Added assembly for macosx

10 files changed, 7073 insertions, 16 deletions

diff --git a/cfg.mk b/cfg.mk
index 21b59739a2..ac64acf362 100644
--- a/cfg.mk
+++ b/cfg.mk
@@ -157,20 +157,28 @@ upload-web:
 		cvs commit -m "Update." manual/ reference/ \
 			doxygen/ devel/ cyclo/
 
-ASM_SOURCES:= lib/accelerated/x86/asm-coff/cpuid-x86-64-coff.s \
+ASM_SOURCES:= \
 	lib/accelerated/x86/asm/cpuid-x86-64.s \
-	lib/accelerated/x86/asm-coff/cpuid-x86-coff.s \
 	lib/accelerated/x86/asm/cpuid-x86.s \
 	lib/accelerated/x86/asm/appro-aes-gcm-x86-64.s \
 	lib/accelerated/x86/asm/appro-aes-x86-64.s \
 	lib/accelerated/x86/asm/appro-aes-x86.s \
 	lib/accelerated/x86/asm/padlock-x86-64.s \
 	lib/accelerated/x86/asm/padlock-x86.s \
+	lib/accelerated/x86/asm-coff/cpuid-x86-coff.s \
+	lib/accelerated/x86/asm-coff/cpuid-x86-64-coff.s \
 	lib/accelerated/x86/asm-coff/appro-aes-gcm-x86-64-coff.s \
 	lib/accelerated/x86/asm-coff/appro-aes-x86-64-coff.s \
 	lib/accelerated/x86/asm-coff/appro-aes-x86-coff.s \
 	lib/accelerated/x86/asm-coff/padlock-x86-64-coff.s \
-	lib/accelerated/x86/asm-coff/padlock-x86-coff.s
+	lib/accelerated/x86/asm-coff/padlock-x86-coff.s \
+	lib/accelerated/x86/asm-macosx/cpuid-x86-64-macosx.s \
+	lib/accelerated/x86/asm-macosx/cpuid-x86-macosx.s \
+	lib/accelerated/x86/asm-macosx/appro-aes-gcm-x86-64-macosx.s \
+	lib/accelerated/x86/asm-macosx/appro-aes-x86-64-macosx.s \
+	lib/accelerated/x86/asm-macosx/appro-aes-x86-macosx.s \
+	lib/accelerated/x86/asm-macosx/padlock-x86-64-macosx.s \
+	lib/accelerated/x86/asm-macosx/padlock-x86-macosx.s
 
 asm-sources: $(ASM_SOURCES)
 
@@ -247,3 +255,31 @@ lib/accelerated/x86/asm-coff/cpuid-x86-64-coff.s: devel/perlasm/cpuid-x86_64.pl
 lib/accelerated/x86/asm-coff/cpuid-x86-coff.s: devel/perlasm/cpuid-x86.pl
 	cat devel/perlasm/license-gnutls.txt > $@
 	perl $< coff >> $@
+
+lib/accelerated/x86/asm-macosx/appro-aes-gcm-x86-64-macosx.s: devel/perlasm/ghash-x86_64.pl
+	cat devel/perlasm/license.txt > $@
+	perl $< macosx >> $@
+
+lib/accelerated/x86/asm-macosx/appro-aes-x86-64-macosx.s: devel/perlasm/aesni-x86_64.pl
+	cat devel/perlasm/license.txt > $@
+	perl $< macosx >> $@
+
+lib/accelerated/x86/asm-macosx/appro-aes-x86-macosx.s: devel/perlasm/aesni-x86.pl
+	cat devel/perlasm/license.txt > $@
+	perl $< macosx >> $@
+
+lib/accelerated/x86/asm-macosx/padlock-x86-64-macosx.s: devel/perlasm/e_padlock-x86_64.pl
+	cat devel/perlasm/license.txt > $@
+	perl $< macosx >> $@
+
+lib/accelerated/x86/asm-macosx/padlock-x86-macosx.s: devel/perlasm/e_padlock-x86.pl
+	cat devel/perlasm/license.txt > $@
+	perl $< macosx >> $@
+
+lib/accelerated/x86/asm-macosx/cpuid-x86-64-macosx.s: devel/perlasm/cpuid-x86_64.pl
+	cat devel/perlasm/license-gnutls.txt > $@
+	perl $< macosx >> $@
+
+lib/accelerated/x86/asm-macosx/cpuid-x86-macosx.s: devel/perlasm/cpuid-x86.pl
+	cat devel/perlasm/license-gnutls.txt > $@
+	perl $< macosx >> $@
diff --git a/configure.ac b/configure.ac
index 6fdf1d4bbe..f1a602e34c 100644
--- a/configure.ac
+++ b/configure.ac
@@ -67,11 +67,15 @@ case "$host" in
   *mingw32* | *mingw64*)
     have_win=yes
   ;;
+  *darwin*)
+    have_macosx=yes
+  ;;
   *)
   ;;
 esac
 
