b,v0.16b
+	aesmc		v1.16b,v1.16b
+	aese		v0.16b,v19.16b
+	aese		v1.16b,v19.16b
+	aesmc		v0.16b,v0.16b
+	aesmc		v1.16b,v1.16b
+	aese		v0.16b,v20.16b
+	aese		v1.16b,v20.16b
+	aesmc		v0.16b,v0.16b
+	aesmc		v1.16b,v1.16b
+	aese		v0.16b,v21.16b
+	aese		v1.16b,v21.16b
+	aesmc		v0.16b,v0.16b
+	aesmc		v1.16b,v1.16b
+	aese		v0.16b,v22.16b
+	aese		v1.16b,v22.16b
+	aesmc		v0.16b,v0.16b
+	aesmc		v1.16b,v1.16b
+	 eor		v2.16b,v2.16b,v7.16b
+	aese		v0.16b,v23.16b
+	 eor		v3.16b,v3.16b,v7.16b
+	aese		v1.16b,v23.16b
+
+	eor		v2.16b,v2.16b,v0.16b
+	orr		v0.16b,v6.16b,v6.16b
+	eor		v3.16b,v3.16b,v1.16b
+	orr		v1.16b,v6.16b,v6.16b
+	st1		{v2.16b},[x1],#16
+	mov		v0.s[3], w9
+	st1		{v3.16b},[x1],#16
+	mov		v1.s[3], w10
+	b.hs		.Loop2x_ctr32_128
+
+	adds		x2,x2,#2
+	b.eq		.Lctr32_done
+
+.Lctr32_tail:
+	aese		v0.16b,v16.16b
+	ld1		{v16.4s},[x7],#16
+	aesmc		v0.16b,v0.16b
+	subs		w6,w6,#2
+	aese		v0.16b,v17.16b
+	ld1		{v17.4s},[x7],#16
+	aesmc		v0.16b,v0.16b
+	b.gt		.Lctr32_tail
+
+	aese		v0.16b,v16.16b
+	aesmc		v0.16b,v0.16b
+	aese		v0.16b,v17.16b
+	aesmc		v0.16b,v0.16b
+	 ld1		{v2.16b},[x0]
+	aese		v0.16b,v18.16b
+	aesmc		v0.16b,v0.16b
+	aese		v0.16b,v19.16b
+	aesmc		v0.16b,v0.16b
+	aese		v0.16b,v20.16b
+	aesmc		v0.16b,v0.16b
+	aese		v0.16b,v21.16b
+	aesmc		v0.16b,v0.16b
+	aese		v0.16b,v22.16b
+	aesmc		v0.16b,v0.16b
+	 eor		v2.16b,v2.16b,v7.16b
+	aese		v0.16b,v23.16b
+
+	eor		v2.16b,v2.16b,v0.16b
+	st1		{v2.16b},[x1]
+
+.Lctr32_done:
+	ldr		x29,[sp],#16
+	ret
+.size	aes_v8_ctr32_encrypt_blocks,.-aes_v8_ctr32_encrypt_blocks
+#endif

jni/openssl/crypto/aes/asm/aesv8-armx.S

@@ -0,0 +1,767 @@
+#include "arm_arch.h"
+
+#if __ARM_ARCH__>=7
+.text
+.fpu	neon
+.code	32
+.align	5
+rcon:
+.long	0x01,0x01,0x01,0x01
+.long	0x0c0f0e0d,0x0c0f0e0d,0x0c0f0e0d,0x0c0f0e0d	@ rotate-n-splat
+.long	0x1b,0x1b,0x1b,0x1b
+
+.globl	aes_v8_set_encrypt_key
+.type	aes_v8_set_encrypt_key,%function
+.align	5
+aes_v8_set_encrypt_key:
+.Lenc_key:
+	adr	r3,rcon
+	cmp	r1,#192
+
+	veor	q0,q0,q0
+	vld1.8	{q3},[r0]!
+	mov	r1,#8		@ reuse r1
+	vld1.32	{q1,q2},[r3]!
+
+	blt	.Loop128
+	beq	.L192
+	b	.L256
+
+.align	4
+.Loop128:
+	vtbl.8	d20,{q3},d4
+	vtbl.8	d21,{q3},d5
+	vext.8	q9,q0,q3,#12
+	vst1.32	{q3},[r2]!
+	.byte	0x00,0x43,0xf0,0xf3	@ aese q10,q0
+	subs	r1,r1,#1
+
+	veor	q3,q3,q9
+	vext.8	q9,q0,q9,#12
+	veor	q3,q3,q9
+	vext.8	q9,q0,q9,#12
+	 veor	q10,q10,q1
+	veor	q3,q3,q9
+	vshl.u8	q1,q1,#1
+	veor	q3,q3,q10
+	bne	.Loop128
+
+	vld1.32	{q1},[r3]
+
+	vtbl.8	d20,{q3},d4
+	vtbl.8	d21,{q3},d5
+	vext.8	q9,q0,q3,#12
+	vst1.32	{q3},[r2]!
+	.byte	0x00,0x43,0xf0,0xf3	@ aese q10,q0
+
+	veor	q3,q3,q9
+	vext.8	q9,q0,q9,#12
+	veor	q3,q3,q9
+	vext.8	q9,q0,q9,#12
+	 veor	q10,q10,q1
+	veor	q3,q3,q9
+	vshl.u8	q1,q1,#1
+	veor	q3,q3,q10
+
+	vtbl.8	d20,{q3},d4
+	vtbl.8	d21,{q3},d5
+	vext.8	q9,q0,q3,#12
+	vst1.32	{q3},[r2]!
+	.byte	0x00,0x43,0xf0,0xf3	@ aese q10,q0
+
+	veor	q3,q3,q9
+	vext.8	q9,q0,q9,#12
+	veor	q3,q3,q9
+	vext.8	q9,q0,q9,#12
+	 veor	q10,q10,q1
+	veor	q3,q3,q9
+	veor	q3,q3,q10
+	vst1.32	{q3},[r2]
+	add	r2,r2,#0x50
+
+	mov	r12,#10
+	b	.Ldone
+
+.align	4
+.L192:
+	vld1.8	{d16},[r0]!
+	vmov.i8	q10,#8			@ borrow q10
+	vst1.32	{q3},[r2]!
+	vsub.i8	q2,q2,q10	@ adjust the mask
+
+.Loop192:
+	vtbl.8	d20,{q8},d4
+	vtbl.8	d21,{q8},d5
+	vext.8	q9,q0,q3,#12
+	vst1.32	{d16},[r2]!
+	.byte	0x00,0x43,0xf0,0xf3	@ aese q10,q0
+	subs	r1,r1,#1
+
+	veor	q3,q3,q9
+	vext.8	q9,q0,q9,#12
+	veor	q3,q3,q9
+	vext.8	q9,q0,q9,#12
+	veor	q3,q3,q9
+
+	vdup.32	q9,d7[1]
+	veor	q9,q9,q8
+	 veor	q10,q10,q1
+	vext.8	q8,q0,q8,#12
+	vshl.u8	q1,q1,#1
+	veor	q8,q8,q9
+	veor	q3,q3,q10
+	veor	q8,q8,q10
+	vst1.32	{q3},[r2]!
+	bne	.Loop192
+
+	mov	r12,#12
+	add	r2,r2,#0x20
+	b	.Ldone
+
+.align	4
+.L256:
+	vld1.8	{q8},[r0]
+	mov	r1,#7
+	mov	r12,#14
+	vst1.32	{q3},[r2]!
+
+.Loop256:
+	vtbl.8	d20,{q8},d4
+	vtbl.8	d21,{q8},d5
+	vext.8	q9,q0,q3,#12
+	vst1.32	{q8},[r2]!
+	.byte	0x00,0x43,0xf0,0xf3	@ aese q10,q0
+	subs	r1,r1,#1
+
+	veor	q3,q3,q9
+	vext.8	q9,q0,q9,#12
+	veor	q3,q3,q9
+	vext.8	q9,q0,q9,#12
+	 veor	q10,q10,q1
+	veor	q3,q3,q9
+	vshl.u8	q1,q1,#1
+	veor	q3,q3,q10
+	vst1.32	{q3},[r2]!
+	beq	.Ldone
+
+	vdup.32	q10,d7[1]
+	vext.8	q9,q0,q8,#12
+	.byte	0x00,0x43,0xf0,0xf3	@ aese q10,q0
+
+	veor	q8,q8,q9
+	vext.8	q9,q0,q9,#12
+	veor	q8,q8,q9
+	vext.8	q9,q0,q9,#12
+	veor	q8,q8,q9
+
+	veor	q8,q8,q10
+	b	.Loop256
+
+.Ldone:
+	str	r12,[r2]
+
+	eor	r0,r0,r0		@ return value
+	
+	bx	lr
+.size	aes_v8_set_encrypt_key,.-aes_v8_set_encrypt_key
+
+.globl	aes_v8_set_decrypt_key
+.type	aes_v8_set_decrypt_key,%function
+.align	5
+aes_v8_set_decrypt_key:
+	stmdb	sp!,{r4,lr}
+	bl	.Lenc_key
+
+	sub	r2,r2,#240		@ restore original r2
+	mov	r4,#-16
+	add	r0,r2,r12,lsl#4	@ end of key schedule
+
+	vld1.32	{q0},[r2]
+	vld1.32	{q1},[r0]
+	vst1.32	{q0},[r0],r4
+	vst1.32	{q1},[r2]!
+
+.Loop_imc:
+	vld1.32	{q0},[r2]
+	vld1.32	{q1},[r0]
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	vst1.32	{q0},[r0],r4
+	vst1.32	{q1},[r2]!
+	cmp	r0,r2
+	bhi	.Loop_imc
+
+	vld1.32	{q0},[r2]
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	vst1.32	{q0},[r0]
+
+	eor	r0,r0,r0		@ return value
+	ldmia	sp!,{r4,pc}
+.size	aes_v8_set_decrypt_key,.-aes_v8_set_decrypt_key
+.globl	aes_v8_encrypt
+.type	aes_v8_encrypt,%function
+.align	5
+aes_v8_encrypt:
+	ldr	r3,[r2,#240]
+	vld1.32	{q0},[r2]!
+	vld1.8	{q2},[r0]
+	sub	r3,r3,#2
+	vld1.32	{q1},[r2]!
+
+.Loop_enc:
+	.byte	0x00,0x43,0xb0,0xf3	@ aese q2,q0
+	vld1.32	{q0},[r2]!
+	.byte	0x84,0x43,0xb0,0xf3	@ aesmc q2,q2
+	subs	r3,r3,#2
+	.byte	0x02,0x43,0xb0,0xf3	@ aese q2,q1
+	vld1.32	{q1},[r2]!
+	.byte	0x84,0x43,0xb0,0xf3	@ aesmc q2,q2
+	bgt	.Loop_enc
+
+	.byte	0x00,0x43,0xb0,0xf3	@ aese q2,q0
+	vld1.32	{q0},[r2]
+	.byte	0x84,0x43,0xb0,0xf3	@ aesmc q2,q2
+	.byte	0x02,0x43,0xb0,0xf3	@ aese q2,q1
+	veor	q2,q2,q0
+
+	vst1.8	{q2},[r1]
+	bx	lr
+.size	aes_v8_encrypt,.-aes_v8_encrypt
+.globl	aes_v8_decrypt
+.type	aes_v8_decrypt,%function
+.align	5
+aes_v8_decrypt:
+	ldr	r3,[r2,#240]
+	vld1.32	{q0},[r2]!
+	vld1.8	{q2},[r0]
+	sub	r3,r3,#2
+	vld1.32	{q1},[r2]!
+
+.Loop_dec:
+	.byte	0x40,0x43,0xb0,0xf3	@ aesd q2,q0
+	vld1.32	{q0},[r2]!
+	.byte	0xc4,0x43,0xb0,0xf3	@ aesimc q2,q2
+	subs	r3,r3,#2
+	.byte	0x42,0x43,0xb0,0xf3	@ aesd q2,q1
+	vld1.32	{q1},[r2]!
+	.byte	0xc4,0x43,0xb0,0xf3	@ aesimc q2,q2
+	bgt	.Loop_dec
+
+	.byte	0x40,0x43,0xb0,0xf3	@ aesd q2,q0
+	vld1.32	{q0},[r2]
+	.byte	0xc4,0x43,0xb0,0xf3	@ aesimc q2,q2
+	.byte	0x42,0x43,0xb0,0xf3	@ aesd q2,q1
+	veor	q2,q2,q0
+
+	vst1.8	{q2},[r1]
+	bx	lr
+.size	aes_v8_decrypt,.-aes_v8_decrypt
+.globl	aes_v8_cbc_encrypt
+.type	aes_v8_cbc_encrypt,%function
+.align	5
+aes_v8_cbc_encrypt:
+	mov	ip,sp
+	stmdb	sp!,{r4-r8,lr}
+	vstmdb	sp!,{d8-d15}            @ ABI specification says so
+	ldmia	ip,{r4-r5}		@ load remaining args
+	subs	r2,r2,#16
+	mov	r8,#16
+	blo	.Lcbc_abort
+	moveq	r8,#0
+
+	cmp	r5,#0			@ en- or decrypting?
+	ldr	r5,[r3,#240]
+	and	r2,r2,#-16
+	vld1.8	{q6},[r4]
+	vld1.8	{q0},[r0],r8
+
+	vld1.32	{q8-q9},[r3]		@ load key schedule...
+	sub	r5,r5,#6
+	add	r7,r3,r5,lsl#4	@ pointer to last 7 round keys
+	sub	r5,r5,#2
+	vld1.32	{q10-q11},[r7]!
+	vld1.32	{q12-q13},[r7]!
+	vld1.32	{q14-q15},[r7]!
+	vld1.32	{q7},[r7]
+
+	add	r7,r3,#32
+	mov	r6,r5
+	beq	.Lcbc_dec
+
+	cmp	r5,#2
+	veor	q0,q0,q6
+	veor	q5,q8,q7
+	beq	.Lcbc_enc128
+
+.Loop_cbc_enc:
+	.byte	0x20,0x03,0xb0,0xf3	@ aese q0,q8
+	vld1.32	{q8},[r7]!
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	subs	r6,r6,#2
+	.byte	0x22,0x03,0xb0,0xf3	@ aese q0,q9
+	vld1.32	{q9},[r7]!
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	bgt	.Loop_cbc_enc
+
+	.byte	0x20,0x03,0xb0,0xf3	@ aese q0,q8
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	 subs	r2,r2,#16
+	.byte	0x22,0x03,0xb0,0xf3	@ aese q0,q9
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	 moveq	r8,#0
+	.byte	0x24,0x03,0xb0,0xf3	@ aese q0,q10
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	 add	r7,r3,#16
+	.byte	0x26,0x03,0xb0,0xf3	@ aese q0,q11
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	 vld1.8	{q8},[r0],r8
+	.byte	0x28,0x03,0xb0,0xf3	@ aese q0,q12
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	 veor	q8,q8,q5
+	.byte	0x2a,0x03,0xb0,0xf3	@ aese q0,q13
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	 vld1.32 {q9},[r7]!	@ re-pre-load rndkey[1]
+	.byte	0x2c,0x03,0xb0,0xf3	@ aese q0,q14
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x2e,0x03,0xb0,0xf3	@ aese q0,q15
+
+	 mov	r6,r5
+	veor	q6,q0,q7
+	vst1.8	{q6},[r1]!
+	bhs	.Loop_cbc_enc
+
+	b	.Lcbc_done
+
+.align	5
+.Lcbc_enc128:
+	vld1.32	{q2-q3},[r7]
+	.byte	0x20,0x03,0xb0,0xf3	@ aese q0,q8
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	b	.Lenter_cbc_enc128
+.Loop_cbc_enc128:
+	.byte	0x20,0x03,0xb0,0xf3	@ aese q0,q8
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	 vst1.8	{q6},[r1]!
+.Lenter_cbc_enc128:
+	.byte	0x22,0x03,0xb0,0xf3	@ aese q0,q9
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	 subs	r2,r2,#16
+	.byte	0x04,0x03,0xb0,0xf3	@ aese q0,q2
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	 moveq	r8,#0
+	.byte	0x06,0x03,0xb0,0xf3	@ aese q0,q3
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x24,0x03,0xb0,0xf3	@ aese q0,q10
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x26,0x03,0xb0,0xf3	@ aese q0,q11
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	 vld1.8	{q8},[r0],r8
+	.byte	0x28,0x03,0xb0,0xf3	@ aese q0,q12
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x2a,0x03,0xb0,0xf3	@ aese q0,q13
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x2c,0x03,0xb0,0xf3	@ aese q0,q14
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	 veor	q8,q8,q5
+	.byte	0x2e,0x03,0xb0,0xf3	@ aese q0,q15
+	veor	q6,q0,q7
+	bhs	.Loop_cbc_enc128
+
+	vst1.8	{q6},[r1]!
+	b	.Lcbc_done
+
+.align	5
+.Lcbc_dec128:
+	vld1.32	{q4-q5},[r7]
+	veor	q6,q6,q7
+	veor	q2,q0,q7
+	mov	r12,r8
+
+.Loop2x_cbc_dec128:
+	.byte	0x60,0x03,0xb0,0xf3	@ aesd q0,q8
+	.byte	0x60,0x23,0xb0,0xf3	@ aesd q1,q8
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	 subs	r2,r2,#32
+	.byte	0x62,0x03,0xb0,0xf3	@ aesd q0,q9
+	.byte	0x62,0x23,0xb0,0xf3	@ aesd q1,q9
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	 movlo	r8,#0
+	.byte	0x48,0x03,0xb0,0xf3	@ aesd q0,q4
+	.byte	0x48,0x23,0xb0,0xf3	@ aesd q1,q4
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	 movls	r12,#0
+	.byte	0x4a,0x03,0xb0,0xf3	@ aesd q0,q5
+	.byte	0x4a,0x23,0xb0,0xf3	@ aesd q1,q5
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	.byte	0x64,0x03,0xb0,0xf3	@ aesd q0,q10
+	.byte	0x64,0x23,0xb0,0xf3	@ aesd q1,q10
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	.byte	0x66,0x03,0xb0,0xf3	@ aesd q0,q11
+	.byte	0x66,0x23,0xb0,0xf3	@ aesd q1,q11
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	.byte	0x68,0x03,0xb0,0xf3	@ aesd q0,q12
+	.byte	0x68,0x23,0xb0,0xf3	@ aesd q1,q12
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	.byte	0x6a,0x03,0xb0,0xf3	@ aesd q0,q13
+	.byte	0x6a,0x23,0xb0,0xf3	@ aesd q1,q13
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	.byte	0x6c,0x03,0xb0,0xf3	@ aesd q0,q14
+	.byte	0x6c,0x23,0xb0,0xf3	@ aesd q1,q14
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	.byte	0x6e,0x03,0xb0,0xf3	@ aesd q0,q15
+	.byte	0x6e,0x23,0xb0,0xf3	@ aesd q1,q15
+
+	veor	q6,q6,q0
+	vld1.8	{q0},[r0],r8
+	veor	q2,q2,q1
+	vld1.8	{q1},[r0],r12
+	vst1.8	{q6},[r1]!
+	veor	q6,q3,q7
+	vst1.8	{q2},[r1]!
+	veor	q2,q0,q7
+	vorr	q3,q1,q1
+	bhs	.Loop2x_cbc_dec128
+
+	adds	r2,r2,#32
+	veor	q6,q6,q7
+	beq	.Lcbc_done
+	veor	q2,q2,q7
+	b	.Lcbc_dec_tail
+
+.align	5
+.Lcbc_dec:
+	subs	r2,r2,#16
+	vorr	q2,q0,q0
+	blo	.Lcbc_dec_tail
+
+	moveq	r8,#0
+	cmp	r5,#2
+	vld1.8	{q1},[r0],r8
+	vorr	q3,q1,q1
+	beq	.Lcbc_dec128
+
+.Loop2x_cbc_dec:
+	.byte	0x60,0x03,0xb0,0xf3	@ aesd q0,q8
+	.byte	0x60,0x23,0xb0,0xf3	@ aesd q1,q8
+	vld1.32	{q8},[r7]!
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	subs	r6,r6,#2
+	.byte	0x62,0x03,0xb0,0xf3	@ aesd q0,q9
+	.byte	0x62,0x23,0xb0,0xf3	@ aesd q1,q9
+	vld1.32	{q9},[r7]!