 AM_CONDITIONAL(WINDOWS, test "$have_win" = yes)
+AM_CONDITIONAL(MACOSX, test "$have_macosx" = yes)
 
 dnl Hardware Acceleration
 AC_ARG_ENABLE(hardware-acceleration,
diff --git a/lib/accelerated/x86/Makefile.am b/lib/accelerated/x86/Makefile.am
index 3ddd066e27..7040847af4 100644
--- a/lib/accelerated/x86/Makefile.am
+++ b/lib/accelerated/x86/Makefile.am
@@ -38,22 +38,34 @@ noinst_LTLIBRARIES = libx86.la
 libx86_la_SOURCES = sha-padlock.c hmac-padlock.c aes-x86.c aes-padlock.c aes-gcm-padlock.c aes-padlock.h aes-x86.h x86.h sha-padlock.h
 
 if ASM_X86_64
-AM_CFLAGS += -DASM_X86_64 -DASM_X86
-libx86_la_SOURCES += aes-gcm-x86.c
+ AM_CFLAGS += -DASM_X86_64 -DASM_X86
+ libx86_la_SOURCES += aes-gcm-x86.c
 
-if WINDOWS
-libx86_la_SOURCES += asm-coff/appro-aes-x86-64-coff.s asm-coff/padlock-x86-64-coff.s asm-coff/cpuid-x86-64-coff.s asm-coff/appro-aes-gcm-x86-64-coff.s
-else
-libx86_la_SOURCES += asm/appro-aes-x86-64.s asm/appro-aes-gcm-x86-64.s asm/padlock-x86-64.s asm/cpuid-x86-64.s
-endif
+ if WINDOWS
+  libx86_la_SOURCES += asm-coff/appro-aes-x86-64-coff.s asm-coff/padlock-x86-64-coff.s asm-coff/cpuid-x86-64-coff.s asm-coff/appro-aes-gcm-x86-64-coff.s
+ endif
 
-else
-AM_CFLAGS += -DASM_X86_32 -DASM_X86
+ if MACOSX
+  libx86_la_SOURCES += asm-macosx/appro-aes-x86-64-macosx.s asm-macosx/padlock-x86-64-macosx.s asm-macosx/cpuid-x86-64-macosx.s asm-macosx/appro-aes-gcm-x86-64-macosx.s
+ endif
+
+ if ELF
+  libx86_la_SOURCES += asm/appro-aes-x86-64.s asm/appro-aes-gcm-x86-64.s asm/padlock-x86-64.s asm/cpuid-x86-64.s
+ endif
 
-if WINDOWS
-libx86_la_SOURCES += asm-coff/appro-aes-x86-coff.s asm-coff/padlock-x86-coff.s asm-coff/cpuid-x86-coff.s
 else
-libx86_la_SOURCES += asm/appro-aes-x86.s asm/padlock-x86.s asm/cpuid-x86.s
-endif
+ AM_CFLAGS += -DASM_X86_32 -DASM_X86
+
+ if WINDOWS
+  libx86_la_SOURCES += asm-coff/appro-aes-x86-coff.s asm-coff/padlock-x86-coff.s asm-coff/cpuid-x86-coff.s
+ endif
+
+ if MACOSX
+  libx86_la_SOURCES += asm-macosx/appro-aes-x86-macosx.s asm-macosx/padlock-x86-macosx.s asm-macosx/cpuid-x86-macosx.s
+ endif
+
+ if ELF
+  libx86_la_SOURCES += asm/appro-aes-x86.s asm/padlock-x86.s asm/cpuid-x86.s
+ endif
 