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	bgt	.Loop2x_cbc_dec
+
+	.byte	0x60,0x03,0xb0,0xf3	@ aesd q0,q8
+	.byte	0x60,0x23,0xb0,0xf3	@ aesd q1,q8
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	 veor	q4,q6,q7
+	 veor	q5,q2,q7
+	.byte	0x62,0x03,0xb0,0xf3	@ aesd q0,q9
+	.byte	0x62,0x23,0xb0,0xf3	@ aesd q1,q9
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	 vorr	q6,q3,q3
+	 subs	r2,r2,#32
+	.byte	0x64,0x03,0xb0,0xf3	@ aesd q0,q10
+	.byte	0x64,0x23,0xb0,0xf3	@ aesd q1,q10
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	 movlo	r8,#0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	 mov	r7,r3
+	.byte	0x66,0x03,0xb0,0xf3	@ aesd q0,q11
+	.byte	0x66,0x23,0xb0,0xf3	@ aesd q1,q11
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	 vld1.8	{q2},[r0],r8
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	 movls	r8,#0
+	.byte	0x68,0x03,0xb0,0xf3	@ aesd q0,q12
+	.byte	0x68,0x23,0xb0,0xf3	@ aesd q1,q12
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	 vld1.8	{q3},[r0],r8
+	.byte	0x6a,0x03,0xb0,0xf3	@ aesd q0,q13
+	.byte	0x6a,0x23,0xb0,0xf3	@ aesd q1,q13
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	 vld1.32 {q8},[r7]!	@ re-pre-load rndkey[0]
+	.byte	0x6c,0x03,0xb0,0xf3	@ aesd q0,q14
+	.byte	0x6c,0x23,0xb0,0xf3	@ aesd q1,q14
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0xc2,0x23,0xb0,0xf3	@ aesimc q1,q1
+	 vld1.32 {q9},[r7]!	@ re-pre-load rndkey[1]
+	.byte	0x6e,0x03,0xb0,0xf3	@ aesd q0,q15
+	.byte	0x6e,0x23,0xb0,0xf3	@ aesd q1,q15
+
+	 mov	r6,r5
+	veor	q4,q4,q0
+	veor	q5,q5,q1
+	 vorr	q0,q2,q2
+	vst1.8	{q4},[r1]!
+	 vorr	q1,q3,q3
+	vst1.8	{q5},[r1]!
+	bhs	.Loop2x_cbc_dec
+
+	adds	r2,r2,#32
+	beq	.Lcbc_done
+
+.Lcbc_dec_tail:
+	.byte	0x60,0x03,0xb0,0xf3	@ aesd q0,q8
+	vld1.32	{q8},[r7]!
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	subs	r6,r6,#2
+	.byte	0x62,0x03,0xb0,0xf3	@ aesd q0,q9
+	vld1.32	{q9},[r7]!
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	bgt	.Lcbc_dec_tail
+
+	.byte	0x60,0x03,0xb0,0xf3	@ aesd q0,q8
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0x62,0x03,0xb0,0xf3	@ aesd q0,q9
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	 veor	q4,q6,q7
+	.byte	0x64,0x03,0xb0,0xf3	@ aesd q0,q10
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	 vorr	q6,q2,q2
+	.byte	0x66,0x03,0xb0,0xf3	@ aesd q0,q11
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0x68,0x03,0xb0,0xf3	@ aesd q0,q12
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0x6a,0x03,0xb0,0xf3	@ aesd q0,q13
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0x6c,0x03,0xb0,0xf3	@ aesd q0,q14
+	.byte	0xc0,0x03,0xb0,0xf3	@ aesimc q0,q0
+	.byte	0x6e,0x03,0xb0,0xf3	@ aesd q0,q15
+
+	veor	q4,q4,q0
+	vst1.8	{q4},[r1]!
+
+.Lcbc_done:
+	vst1.8	{q6},[r4]
+.Lcbc_abort:
+	vldmia	sp!,{d8-d15}
+	ldmia	sp!,{r4-r8,pc}
+.size	aes_v8_cbc_encrypt,.-aes_v8_cbc_encrypt
+.globl	aes_v8_ctr32_encrypt_blocks
+.type	aes_v8_ctr32_encrypt_blocks,%function
+.align	5
+aes_v8_ctr32_encrypt_blocks:
+	mov		ip,sp
+	stmdb		sp!,{r4-r10,lr}
+	vstmdb		sp!,{d8-d15}            @ ABI specification says so
+	ldr		r4, [ip]		@ load remaining arg
+	ldr		r5,[r3,#240]
+
+	ldr		r8, [r4, #12]
+	vld1.32		{q0},[r4]
+
+	vld1.32		{q8-q9},[r3]		@ load key schedule...
+	sub		r5,r5,#6
+	add		r7,r3,r5,lsl#4	@ pointer to last 7 round keys
+	sub		r5,r5,#2
+	vld1.32		{q10-q11},[r7]!
+	vld1.32		{q12-q13},[r7]!
+	vld1.32		{q14-q15},[r7]!
+	vld1.32		{q7},[r7]
+
+	add		r7,r3,#32
+	mov		r6,r5
+
+	subs		r2,r2,#2
+	blo		.Lctr32_tail
+
+#ifndef __ARMEB__
+	rev		r8, r8
+#endif
+	vorr		q1,q0,q0
+	add		r8, r8, #1
+	vorr		q6,q0,q0
+	rev		r10, r8
+	cmp		r5,#2
+	vmov.32	d3[1],r10
+	beq		.Lctr32_128
+
+.Loop2x_ctr32:
+	.byte	0x20,0x03,0xb0,0xf3	@ aese q0,q8
+	.byte	0x20,0x23,0xb0,0xf3	@ aese q1,q8
+	vld1.32		{q8},[r7]!
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x82,0x23,0xb0,0xf3	@ aesmc q1,q1
+	subs		r6,r6,#2
+	.byte	0x22,0x03,0xb0,0xf3	@ aese q0,q9
+	.byte	0x22,0x23,0xb0,0xf3	@ aese q1,q9
+	vld1.32		{q9},[r7]!
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x82,0x23,0xb0,0xf3	@ aesmc q1,q1
+	bgt		.Loop2x_ctr32
+
+	.byte	0x20,0x03,0xb0,0xf3	@ aese q0,q8
+	.byte	0x20,0x23,0xb0,0xf3	@ aese q1,q8
+	.byte	0x80,0x83,0xb0,0xf3	@ aesmc q4,q0
+	 vorr		q0,q6,q6
+	.byte	0x82,0xa3,0xb0,0xf3	@ aesmc q5,q1
+	 vorr		q1,q6,q6
+	.byte	0x22,0x83,0xb0,0xf3	@ aese q4,q9
+	.byte	0x22,0xa3,0xb0,0xf3	@ aese q5,q9
+	 vld1.8		{q2},[r0]!
+	.byte	0x88,0x83,0xb0,0xf3	@ aesmc q4,q4
+	 vld1.8		{q3},[r0]!
+	.byte	0x8a,0xa3,0xb0,0xf3	@ aesmc q5,q5
+	 add		r8,r8,#1
+	.byte	0x24,0x83,0xb0,0xf3	@ aese q4,q10
+	.byte	0x24,0xa3,0xb0,0xf3	@ aese q5,q10
+	 rev		r9,r8
+	.byte	0x88,0x83,0xb0,0xf3	@ aesmc q4,q4
+	.byte	0x8a,0xa3,0xb0,0xf3	@ aesmc q5,q5
+	 add		r8,r8,#1
+	.byte	0x26,0x83,0xb0,0xf3	@ aese q4,q11
+	.byte	0x26,0xa3,0xb0,0xf3	@ aese q5,q11
+	 veor		q2,q2,q7
+	 rev		r10,r8
+	.byte	0x88,0x83,0xb0,0xf3	@ aesmc q4,q4
+	.byte	0x8a,0xa3,0xb0,0xf3	@ aesmc q5,q5
+	 veor		q3,q3,q7
+	 mov		r7,r3
+	.byte	0x28,0x83,0xb0,0xf3	@ aese q4,q12
+	.byte	0x28,0xa3,0xb0,0xf3	@ aese q5,q12
+	 subs		r2,r2,#2
+	.byte	0x88,0x83,0xb0,0xf3	@ aesmc q4,q4
+	.byte	0x8a,0xa3,0xb0,0xf3	@ aesmc q5,q5
+	 vld1.32	 {q8-q9},[r7]!	@ re-pre-load rndkey[0-1]
+	.byte	0x2a,0x83,0xb0,0xf3	@ aese q4,q13
+	.byte	0x2a,0xa3,0xb0,0xf3	@ aese q5,q13
+	.byte	0x88,0x83,0xb0,0xf3	@ aesmc q4,q4
+	.byte	0x8a,0xa3,0xb0,0xf3	@ aesmc q5,q5
+	.byte	0x2c,0x83,0xb0,0xf3	@ aese q4,q14
+	.byte	0x2c,0xa3,0xb0,0xf3	@ aese q5,q14
+	 vmov.32	d1[1], r9
+	.byte	0x88,0x83,0xb0,0xf3	@ aesmc q4,q4
+	 vmov.32	d3[1], r10
+	.byte	0x8a,0xa3,0xb0,0xf3	@ aesmc q5,q5
+	.byte	0x2e,0x83,0xb0,0xf3	@ aese q4,q15
+	.byte	0x2e,0xa3,0xb0,0xf3	@ aese q5,q15
+
+	 mov		r6,r5
+	veor		q2,q2,q4
+	veor		q3,q3,q5
+	vst1.8		{q2},[r1]!
+	vst1.8		{q3},[r1]!
+	bhs		.Loop2x_ctr32
+
+	adds		r2,r2,#2
+	beq		.Lctr32_done
+	b		.Lctr32_tail
+
+.Lctr32_128:
+	vld1.32		{q4-q5},[r7]
+
+.Loop2x_ctr32_128:
+	.byte	0x20,0x03,0xb0,0xf3	@ aese q0,q8
+	.byte	0x20,0x23,0xb0,0xf3	@ aese q1,q8
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	 vld1.8		{q2},[r0]!
+	.byte	0x82,0x23,0xb0,0xf3	@ aesmc q1,q1
+	 vld1.8		{q3},[r0]!
+	.byte	0x22,0x03,0xb0,0xf3	@ aese q0,q9
+	.byte	0x22,0x23,0xb0,0xf3	@ aese q1,q9
+	 add		r8,r8,#1
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x82,0x23,0xb0,0xf3	@ aesmc q1,q1
+	 rev		r9,r8
+	.byte	0x08,0x03,0xb0,0xf3	@ aese q0,q4
+	.byte	0x08,0x23,0xb0,0xf3	@ aese q1,q4
+	 add		r8,r8,#1
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x82,0x23,0xb0,0xf3	@ aesmc q1,q1
+	 rev		r10,r8
+	.byte	0x0a,0x03,0xb0,0xf3	@ aese q0,q5
+	.byte	0x0a,0x23,0xb0,0xf3	@ aese q1,q5
+	 subs		r2,r2,#2
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x82,0x23,0xb0,0xf3	@ aesmc q1,q1
+	.byte	0x24,0x03,0xb0,0xf3	@ aese q0,q10
+	.byte	0x24,0x23,0xb0,0xf3	@ aese q1,q10
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x82,0x23,0xb0,0xf3	@ aesmc q1,q1
+	.byte	0x26,0x03,0xb0,0xf3	@ aese q0,q11
+	.byte	0x26,0x23,0xb0,0xf3	@ aese q1,q11
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x82,0x23,0xb0,0xf3	@ aesmc q1,q1
+	.byte	0x28,0x03,0xb0,0xf3	@ aese q0,q12
+	.byte	0x28,0x23,0xb0,0xf3	@ aese q1,q12
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x82,0x23,0xb0,0xf3	@ aesmc q1,q1
+	.byte	0x2a,0x03,0xb0,0xf3	@ aese q0,q13
+	.byte	0x2a,0x23,0xb0,0xf3	@ aese q1,q13
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x82,0x23,0xb0,0xf3	@ aesmc q1,q1
+	.byte	0x2c,0x03,0xb0,0xf3	@ aese q0,q14
+	.byte	0x2c,0x23,0xb0,0xf3	@ aese q1,q14
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x82,0x23,0xb0,0xf3	@ aesmc q1,q1
+	 veor		q2,q2,q7
+	.byte	0x2e,0x03,0xb0,0xf3	@ aese q0,q15
+	 veor		q3,q3,q7
+	.byte	0x2e,0x23,0xb0,0xf3	@ aese q1,q15
+
+	veor		q2,q2,q0
+	vorr		q0,q6,q6
+	veor		q3,q3,q1
+	vorr		q1,q6,q6
+	vst1.8		{q2},[r1]!
+	vmov.32	d1[1], r9
+	vst1.8		{q3},[r1]!
+	vmov.32	d3[1], r10
+	bhs		.Loop2x_ctr32_128
+
+	adds		r2,r2,#2
+	beq		.Lctr32_done
+
+.Lctr32_tail:
+	.byte	0x20,0x03,0xb0,0xf3	@ aese q0,q8
+	vld1.32		{q8},[r7]!
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	subs		r6,r6,#2
+	.byte	0x22,0x03,0xb0,0xf3	@ aese q0,q9
+	vld1.32		{q9},[r7]!
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	bgt		.Lctr32_tail
+
+	.byte	0x20,0x03,0xb0,0xf3	@ aese q0,q8
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x22,0x03,0xb0,0xf3	@ aese q0,q9
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	 vld1.8		{q2},[r0]
+	.byte	0x24,0x03,0xb0,0xf3	@ aese q0,q10
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x26,0x03,0xb0,0xf3	@ aese q0,q11
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x28,0x03,0xb0,0xf3	@ aese q0,q12
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x2a,0x03,0xb0,0xf3	@ aese q0,q13
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	.byte	0x2c,0x03,0xb0,0xf3	@ aese q0,q14
+	.byte	0x80,0x03,0xb0,0xf3	@ aesmc q0,q0
+	 veor		q2,q2,q7
+	.byte	0x2e,0x03,0xb0,0xf3	@ aese q0,q15
+
+	veor		q2,q2,q0
+	vst1.8		{q2},[r1]
+
+.Lctr32_done:
+	vldmia		sp!,{d8-d15}
+	ldmia		sp!,{r4-r10,pc}
+.size	aes_v8_ctr32_encrypt_blocks,.-aes_v8_ctr32_encrypt_blocks
+#endif

jni/openssl/crypto/aes/asm/aesv8-armx.pl

@@ -0,0 +1,980 @@
+#!/usr/bin/env perl
+#
+# ====================================================================
+# Written by Andy Polyakov <appro@openssl.org> for the OpenSSL
+# project. The module is, however, dual licensed under OpenSSL and
+# CRYPTOGAMS licenses depending on where you obtain it. For further
+# details see http://www.openssl.org/~appro/cryptogams/.
+# ====================================================================
+#
+# This module implements support for ARMv8 AES instructions. The
+# module is endian-agnostic in sense that it supports both big- and
+# little-endian cases. As does it support both 32- and 64-bit modes
+# of operation. Latter is achieved by limiting amount of utilized
+# registers to 16, which implies additional instructions. This has
+# no effect on mighty Apple A7, as results are literally equal to
+# the theoretical estimates based on instruction latencies and issue
+# rate. It remains to be seen how does it affect other platforms...
+#
+# Performance in cycles per byte processed with 128-bit key:
+#
+#		CBC enc		CBC dec		CTR
+# Apple A7	2.39		1.20		1.20
+# Cortex-A5x	n/a		n/a		n/a
+
+$flavour = shift;
+open STDOUT,">".shift;
+
+$prefix="aes_v8";
+
+$code=<<___;
+#include "arm_arch.h"
+
+#if __ARM_ARCH__>=7
+.text
+___
+$code.=".arch	armv8-a+crypto\n"	if ($flavour =~ /64/);
+$code.=".fpu	neon\n.code	32\n"	if ($flavour !~ /64/);
+
+# Assembler mnemonics are an eclectic mix of 32- and 64-bit syntax,
+# NEON is mostly 32-bit mnemonics, integer - mostly 64. Goal is to
+# maintain both 32- and 64-bit codes within single module and
+# transliterate common code to either flavour with regex vodoo.
+#
+{{{
+my ($inp,$bits,$out,$ptr,$rounds)=("x0","w1","x2","x3","w12");
+my ($zero,$rcon,$mask,$in0,$in1,$tmp,$key)=
+	$flavour=~/64/? map("q$_",(0..6)) : map("q$_",(0..3,8..10));
+
+
+$code.=<<___;
+.align	5
+rcon:
+.long	0x01,0x01,0x01,0x01
+.long	0x0c0f0e0d,0x0c0f0e0d,0x0c0f0e0d,0x0c0f0e0d	// rotate-n-splat
+.long	0x1b,0x1b,0x1b,0x1b
+
+.globl	${prefix}_set_encrypt_key
+.type	${prefix}_set_encrypt_key,%function
+.align	5
+${prefix}_set_encrypt_key:
+.Lenc_key:
+___
+$code.=<<___	if ($flavour =~ /64/);
+	stp	x29,x30,[sp,#-16]!
+	add	x29,sp,#0
+___
+$code.=<<___;
+	adr	$ptr,rcon
+	cmp	$bits,#192
+
+	veor	$zero,$zero,$zero
+	vld1.8	{$in0},[$inp],#16
+	mov	$bits,#8		// reuse $bits
+	vld1.32	{$rcon,$mask},[$ptr],#32
+
+	b.lt	.Loop128
+	b.eq	.L192
+	b	.L256
+
+.align	4
+.Loop128:
+	vtbl.8	$key,{$in0},$mask
+	vext.8	$tmp,$zero,$in0,#12
+	vst1.32	{$in0},[$out],#16
+	aese	$key,$zero
+	subs	$bits,$bits,#1
+
+	veor	$in0,$in0,$tmp
+	vext.8	$tmp,$zero,$tmp,#12
+	veor	$in0,$in0,$tmp
+	vext.8	$tmp,$zero,$tmp,#12
+	 veor	$key,$key,$rcon
+	veor	$in0,$in0,$tmp
+	vshl.u8	$rcon,$rcon,#1
+	veor	$in0,$in0,$key
+	b.ne	.Loop128
+
+	vld1.32	{$rcon},[$ptr]
+
+	vtbl.8	$key,{$in0},$mask
+	vext.8	$tmp,$zero,$in0,#12
+	vst1.32	{$in0},[$out],#16
+	aese	$key,$zero
+
+	veor	$in0,$in0,$tmp
+	vext.8	$tmp,$zero,$tmp,#12
+	veor	$in0,$in0,$tmp
+	vext.8	$tmp,$zero,$tmp,#12
+	 veor	$key,$key,$rcon
+	veor	$in0,$in0,$tmp
+	vshl.u8	$rcon,$rcon,#1
+	veor	$in0,$in0,$key
+
+	vtbl.8	$key,{$in0},$mask
+	vext.8	$tmp,$zero,$in0,#12
+	vst1.32	{$in0},[$out],#16
+	aese	$key,$zero
+
+	veor	$in0,$in0,$tmp
+	vext.8	$tmp,$zero,$tmp,#12
+	veor	$in0,$in0,$tmp
+	vext.8	$tmp,$zero,$tmp,#12
+	 veor	$key,$key,$rcon
+	veor	$in0,$in0,$tmp
+	veor	$in0,$in0,$key
+	vst1.32	{$in0},[$out]
+	add	$out,$out,#0x50
+
+	mov	$rounds,#10
+	b	.Ldone
+
+.align	4
+.L192:
+	vld1.8	{$in1},[$inp],#8
+	vmov.i8	$key,#8			// borrow $key
+	vst1.32	{$in0},[$out],#16
+	vsub.i8	$mask,$mask,$key	// adjust the mask
+
+.Loop192:
+	vtbl.8	$key,{$in1},$mask
+	vext.8	$tmp,$zero,$in0,#12
+	vst1.32	{$in1},[$out],#8
+	aese	$key,$zero
+	subs	$bits,$bits,#1
+
+	veor	$in0,$in0,$tmp
+	vext.8	$tmp,$zero,$tmp,#12
+	veor	$in0,$in0,$tmp
+	vext.8	$tmp,$zero,$tmp,#12
+	veor	$in0,$in0,$tmp
+
+	vdup.32	$tmp,${in0}[3]
+	veor	$tmp,$tmp,$in1
+	 veor	$key,$key,$rcon
+	vext.8	$in1,$zero,$in1,#12
+	vshl.u8	$rcon,$rcon,#1
+	veor	$in1,$in1,$tmp
+	veor	$in0,$in0,$key
+	veor	$in1,$in1,$key
+	vst1.32	{$in0},[$out],#16
+	b.ne	.Loop192
+
+	mov	$rounds,#12
+	add	$out,$out,#0x20
+	b	.Ldone
+
+.align	4
+.L256:
+	vld1.8	{$in1},[$inp]
+	mov	$bits,#7
+	mov	$rounds,#14
+	vst1.32	{$in0},[$out],#16
+
+.Loop256:
+	vtbl.8	$key,{$in1},$mask
+	vext.8	$tmp,$zero,$in0,#12
+	vst1.32	{$in1},[$out],#16
+	aese	$key,$zero
+	subs	$bits,$bits,#1
+
+	veor	$in0,$in0,$tmp
+	vext.8	$tmp,$zero,$tmp,#12
+	veor	$in0,$in0,$tmp
+	vext.8	$tmp,$zero,$tmp,#12
+	 veor	$key,$key,$rcon
+	veor	$in0,$in0,$tmp
+	vshl.u8	$rcon,$rcon,#1
+	veor	$in0,$in0,$key
+	vst1.32	{$in0},[$out],#16
+	b.eq	.Ldone
+
+	vdup.32	$key,${in0}[3]		// just splat
+	vext.8	$tmp,$zero,$in1,#12
+	aese	$key,$zero
+
+	veor	$in1,$in1,$tmp
+	vext.8	$tmp,$zero,$tmp,#12
+	veor	$in1,$in1,$tmp
+	vext.8	$tmp,$zero,$tmp,#12
+	veor	$in1,$in1,$tmp
+
+	veor	$in1,$in1,$key
+	b	.Loop256
+
+.Ldone:
+	str	$rounds,[$out]
+
+	eor	x0,x0,x0		// return value
+	`"ldr	x29,[sp],#16"		if ($flavour =~ /64/)`
+	ret
+.size	${prefix}_set_encrypt_key,.-${prefix}_set_encrypt_key
+
+.globl	${prefix}_set_decrypt_key
+.type	${prefix}_set_decrypt_key,%function
+.align	5
+${prefix}_set_decrypt_key:
+___
+$code.=<<___	if ($flavour =~ /64/);
+	stp	x29,x30,[sp,#-16]!