 endif
diff --git a/lib/accelerated/x86/asm-macosx/appro-aes-gcm-x86-64-macosx.s b/lib/accelerated/x86/asm-macosx/appro-aes-gcm-x86-64-macosx.s
new file mode 100644
index 0000000000..732332ba28
--- /dev/null
+++ b/lib/accelerated/x86/asm-macosx/appro-aes-gcm-x86-64-macosx.s
@@ -0,0 +1,1063 @@
+# Copyright (c) 2011, Andy Polyakov by <appro@openssl.org>
+# All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions
+# are met:
+# 
+#     * Redistributions of source code must retain copyright notices,
+#      this list of conditions and the following disclaimer.
+#
+#     * 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.
+#
+#     * Neither the name of the Andy Polyakov nor the names of its
+#      copyright holder and contributors may be used to endorse or
+#      promote products derived from this software without specific
+#      prior written permission.
+#
+# ALTERNATIVELY, provided that this notice is retained in full, this
+# product may be distributed under the terms of the GNU General Public
+# License (GPL), in which case the provisions of the GPL apply INSTEAD OF
+# those given above.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDER AND CONTRIBUTORS
+# "AS IS" AND ANY EXPRESS 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 COPYRIGHT
+# OWNER OR 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.
+
+.text	
+
+.globl	_gcm_gmult_4bit
+
+.p2align	4
+_gcm_gmult_4bit:
+	pushq	%rbx
+	pushq	%rbp
+	pushq	%r12
+L$gmult_prologue:
+
+	movzbq	15(%rdi),%r8
+	leaq	L$rem_4bit(%rip),%r11
+	xorq	%rax,%rax
+	xorq	%rbx,%rbx
+	movb	%r8b,%al
+	movb	%r8b,%bl
+	shlb	$4,%al
+	movq	$14,%rcx
+	movq	8(%rsi,%rax,1),%r8
+	movq	(%rsi,%rax,1),%r9
+	andb	$240,%bl
+	movq	%r8,%rdx
+	jmp	L$oop1
+
+.p2align	4
+L$oop1:
+	shrq	$4,%r8
+	andq	$15,%rdx
+	movq	%r9,%r10
+	movb	(%rdi,%rcx,1),%al
+	shrq	$4,%r9
+	xorq	8(%rsi,%rbx,1),%r8
+	shlq	$60,%r10
+	xorq	(%rsi,%rbx,1),%r9
+	movb	%al,%bl
+	xorq	(%r11,%rdx,8),%r9
+	movq	%r8,%rdx
+	shlb	$4,%al
+	xorq	%r10,%r8
+	decq	%rcx
+	js	L$break1
+
+	shrq	$4,%r8
+	andq	$15,%rdx
+	movq	%r9,%r10
+	shrq	$4,%r9
+	xorq	8(%rsi,%rax,1),%r8
+	shlq	$60,%r10
+	xorq	(%rsi,%rax,1),%r9
+	andb	$240,%bl
+	xorq	(%r11,%rdx,8),%r9
+	movq	%r8,%rdx
+	xorq	%r10,%r8
+	jmp	L$oop1
+
+.p2align	4
+L$break1:
+	shrq	$4,%r8
+	andq	$15,%rdx
+	movq	%r9,%r10