+	add	x29,sp,#0
+___
+$code.=<<___	if ($flavour !~ /64/);
+	stmdb	sp!,{r4,lr}
+___
+$code.=<<___;
+	bl	.Lenc_key
+
+	sub	$out,$out,#240		// restore original $out
+	mov	x4,#-16
+	add	$inp,$out,x12,lsl#4	// end of key schedule
+
+	vld1.32	{v0.16b},[$out]
+	vld1.32	{v1.16b},[$inp]
+	vst1.32	{v0.16b},[$inp],x4
+	vst1.32	{v1.16b},[$out],#16
+
+.Loop_imc:
+	vld1.32	{v0.16b},[$out]
+	vld1.32	{v1.16b},[$inp]
+	aesimc	v0.16b,v0.16b
+	aesimc	v1.16b,v1.16b
+	vst1.32	{v0.16b},[$inp],x4
+	vst1.32	{v1.16b},[$out],#16
+	cmp	$inp,$out
+	b.hi	.Loop_imc
+
+	vld1.32	{v0.16b},[$out]
+	aesimc	v0.16b,v0.16b
+	vst1.32	{v0.16b},[$inp]
+
+	eor	x0,x0,x0		// return value
+___
+$code.=<<___	if ($flavour !~ /64/);
+	ldmia	sp!,{r4,pc}
+___
+$code.=<<___	if ($flavour =~ /64/);
+	ldp	x29,x30,[sp],#16
+	ret
+___
+$code.=<<___;
+.size	${prefix}_set_decrypt_key,.-${prefix}_set_decrypt_key
+___
+}}}
+{{{
+sub gen_block () {
+my $dir = shift;
+my ($e,$mc) = $dir eq "en" ? ("e","mc") : ("d","imc");
+my ($inp,$out,$key)=map("x$_",(0..2));
+my $rounds="w3";
+my ($rndkey0,$rndkey1,$inout)=map("q$_",(0..3));
+
+$code.=<<___;
+.globl	${prefix}_${dir}crypt
+.type	${prefix}_${dir}crypt,%function
+.align	5
+${prefix}_${dir}crypt:
+	ldr	$rounds,[$key,#240]
+	vld1.32	{$rndkey0},[$key],#16
+	vld1.8	{$inout},[$inp]
+	sub	$rounds,$rounds,#2
+	vld1.32	{$rndkey1},[$key],#16
+
+.Loop_${dir}c:
+	aes$e	$inout,$rndkey0
+	vld1.32	{$rndkey0},[$key],#16
+	aes$mc	$inout,$inout
+	subs	$rounds,$rounds,#2
+	aes$e	$inout,$rndkey1
+	vld1.32	{$rndkey1},[$key],#16
+	aes$mc	$inout,$inout
+	b.gt	.Loop_${dir}c
+
+	aes$e	$inout,$rndkey0
+	vld1.32	{$rndkey0},[$key]
+	aes$mc	$inout,$inout
+	aes$e	$inout,$rndkey1
+	veor	$inout,$inout,$rndkey0
+
+	vst1.8	{$inout},[$out]
+	ret
+.size	${prefix}_${dir}crypt,.-${prefix}_${dir}crypt
+___
+}
+&gen_block("en");
+&gen_block("de");
+}}}
+{{{
+my ($inp,$out,$len,$key,$ivp)=map("x$_",(0..4)); my $enc="w5";
+my ($rounds,$cnt,$key_,$step,$step1)=($enc,"w6","x7","x8","x12");
+my ($dat0,$dat1,$in0,$in1,$tmp0,$tmp1,$ivec,$rndlast)=map("q$_",(0..7));
+
+my ($dat,$tmp,$rndzero_n_last)=($dat0,$tmp0,$tmp1);
+
+### q8-q15	preloaded key schedule
+
+$code.=<<___;
+.globl	${prefix}_cbc_encrypt
+.type	${prefix}_cbc_encrypt,%function
+.align	5
+${prefix}_cbc_encrypt:
+___
+$code.=<<___	if ($flavour =~ /64/);
+	stp	x29,x30,[sp,#-16]!
+	add	x29,sp,#0
+___
+$code.=<<___	if ($flavour !~ /64/);
+	mov	ip,sp
+	stmdb	sp!,{r4-r8,lr}
+	vstmdb	sp!,{d8-d15}            @ ABI specification says so
+	ldmia	ip,{r4-r5}		@ load remaining args
+___
+$code.=<<___;
+	subs	$len,$len,#16
+	mov	$step,#16
+	b.lo	.Lcbc_abort
+	cclr	$step,eq
+
+	cmp	$enc,#0			// en- or decrypting?
+	ldr	$rounds,[$key,#240]
+	and	$len,$len,#-16
+	vld1.8	{$ivec},[$ivp]
+	vld1.8	{$dat},[$inp],$step
+
+	vld1.32	{q8-q9},[$key]		// load key schedule...
+	sub	$rounds,$rounds,#6
+	add	$key_,$key,x5,lsl#4	// pointer to last 7 round keys
+	sub	$rounds,$rounds,#2
+	vld1.32	{q10-q11},[$key_],#32
+	vld1.32	{q12-q13},[$key_],#32
+	vld1.32	{q14-q15},[$key_],#32
+	vld1.32	{$rndlast},[$key_]
+
+	add	$key_,$key,#32
+	mov	$cnt,$rounds
+	b.eq	.Lcbc_dec
+
+	cmp	$rounds,#2
+	veor	$dat,$dat,$ivec
+	veor	$rndzero_n_last,q8,$rndlast
+	b.eq	.Lcbc_enc128
+
+.Loop_cbc_enc:
+	aese	$dat,q8
+	vld1.32	{q8},[$key_],#16
+	aesmc	$dat,$dat
+	subs	$cnt,$cnt,#2
+	aese	$dat,q9
+	vld1.32	{q9},[$key_],#16
+	aesmc	$dat,$dat
+	b.gt	.Loop_cbc_enc
+
+	aese	$dat,q8
+	aesmc	$dat,$dat
+	 subs	$len,$len,#16
+	aese	$dat,q9
+	aesmc	$dat,$dat
+	 cclr	$step,eq
+	aese	$dat,q10
+	aesmc	$dat,$dat
+	 add	$key_,$key,#16
+	aese	$dat,q11
+	aesmc	$dat,$dat
+	 vld1.8	{q8},[$inp],$step
+	aese	$dat,q12
+	aesmc	$dat,$dat
+	 veor	q8,q8,$rndzero_n_last
+	aese	$dat,q13
+	aesmc	$dat,$dat
+	 vld1.32 {q9},[$key_],#16	// re-pre-load rndkey[1]
+	aese	$dat,q14
+	aesmc	$dat,$dat
+	aese	$dat,q15
+
+	 mov	$cnt,$rounds
+	veor	$ivec,$dat,$rndlast
+	vst1.8	{$ivec},[$out],#16
+	b.hs	.Loop_cbc_enc
+
+	b	.Lcbc_done
+
+.align	5
+.Lcbc_enc128:
+	vld1.32	{$in0-$in1},[$key_]
+	aese	$dat,q8
+	aesmc	$dat,$dat
+	b	.Lenter_cbc_enc128
+.Loop_cbc_enc128:
+	aese	$dat,q8
+	aesmc	$dat,$dat
+	 vst1.8	{$ivec},[$out],#16
+.Lenter_cbc_enc128:
+	aese	$dat,q9
+	aesmc	$dat,$dat
+	 subs	$len,$len,#16
+	aese	$dat,$in0
+	aesmc	$dat,$dat
+	 cclr	$step,eq
+	aese	$dat,$in1
+	aesmc	$dat,$dat
+	aese	$dat,q10
+	aesmc	$dat,$dat
+	aese	$dat,q11
+	aesmc	$dat,$dat
+	 vld1.8	{q8},[$inp],$step
+	aese	$dat,q12
+	aesmc	$dat,$dat
+	aese	$dat,q13
+	aesmc	$dat,$dat
+	aese	$dat,q14
+	aesmc	$dat,$dat
+	 veor	q8,q8,$rndzero_n_last
+	aese	$dat,q15
+	veor	$ivec,$dat,$rndlast
+	b.hs	.Loop_cbc_enc128
+
+	vst1.8	{$ivec},[$out],#16
+	b	.Lcbc_done
+
+.align	5
+.Lcbc_dec128:
+	vld1.32	{$tmp0-$tmp1},[$key_]
+	veor	$ivec,$ivec,$rndlast
+	veor	$in0,$dat0,$rndlast
+	mov	$step1,$step
+
+.Loop2x_cbc_dec128:
+	aesd	$dat0,q8
+	aesd	$dat1,q8
+	aesimc	$dat0,$dat0
+	aesimc	$dat1,$dat1
+	 subs	$len,$len,#32
+	aesd	$dat0,q9
+	aesd	$dat1,q9
+	aesimc	$dat0,$dat0
+	aesimc	$dat1,$dat1
+	 cclr	$step,lo
+	aesd	$dat0,$tmp0
+	aesd	$dat1,$tmp0
+	aesimc	$dat0,$dat0
+	aesimc	$dat1,$dat1
+	 cclr	$step1,ls
+	aesd	$dat0,$tmp1
+	aesd	$dat1,$tmp1
+	aesimc	$dat0,$dat0
+	aesimc	$dat1,$dat1
+	aesd	$dat0,q10
+	aesd	$dat1,q10
+	aesimc	$dat0,$dat0
+	aesimc	$dat1,$dat1
+	aesd	$dat0,q11
+	aesd	$dat1,q11
+	aesimc	$dat0,$dat0
+	aesimc	$dat1,$dat1
+	aesd	$dat0,q12
+	aesd	$dat1,q12
+	aesimc	$dat0,$dat0
+	aesimc	$dat1,$dat1
+	aesd	$dat0,q13
+	aesd	$dat1,q13
+	aesimc	$dat0,$dat0
+	aesimc	$dat1,$dat1
+	aesd	$dat0,q14
+	aesd	$dat1,q14
+	aesimc	$dat0,$dat0
+	aesimc	$dat1,$dat1
+	aesd	$dat0,q15
+	aesd	$dat1,q15
+
+	veor	$ivec,$ivec,$dat0
+	vld1.8	{$dat0},[$inp],$step
+	veor	$in0,$in0,$dat1
+	vld1.8	{$dat1},[$inp],$step1
+	vst1.8	{$ivec},[$out],#16
+	veor	$ivec,$in1,$rndlast
+	vst1.8	{$in0},[$out],#16
+	veor	$in0,$dat0,$rndlast
+	vorr	$in1,$dat1,$dat1
+	b.hs	.Loop2x_cbc_dec128
+
+	adds	$len,$len,#32
+	veor	$ivec,$ivec,$rndlast
+	b.eq	.Lcbc_done
+	veor	$in0,$in0,$rndlast
+	b	.Lcbc_dec_tail
+
+.align	5
+.Lcbc_dec:
+	subs	$len,$len,#16
+	vorr	$in0,$dat,$dat
+	b.lo	.Lcbc_dec_tail
+
+	cclr	$step,eq
+	cmp	$rounds,#2
+	vld1.8	{$dat1},[$inp],$step
+	vorr	$in1,$dat1,$dat1
+	b.eq	.Lcbc_dec128
+
+.Loop2x_cbc_dec:
+	aesd	$dat0,q8
+	aesd	$dat1,q8
+	vld1.32	{q8},[$key_],#16
+	aesimc	$dat0,$dat0
+	aesimc	$dat1,$dat1
+	subs	$cnt,$cnt,#2
+	aesd	$dat0,q9
+	aesd	$dat1,q9
+	vld1.32	{q9},[$key_],#16
+	aesimc	$dat0,$dat0
+	aesimc	$dat1,$dat1
+	b.gt	.Loop2x_cbc_dec
+
+	aesd	$dat0,q8
+	aesd	$dat1,q8
+	aesimc	$dat0,$dat0
+	aesimc	$dat1,$dat1
+	 veor	$tmp0,$ivec,$rndlast
+	 veor	$tmp1,$in0,$rndlast
+	aesd	$dat0,q9
+	aesd	$dat1,q9
+	aesimc	$dat0,$dat0
+	aesimc	$dat1,$dat1
+	 vorr	$ivec,$in1,$in1
+	 subs	$len,$len,#32
+	aesd	$dat0,q10
+	aesd	$dat1,q10
+	aesimc	$dat0,$dat0
+	 cclr	$step,lo
+	aesimc	$dat1,$dat1
+	 mov	$key_,$key
+	aesd	$dat0,q11
+	aesd	$dat1,q11
+	aesimc	$dat0,$dat0
+	 vld1.8	{$in0},[$inp],$step
+	aesimc	$dat1,$dat1
+	 cclr	$step,ls
+	aesd	$dat0,q12
+	aesd	$dat1,q12
+	aesimc	$dat0,$dat0
+	aesimc	$dat1,$dat1
+	 vld1.8	{$in1},[$inp],$step
+	aesd	$dat0,q13
+	aesd	$dat1,q13
+	aesimc	$dat0,$dat0
+	aesimc	$dat1,$dat1
+	 vld1.32 {q8},[$key_],#16	// re-pre-load rndkey[0]
+	aesd	$dat0,q14
+	aesd	$dat1,q14
+	aesimc	$dat0,$dat0
+	aesimc	$dat1,$dat1
+	 vld1.32 {q9},[$key_],#16	// re-pre-load rndkey[1]
+	aesd	$dat0,q15
+	aesd	$dat1,q15
+
+	 mov	$cnt,$rounds
+	veor	$tmp0,$tmp0,$dat0
+	veor	$tmp1,$tmp1,$dat1
+	 vorr	$dat0,$in0,$in0
+	vst1.8	{$tmp0},[$out],#16
+	 vorr	$dat1,$in1,$in1
+	vst1.8	{$tmp1},[$out],#16
+	b.hs	.Loop2x_cbc_dec
+
+	adds	$len,$len,#32
+	b.eq	.Lcbc_done
+
+.Lcbc_dec_tail:
+	aesd	$dat,q8
+	vld1.32	{q8},[$key_],#16
+	aesimc	$dat,$dat
+	subs	$cnt,$cnt,#2
+	aesd	$dat,q9
+	vld1.32	{q9},[$key_],#16
+	aesimc	$dat,$dat
+	b.gt	.Lcbc_dec_tail
+
+	aesd	$dat,q8
+	aesimc	$dat,$dat
+	aesd	$dat,q9
+	aesimc	$dat,$dat
+	 veor	$tmp,$ivec,$rndlast
+	aesd	$dat,q10
+	aesimc	$dat,$dat
+	 vorr	$ivec,$in0,$in0
+	aesd	$dat,q11
+	aesimc	$dat,$dat
+	aesd	$dat,q12
+	aesimc	$dat,$dat
+	aesd	$dat,q13
+	aesimc	$dat,$dat
+	aesd	$dat,q14
+	aesimc	$dat,$dat
+	aesd	$dat,q15
+
+	veor	$tmp,$tmp,$dat
+	vst1.8	{$tmp},[$out],#16
+
+.Lcbc_done:
+	vst1.8	{$ivec},[$ivp]
+.Lcbc_abort:
+___
+$code.=<<___	if ($flavour !~ /64/);
+	vldmia	sp!,{d8-d15}
+	ldmia	sp!,{r4-r8,pc}
+___
+$code.=<<___	if ($flavour =~ /64/);
+	ldr	x29,[sp],#16
+	ret
+___
+$code.=<<___;
+.size	${prefix}_cbc_encrypt,.-${prefix}_cbc_encrypt
+___
+}}}
+{{{
+my ($inp,$out,$len,$key,$ivp)=map("x$_",(0..4));
+my ($rounds,$cnt,$key_,$ctr,$tctr,$tctr1)=("w5","w6","x7","w8","w9","w10");
+my ($dat0,$dat1,$in0,$in1,$tmp0,$tmp1,$ivec,$rndlast)=map("q$_",(0..7));
+
+my ($dat,$tmp)=($dat0,$tmp0);
+
+### q8-q15	preloaded key schedule
+
+$code.=<<___;
+.globl	${prefix}_ctr32_encrypt_blocks
+.type	${prefix}_ctr32_encrypt_blocks,%function
+.align	5
+${prefix}_ctr32_encrypt_blocks:
+___
+$code.=<<___	if ($flavour =~ /64/);
+	stp		x29,x30,[sp,#-16]!
+	add		x29,sp,#0
+___
+$code.=<<___	if ($flavour !~ /64/);
+	mov		ip,sp
+	stmdb		sp!,{r4-r10,lr}
+	vstmdb		sp!,{d8-d15}            @ ABI specification says so
+	ldr		r4, [ip]		@ load remaining arg
+___
+$code.=<<___;
+	ldr		$rounds,[$key,#240]
+
+	ldr		$ctr, [$ivp, #12]
+	vld1.32		{$dat0},[$ivp]
+
+	vld1.32		{q8-q9},[$key]		// load key schedule...
+	sub		$rounds,$rounds,#6
+	add		$key_,$key,x5,lsl#4	// pointer to last 7 round keys
+	sub		$rounds,$rounds,#2
+	vld1.32		{q10-q11},[$key_],#32
+	vld1.32		{q12-q13},[$key_],#32
+	vld1.32		{q14-q15},[$key_],#32
+	vld1.32		{$rndlast},[$key_]
+
+	add		$key_,$key,#32
+	mov		$cnt,$rounds
+
+	subs		$len,$len,#2
+	b.lo		.Lctr32_tail
+
+#ifndef __ARMEB__
+	rev		$ctr, $ctr
+#endif
+	vorr		$dat1,$dat0,$dat0
+	add		$ctr, $ctr, #1
+	vorr		$ivec,$dat0,$dat0
+	rev		$tctr1, $ctr
+	cmp		$rounds,#2
+	vmov.32		${dat1}[3],$tctr1
+	b.eq		.Lctr32_128
+
+.Loop2x_ctr32:
+	aese		$dat0,q8
+	aese		$dat1,q8
+	vld1.32		{q8},[$key_],#16
+	aesmc		$dat0,$dat0
+	aesmc		$dat1,$dat1
+	subs		$cnt,$cnt,#2
+	aese		$dat0,q9
+	aese		$dat1,q9
+	vld1.32		{q9},[$key_],#16
+	aesmc		$dat0,$dat0
+	aesmc		$dat1,$dat1
+	b.gt		.Loop2x_ctr32
+
+	aese		$dat0,q8
+	aese		$dat1,q8
+	aesmc		$tmp0,$dat0
+	 vorr		$dat0,$ivec,$ivec
+	aesmc		$tmp1,$dat1
+	 vorr		$dat1,$ivec,$ivec
+	aese		$tmp0,q9
+	aese		$tmp1,q9
+	 vld1.8		{$in0},[$inp],#16
+	aesmc		$tmp0,$tmp0
+	 vld1.8		{$in1},[$inp],#16
+	aesmc		$tmp1,$tmp1
+	 add		$ctr,$ctr,#1
+	aese		$tmp0,q10
+	aese		$tmp1,q10
+	 rev		$tctr,$ctr
+	aesmc		$tmp0,$tmp0
+	aesmc		$tmp1,$tmp1
+	 add		$ctr,$ctr,#1
+	aese		$tmp0,q11
+	aese		$tmp1,q11
+	 veor		$in0,$in0,$rndlast
+	 rev		$tctr1,$ctr
+	aesmc		$tmp0,$tmp0
+	aesmc		$tmp1,$tmp1
+	 veor		$in1,$in1,$rndlast
+	 mov		$key_,$key
+	aese		$tmp0,q12
+	aese		$tmp1,q12
+	 subs		$len,$len,#2
+	aesmc		$tmp0,$tmp0
+	aesmc		$tmp1,$tmp1
+	 vld1.32	 {q8-q9},[$key_],#32	// re-pre-load rndkey[0-1]
+	aese		$tmp0,q13
+	aese		$tmp1,q13
+	aesmc		$tmp0,$tmp0
+	aesmc		$tmp1,$tmp1
+	aese		$tmp0,q14
+	aese		$tmp1,q14
+	 vmov.32	${dat0}[3], $tctr
+	aesmc		$tmp0,$tmp0
+	 vmov.32	${dat1}[3], $tctr1
+	aesmc		$tmp1,$tmp1
+	aese		$tmp0,q15
+	aese		$tmp1,q15
+
+	 mov		$cnt,$rounds
+	veor		$in0,$in0,$tmp0
+	veor		$in1,$in1,$tmp1
+	vst1.8		{$in0},[$out],#16
+	vst1.8		{$in1},[$out],#16
+	b.hs		.Loop2x_ctr32
+
+	adds		$len,$len,#2
+	b.eq		.Lctr32_done
+	b		.Lctr32_tail
+
+.Lctr32_128:
+	vld1.32		{$tmp0-$tmp1},[$key_]
+
+.Loop2x_ctr32_128:
+	aese		$dat0,q8
+	aese		$dat1,q8
+	aesmc		$dat0,$dat0
+	 vld1.8		{$in0},[$inp],#16
+	aesmc		$dat1,$dat1
+	 vld1.8		{$in1},[$inp],#16
+	aese		$dat0,q9
+	aese		$dat1,q9
+	 add		$ctr,$ctr,#1
+	aesmc		$dat0,$dat0
+	aesmc		$dat1,$dat1
+	 rev		$tctr,$ctr
+	aese		$dat0,$tmp0
+	aese		$dat1,$tmp0
+	 add		$ctr,$ctr,#1
+	aesmc		$dat0,$dat0
+	aesmc		$dat1,$dat1
+	 rev		$tctr1,$ctr
+	aese		$dat0,$tmp1
+	aese		$dat1,$tmp1
+	 subs		$len,$len,#2
+	aesmc		$dat0,$dat0
+	aesmc		$dat1,$dat1
+	aese		$dat0,q10
+	aese		$dat1,q10
+	aesmc		$dat0,$dat0
+	aesmc		$dat1,$dat1
+	aese		$dat0,q11
+	aese		$dat1,q11
+	aesmc		$dat0,$dat0
+	aesmc		$dat1,$dat1
+	aese		$dat0,q12
+	aese		$dat1,q12
+	aesmc		$dat0,$dat0
+	aesmc		$dat1,$dat1
+	aese		$dat0,q13
+	aese		$dat1,q13
+	aesmc		$dat0,$dat0
+	aesmc		$dat1,$dat1
+	aese		$dat0,q14
+	aese		$dat1,q14
+	aesmc		$dat0,$dat0
+	aesmc		$dat1,$dat1
+	 veor		$in0,$in0,$rndlast
+	aese		$dat0,q15
+	 veor		$in1,$in1,$rndlast
+	aese		$dat1,q15
+
+	veor		$in0,$in0,$dat0
+	vorr		$dat0,$ivec,$ivec
+	veor		$in1,$in1,$dat1
+	vorr		$dat1,$ivec,$ivec
+	vst1.8		{$in0},[$out],#16
+	vmov.32		${dat0}[3], $tctr
+	vst1.8		{$in1},[$out],#16
+	vmov.32		${dat1}[3], $tctr1
+	b.hs		.Loop2x_ctr32_128
+
+	adds		$len,$len,#2
+	b.eq		.Lctr32_done
+
+.Lctr32_tail:
+	aese		$dat,q8
+	vld1.32		{q8},[$key_],#16
+	aesmc		$dat,$dat
+	subs		$cnt,$cnt,#2
+	aese		$dat,q9
+	vld1.32		{q9},[$key_],#16
+	aesmc		$dat,$dat
+	b.gt		.Lctr32_tail
+
+	aese		$dat,q8
+	aesmc		$dat,$dat
+	aese		$dat,q9
+	aesmc		$dat,$dat
+	 vld1.8		{$in0},[$inp]
+	aese		$dat,q10
+	aesmc		$dat,$dat
+	aese		$dat,q11
+	aesmc		$dat,$dat
+	aese		$dat,q12
+	aesmc		$dat,$dat
+	aese		$dat,q13
+	aesmc		$dat,$dat
+	aese		$dat,q14
+	aesmc		$dat,$dat
+	 veor		$in0,$in0,$rndlast
+	aese		$dat,q15
+
+	veor		$in0,$in0,$dat
+	vst1.8		{$in0},[$out]
+
+.Lctr32_done:
+___
+$code.=<<___	if ($flavour !~ /64/);
+	vldmia		sp!,{d8-d15}
+	ldmia		sp!,{r4-r10,pc}
+___
+$code.=<<___	if ($flavour =~ /64/);
+	ldr		x29,[sp],#16
+	ret
+___
+$code.=<<___;
+.size	${prefix}_ctr32_encrypt_blocks,.-${prefix}_ctr32_encrypt_blocks
+___
+}}}
+$code.=<<___;
+#endif
+___
+########################################
+if ($flavour =~ /64/) {			######## 64-bit code
+    my %opcode = (
+	"aesd"	=>	0x4e285800,	"aese"	=>	0x4e284800,
+	"aesimc"=>	0x4e287800,	"aesmc"	=>	0x4e286800	);
+
+    local *unaes = sub {
+	my ($mnemonic,$arg)=@_;
+
+	$arg =~ m/[qv]([0-9]+)[^,]*,\s*[qv]([0-9]+)/o	&&
+	sprintf ".inst\t0x%08x\t//%s %s",
+			$opcode{$mnemonic}|$1|($2<<5),
+			$mnemonic,$arg;
+    };
+
+    foreach(split("\n",$code)) {
+        s/\`([^\`]*)\`/eval($1)/geo;
+
+	s/\bq([0-9]+)\b/"v".($1<8?$1:$1+8).".16b"/geo;	# old->new registers
+        s/@\s/\/\//o;			# old->new style commentary
+
+	#s/[v]?(aes\w+)\s+([qv].*)/unaes($1,$2)/geo	or
+	s/cclr\s+([wx])([^,]+),\s*([a-z]+)/csel	$1$2,$1zr,$1$2,$3/o	or
+        s/vmov\.i8/movi/o	or	# fix up legacy mnemonics
+        s/vext\.8/ext/o		or
+        s/vrev32\.8/rev32/o	or
+        s/vtst\.8/cmtst/o	or
+        s/vshr/ushr/o		or
+        s/^(\s+)v/$1/o		or	# strip off v prefix
+	s/\bbx\s+lr\b/ret/o;
+
+	# fix up remainig legacy suffixes
+	s/\.[ui]?8//o;
+	m/\],#8/o and s/\.16b/\.8b/go;
+        s/\.[ui]?32//o and s/\.16b/\.4s/go;
+        s/\.[ui]?64//o and s/\.16b/\.2d/go;
+	s/\.[42]([sd])\[([0-3])\]/\.$1\[$2\]/o;
+
+        print $_,"\n";
+    }
+} else {				######## 32-bit code
+    my %opcode = (
+	"aesd"	=>	0xf3b00340,	"aese"	=>	0xf3b00300,
+	"aesimc"=>	0xf3b003c0,	"aesmc"	=>	0xf3b00380	);
+
+    local *unaes = sub {
+	my ($mnemonic,$arg)=@_;
+
+	if ($arg =~ m/[qv]([0-9]+)[^,]*,\s*[qv]([0-9]+)/o) {
+	    my $word = $opcode{$mnemonic}|(($1&7)<<13)|(($1&8)<<19)
+					 |(($2&7)<<1) |(($2&8)<<2);
+	    # since ARMv7 instructions are always encoded little-endian.
+	    # correct solution is to use .inst directive, but older
+	    # assemblers don't implement it:-(
+	    sprintf ".byte\t0x%02x,0x%02x,0x%02x,0x%02x\t@ %s %s",
+			$word&0xff,($word>>8)&0xff,
+			($word>>16)&0xff,($word>>24)&0xff,
+			$mnemonic,$arg;
+	}
+    };
+
+    sub unvtbl {
+	my $arg=shift;
+
+	$arg =~ m/q([0-9]+),\s*\{q([0-9]+)\},\s*q([0-9]+)/o &&
+	sprintf	"vtbl.8	d%d,{q%d},d%d\n\t".
+		"vtbl.8	d%d,{q%d},d%d", 2*$1,$2,2*$3, 2*$1+1,$2,2*$3+1;	
+    }
+
+    sub unvdup32 {
+	my $arg=shift;
+
+	$arg =~ m/q([0-9]+),\s*q([0-9]+)\[([0-3])\]/o &&
+	sprintf	"vdup.32	q%d,d%d[%d]",$1,2*$2+($3>>1),$3&1;	
+    }
+
+    sub unvmov32 {
+	my $arg=shift;
+
+	$arg =~ m/q([0-9]+)\[([0-3])\],(.*)/o &&
+	sprintf	"vmov.32	d%d[%d],%s",2*$1+($2>>1),$2&1,$3;	
+    }
+
+    foreach(split("\n",$code)) {
+        s/\`([^\`]*)\`/eval($1)/geo;
+
+	s/\b[wx]([0-9]+)\b/r$1/go;		# new->old registers
+	s/\bv([0-9])\.[12468]+[bsd]\b/q$1/go;	# new->old registers
+        s/\/\/\s?/@ /o;				# new->old style commentary
+
+	# fix up remainig new-style suffixes
+	s/\{q([0-9]+)\},\s*\[(.+)\],#8/sprintf "{d%d},[$2]!",2*$1/eo	or
+	s/\],#[0-9]+/]!/o;
+
+	s/[v]?(aes\w+)\s+([qv].*)/unaes($1,$2)/geo	or
+	s/cclr\s+([^,]+),\s*([a-z]+)/mov$2	$1,#0/o	or
+	s/vtbl\.8\s+(.*)/unvtbl($1)/geo			or
+	s/vdup\.32\s+(.*)/unvdup32($1)/geo		or
+	s/vmov\.32\s+(.*)/unvmov32($1)/geo		or
+	s/^(\s+)b\./$1b/o				or
+	s/^(\s+)ret/$1bx\tlr/o;
+
+        print $_,"\n";
+    }
+}
+
+close STDOUT;

jni/openssl/crypto/aes/asm/vpaes-x86.S

@@ -0,0 +1,661 @@
+.file	"vpaes-x86.s"
+.text
+.align	64
+.L_vpaes_consts:
+.long	218628480,235210255,168496130,67568393
+.long	252381056,17041926,33884169,51187212
+.long	252645135,252645135,252645135,252645135
+.long	1512730624,3266504856,1377990664,3401244816
+.long	830229760,1275146365,2969422977,3447763452
+.long	3411033600,2979783055,338359620,2782886510
+.long	4209124096,907596821,221174255,1006095553
+.long	191964160,3799684038,3164090317,1589111125
+.long	182528256,1777043520,2877432650,3265356744
+.long	1874708224,3503451415,3305285752,363511674
+.long	1606117888,3487855781,1093350906,2384367825
+.long	197121,67569157,134941193,202313229
+.long	67569157,134941193,202313229,197121
+.long	134941193,202313229,197121,67569157
+.long	202313229,197121,67569157,134941193
+.long	33619971,100992007,168364043,235736079
+.long	235736079,33619971,100992007,168364043
+.long	168364043,235736079,33619971,100992007
+.long	100992007,168364043,235736079,33619971
+.long	50462976,117835012,185207048,252579084
+.long	252314880,51251460,117574920,184942860
+.long	184682752,252054788,50987272,118359308
+.long	118099200,185467140,251790600,50727180
+.long	2946363062,528716217,1300004225,1881839624
+.long	1532713819,1532713819,1532713819,1532713819
+.long	3602276352,4288629033,3737020424,4153884961
+.long	1354558464,32357713,2958822624,3775749553
+.long	1201988352,132424512,1572796698,503232858
+.long	2213177600,1597421020,4103937655,675398315
+.long	2749646592,4273543773,1511898873,121693092
+.long	3040248576,1103263732,2871565598,1608280554
+.long	2236667136,2588920351,482954393,64377734
+.long	3069987328,291237287,2117370568,3650299247
+.long	533321216,3573750986,2572112006,1401264716
+.long	1339849704,2721158661,548607111,3445553514
+.long	2128193280,3054596040,2183486460,1257083700
+.long	655635200,1165381986,3923443150,2344132524
+.long	190078720,256924420,290342170,357187870
+.long	1610966272,2263057382,4103205268,309794674
+.long	2592527872,2233205587,1335446729,3402964816
+.long	3973531904,3225098121,3002836325,1918774430
+.long	3870401024,2102906079,2284471353,4117666579
+.long	617007872,1021508343,366931923,691083277
+.long	2528395776,3491914898,2968704004,1613121270
+.long	3445188352,3247741094,844474987,4093578302
+.long	651481088,1190302358,1689581232,574775300
+.long	4289380608,206939853,2555985458,2489840491
+.long	2130264064,327674451,3566485037,3349835193
+.long	2470714624,316102159,3636825756,3393945945
+.byte	86,101,99,116,111,114,32,80,101,114,109,117,116,97,116,105
+.byte	111,110,32,65,69,83,32,102,111,114,32,120,56,54,47,83
+.byte	83,83,69,51,44,32,77,105,107,101,32,72,97,109,98,117
+.byte	114,103,32,40,83,116,97,110,102,111,114,100,32,85,110,105
+.byte	118,101,114,115,105,116,121,41,0
+.align	64
+.type	_vpaes_preheat,@function
+.align	16
+_vpaes_preheat:
+	addl	(%esp),%ebp
+	movdqa	-48(%ebp),%xmm7
+	movdqa	-16(%ebp),%xmm6
+	ret
+.size	_vpaes_preheat,.-_vpaes_preheat
+.type	_vpaes_encrypt_core,@function
+.align	16
+_vpaes_encrypt_core:
+	movl	$16,%ecx
+	movl	240(%edx),%eax
+	movdqa	%xmm6,%xmm1
+	movdqa	(%ebp),%xmm2
+	pandn	%xmm0,%xmm1
+	movdqu	(%edx),%xmm5
+	psrld	$4,%xmm1
+	pand	%xmm6,%xmm0
+.byte	102,15,56,0,208
+	movdqa	16(%ebp),%xmm0
+.byte	102,15,56,0,193
+	pxor	%xmm5,%xmm2
+	pxor	%xmm2,%xmm0
+	addl	$16,%edx
+	leal	192(%ebp),%ebx
+	jmp	.L000enc_entry
+.align	16
+.L001enc_loop:
+	movdqa	32(%ebp),%xmm4
+.byte	102,15,56,0,226
+	pxor	%xmm5,%xmm4
+	movdqa	48(%ebp),%xmm0
+.byte	102,15,56,0,195
+	pxor	%xmm4,%xmm0
+	movdqa	64(%ebp),%xmm5
+.byte	102,15,56,0,234
+	movdqa	-64(%ebx,%ecx,1),%xmm1
+	movdqa	80(%ebp),%xmm2
+.byte	102,15,56,0,211
+	pxor	%xmm5,%xmm2
+	movdqa	(%ebx,%ecx,1),%xmm4
+	movdqa	%xmm0,%xmm3
+.byte	102,15,56,0,193
+	addl	$16,%edx
+	pxor	%xmm2,%xmm0
+.byte	102,15,56,0,220
+	addl	$16,%ecx
+	pxor	%xmm0,%xmm3
+.byte	102,15,56,0,193
+	andl	$48,%ecx
+	pxor	%xmm3,%xmm0
+	subl	$1,%eax
+.L000enc_entry:
+	movdqa	%xmm6,%xmm1
+	pandn	%xmm0,%xmm1
+	psrld	$4,%xmm1
+	pand	%xmm6,%xmm0
+	movdqa	-32(%ebp),%xmm5
+.byte	102,15,56,0,232
+	pxor	%xmm1,%xmm0
+	movdqa	%xmm7,%xmm3
+.byte	102,15,56,0,217
+	pxor	%xmm5,%xmm3
+	movdqa	%xmm7,%xmm4
+.byte	102,15,56,0,224
+	pxor	%xmm5,%xmm4
+	movdqa	%xmm7,%xmm2
+.byte	102,15,56,0,211
+	pxor	%xmm0,%xmm2
+	movdqa	%xmm7,%xmm3
+	movdqu	(%edx),%xmm5
+.byte	102,15,56,0,220
+	pxor	%xmm1,%xmm3
+	jnz	.L001enc_loop
+	movdqa	96(%ebp),%xmm4
+	movdqa	112(%ebp),%xmm0
+.byte	102,15,56,0,226
+	pxor	%xmm5,%xmm4
+.byte	102,15,56,0,195
+	movdqa	64(%ebx,%ecx,1),%xmm1
+	pxor	%xmm4,%xmm0
+.byte	102,15,56,0,193
+	ret
+.size	_vpaes_encrypt_core,.-_vpaes_encrypt_core
+.type	_vpaes_decrypt_core,@function
+.align	16
+_vpaes_decrypt_core:
+	movl	240(%edx),%eax
+	leal	608(%ebp),%ebx
+	movdqa	%xmm6,%xmm1
+	movdqa	-64(%ebx),%xmm2
+	pandn	%xmm0,%xmm1
+	movl	%eax,%ecx
+	psrld	$4,%xmm1
+	movdqu	(%edx),%xmm5
+	shll	$4,%ecx
+	pand	%xmm6,%xmm0
+.byte	102,15,56,0,208
+	movdqa	-48(%ebx),%xmm0
+	xorl	$48,%ecx
+.byte	102,15,56,0,193
+	andl	$48,%ecx
+	pxor	%xmm5,%xmm2
+	movdqa	176(%ebp),%xmm5
+	pxor	%xmm2,%xmm0
+	addl	$16,%edx
+	leal	-352(%ebx,%ecx,1),%ecx
+	jmp	.L002dec_entry
+.align	16
+.L003dec_loop:
+	movdqa	-32(%ebx),%xmm4
+.byte	102,15,56,0,226
+	pxor	%xmm0,%xmm4
+	movdqa	-16(%ebx),%xmm0
+.byte	102,15,56,0,195
+	pxor	%xmm4,%xmm0
+	addl	$16,%edx
+.byte	102,15,56,0,197
+	movdqa	(%ebx),%xmm4
+.byte	102,15,56,0,226
+	pxor	%xmm0,%xmm4
+	movdqa	16(%ebx),%xmm0
+.byte	102,15,56,0,195
+	pxor	%xmm4,%xmm0
+	subl	$1,%eax
+.byte	102,15,56,0,197
+	movdqa	32(%ebx),%xmm4
+.byte	102,15,56,0,226
+	pxor	%xmm0,%xmm4
+	movdqa	48(%ebx),%xmm0
+.byte	102,15,56,0,195
+	pxor	%xmm4,%xmm0
+.byte	102,15,56,0,197
+	movdqa	64(%ebx),%xmm4
+.byte	102,15,56,0,226
+	pxor	%xmm0,%xmm4
+	movdqa	80(%ebx),%xmm0
+.byte	102,15,56,0,195
+	pxor	%xmm4,%xmm0
+.byte	102,15,58,15,237,12
+.L002dec_entry:
+	movdqa	%xmm6,%xmm1
+	pandn	%xmm0,%xmm1
+	psrld	$4,%xmm1
+	pand	%xmm6,%xmm0
+	movdqa	-32(%ebp),%xmm2
+.byte	102,15,56,0,208
+	pxor	%xmm1,%xmm0
+	movdqa	%xmm7,%xmm3
+.byte	102,15,56,0,217
+	pxor	%xmm2,%xmm3
+	movdqa	%xmm7,%xmm4
+.byte	102,15,56,0,224
+	pxor	%xmm2,%xmm4
+	movdqa	%xmm7,%xmm2
+.byte	102,15,56,0,211
+	pxor	%xmm0,%xmm2
+	movdqa	%xmm7,%xmm3
+.byte	102,15,56,0,220
+	pxor	%xmm1,%xmm3
+	movdqu	(%edx),%xmm0
+	jnz	.L003dec_loop
+	movdqa	96(%ebx),%xmm4
+.byte	102,15,56,0,226
+	pxor	%xmm0,%xmm4
+	movdqa	112(%ebx),%xmm0
+	movdqa	(%ecx),%xmm2
+.byte	102,15,56,0,195
+	pxor	%xmm4,%xmm0
+.byte	102,15,56,0,194
+	ret
+.size	_vpaes_decrypt_core,.-_vpaes_decrypt_core
+.type	_vpaes_schedule_core,@function
+.align	16
+_vpaes_schedule_core:
+	addl	(%esp),%ebp
+	movdqu	(%esi),%xmm0
+	movdqa	320(%ebp),%xmm2
+	movdqa	%xmm0,%xmm3
+	leal	(%ebp),%ebx
+	movdqa	%xmm2,4(%esp)
+	call	_vpaes_schedule_transform
+	movdqa	%xmm0,%xmm7
+	testl	%edi,%edi
+	jnz	.L004schedule_am_decrypting
+	movdqu	%xmm0,(%edx)
+	jmp	.L005schedule_go
+.L004schedule_am_decrypting:
+	movdqa	256(%ebp,%ecx,1),%xmm1
+.byte	102,15,56,0,217
+	movdqu	%xmm3,(%edx)
+	xorl	$48,%ecx
+.L005schedule_go:
+	cmpl	$192,%eax
+	ja	.L006schedule_256
+	je	.L007schedule_192
+.L008schedule_128:
+	movl	$10,%eax
+.L009loop_schedule_128:
+	call	_vpaes_schedule_round
+	decl	%eax
+	jz	.L010schedule_mangle_last
+	call	_vpaes_schedule_mangle
+	jmp	.L009loop_schedule_128
+.align	16
+.L007schedule_192:
+	movdqu	8(%esi),%xmm0
+	call	_vpaes_schedule_transform
+	movdqa	%xmm0,%xmm6
+	pxor	%xmm4,%xmm4
+	movhlps	%xmm4,%xmm6
+	movl	$4,%eax
+.L011loop_schedule_192:
+	call	_vpaes_schedule_round
+.byte	102,15,58,15,198,8
+	call	_vpaes_schedule_mangle
+	call	_vpaes_schedule_192_smear
+	call	_vpaes_schedule_mangle
+	call	_vpaes_schedule_round
+	decl	%eax
+	jz	.L010schedule_mangle_last
+	call	_vpaes_schedule_mangle
+	call	_vpaes_schedule_192_smear
+	jmp	.L011loop_schedule_192
+.align	16
+.L006schedule_256:
+	movdqu	16(%esi),%xmm0
+	call	_vpaes_schedule_transform
+	movl	$7,%eax
+.L012loop_schedule_256:
+	call	_vpaes_schedule_mangle
+	movdqa	%xmm0,%xmm6
+	call	_vpaes_schedule_round
+	decl	%eax
+	jz	.L010schedule_mangle_last
+	call	_vpaes_schedule_mangle
+	pshufd	$255,%xmm0,%xmm0
+	movdqa	%xmm7,20(%esp)
+	movdqa	%xmm6,%xmm7
+	call	.L_vpaes_schedule_low_round
+	movdqa	20(%esp),%xmm7
+	jmp	.L012loop_schedule_256
+.align	16
+.L010schedule_mangle_last:
+	leal	384(%ebp),%ebx
+	testl	%edi,%edi
+	jnz	.L013schedule_mangle_last_dec
+	movdqa	256(%ebp,%ecx,1),%xmm1
+.byte	102,15,56,0,193
+	leal	352(%ebp),%ebx
+	addl	$32,%edx
+.L013schedule_mangle_last_dec:
+	addl	$-16,%edx
+	pxor	336(%ebp),%xmm0
+	call	_vpaes_schedule_transform
+	movdqu	%xmm0,(%edx)
+	pxor	%xmm0,%xmm0
+	pxor	%xmm1,%xmm1
+	pxor	%xmm2,%xmm2
+	pxor	%xmm3,%xmm3
+	pxor	%xmm4,%xmm4
+	pxor	%xmm5,%xmm5
+	pxor	%xmm6,%xmm6
+	pxor	%xmm7,%xmm7
+	ret
+.size	_vpaes_schedule_core,.-_vpaes_schedule_core
+.type	_vpaes_schedule_192_smear,@function
+.align	16
+_vpaes_schedule_192_smear:
+	pshufd	$128,%xmm6,%xmm0
+	pxor	%xmm0,%xmm6
+	pshufd	$254,%xmm7,%xmm0
+	pxor	%xmm0,%xmm6
+	movdqa	%xmm6,%xmm0
+	pxor	%xmm1,%xmm1
+	movhlps	%xmm1,%xmm6
+	ret
+.size	_vpaes_schedule_192_smear,.-_vpaes_schedule_192_smear
+.type	_vpaes_schedule_round,@function
+.align	16
+_vpaes_schedule_round:
+	movdqa	8(%esp),%xmm2
+	pxor	%xmm1,%xmm1
+.byte	102,15,58,15,202,15
+.byte	102,15,58,15,210,15
+	pxor	%xmm1,%xmm7
+	pshufd	$255,%xmm0,%xmm0
+.byte	102,15,58,15,192,1
+	movdqa	%xmm2,8(%esp)
+.L_vpaes_schedule_low_round:
+	movdqa	%xmm7,%xmm1
+	pslldq	$4,%xmm7
+	pxor	%xmm1,%xmm7
+	movdqa	%xmm7,%xmm1
+	pslldq	$8,%xmm7
+	pxor	%xmm1,%xmm7
+	pxor	336(%ebp),%xmm7
+	movdqa	-16(%ebp),%xmm4
+	movdqa	-48(%ebp),%xmm5
+	movdqa	%xmm4,%xmm1
+	pandn	%xmm0,%xmm1
+	psrld	$4,%xmm1
+	pand	%xmm4,%xmm0
+	movdqa	-32(%ebp),%xmm2
+.byte	102,15,56,0,208
+	pxor	%xmm1,%xmm0
+	movdqa	%xmm5,%xmm3
+.byte	102,15,56,0,217
+	pxor	%xmm2,%xmm3
+	movdqa	%xmm5,%xmm4
+.byte	102,15,56,0,224
+	pxor	%xmm2,%xmm4
+	movdqa	%xmm5,%xmm2
+.byte	102,15,56,0,211
+	pxor	%xmm0,%xmm2
+	movdqa	%xmm5,%xmm3
+.byte	102,15,56,0,220
+	pxor	%xmm1,%xmm3
+	movdqa	32(%ebp),%xmm4
+.byte	102,15,56,0,226
+	movdqa	48(%ebp),%xmm0
+.byte	102,15,56,0,195
+	pxor	%xmm4,%xmm0
+	pxor	%xmm7,%xmm0
+	movdqa	%xmm0,%xmm7
+	ret
+.size	_vpaes_schedule_round,.-_vpaes_schedule_round
+.type	_vpaes_schedule_transform,@function
+.align	16
+_vpaes_schedule_transform:
+	movdqa	-16(%ebp),%xmm2
+	movdqa	%xmm2,%xmm1
+	pandn	%xmm0,%xmm1
+	psrld	$4,%xmm1
+	pand	%xmm2,%xmm0
+	movdqa	(%ebx),%xmm2
+.byte	102,15,56,0,208
+	movdqa	16(%ebx),%xmm0
+.byte	102,15,56,0,193
+	pxor	%xmm2,%xmm0
+	ret
+.size	_vpaes_schedule_transform,.-_vpaes_schedule_transform
+.type	_vpaes_schedule_mangle,@function
+.align	16
+_vpaes_schedule_mangle:
+	movdqa	%xmm0,%xmm4
+	movdqa	128(%ebp),%xmm5
+	testl	%edi,%edi
+	jnz	.L014schedule_mangle_dec
+	addl	$16,%edx
+	pxor	336(%ebp),%xmm4
+.byte	102,15,56,0,229
+	movdqa	%xmm4,%xmm3
+.byte	102,15,56,0,229
+	pxor	%xmm4,%xmm3
+.byte	102,15,56,0,229
+	pxor	%xmm4,%xmm3
+	jmp	.L015schedule_mangle_both
+.align	16
+.L014schedule_mangle_dec:
+	movdqa	-16(%ebp),%xmm2
+	leal	416(%ebp),%esi
+	movdqa	%xmm2,%xmm1
+	pandn	%xmm4,%xmm1
+	psrld	$4,%xmm1
+	pand	%xmm2,%xmm4
+	movdqa	(%esi),%xmm2
+.byte	102,15,56,0,212
+	movdqa	16(%esi),%xmm3
+.byte	102,15,56,0,217
+	pxor	%xmm2,%xmm3
+.byte	102,15,56,0,221
+	movdqa	32(%esi),%xmm2
+.byte	102,15,56,0,212
+	pxor	%xmm3,%xmm2
+	movdqa	48(%esi),%xmm3
+.byte	102,15,56,0,217
+	pxor	%xmm2,%xmm3
+.byte	102,15,56,0,221
+	movdqa	64(%esi),%xmm2
+.byte	102,15,56,0,212
+	pxor	%xmm3,%xmm2
+	movdqa	80(%esi),%xmm3
+.byte	102,15,56,0,217
+	pxor	%xmm2,%xmm3
+.byte	102,15,56,0,221
+	movdqa	96(%esi),%xmm2
+.byte	102,15,56,0,212
+	pxor	%xmm3,%xmm2
+	movdqa	112(%esi),%xmm3
+.byte	102,15,56,0,217
+	pxor	%xmm2,%xmm3
+	addl	$-16,%edx
+.L015schedule_mangle_both:
+	movdqa	256(%ebp,%ecx,1),%xmm1
+.byte	102,15,56,0,217
+	addl	$-16,%ecx
+	andl	$48,%ecx
+	movdqu	%xmm3,(%edx)
+	ret
+.size	_vpaes_schedule_mangle,.-_vpaes_schedule_mangle
+.globl	vpaes_set_encrypt_key
+.type	vpaes_set_encrypt_key,@function
+.align	16
+vpaes_set_encrypt_key:
+.L_vpaes_set_encrypt_key_begin:
+	pushl	%ebp
+	pushl	%ebx
+	pushl	%esi
+	pushl	%edi
+	movl	20(%esp),%esi
+	leal	-56(%esp),%ebx
+	movl	24(%esp),%eax
+	andl	$-16,%ebx
+	movl	28(%esp),%edx
+	xchgl	%esp,%ebx
+	movl	%ebx,48(%esp)
+	movl	%eax,%ebx
+	shrl	$5,%ebx
+	addl	$5,%ebx
+	movl	%ebx,240(%edx)
+	movl	$48,%ecx
+	movl	$0,%edi
+	leal	.L_vpaes_consts+0x30-.L016pic_point,%ebp
+	call	_vpaes_schedule_core
+.L016pic_point:
+	movl	48(%esp),%esp
+	xorl	%eax,%eax
+	popl	%edi
+	popl	%esi
+	popl	%ebx
+	popl	%ebp
+	ret
+.size	vpaes_set_encrypt_key,.-.L_vpaes_set_encrypt_key_begin
+.globl	vpaes_set_decrypt_key
+.type	vpaes_set_decrypt_key,@function
+.align	16
+vpaes_set_decrypt_key:
+.L_vpaes_set_decrypt_key_begin:
+	pushl	%ebp
+	pushl	%ebx
+	pushl	%esi
+	pushl	%edi
+	movl	20(%esp),%esi
+	leal	-56(%esp),%ebx
+	movl	24(%esp),%eax
+	andl	$-16,%ebx
+	movl	28(%esp),%edx
+	xchgl	%esp,%ebx
+	movl	%ebx,48(%esp)
+	movl	%eax,%ebx
+	shrl	$5,%ebx
+	addl	$5,%ebx
+	movl	%ebx,240(%edx)
+	shll	$4,%ebx
+	leal	16(%edx,%ebx,1),%edx
+	movl	$1,%edi
+	movl	%eax,%ecx
+	shrl	$1,%ecx
+	andl	$32,%ecx
+	xorl	$32,%ecx
+	leal	.L_vpaes_consts+0x30-.L017pic_point,%ebp
+	call	_vpaes_schedule_core
+.L017pic_point:
+	movl	48(%esp),%esp
+	xorl	%eax,%eax
+	popl	%edi
+	popl	%esi
+	popl	%ebx
+	popl	%ebp
+	ret
+.size	vpaes_set_decrypt_key,.-.L_vpaes_set_decrypt_key_begin
+.globl	vpaes_encrypt
+.type	vpaes_encrypt,@function
+.align	16
+vpaes_encrypt:
+.L_vpaes_encrypt_begin:
+	pushl	%ebp
+	pushl	%ebx
+	pushl	%esi
+	pushl	%edi
+	leal	.L_vpaes_consts+0x30-.L018pic_point,%ebp
+	call	_vpaes_preheat
+.L018pic_point:
+	movl	20(%esp),%esi
+	leal	-56(%esp),%ebx
+	movl	24(%esp),%edi
+	andl	$-16,%ebx
+	movl	28(%esp),%edx
+	xchgl	%esp,%ebx
+	movl	%ebx,48(%esp)
+	movdqu	(%esi),%xmm0
+	call	_vpaes_encrypt_core
+	movdqu	%xmm0,(%edi)
+	movl	48(%esp),%esp
+	popl	%edi
+	popl	%esi
+	popl	%ebx
+	popl	%ebp
+	ret
+.size	vpaes_encrypt,.-.L_vpaes_encrypt_begin
+.globl	vpaes_decrypt
+.type	vpaes_decrypt,@function
+.align	16
+vpaes_decrypt:
+.L_vpaes_decrypt_begin:
+	pushl	%ebp
+	pushl	%ebx
+	pushl	%esi
+	pushl	%edi
+	leal	.L_vpaes_consts+0x30-.L019pic_point,%ebp
+	call	_vpaes_preheat
+.L019pic_point:
+	movl	20(%esp),%esi
+	leal	-56(%esp),%ebx
+	movl	24(%esp),%edi
+	andl	$-16,%ebx
+	movl	28(%esp),%edx
+	xchgl	%esp,%ebx
+	movl	%ebx,48(%esp)
+	movdqu	(%esi),%xmm0
+	call	_vpaes_decrypt_core
+	movdqu	%xmm0,(%edi)
+	movl	48(%esp),%esp
+	popl	%edi
+	popl	%esi
+	popl	%ebx
+	popl	%ebp
+	ret
+.size	vpaes_decrypt,.-.L_vpaes_decrypt_begin
+.globl	vpaes_cbc_encrypt
+.type	vpaes_cbc_encrypt,@function
+.align	16
+vpaes_cbc_encrypt:
+.L_vpaes_cbc_encrypt_begin:
+	pushl	%ebp
+	pushl	%ebx
+	pushl	%esi
+	pushl	%edi
+	movl	20(%esp),%esi
+	movl	24(%esp),%edi
+	movl	28(%esp),%eax
+	movl	32(%esp),%edx
+	subl	$16,%eax
+	jc	.L020cbc_abort
+	leal	-56(%esp),%ebx
+	movl	36(%esp),%ebp
+	andl	$-16,%ebx
+	movl	40(%esp),%ecx
+	xchgl	%esp,%ebx
+	movdqu	(%ebp),%xmm1
+	subl	%esi,%edi
+	movl	%ebx,48(%esp)
+	movl	%edi,(%esp)
+	movl	%edx,4(%esp)
+	movl	%ebp,8(%esp)
+	movl	%eax,%edi
+	leal	.L_vpaes_consts+0x30-.L021pic_point,%ebp
+	call	_vpaes_preheat
+.L021pic_point:
+	cmpl	$0,%ecx
+	je	.L022cbc_dec_loop
+	jmp	.L023cbc_enc_loop
+.align	16
+.L023cbc_enc_loop:
+	movdqu	(%esi),%xmm0
+	pxor	%xmm1,%xmm0
+	call	_vpaes_encrypt_core
+	movl	(%esp),%ebx
+	movl	4(%esp),%edx
+	movdqa	%xmm0,%xmm1
+	movdqu	%xmm0,(%ebx,%esi,1)
+	leal	16(%esi),%esi
+	subl	$16,%edi
+	jnc	.L023cbc_enc_loop
+	jmp	.L024cbc_done
+.align	16
+.L022cbc_dec_loop:
+	movdqu	(%esi),%xmm0
+	movdqa	%xmm1,16(%esp)
+	movdqa	%xmm0,32(%esp)
+	call	_vpaes_decrypt_core
+	movl	(%esp),%ebx
+	movl	4(%esp),%edx
+	pxor	16(%esp),%xmm0
+	movdqa	32(%esp),%xmm1
+	movdqu	%xmm0,(%ebx,%esi,1)
+	leal	16(%esi),%esi
+	subl	$16,%edi
+	jnc	.L022cbc_dec_loop
+.L024cbc_done:
+	movl	8(%esp),%ebx
+	movl	48(%esp),%esp
+	movdqu	%xmm1,(%ebx)
+.L020cbc_abort:
+	popl	%edi
+	popl	%esi
+	popl	%ebx
+	popl	%ebp
+	ret
+.size	vpaes_cbc_encrypt,.-.L_vpaes_cbc_encrypt_begin

jni/openssl/crypto/aes/asm/vpaes-x86.pl

@@ -0,0 +1,903 @@
+#!/usr/bin/env perl
+
+######################################################################
+## Constant-time SSSE3 AES core implementation.
+## version 0.1
+##
+## By Mike Hamburg (Stanford University), 2009
+## Public domain.
+##
+## For details see http://shiftleft.org/papers/vector_aes/ and
+## http://crypto.stanford.edu/vpaes/.
+
+######################################################################
+# September 2011.
+#
+# Port vpaes-x86_64.pl as 32-bit "almost" drop-in replacement for
+# aes-586.pl. "Almost" refers to the fact that AES_cbc_encrypt
+# doesn't handle partial vectors (doesn't have to if called from
+# EVP only). "Drop-in" implies that this module doesn't share key
+# schedule structure with the original nor does it make assumption
+# about its alignment...
+#
+# Performance summary. aes-586.pl column lists large-block CBC
+# encrypt/decrypt/with-hyper-threading-off(*) results in cycles per
+# byte processed with 128-bit key, and vpaes-x86.pl column - [also
+# large-block CBC] encrypt/decrypt.
+#
+#		aes-586.pl		vpaes-x86.pl
+#
+# Core 2(**)	29.1/42.3/18.3		22.0/25.6(***)
+# Nehalem	27.9/40.4/18.1		10.3/12.0
+# Atom		102./119./60.1		64.5/85.3(***)
+#
+# (*)	"Hyper-threading" in the context refers rather to cache shared
+#	among multiple cores, than to specifically Intel HTT. As vast
+#	majority of contemporary cores share cache, slower code path
+#	is common place. In other words "with-hyper-threading-off"
+#	results are presented mostly for reference purposes.
+#
+# (**)	"Core 2" refers to initial 65nm design, a.k.a. Conroe.
+#
+# (***)	Less impressive improvement on Core 2 and Atom is due to slow
+#	pshufb,	yet it's respectable +32%/65%  improvement on Core 2
+#	and +58%/40% on Atom (as implied, over "hyper-threading-safe"
+#	code path).
+#
+#						<appro@openssl.org>
+
+$0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
+push(@INC,"${dir}","${dir}../../perlasm");
+require "x86asm.pl";
+
+&asm_init($ARGV[0],"vpaes-x86.pl",$x86only = $ARGV[$#ARGV] eq "386");
+
+$PREFIX="vpaes";
+
+my  ($round, $base, $magic, $key, $const, $inp, $out)=
+    ("eax",  "ebx", "ecx",  "edx","ebp",  "esi","edi");
+
+&static_label("_vpaes_consts");
+&static_label("_vpaes_schedule_low_round");
+
+&set_label("_vpaes_consts",64);
+$k_inv=-0x30;		# inv, inva
+	&data_word(0x0D080180,0x0E05060F,0x0A0B0C02,0x04070309);
+	&data_word(0x0F0B0780,0x01040A06,0x02050809,0x030D0E0C);
+
+$k_s0F=-0x10;		# s0F
+	&data_word(0x0F0F0F0F,0x0F0F0F0F,0x0F0F0F0F,0x0F0F0F0F);
+
+$k_ipt=0x00;		# input transform (lo, hi)
+	&data_word(0x5A2A7000,0xC2B2E898,0x52227808,0xCABAE090);
+	&data_word(0x317C4D00,0x4C01307D,0xB0FDCC81,0xCD80B1FC);
+
+$k_sb1=0x20;		# sb1u, sb1t
+	&data_word(0xCB503E00,0xB19BE18F,0x142AF544,0xA5DF7A6E);
+	&data_word(0xFAE22300,0x3618D415,0x0D2ED9EF,0x3BF7CCC1);
+$k_sb2=0x40;		# sb2u, sb2t
+	&data_word(0x0B712400,0xE27A93C6,0xBC982FCD,0x5EB7E955);
+	&data_word(0x0AE12900,0x69EB8840,0xAB82234A,0xC2A163C8);
+$k_sbo=0x60;		# sbou, sbot
+	&data_word(0x6FBDC700,0xD0D26D17,0xC502A878,0x15AABF7A);
+	&data_word(0x5FBB6A00,0xCFE474A5,0x412B35FA,0x8E1E90D1);
+
+$k_mc_forward=0x80;	# mc_forward
+	&data_word(0x00030201,0x04070605,0x080B0A09,0x0C0F0E0D);
+	&data_word(0x04070605,0x080B0A09,0x0C0F0E0D,0x00030201);
+	&data_word(0x080B0A09,0x0C0F0E0D,0x00030201,0x04070605);
+	&data_word(0x0C0F0E0D,0x00030201,0x04070605,0x080B0A09);
+
+$k_mc_backward=0xc0;	# mc_backward
+	&data_word(0x02010003,0x06050407,0x0A09080B,0x0E0D0C0F);
+	&data_word(0x0E0D0C0F,0x02010003,0x06050407,0x0A09080B);
+	&data_word(0x0A09080B,0x0E0D0C0F,0x02010003,0x06050407);
+	&data_word(0x06050407,0x0A09080B,0x0E0D0C0F,0x02010003);
+
+$k_sr=0x100;		# sr
+	&data_word(0x03020100,0x07060504,0x0B0A0908,0x0F0E0D0C);
+	&data_word(0x0F0A0500,0x030E0904,0x07020D08,0x0B06010C);
+	&data_word(0x0B020900,0x0F060D04,0x030A0108,0x070E050C);
+	&data_word(0x070A0D00,0x0B0E0104,0x0F020508,0x0306090C);
+
+$k_rcon=0x140;		# rcon
+	&data_word(0xAF9DEEB6,0x1F8391B9,0x4D7C7D81,0x702A9808);
+
+$k_s63=0x150;		# s63: all equal to 0x63 transformed
+	&data_word(0x5B5B5B5B,0x5B5B5B5B,0x5B5B5B5B,0x5B5B5B5B);
+
+$k_opt=0x160;		# output transform
+	&data_word(0xD6B66000,0xFF9F4929,0xDEBE6808,0xF7974121);
+	&data_word(0x50BCEC00,0x01EDBD51,0xB05C0CE0,0xE10D5DB1);
+
+$k_deskew=0x180;	# deskew tables: inverts the sbox's "skew"
+	&data_word(0x47A4E300,0x07E4A340,0x5DBEF91A,0x1DFEB95A);
+	&data_word(0x83EA6900,0x5F36B5DC,0xF49D1E77,0x2841C2AB);
+##
+##  Decryption stuff
+##  Key schedule constants
+##
+$k_dksd=0x1a0;		# decryption key schedule: invskew x*D
+	&data_word(0xA3E44700,0xFEB91A5D,0x5A1DBEF9,0x0740E3A4);
+	&data_word(0xB5368300,0x41C277F4,0xAB289D1E,0x5FDC69EA);
+$k_dksb=0x1c0;		# decryption key schedule: invskew x*B
+	&data_word(0x8550D500,0x9A4FCA1F,0x1CC94C99,0x03D65386);
+	&data_word(0xB6FC4A00,0x115BEDA7,0x7E3482C8,0xD993256F);
+$k_dkse=0x1e0;		# decryption key schedule: invskew x*E + 0x63
+	&data_word(0x1FC9D600,0xD5031CCA,0x994F5086,0x53859A4C);
+	&data_word(0x4FDC7BE8,0xA2319605,0x20B31487,0xCD5EF96A);
+$k_dks9=0x200;		# decryption key schedule: invskew x*9
+	&data_word(0x7ED9A700,0xB6116FC8,0x82255BFC,0x4AED9334);
+	&data_word(0x27143300,0x45765162,0xE9DAFDCE,0x8BB89FAC);
+
+##
+##  Decryption stuff
+##  Round function constants
+##
+$k_dipt=0x220;		# decryption input transform
+	&data_word(0x0B545F00,0x0F505B04,0x114E451A,0x154A411E);
+	&data_word(0x60056500,0x86E383E6,0xF491F194,0x12771772);
+
+$k_dsb9=0x240;		# decryption sbox output *9*u, *9*t
+	&data_word(0x9A86D600,0x851C0353,0x4F994CC9,0xCAD51F50);
+	&data_word(0xECD74900,0xC03B1789,0xB2FBA565,0x725E2C9E);
+$k_dsbd=0x260;		# decryption sbox output *D*u, *D*t
+	&data_word(0xE6B1A200,0x7D57CCDF,0x882A4439,0xF56E9B13);
+	&data_word(0x24C6CB00,0x3CE2FAF7,0x15DEEFD3,0x2931180D);
+$k_dsbb=0x280;		# decryption sbox output *B*u, *B*t
+	&data_word(0x96B44200,0xD0226492,0xB0F2D404,0x602646F6);
+	&data_word(0xCD596700,0xC19498A6,0x3255AA6B,0xF3FF0C3E);
+$k_dsbe=0x2a0;		# decryption sbox output *E*u, *E*t
+	&data_word(0x26D4D000,0x46F29296,0x64B4F6B0,0x22426004);
+	&data_word(0xFFAAC100,0x0C55A6CD,0x98593E32,0x9467F36B);
+$k_dsbo=0x2c0;		# decryption sbox final output
+	&data_word(0x7EF94000,0x1387EA53,0xD4943E2D,0xC7AA6DB9);
+	&data_word(0x93441D00,0x12D7560F,0xD8C58E9C,0xCA4B8159);
+&asciz	("Vector Permutation AES for x86/SSSE3, Mike Hamburg (Stanford University)");
+&align	(64);
+
+&function_begin_B("_vpaes_preheat");
+	&add	($const,&DWP(0,"esp"));
+	&movdqa	("xmm7",&QWP($k_inv,$const));
+	&movdqa	("xmm6",&QWP($k_s0F,$const));
+	&ret	();
+&function_end_B("_vpaes_preheat");
+
+##
+##  _aes_encrypt_core
+##
+##  AES-encrypt %xmm0.
+##
+##  Inputs:
+##     %xmm0 = input
+##     %xmm6-%xmm7 as in _vpaes_preheat
+##    (%edx) = scheduled keys
+##
+##  Output in %xmm0
+##  Clobbers  %xmm1-%xmm5, %eax, %ebx, %ecx, %edx
+##
+##
+&function_begin_B("_vpaes_encrypt_core");
+	&mov	($magic,16);
+	&mov	($round,&DWP(240,$key));
+	&movdqa	("xmm1","xmm6")
+	&movdqa	("xmm2",&QWP($k_ipt,$const));
+	&pandn	("xmm1","xmm0");
+	&movdqu	("xmm5",&QWP(0,$key));
+	&psrld	("xmm1",4);
+	&pand	("xmm0","xmm6");
+	&pshufb	("xmm2","xmm0");
+	&movdqa	("xmm0",&QWP($k_ipt+16,$const));
+	&pshufb	("xmm0","xmm1");
+	&pxor	("xmm2","xmm5");
+	&pxor	("xmm0","xmm2");
+	&add	($key,16);
+	&lea	($base,&DWP($k_mc_backward,$const));
+	&jmp	(&label("enc_entry"));
+
+
+&set_label("enc_loop",16);
+	# middle of middle round
+	&movdqa	("xmm4",&QWP($k_sb1,$const));	# 4 : sb1u
+	&pshufb	("xmm4","xmm2");		# 4 = sb1u
+	&pxor	("xmm4","xmm5");		# 4 = sb1u + k
+	&movdqa	("xmm0",&QWP($k_sb1+16,$const));# 0 : sb1t
+	&pshufb	("xmm0","xmm3");		# 0 = sb1t
+	&pxor	("xmm0","xmm4");		# 0 = A
+	&movdqa	("xmm5",&QWP($k_sb2,$const));	# 4 : sb2u
+	&pshufb	("xmm5","xmm2");		# 4 = sb2u
+	&movdqa	("xmm1",&QWP(-0x40,$base,$magic));# .Lk_mc_forward[]
+	&movdqa	("xmm2",&QWP($k_sb2+16,$const));# 2 : sb2t
+	&pshufb	("xmm2","xmm3");		# 2 = sb2t
+	&pxor	("xmm2","xmm5");		# 2 = 2A
+	&movdqa	("xmm4",&QWP(0,$base,$magic));	# .Lk_mc_backward[]
+	&movdqa	("xmm3","xmm0");		# 3 = A
+	&pshufb	("xmm0","xmm1");		# 0 = B
+	&add	($key,16);			# next key
+	&pxor	("xmm0","xmm2");		# 0 = 2A+B
+	&pshufb	("xmm3","xmm4");		# 3 = D
+	&add	($magic,16);			# next mc
+	&pxor	("xmm3","xmm0");		# 3 = 2A+B+D
+	&pshufb	("xmm0","xmm1");		# 0 = 2B+C
+	&and	($magic,0x30);			# ... mod 4
+	&pxor	("xmm0","xmm3");		# 0 = 2A+3B+C+D
+	&sub	($round,1);			# nr--
+
+&set_label("enc_entry");
+	# top of round
+	&movdqa	("xmm1","xmm6");		# 1 : i
+	&pandn	("xmm1","xmm0");		# 1 = i<<4
+	&psrld	("xmm1",4);			# 1 = i
+	&pand	("xmm0","xmm6");		# 0 = k
+	&movdqa	("xmm5",&QWP($k_inv+16,$const));# 2 : a/k
+	&pshufb	("xmm5","xmm0");		# 2 = a/k
+	&pxor	("xmm0","xmm1");		# 0 = j
+	&movdqa	("xmm3","xmm7");		# 3 : 1/i
+	&pshufb	("xmm3","xmm1");		# 3 = 1/i
+	&pxor	("xmm3","xmm5");		# 3 = iak = 1/i + a/k
+	&movdqa	("xmm4","xmm7");		# 4 : 1/j
+	&pshufb	("xmm4","xmm0");		# 4 = 1/j
+	&pxor	("xmm4","xmm5");		# 4 = jak = 1/j + a/k
+	&movdqa	("xmm2","xmm7");		# 2 : 1/iak
+	&pshufb	("xmm2","xmm3");		# 2 = 1/iak
+	&pxor	("xmm2","xmm0");		# 2 = io
+	&movdqa	("xmm3","xmm7");		# 3 : 1/jak
+	&movdqu	("xmm5",&QWP(0,$key));
+	&pshufb	("xmm3","xmm4");		# 3 = 1/jak
+	&pxor	("xmm3","xmm1");		# 3 = jo
+	&jnz	(&label("enc_loop"));
+
+	# middle of last round
+	&movdqa	("xmm4",&QWP($k_sbo,$const));	# 3 : sbou      .Lk_sbo
+	&movdqa	("xmm0",&QWP($k_sbo+16,$const));# 3 : sbot      .Lk_sbo+16
+	&pshufb	("xmm4","xmm2");		# 4 = sbou
+	&pxor	("xmm4","xmm5");		# 4 = sb1u + k
+	&pshufb	("xmm0","xmm3");		# 0 = sb1t
+	&movdqa	("xmm1",&QWP(0x40,$base,$magic));# .Lk_sr[]
+	&pxor	("xmm0","xmm4");		# 0 = A
+	&pshufb	("xmm0","xmm1");
+	&ret	();
+&function_end_B("_vpaes_encrypt_core");
+
+##
+##  Decryption core
+##
+##  Same API as encryption core.
+##
+&function_begin_B("_vpaes_decrypt_core");
+	&mov	($round,&DWP(240,$key));
+	&lea	($base,&DWP($k_dsbd,$const));
+	&movdqa	("xmm1","xmm6");
+	&movdqa	("xmm2",&QWP($k_dipt-$k_dsbd,$base));
+	&pandn	("xmm1","xmm0");
+	&mov	($magic,$round);
+	&psrld	("xmm1",4)
+	&movdqu	("xmm5",&QWP(0,$key));
+	&shl	($magic,4);
+	&pand	("xmm0","xmm6");
+	&pshufb	("xmm2","xmm0");
+	&movdqa	("xmm0",&QWP($k_dipt-$k_dsbd+16,$base));
+	&xor	($magic,0x30);
+	&pshufb	("xmm0","xmm1");
+	&and	($magic,0x30);
+	&pxor	("xmm2","xmm5");
+	&movdqa	("xmm5",&QWP($k_mc_forward+48,$const));
+	&pxor	("xmm0","xmm2");
+	&add	($key,16);
+	&lea	($magic,&DWP($k_sr-$k_dsbd,$base,$magic));
+	&jmp	(&label("dec_entry"));
+
+&set_label("dec_loop",16);
+##
+##  Inverse mix columns
+##
+	&movdqa	("xmm4",&QWP(-0x20,$base));	# 4 : sb9u
+	&pshufb	("xmm4","xmm2");		# 4 = sb9u
+	&pxor	("xmm4","xmm0");
+	&movdqa	("xmm0",&QWP(-0x10,$base));	# 0 : sb9t
+	&pshufb	("xmm0","xmm3");		# 0 = sb9t
+	&pxor	("xmm0","xmm4");		# 0 = ch
+	&add	($key,16);			# next round key
+
+	&pshufb	("xmm0","xmm5");		# MC ch
+	&movdqa	("xmm4",&QWP(0,$base));		# 4 : sbdu
+	&pshufb	("xmm4","xmm2");		# 4 = sbdu
+	&pxor	("xmm4","xmm0");		# 4 = ch
+	&movdqa	("xmm0",&QWP(0x10,$base));	# 0 : sbdt
+	&pshufb	("xmm0","xmm3");		# 0 = sbdt
+	&pxor	("xmm0","xmm4");		# 0 = ch
+	&sub	($round,1);			# nr--
+
+	&pshufb	("xmm0","xmm5");		# MC ch
+	&movdqa	("xmm4",&QWP(0x20,$base));	# 4 : sbbu
+	&pshufb	("xmm4","xmm2");		# 4 = sbbu
+	&pxor	("xmm4","xmm0");		# 4 = ch
+	&movdqa	("xmm0",&QWP(0x30,$base));	# 0 : sbbt
+	&pshufb	("xmm0","xmm3");		# 0 = sbbt
+	&pxor	("xmm0","xmm4");		# 0 = ch
+
+	&pshufb	("xmm0","xmm5");		# MC ch
+	&movdqa	("xmm4",&QWP(0x40,$base));	# 4 : sbeu
+	&pshufb	("xmm4","xmm2");		# 4 = sbeu
+	&pxor	("xmm4","xmm0");		# 4 = ch
+	&movdqa	("xmm0",&QWP(0x50,$base));	# 0 : sbet
+	&pshufb	("xmm0","xmm3");		# 0 = sbet
+	&pxor	("xmm0","xmm4");		# 0 = ch
+
+	&palignr("xmm5","xmm5",12);
+
+&set_label("dec_entry");
+	# top of round
+	&movdqa	("xmm1","xmm6");		# 1 : i
+	&pandn	("xmm1","xmm0");		# 1 = i<<4
+	&psrld	("xmm1",4);			# 1 = i
+	&pand	("xmm0","xmm6");		# 0 = k
+	&movdqa	("xmm2",&QWP($k_inv+16,$const));# 2 : a/k
+	&pshufb	("xmm2","xmm0");		# 2 = a/k
+	&pxor	("xmm0","xmm1");		# 0 = j
+	&movdqa	("xmm3","xmm7");		# 3 : 1/i
+	&pshufb	("xmm3","xmm1");		# 3 = 1/i
+	&pxor	("xmm3","xmm2");		# 3 = iak = 1/i + a/k
+	&movdqa	("xmm4","xmm7");		# 4 : 1/j
+	&pshufb	("xmm4","xmm0");		# 4 = 1/j
+	&pxor	("xmm4","xmm2");		# 4 = jak = 1/j + a/k
+	&movdqa	("xmm2","xmm7");		# 2 : 1/iak
+	&pshufb	("xmm2","xmm3");		# 2 = 1/iak
+	&pxor	("xmm2","xmm0");		# 2 = io
+	&movdqa	("xmm3","xmm7");		# 3 : 1/jak
+	&pshufb	("xmm3","xmm4");		# 3 = 1/jak
+	&pxor	("xmm3","xmm1");		# 3 = jo
+	&movdqu	("xmm0",&QWP(0,$key));
+	&jnz	(&label("dec_loop"));
+
+	# middle of last round
+	&movdqa	("xmm4",&QWP(0x60,$base));	# 3 : sbou
+	&pshufb	("xmm4","xmm2");		# 4 = sbou
+	&pxor	("xmm4","xmm0");		# 4 = sb1u + k
+	&movdqa	("xmm0",&QWP(0x70,$base));	# 0 : sbot
+	&movdqa	("xmm2",&QWP(0,$magic));
+	&pshufb	("xmm0","xmm3");		# 0 = sb1t
+	&pxor	("xmm0","xmm4");		# 0 = A
+	&pshufb	("xmm0","xmm2");
+	&ret	();
+&function_end_B("_vpaes_decrypt_core");
+
+########################################################
+##                                                    ##
+##                  AES key schedule                  ##
+##                                                    ##
+########################################################
+&function_begin_B("_vpaes_schedule_core");
+	&add	($const,&DWP(0,"esp"));
+	&movdqu	("xmm0",&QWP(0,$inp));		# load key (unaligned)
+	&movdqa	("xmm2",&QWP($k_rcon,$const));	# load rcon
+
+	# input transform
+	&movdqa	("xmm3","xmm0");
+	&lea	($base,&DWP($k_ipt,$const));
+	&movdqa	(&QWP(4,"esp"),"xmm2");		# xmm8
+	&call	("_vpaes_schedule_transform");
+	&movdqa	("xmm7","xmm0");
+
+	&test	($out,$out);
+	&jnz	(&label("schedule_am_decrypting"));
+
+	# encrypting, output zeroth round key after transform
+	&movdqu	(&QWP(0,$key),"xmm0");
+	&jmp	(&label("schedule_go"));
+
+&set_label("schedule_am_decrypting");
+	# decrypting, output zeroth round key after shiftrows
+	&movdqa	("xmm1",&QWP($k_sr,$const,$magic));
+	&pshufb	("xmm3","xmm1");
+	&movdqu	(&QWP(0,$key),"xmm3");
+	&xor	($magic,0x30);
+
+&set_label("schedule_go");
+	&cmp	($round,192);
+	&ja	(&label("schedule_256"));
+	&je	(&label("schedule_192"));
+	# 128: fall though
+
+##
+##  .schedule_128
+##
+##  128-bit specific part of key schedule.
+##
+##  This schedule is really simple, because all its parts
+##  are accomplished by the subroutines.
+##
+&set_label("schedule_128");
+	&mov	($round,10);
+
+&set_label("loop_schedule_128");
+	&call	("_vpaes_schedule_round");
+	&dec	($round);
+	&jz	(&label("schedule_mangle_last"));
+	&call	("_vpaes_schedule_mangle");	# write output
+	&jmp	(&label("loop_schedule_128"));
+
+##
+##  .aes_schedule_192
+##
+##  192-bit specific part of key schedule.
+##
+##  The main body of this schedule is the same as the 128-bit
+##  schedule, but with more smearing.  The long, high side is
+##  stored in %xmm7 as before, and the short, low side is in
+##  the high bits of %xmm6.
+##
+##  This schedule is somewhat nastier, however, because each
+##  round produces 192 bits of key material, or 1.5 round keys.
+##  Therefore, on each cycle we do 2 rounds and produce 3 round
+##  keys.
+##
+&set_label("schedule_192",16);
+	&movdqu	("xmm0",&QWP(8,$inp));		# load key part 2 (very unaligned)
+	&call	("_vpaes_schedule_transform");	# input transform	
+	&movdqa	("xmm6","xmm0");		# save short part
+	&pxor	("xmm4","xmm4");		# clear 4
+	&movhlps("xmm6","xmm4");		# clobber low side with zeros
+	&mov	($round,4);
+
+&set_label("loop_schedule_192");
+	&call	("_vpaes_schedule_round");
+	&palignr("xmm0","xmm6",8);
+	&call	("_vpaes_schedule_mangle");	# save key n
+	&call	("_vpaes_schedule_192_smear");
+	&call	("_vpaes_schedule_mangle");	# save key n+1
+	&call	("_vpaes_schedule_round");
+	&dec	($round);
+	&jz	(&label("schedule_mangle_last"));
+	&call	("_vpaes_schedule_mangle");	# save key n+2
+	&call	("_vpaes_schedule_192_smear");
+	&jmp	(&label("loop_schedule_192"));
+
+##
+##  .aes_schedule_256
+##
+##  256-bit specific part of key schedule.
+##
+##  The structure here is very similar to the 128-bit
+##  schedule, but with an additional "low side" in
+##  %xmm6.  The low side's rounds are the same as the
+##  high side's, except no rcon and no rotation.
+##
+&set_label("schedule_256",16);
+	&movdqu	("xmm0",&QWP(16,$inp));		# load key part 2 (unaligned)
+	&call	("_vpaes_schedule_transform");	# input transform	
+	&mov	($round,7);
+
+&set_label("loop_schedule_256");
+	&call	("_vpaes_schedule_mangle");	# output low result
+	&movdqa	("xmm6","xmm0");		# save cur_lo in xmm6
+
+	# high round
+	&call	("_vpaes_schedule_round");
+	&dec	($round);
+	&jz	(&label("schedule_mangle_last"));
+	&call	("_vpaes_schedule_mangle");	
+
+	# low round. swap xmm7 and xmm6
+	&pshufd	("xmm0","xmm0",0xFF);
+	&movdqa	(&QWP(20,"esp"),"xmm7");
+	&movdqa	("xmm7","xmm6");
+	&call	("_vpaes_schedule_low_round");
+	&movdqa	("xmm7",&QWP(20,"esp"));
+
+	&jmp	(&label("loop_schedule_256"));
+
+##
+##  .aes_schedule_mangle_last
+##
+##  Mangler for last round of key schedule
+##  Mangles %xmm0
+##    when encrypting, outputs out(%xmm0) ^ 63
+##    when decrypting, outputs unskew(%xmm0)
+##
+##  Always called right before return... jumps to cleanup and exits
+##
+&set_label("schedule_mangle_last",16);
+	# schedule last round key from xmm0
+	&lea	($base,&DWP($k_deskew,$const));
+	&test	($out,$out);
+	&jnz	(&label("schedule_mangle_last_dec"));
+
+	# encrypting
+	&movdqa	("xmm1",&QWP($k_sr,$const,$magic));
+	&pshufb	("xmm0","xmm1");		# output permute
+	&lea	($base,&DWP($k_opt,$const));	# prepare to output transform
+	&add	($key,32);
+
+&set_label("schedule_mangle_last_dec");
+	&add	($key,-16);
+	&pxor	("xmm0",&QWP($k_s63,$const));
+	&call	("_vpaes_schedule_transform");	# output transform
+	&movdqu	(&QWP(0,$key),"xmm0");		# save last key
+
+	# cleanup
+	&pxor	("xmm0","xmm0");
+	&pxor	("xmm1","xmm1");
+	&pxor	("xmm2","xmm2");
+	&pxor	("xmm3","xmm3");
+	&pxor	("xmm4","xmm4");
+	&pxor	("xmm5","xmm5");
+	&pxor	("xmm6","xmm6");
+	&pxor	("xmm7","xmm7");
+	&ret	();
+&function_end_B("_vpaes_schedule_core");
+
+##
+##  .aes_schedule_192_smear
+##
+##  Smear the short, low side in the 192-bit key schedule.
+##
+##  Inputs:
+##    %xmm7: high side, b  a  x  y
+##    %xmm6:  low side, d  c  0  0
+##    %xmm13: 0
+##
+##  Outputs:
+##    %xmm6: b+c+d  b+c  0  0
+##    %xmm0: b+c+d  b+c  b  a
+##
+&function_begin_B("_vpaes_schedule_192_smear");
+	&pshufd	("xmm0","xmm6",0x80);		# d c 0 0 -> c 0 0 0
+	&pxor	("xmm6","xmm0");		# -> c+d c 0 0
+	&pshufd	("xmm0","xmm7",0xFE);		# b a _ _ -> b b b a
+	&pxor	("xmm6","xmm0");		# -> b+c+d b+c b a
+	&movdqa	("xmm0","xmm6");
+	&pxor	("xmm1","xmm1");
+	&movhlps("xmm6","xmm1");		# clobber low side with zeros
+	&ret	();
+&function_end_B("_vpaes_schedule_192_smear");
+
+##
+##  .aes_schedule_round
+##
+##  Runs one main round of the key schedule on %xmm0, %xmm7
+##
+##  Specifically, runs subbytes on the high dword of %xmm0
+##  then rotates it by one byte and xors into the low dword of
+##  %xmm7.
+##
+##  Adds rcon from low byte of %xmm8, then rotates %xmm8 for
+##  next rcon.
+##
+##  Smears the dwords of %xmm7 by xoring the low into the
+##  second low, result into third, result into highest.
+##
+##  Returns results in %xmm7 = %xmm0.
+##  Clobbers %xmm1-%xmm5.
+##
+&function_begin_B("_vpaes_schedule_round");
+	# extract rcon from xmm8
+	&movdqa	("xmm2",&QWP(8,"esp"));		# xmm8
+	&pxor	("xmm1","xmm1");
+	&palignr("xmm1","xmm2",15);
+	&palignr("xmm2","xmm2",15);
+	&pxor	("xmm7","xmm1");
+
+	# rotate
+	&pshufd	("xmm0","xmm0",0xFF);
+	&palignr("xmm0","xmm0",1);
+
+	# fall through...
+	&movdqa	(&QWP(8,"esp"),"xmm2");		# xmm8
+
+	# low round: same as high round, but no rotation and no rcon.
+&set_label("_vpaes_schedule_low_round");
+	# smear xmm7
+	&movdqa	("xmm1","xmm7");
+	&pslldq	("xmm7",4);
+	&pxor	("xmm7","xmm1");
+	&movdqa	("xmm1","xmm7");
+	&pslldq	("xmm7",8);
+	&pxor	("xmm7","xmm1");
+	&pxor	("xmm7",&QWP($k_s63,$const));
+
+	# subbyte
+	&movdqa	("xmm4",&QWP($k_s0F,$const));
+	&movdqa	("xmm5",&QWP($k_inv,$const));	# 4 : 1/j
+	&movdqa	("xmm1","xmm4");	
+	&pandn	("xmm1","xmm0");
+	&psrld	("xmm1",4);			# 1 = i
+	&pand	("xmm0","xmm4");		# 0 = k
+	&movdqa	("xmm2",&QWP($k_inv+16,$const));# 2 : a/k
+	&pshufb	("xmm2","xmm0");		# 2 = a/k
+	&pxor	("xmm0","xmm1");		# 0 = j
+	&movdqa	("xmm3","xmm5");		# 3 : 1/i
+	&pshufb	("xmm3","xmm1");		# 3 = 1/i
+	&pxor	("xmm3","xmm2");		# 3 = iak = 1/i + a/k
+	&movdqa	("xmm4","xmm5");		# 4 : 1/j
+	&pshufb	("xmm4","xmm0");		# 4 = 1/j
+	&pxor	("xmm4","xmm2");		# 4 = jak = 1/j + a/k
+	&movdqa	("xmm2","xmm5");		# 2 : 1/iak
+	&pshufb	("xmm2","xmm3");		# 2 = 1/iak
+	&pxor	("xmm2","xmm0");		# 2 = io
+	&movdqa	("xmm3","xmm5");		# 3 : 1/jak
+	&pshufb	("xmm3","xmm4");		# 3 = 1/jak
+	&pxor	("xmm3","xmm1");		# 3 = jo
+	&movdqa	("xmm4",&QWP($k_sb1,$const));	# 4 : sbou
+	&pshufb	("xmm4","xmm2");		# 4 = sbou
+	&movdqa	("xmm0",&QWP($k_sb1+16,$const));# 0 : sbot
+	&pshufb	("xmm0","xmm3");		# 0 = sb1t
+	&pxor	("xmm0","xmm4");		# 0 = sbox output
+
+	# add in smeared stuff
+	&pxor	("xmm0","xmm7");
+	&movdqa	("xmm7","xmm0");
+	&ret	();
+&function_end_B("_vpaes_schedule_round");
+
+##
+##  .aes_schedule_transform
+##
+##  Linear-transform %xmm0 according to tables at (%ebx)
+##
+##  Output in %xmm0
+##  Clobbers %xmm1, %xmm2
+##
+&function_begin_B("_vpaes_schedule_transform");
+	&movdqa	("xmm2",&QWP($k_s0F,$const));
+	&movdqa	("xmm1","xmm2");
+	&pandn	("xmm1","xmm0");
+	&psrld	("xmm1",4);
+	&pand	("xmm0","xmm2");
+	&movdqa	("xmm2",&QWP(0,$base));
+	&pshufb	("xmm2","xmm0");
+	&movdqa	("xmm0",&QWP(16,$base));
+	&pshufb	("xmm0","xmm1");
+	&pxor	("xmm0","xmm2");
+	&ret	();
+&function_end_B("_vpaes_schedule_transform");
+
+##
+##  .aes_schedule_mangle
+##
+##  Mangle xmm0 from (basis-transformed) standard version
+##  to our version.
+##
+##  On encrypt,
+##    xor with 0x63
+##    multiply by circulant 0,1,1,1
+##    apply shiftrows transform
+##
+##  On decrypt,
+##    xor with 0x63
+##    multiply by "inverse mixcolumns" circulant E,B,D,9
+##    deskew
+##    apply shiftrows transform
+##
+##
+##  Writes out to (%edx), and increments or decrements it
+##  Keeps track of round number mod 4 in %ecx
+##  Preserves xmm0
+##  Clobbers xmm1-xmm5
+##
+&function_begin_B("_vpaes_schedule_mangle");
+	&movdqa	("xmm4","xmm0");	# save xmm0 for later
+	&movdqa	("xmm5",&QWP($k_mc_forward,$const));
+	&test	($out,$out);
+	&jnz	(&label("schedule_mangle_dec"));
+
+	# encrypting
+	&add	($key,16);
+	&pxor	("xmm4",&QWP($k_s63,$const));
+	&pshufb	("xmm4","xmm5");
+	&movdqa	("xmm3","xmm4");
+	&pshufb	("xmm4","xmm5");
+	&pxor	("xmm3","xmm4");
+	&pshufb	("xmm4","xmm5");
+	&pxor	("xmm3","xmm4");
+
+	&jmp	(&label("schedule_mangle_both"));
+
+&set_label("schedule_mangle_dec",16);
+	# inverse mix columns
+	&movdqa	("xmm2",&QWP($k_s0F,$const));
+	&lea	($inp,&DWP($k_dksd,$const));
+	&movdqa	("xmm1","xmm2");
+	&pandn	("xmm1","xmm4");
+	&psrld	("xmm1",4);			# 1 = hi
+	&pand	("xmm4","xmm2");		# 4 = lo
+
+	&movdqa	("xmm2",&QWP(0,$inp));
+	&pshufb	("xmm2","xmm4");
+	&movdqa	("xmm3",&QWP(0x10,$inp));
+	&pshufb	("xmm3","xmm1");
+	&pxor	("xmm3","xmm2");
+	&pshufb	("xmm3","xmm5");
+
+	&movdqa	("xmm2",&QWP(0x20,$inp));
+	&pshufb	("xmm2","xmm4");
+	&pxor	("xmm2","xmm3");
+	&movdqa	("xmm3",&QWP(0x30,$inp));
+	&pshufb	("xmm3","xmm1");
+	&pxor	("xmm3","xmm2");
+	&pshufb	("xmm3","xmm5");
+
+	&movdqa	("xmm2",&QWP(0x40,$inp));
+	&pshufb	("xmm2","xmm4");
+	&pxor	("xmm2","xmm3");
+	&movdqa	("xmm3",&QWP(0x50,$inp));
+	&pshufb	("xmm3","xmm1");
+	&pxor	("xmm3","xmm2");
+	&pshufb	("xmm3","xmm5");
+
+	&movdqa	("xmm2",&QWP(0x60,$inp));
+	&pshufb	("xmm2","xmm4");
+	&pxor	("xmm2","xmm3");
+	&movdqa	("xmm3",&QWP(0x70,$inp));
+	&pshufb	("xmm3","xmm1");
+	&pxor	("xmm3","xmm2");
+
+	&add	($key,-16);
+
+&set_label("schedule_mangle_both");
+	&movdqa	("xmm1",&QWP($k_sr,$const,$magic));
+	&pshufb	("xmm3","xmm1");
+	&add	($magic,-16);
+	&and	($magic,0x30);
+	&movdqu	(&QWP(0,$key),"xmm3");
+	&ret	();
+&function_end_B("_vpaes_schedule_mangle");
+
+#
+# Interface to OpenSSL
+#
+&function_begin("${PREFIX}_set_encrypt_key");
+	&mov	($inp,&wparam(0));		# inp
+	&lea	($base,&DWP(-56,"esp"));
+	&mov	($round,&wparam(1));		# bits
+	&and	($base,-16);
+	&mov	($key,&wparam(2));		# key
+	&xchg	($base,"esp");			# alloca
+	&mov	(&DWP(48,"esp"),$base);
+
+	&mov	($base,$round);
+	&shr	($base,5);
+	&add	($base,5);
+	&mov	(&DWP(240,$key),$base);		# AES_KEY->rounds = nbits/32+5;
+	&mov	($magic,0x30);
+	&mov	($out,0);
+
+	&lea	($const,&DWP(&label("_vpaes_consts")."+0x30-".&label("pic_point")));
+	&call	("_vpaes_schedule_core");
+&set_label("pic_point");
+
+	&mov	("esp",&DWP(48,"esp"));
+	&xor	("eax","eax");
+&function_end("${PREFIX}_set_encrypt_key");
+
+&function_begin("${PREFIX}_set_decrypt_key");
+	&mov	($inp,&wparam(0));		# inp
+	&lea	($base,&DWP(-56,"esp"));
+	&mov	($round,&wparam(1));		# bits
+	&and	($base,-16);
+	&mov	($key,&wparam(2));		# key
+	&xchg	($base,"esp");			# alloca
+	&mov	(&DWP(48,"esp"),$base);
+
+	&mov	($base,$round);
+	&shr	($base,5);
+	&add	($base,5);
+	&mov	(&DWP(240,$key),$base);	# AES_KEY->rounds = nbits/32+5;
+	&shl	($base,4);
+	&lea	($key,&DWP(16,$key,$base));
+
+	&mov	($out,1);
+	&mov	($magic,$round);
+	&shr	($magic,1);
+	&and	($magic,32);
+	&xor	($magic,32);			# nbist==192?0:32;
+
+	&lea	($const,&DWP(&label("_vpaes_consts")."+0x30-".&label("pic_point")));
+	&call	("_vpaes_schedule_core");
+&set_label("pic_point");
+
+	&mov	("esp",&DWP(48,"esp"));
+	&xor	("eax","eax");
+&function_end("${PREFIX}_set_decrypt_key");
+
+&function_begin("${PREFIX}_encrypt");
+	&lea	($const,&DWP(&label("_vpaes_consts")."+0x30-".&label("pic_point")));
+	&call	("_vpaes_preheat");
+&set_label("pic_point");
+	&mov	($inp,&wparam(0));		# inp
+	&lea	($base,&DWP(-56,"esp"));
+	&mov	($out,&wparam(1));		# out
+	&and	($base,-16);
+	&mov	($key,&wparam(2));		# key
+	&xchg	($base,"esp");			# alloca
+	&mov	(&DWP(48,"esp"),$base);
+
+	&movdqu	("xmm0",&QWP(0,$inp));
+	&call	("_vpaes_encrypt_core");
+	&movdqu	(&QWP(0,$out),"xmm0");
+
+	&mov	("esp",&DWP(48,"esp"));
+&function_end("${PREFIX}_encrypt");
+
+&function_begin("${PREFIX}_decrypt");
+	&lea	($const,&DWP(&label("_vpaes_consts")."+0x30-".&label("pic_point")));
+	&call	("_vpaes_preheat");
+&set_label("pic_point");
+	&mov	($inp,&wparam(0));		# inp
+	&lea	($base,&DWP(-56,"esp"));
+	&mov	($out,&wparam(1));		# out
+	&and	($base,-16);
+	&mov	($key,&wparam(2));		# key
+	&xchg	($base,"esp");			# alloca
+	&mov	(&DWP(48,"esp"),$base);
+
+	&movdqu	("xmm0",&QWP(0,$inp));
+	&call	("_vpaes_decrypt_core");
+	&movdqu	(&QWP(0,$out),"xmm0");
+
+	&mov	("esp",&DWP(48,"esp"));
+&function_end("${PREFIX}_decrypt");
+
+&function_begin("${PREFIX}_cbc_encrypt");
+	&mov	($inp,&wparam(0));		# inp
+	&mov	($out,&wparam(1));		# out
+	&mov	($round,&wparam(2));		# len
+	&mov	($key,&wparam(3));		# key
+	&sub	($round,16);
+	&jc	(&label("cbc_abort"));
+	&lea	($base,&DWP(-56,"esp"));
+	&mov	($const,&wparam(4));		# ivp
+	&and	($base,-16);
+	&mov	($magic,&wparam(5));		# enc
+	&xchg	($base,"esp");			# alloca
+	&movdqu	("xmm1",&QWP(0,$const));	# load IV
+	&sub	($out,$inp);
+	&mov	(&DWP(48,"esp"),$base);
+
+	&mov	(&DWP(0,"esp"),$out);		# save out
+	&mov	(&DWP(4,"esp"),$key)		# save key
+	&mov	(&DWP(8,"esp"),$const);		# save ivp
+	&mov	($out,$round);			# $out works as $len
+
+	&lea	($const,&DWP(&label("_vpaes_consts")."+0x30-".&label("pic_point")));
+	&call	("_vpaes_preheat");
+&set_label("pic_point");
+	&cmp	($magic,0);
+	&je	(&label("cbc_dec_loop"));
+	&jmp	(&label("cbc_enc_loop"));
+
+&set_label("cbc_enc_loop",16);
+	&movdqu	("xmm0",&QWP(0,$inp));		# load input
+	&pxor	("xmm0","xmm1");		# inp^=iv
+	&call	("_vpaes_encrypt_core");
+	&mov	($base,&DWP(0,"esp"));		# restore out
+	&mov	($key,&DWP(4,"esp"));		# restore key
+	&movdqa	("xmm1","xmm0");
+	&movdqu	(&QWP(0,$base,$inp),"xmm0");	# write output
+	&lea	($inp,&DWP(16,$inp));
+	&sub	($out,16);
+	&jnc	(&label("cbc_enc_loop"));
+	&jmp	(&label("cbc_done"));
+
+&set_label("cbc_dec_loop",16);
+	&movdqu	("xmm0",&QWP(0,$inp));		# load input
+	&movdqa	(&QWP(16,"esp"),"xmm1");	# save IV
+	&movdqa	(&QWP(32,"esp"),"xmm0");	# save future IV
+	&call	("_vpaes_decrypt_core");
+	&mov	($base,&DWP(0,"esp"));		# restore out
+	&mov	($key,&DWP(4,"esp"));		# restore key
+	&pxor	("xmm0",&QWP(16,"esp"));	# out^=iv
+	&movdqa	("xmm1",&QWP(32,"esp"));	# load next IV
+	&movdqu	(&QWP(0,$base,$inp),"xmm0");	# write output
+	&lea	($inp,&DWP(16,$inp));
+	&sub	($out,16);
+	&jnc	(&label("cbc_dec_loop"));
+
+&set_label("cbc_done");
+	&mov	($base,&DWP(8,"esp"));		# restore ivp
+	&mov	("esp",&DWP(48,"esp"));
+	&movdqu	(&QWP(0,$base),"xmm1");		# write IV
+&set_label("cbc_abort");
+&function_end("${PREFIX}_cbc_encrypt");
+
+&asm_finish();

jni/openssl/crypto/aes/asm/vpaes-x86_64.S

@@ -0,0 +1,828 @@
+.text	
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+.type	_vpaes_encrypt_core,@function
+.align	16
+_vpaes_encrypt_core:
+	movq	%rdx,%r9
+	movq	$16,%r11
+	movl	240(%rdx),%eax
+	movdqa	%xmm9,%xmm1
+	movdqa	.Lk_ipt(%rip),%xmm2
+	pandn	%xmm0,%xmm1
+	movdqu	(%r9),%xmm5
+	psrld	$4,%xmm1
+	pand	%xmm9,%xmm0
+.byte	102,15,56,0,208
+	movdqa	.Lk_ipt+16(%rip),%xmm0
+.byte	102,15,56,0,193
+	pxor	%xmm5,%xmm2
+	pxor	%xmm2,%xmm0
+	addq	$16,%r9
+	leaq	.Lk_mc_backward(%rip),%r10
+	jmp	.Lenc_entry
+
+.align	16
+.Lenc_loop:
+
+	movdqa	%xmm13,%xmm4
+.byte	102,15,56,0,226
+	pxor	%xmm5,%xmm4
+	movdqa	%xmm12,%xmm0
+.byte	102,15,56,0,195
+	pxor	%xmm4,%xmm0
+	movdqa	%xmm15,%xmm5
+.byte	102,15,56,0,234
+	movdqa	-64(%r11,%r10,1),%xmm1
+	movdqa	%xmm14,%xmm2
+.byte	102,15,56,0,211
+	pxor	%xmm5,%xmm2
+	movdqa	(%r11,%r10,1),%xmm4
+	movdqa	%xmm0,%xmm3
+.byte	102,15,56,0,193
+	addq	$16,%r9
+	pxor	%xmm2,%xmm0
+.byte	102,15,56,0,220
+	addq	$16,%r11
+	pxor	%xmm0,%xmm3
+.byte	102,15,56,0,193
+	andq	$48,%r11
+	pxor	%xmm3,%xmm0
+	subq	$1,%rax
+
+.Lenc_entry:
+
+	movdqa	%xmm9,%xmm1
+	pandn	%xmm0,%xmm1
+	psrld	$4,%xmm1
+	pand	%xmm9,%xmm0
+	movdqa	%xmm11,%xmm5
+.byte	102,15,56,0,232
+	pxor	%xmm1,%xmm0
+	movdqa	%xmm10,%xmm3
+.byte	102,15,56,0,217
+	pxor	%xmm5,%xmm3
+	movdqa	%xmm10,%xmm4
+.byte	102,15,56,0,224
+	pxor	%xmm5,%xmm4
+	movdqa	%xmm10,%xmm2
+.byte	102,15,56,0,211
+	pxor	%xmm0,%xmm2
+	movdqa	%xmm10,%xmm3
+	movdqu	(%r9),%xmm5
+.byte	102,15,56,0,220
+	pxor	%xmm1,%xmm3
+	jnz	.Lenc_loop
+
+
+	movdqa	-96(%r10),%xmm4
+	movdqa	-80(%r10),%xmm0
+.byte	102,15,56,0,226
+	pxor	%xmm5,%xmm4
+.byte	102,15,56,0,195
+	movdqa	64(%r11,%r10,1),%xmm1
+	pxor	%xmm4,%xmm0
+.byte	102,15,56,0,193
+	.byte	0xf3,0xc3
+.size	_vpaes_encrypt_core,.-_vpaes_encrypt_core
+
+
+
+
+
+
+.type	_vpaes_decrypt_core,@function
+.align	16
+_vpaes_decrypt_core:
+	movq	%rdx,%r9
+	movl	240(%rdx),%eax
+	movdqa	%xmm9,%xmm1
+	movdqa	.Lk_dipt(%rip),%xmm2
+	pandn	%xmm0,%xmm1
+	movq	%rax,%r11
+	psrld	$4,%xmm1
+	movdqu	(%r9),%xmm5
+	shlq	$4,%r11
+	pand	%xmm9,%xmm0
+.byte	102,15,56,0,208
+	movdqa	.Lk_dipt+16(%rip),%xmm0
+	xorq	$48,%r11
+	leaq	.Lk_dsbd(%rip),%r10
+.byte	102,15,56,0,193
+	andq	$48,%r11
+	pxor	%xmm5,%xmm2
+	movdqa	.Lk_mc_forward+48(%rip),%xmm5
+	pxor	%xmm2,%xmm0
+	addq	$16,%r9
+	addq	%r10,%r11
+	jmp	.Ldec_entry
+
+.align	16
+.Ldec_loop:
+
+
+
+	movdqa	-32(%r10),%xmm4
+.byte	102,15,56,0,226
+	pxor	%xmm0,%xmm4
+	movdqa	-16(%r10),%xmm0
+.byte	102,15,56,0,195
+	pxor	%xmm4,%xmm0
+	addq	$16,%r9
+
+.byte	102,15,56,0,197
+	movdqa	0(%r10),%xmm4
+.byte	102,15,56,0,226
+	pxor	%xmm0,%xmm4
+	movdqa	16(%r10),%xmm0
+.byte	102,15,56,0,195
+	pxor	%xmm4,%xmm0
+	subq	$1,%rax
+
+.byte	102,15,56,0,197
+	movdqa	32(%r10),%xmm4
+.byte	102,15,56,0,226
+	pxor	%xmm0,%xmm4
+	movdqa	48(%r10),%xmm0
+.byte	102,15,56,0,195
+	pxor	%xmm4,%xmm0
+
+.byte	102,15,56,0,197
+	movdqa	64(%r10),%xmm4
+.byte	102,15,56,0,226
+	pxor	%xmm0,%xmm4
+	movdqa	80(%r10),%xmm0
+.byte	102,15,56,0,195
+	pxor	%xmm4,%xmm0
+
+.byte	102,15,58,15,237,12
+
+.Ldec_entry:
+
+	movdqa	%xmm9,%xmm1
+	pandn	%xmm0,%xmm1
+	psrld	$4,%xmm1
+	pand	%xmm9,%xmm0
+	movdqa	%xmm11,%xmm2
+.byte	102,15,56,0,208
+	pxor	%xmm1,%xmm0
+	movdqa	%xmm10,%xmm3
+.byte	102,15,56,0,217
+	pxor	%xmm2,%xmm3
+	movdqa	%xmm10,%xmm4
+.byte	102,15,56,0,224
+	pxor	%xmm2,%xmm4
+	movdqa	%xmm10,%xmm2
+.byte	102,15,56,0,211
+	pxor	%xmm0,%xmm2
+	movdqa	%xmm10,%xmm3
+.byte	102,15,56,0,220
+	pxor	%xmm1,%xmm3
+	movdqu	(%r9),%xmm0
+	jnz	.Ldec_loop
+
+
+	movdqa	96(%r10),%xmm4
+.byte	102,15,56,0,226
+	pxor	%xmm0,%xmm4
+	movdqa	112(%r10),%xmm0
+	movdqa	-352(%r11),%xmm2
+.byte	102,15,56,0,195
+	pxor	%xmm4,%xmm0
+.byte	102,15,56,0,194
+	.byte	0xf3,0xc3
+.size	_vpaes_decrypt_core,.-_vpaes_decrypt_core
+
+
+
+
+
+
+.type	_vpaes_schedule_core,@function
+.align	16
+_vpaes_schedule_core:
+
+
+
+
+
+	call	_vpaes_preheat		
+	movdqa	.Lk_rcon(%rip),%xmm8
+	movdqu	(%rdi),%xmm0
+
+
+	movdqa	%xmm0,%xmm3
+	leaq	.Lk_ipt(%rip),%r11
+	call	_vpaes_schedule_transform
+	movdqa	%xmm0,%xmm7
+
+	leaq	.Lk_sr(%rip),%r10
+	testq	%rcx,%rcx
+	jnz	.Lschedule_am_decrypting
+
+
+	movdqu	%xmm0,(%rdx)
+	jmp	.Lschedule_go
+
+.Lschedule_am_decrypting:
+
+	movdqa	(%r8,%r10,1),%xmm1
+.byte	102,15,56,0,217
+	movdqu	%xmm3,(%rdx)
+	xorq	$48,%r8
+
+.Lschedule_go:
+	cmpl	$192,%esi
+	ja	.Lschedule_256
+	je	.Lschedule_192
+
+
+
+
+
+
+
+
+
+
+.Lschedule_128:
+	movl	$10,%esi
+
+.Loop_schedule_128:
+	call	_vpaes_schedule_round
+	decq	%rsi
+	jz	.Lschedule_mangle_last
+	call	_vpaes_schedule_mangle	
+	jmp	.Loop_schedule_128
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+.align	16
+.Lschedule_192:
+	movdqu	8(%rdi),%xmm0
+	call	_vpaes_schedule_transform	
+	movdqa	%xmm0,%xmm6
+	pxor	%xmm4,%xmm4
+	movhlps	%xmm4,%xmm6
+	movl	$4,%esi
+
+.Loop_schedule_192:
+	call	_vpaes_schedule_round
+.byte	102,15,58,15,198,8
+	call	_vpaes_schedule_mangle	
+	call	_vpaes_schedule_192_smear
+	call	_vpaes_schedule_mangle	
+	call	_vpaes_schedule_round
+	decq	%rsi
+	jz	.Lschedule_mangle_last
+	call	_vpaes_schedule_mangle	
+	call	_vpaes_schedule_192_smear
+	jmp	.Loop_schedule_192
+
+
+
+
+
+
+
+
+
+
+
+.align	16
+.Lschedule_256:
+	movdqu	16(%rdi),%xmm0
+	call	_vpaes_schedule_transform	
+	movl	$7,%esi
+
+.Loop_schedule_256:
+	call	_vpaes_schedule_mangle	
+	movdqa	%xmm0,%xmm6
+
+
+	call	_vpaes_schedule_round
+	decq	%rsi
+	jz	.Lschedule_mangle_last
+	call	_vpaes_schedule_mangle	
+
+
+	pshufd	$255,%xmm0,%xmm0
+	movdqa	%xmm7,%xmm5
+	movdqa	%xmm6,%xmm7
+	call	_vpaes_schedule_low_round
+	movdqa	%xmm5,%xmm7
+
+	jmp	.Loop_schedule_256
+
+
+
+
+
+
+
+
+
+
+
+
+.align	16
+.Lschedule_mangle_last:
+
+	leaq	.Lk_deskew(%rip),%r11
+	testq	%rcx,%rcx
+	jnz	.Lschedule_mangle_last_dec
+
+
+	movdqa	(%r8,%r10,1),%xmm1
+.byte	102,15,56,0,193
+	leaq	.Lk_opt(%rip),%r11
+	addq	$32,%rdx
+
+.Lschedule_mangle_last_dec:
+	addq	$-16,%rdx
+	pxor	.Lk_s63(%rip),%xmm0
+	call	_vpaes_schedule_transform 
+	movdqu	%xmm0,(%rdx)
+
+
+	pxor	%xmm0,%xmm0
+	pxor	%xmm1,%xmm1
+	pxor	%xmm2,%xmm2
+	pxor	%xmm3,%xmm3
+	pxor	%xmm4,%xmm4
+	pxor	%xmm5,%xmm5
+	pxor	%xmm6,%xmm6
+	pxor	%xmm7,%xmm7
+	.byte	0xf3,0xc3
+.size	_vpaes_schedule_core,.-_vpaes_schedule_core
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+.type	_vpaes_schedule_192_smear,@function
+.align	16
+_vpaes_schedule_192_smear:
+	pshufd	$128,%xmm6,%xmm0
+	pxor	%xmm0,%xmm6
+	pshufd	$254,%xmm7,%xmm0
+	pxor	%xmm0,%xmm6
+	movdqa	%xmm6,%xmm0
+	pxor	%xmm1,%xmm1
+	movhlps	%xmm1,%xmm6
+	.byte	0xf3,0xc3
+.size	_vpaes_schedule_192_smear,.-_vpaes_schedule_192_smear
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+.type	_vpaes_schedule_round,@function
+.align	16
+_vpaes_schedule_round:
+
+	pxor	%xmm1,%xmm1
+.byte	102,65,15,58,15,200,15
+.byte	102,69,15,58,15,192,15
+	pxor	%xmm1,%xmm7
+
+
+	pshufd	$255,%xmm0,%xmm0
+.byte	102,15,58,15,192,1
+
+
+
+
+_vpaes_schedule_low_round:
+
+	movdqa	%xmm7,%xmm1
+	pslldq	$4,%xmm7
+	pxor	%xmm1,%xmm7
+	movdqa	%xmm7,%xmm1
+	pslldq	$8,%xmm7
+	pxor	%xmm1,%xmm7
+	pxor	.Lk_s63(%rip),%xmm7
+
+
+	movdqa	%xmm9,%xmm1
+	pandn	%xmm0,%xmm1
+	psrld	$4,%xmm1
+	pand	%xmm9,%xmm0
+	movdqa	%xmm11,%xmm2
+.byte	102,15,56,0,208
+	pxor	%xmm1,%xmm0
+	movdqa	%xmm10,%xmm3
+.byte	102,15,56,0,217
+	pxor	%xmm2,%xmm3
+	movdqa	%xmm10,%xmm4
+.byte	102,15,56,0,224
+	pxor	%xmm2,%xmm4
+	movdqa	%xmm10,%xmm2
+.byte	102,15,56,0,211
+	pxor	%xmm0,%xmm2
+	movdqa	%xmm10,%xmm3
+.byte	102,15,56,0,220
+	pxor	%xmm1,%xmm3
+	movdqa	%xmm13,%xmm4
+.byte	102,15,56,0,226
+	movdqa	%xmm12,%xmm0
+.byte	102,15,56,0,195
+	pxor	%xmm4,%xmm0
+
+
+	pxor	%xmm7,%xmm0
+	movdqa	%xmm0,%xmm7
+	.byte	0xf3,0xc3
+.size	_vpaes_schedule_round,.-_vpaes_schedule_round
+
+
+
+
+
+
+
+
+
+
+.type	_vpaes_schedule_transform,@function
+.align	16
+_vpaes_schedule_transform:
+	movdqa	%xmm9,%xmm1
+	pandn	%xmm0,%xmm1
+	psrld	$4,%xmm1
+	pand	%xmm9,%xmm0
+	movdqa	(%r11),%xmm2
+.byte	102,15,56,0,208
+	movdqa	16(%r11),%xmm0
+.byte	102,15,56,0,193
+	pxor	%xmm2,%xmm0
+	.byte	0xf3,0xc3
+.size	_vpaes_schedule_transform,.-_vpaes_schedule_transform
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+.type	_vpaes_schedule_mangle,@function
+.align	16
+_vpaes_schedule_mangle:
+	movdqa	%xmm0,%xmm4
+	movdqa	.Lk_mc_forward(%rip),%xmm5
+	testq	%rcx,%rcx
+	jnz	.Lschedule_mangle_dec
+
+
+	addq	$16,%rdx
+	pxor	.Lk_s63(%rip),%xmm4
+.byte	102,15,56,0,229
+	movdqa	%xmm4,%xmm3
+.byte	102,15,56,0,229
+	pxor	%xmm4,%xmm3
+.byte	102,15,56,0,229
+	pxor	%xmm4,%xmm3
+
+	jmp	.Lschedule_mangle_both
+.align	16
+.Lschedule_mangle_dec:
+
+	leaq	.Lk_dksd(%rip),%r11
+	movdqa	%xmm9,%xmm1
+	pandn	%xmm4,%xmm1
+	psrld	$4,%xmm1
+	pand	%xmm9,%xmm4
+
+	movdqa	0(%r11),%xmm2
+.byte	102,15,56,0,212
+	movdqa	16(%r11),%xmm3
+.byte	102,15,56,0,217
+	pxor	%xmm2,%xmm3
+.byte	102,15,56,0,221
+
+	movdqa	32(%r11),%xmm2
+.byte	102,15,56,0,212
+	pxor	%xmm3,%xmm2
+	movdqa	48(%r11),%xmm3
+.byte	102,15,56,0,217
+	pxor	%xmm2,%xmm3
+.byte	102,15,56,0,221
+
+	movdqa	64(%r11),%xmm2
+.byte	102,15,56,0,212
+	pxor	%xmm3,%xmm2
+	movdqa	80(%r11),%xmm3
+.byte	102,15,56,0,217
+	pxor	%xmm2,%xmm3
+.byte	102,15,56,0,221
+
+	movdqa	96(%r11),%xmm2
+.byte	102,15,56,0,212
+	pxor	%xmm3,%xmm2
+	movdqa	112(%r11),%xmm3
+.byte	102,15,56,0,217
+	pxor	%xmm2,%xmm3
+
+	addq	$-16,%rdx
+
+.Lschedule_mangle_both:
+	movdqa	(%r8,%r10,1),%xmm1
+.byte	102,15,56,0,217
+	addq	$-16,%r8
+	andq	$48,%r8
+	movdqu	%xmm3,(%rdx)
+	.byte	0xf3,0xc3
+.size	_vpaes_schedule_mangle,.-_vpaes_schedule_mangle
+
+
+
+
+.globl	vpaes_set_encrypt_key
+.type	vpaes_set_encrypt_key,@function
+.align	16
+vpaes_set_encrypt_key:
+	movl	%esi,%eax
+	shrl	$5,%eax
+	addl	$5,%eax
+	movl	%eax,240(%rdx)
+
+	movl	$0,%ecx
+	movl	$48,%r8d
+	call	_vpaes_schedule_core
+	xorl	%eax,%eax
+	.byte	0xf3,0xc3
+.size	vpaes_set_encrypt_key,.-vpaes_set_encrypt_key
+
+.globl	vpaes_set_decrypt_key
+.type	vpaes_set_decrypt_key,@function
+.align	16
+vpaes_set_decrypt_key:
+	movl	%esi,%eax
+	shrl	$5,%eax
+	addl	$5,%eax
+	movl	%eax,240(%rdx)
+	shll	$4,%eax
+	leaq	16(%rdx,%rax,1),%rdx
+
+	movl	$1,%ecx
+	movl	%esi,%r8d
+	shrl	$1,%r8d
+	andl	$32,%r8d
+	xorl	$32,%r8d
+	call	_vpaes_schedule_core
+	xorl	%eax,%eax
+	.byte	0xf3,0xc3
+.size	vpaes_set_decrypt_key,.-vpaes_set_decrypt_key
+
+.globl	vpaes_encrypt
+.type	vpaes_encrypt,@function
+.align	16
+vpaes_encrypt:
+	movdqu	(%rdi),%xmm0
+	call	_vpaes_preheat
+	call	_vpaes_encrypt_core
+	movdqu	%xmm0,(%rsi)
+	.byte	0xf3,0xc3
+.size	vpaes_encrypt,.-vpaes_encrypt
+
+.globl	vpaes_decrypt
+.type	vpaes_decrypt,@function
+.align	16
+vpaes_decrypt:
+	movdqu	(%rdi),%xmm0
+	call	_vpaes_preheat
+	call	_vpaes_decrypt_core
+	movdqu	%xmm0,(%rsi)
+	.byte	0xf3,0xc3
+.size	vpaes_decrypt,.-vpaes_decrypt
+.globl	vpaes_cbc_encrypt
+.type	vpaes_cbc_encrypt,@function
+.align	16
+vpaes_cbc_encrypt:
+	xchgq	%rcx,%rdx
+	subq	$16,%rcx
+	jc	.Lcbc_abort
+	movdqu	(%r8),%xmm6
+	subq	%rdi,%rsi
+	call	_vpaes_preheat
+	cmpl	$0,%r9d
+	je	.Lcbc_dec_loop
+	jmp	.Lcbc_enc_loop
+.align	16
+.Lcbc_enc_loop:
+	movdqu	(%rdi),%xmm0
+	pxor	%xmm6,%xmm0
+	call	_vpaes_encrypt_core
+	movdqa	%xmm0,%xmm6
+	movdqu	%xmm0,(%rsi,%rdi,1)
+	leaq	16(%rdi),%rdi
+	subq	$16,%rcx
+	jnc	.Lcbc_enc_loop
+	jmp	.Lcbc_done
+.align	16
+.Lcbc_dec_loop:
+	movdqu	(%rdi),%xmm0
+	movdqa	%xmm0,%xmm7
+	call	_vpaes_decrypt_core
+	pxor	%xmm6,%xmm0
+	movdqa	%xmm7,%xmm6
+	movdqu	%xmm0,(%rsi,%rdi,1)
+	leaq	16(%rdi),%rdi
+	subq	$16,%rcx
+	jnc	.Lcbc_dec_loop
+.Lcbc_done:
+	movdqu	%xmm6,(%r8)
+.Lcbc_abort:
+	.byte	0xf3,0xc3
+.size	vpaes_cbc_encrypt,.-vpaes_cbc_encrypt
+
+
+
+
+
+
+.type	_vpaes_preheat,@function
+.align	16
+_vpaes_preheat:
+	leaq	.Lk_s0F(%rip),%r10
+	movdqa	-32(%r10),%xmm10
+	movdqa	-16(%r10),%xmm11
+	movdqa	0(%r10),%xmm9
+	movdqa	48(%r10),%xmm13
+	movdqa	64(%r10),%xmm12
+	movdqa	80(%r10),%xmm15
+	movdqa	96(%r10),%xmm14
+	.byte	0xf3,0xc3
+.size	_vpaes_preheat,.-_vpaes_preheat
+
+
+
+
+
+.type	_vpaes_consts,@object
+.align	64
+_vpaes_consts:
+.Lk_inv:
+.quad	0x0E05060F0D080180, 0x040703090A0B0C02
+.quad	0x01040A060F0B0780, 0x030D0E0C02050809
+
+.Lk_s0F:
+.quad	0x0F0F0F0F0F0F0F0F, 0x0F0F0F0F0F0F0F0F
+
+.Lk_ipt:
+.quad	0xC2B2E8985A2A7000, 0xCABAE09052227808
+.quad	0x4C01307D317C4D00, 0xCD80B1FCB0FDCC81
+
+.Lk_sb1:
+.quad	0xB19BE18FCB503E00, 0xA5DF7A6E142AF544
+.quad	0x3618D415FAE22300, 0x3BF7CCC10D2ED9EF
+.Lk_sb2:
+.quad	0xE27A93C60B712400, 0x5EB7E955BC982FCD
+.quad	0x69EB88400AE12900, 0xC2A163C8AB82234A
+.Lk_sbo:
+.quad	0xD0D26D176FBDC700, 0x15AABF7AC502A878
+.quad	0xCFE474A55FBB6A00, 0x8E1E90D1412B35FA
+
+.Lk_mc_forward:
+.quad	0x0407060500030201, 0x0C0F0E0D080B0A09
+.quad	0x080B0A0904070605, 0x000302010C0F0E0D
+.quad	0x0C0F0E0D080B0A09, 0x0407060500030201
+.quad	0x000302010C0F0E0D, 0x080B0A0904070605
+
+.Lk_mc_backward:
+.quad	0x0605040702010003, 0x0E0D0C0F0A09080B
+.quad	0x020100030E0D0C0F, 0x0A09080B06050407
+.quad	0x0E0D0C0F0A09080B, 0x0605040702010003
+.quad	0x0A09080B06050407, 0x020100030E0D0C0F
+
+.Lk_sr:
+.quad	0x0706050403020100, 0x0F0E0D0C0B0A0908
+.quad	0x030E09040F0A0500, 0x0B06010C07020D08
+.quad	0x0F060D040B020900, 0x070E050C030A0108
+.quad	0x0B0E0104070A0D00, 0x0306090C0F020508
+
+.Lk_rcon:
+.quad	0x1F8391B9AF9DEEB6, 0x702A98084D7C7D81
+
+.Lk_s63:
+.quad	0x5B5B5B5B5B5B5B5B, 0x5B5B5B5B5B5B5B5B
+
+.Lk_opt:
+.quad	0xFF9F4929D6B66000, 0xF7974121DEBE6808
+.quad	0x01EDBD5150BCEC00, 0xE10D5DB1B05C0CE0
+
+.Lk_deskew:
+.quad	0x07E4A34047A4E300, 0x1DFEB95A5DBEF91A
+.quad	0x5F36B5DC83EA6900, 0x2841C2ABF49D1E77
+
+
+
+
+
+.Lk_dksd:
+.quad	0xFEB91A5DA3E44700, 0x0740E3A45A1DBEF9
+.quad	0x41C277F4B5368300, 0x5FDC69EAAB289D1E
+.Lk_dksb:
+.quad	0x9A4FCA1F8550D500, 0x03D653861CC94C99
+.quad	0x115BEDA7B6FC4A00, 0xD993256F7E3482C8
+.Lk_dkse:
+.quad	0xD5031CCA1FC9D600, 0x53859A4C994F5086
+.quad	0xA23196054FDC7BE8, 0xCD5EF96A20B31487
+.Lk_dks9:
+.quad	0xB6116FC87ED9A700, 0x4AED933482255BFC
+.quad	0x4576516227143300, 0x8BB89FACE9DAFDCE
+
+
+
+
+
+.Lk_dipt:
+.quad	0x0F505B040B545F00, 0x154A411E114E451A
+.quad	0x86E383E660056500, 0x12771772F491F194
+
+.Lk_dsb9:
+.quad	0x851C03539A86D600, 0xCAD51F504F994CC9
+.quad	0xC03B1789ECD74900, 0x725E2C9EB2FBA565
+.Lk_dsbd:
+.quad	0x7D57CCDFE6B1A200, 0xF56E9B13882A4439
+.quad	0x3CE2FAF724C6CB00, 0x2931180D15DEEFD3
+.Lk_dsbb:
+.quad	0xD022649296B44200, 0x602646F6B0F2D404
+.quad	0xC19498A6CD596700, 0xF3FF0C3E3255AA6B
+.Lk_dsbe:
+.quad	0x46F2929626D4D000, 0x2242600464B4F6B0
+.quad	0x0C55A6CDFFAAC100, 0x9467F36B98593E32
+.Lk_dsbo:
+.quad	0x1387EA537EF94000, 0xC7AA6DB9D4943E2D
+.quad	0x12D7560F93441D00, 0xCA4B8159D8C58E9C
+.byte	86,101,99,116,111,114,32,80,101,114,109,117,116,97,116,105,111,110,32,65,69,83,32,102,111,114,32,120,56,54,95,54,52,47,83,83,83,69,51,44,32,77,105,107,101,32,72,97,109,98,117,114,103,32,40,83,116,97,110,102,111,114,100,32,85,110,105,118,101,114,115,105,116,121,41,0
+.align	64
+.size	_vpaes_consts,.-_vpaes_consts