+	shrq	$4,%r9
+	xorq	8(%rsi,%rax,1),%r8
+	shlq	$60,%r10
+	xorq	(%rsi,%rax,1),%r9
+	andb	$240,%bl
+	xorq	(%r11,%rdx,8),%r9
+	movq	%r8,%rdx
+	xorq	%r10,%r8
+
+	shrq	$4,%r8
+	andq	$15,%rdx
+	movq	%r9,%r10
+	shrq	$4,%r9
+	xorq	8(%rsi,%rbx,1),%r8
+	shlq	$60,%r10
+	xorq	(%rsi,%rbx,1),%r9
+	xorq	%r10,%r8
+	xorq	(%r11,%rdx,8),%r9
+
+	bswapq	%r8
+	bswapq	%r9
+	movq	%r8,8(%rdi)
+	movq	%r9,(%rdi)
+
+	movq	16(%rsp),%rbx
+	leaq	24(%rsp),%rsp
+L$gmult_epilogue:
+	.byte	0xf3,0xc3
+
+.globl	_gcm_ghash_4bit
+
+.p2align	4
+_gcm_ghash_4bit:
+	pushq	%rbx
+	pushq	%rbp
+	pushq	%r12
+	pushq	%r13
+	pushq	%r14
+	pushq	%r15
+	subq	$280,%rsp
+L$ghash_prologue:
+	movq	%rdx,%r14
+	movq	%rcx,%r15
+	subq	$-128,%rsi
+	leaq	16+128(%rsp),%rbp
+	xorl	%edx,%edx
+	movq	0+0-128(%rsi),%r8
+	movq	0+8-128(%rsi),%rax
+	movb	%al,%dl
+	shrq	$4,%rax
+	movq	%r8,%r10
+	shrq	$4,%r8
+	movq	16+0-128(%rsi),%r9
+	shlb	$4,%dl
+	movq	16+8-128(%rsi),%rbx
+	shlq	$60,%r10
+	movb	%dl,0(%rsp)
+	orq	%r10,%rax
+	movb	%bl,%dl
+	shrq	$4,%rbx
+	movq	%r9,%r10
+	shrq	$4,%r9
+	movq	%r8,0(%rbp)
+	movq	32+0-128(%rsi),%r8
+	shlb	$4,%dl
+	movq	%rax,0-128(%rbp)
+	movq	32+8-128(%rsi),%rax
+	shlq	$60,%r10
+	movb	%dl,1(%rsp)
+	orq	%r10,%rbx
+	movb	%al,%dl
+	shrq	$4,%rax
+	movq	%r8,%r10
+	shrq	$4,%r8
+	movq	%r9,8(%rbp)
+	movq	48+0-128(%rsi),%r9
+	shlb	$4,%dl
+	movq	%rbx,8-128(%rbp)
+	movq	48+8-128(%rsi),%rbx
+	shlq	$60,%r10
+	movb	%dl,2(%rsp)
+	orq	%r10,%rax
+	movb	%bl,%dl
+	shrq	$4,%rbx
+	movq	%r9,%r10
+	shrq	$4,%r9
+	movq	%r8,16(%rbp)
+	movq	64+0-128(%rsi),%r8
+	shlb	$4,%dl
+	movq	%rax,16-128(%rbp)
+	movq	64+8-128(%rsi),%rax
+	shlq	$60,%r10
+	movb	%dl,3(%rsp)
+	orq	%r10,%rbx
+	movb	%al,%dl
+	shrq	$4,%rax
+	movq	%r8,%r10
+	shrq	$4,%r8
+	movq	%r9,24(%rbp)
+	movq	80+0-128(%rsi),%r9
+	shlb	$4,%dl
+	movq	%rbx,24-128(%rbp)
+	movq	80+8-128(%rsi),%rbx
+	shlq	$60,%r10
+	movb	%dl,4(%rsp)
+	orq	%r10,%rax
+	movb	%bl,%dl
+	shrq	$4,%rbx
+	movq	%r9,%r10
+	shrq	$4,%r9
+	movq	%r8,32(%rbp)
+	movq	96+0-128(%rsi),%r8
+	shlb	$4,%dl
+	movq	%rax,32-128(%rbp)
+	movq	96+8-128(%rsi),%rax
+	shlq	$60,%r10
+	movb	%dl,5(%rsp)
+	orq	%r10,%rbx
+	movb	%al,%dl
+	shrq	$4,%rax
+	movq	%r8,%r10
+	shrq	$4,%r8
+	movq	%r9,40(%rbp)
+	movq	112+0-128(%rsi),%r9
+	shlb	$4,%dl
+	movq	%rbx,40-128(%rbp)
+	movq	112+8-128(%rsi),%rbx
+	shlq	$60,%r10
+	movb	%dl,6(%rsp)
+	orq	%r10,%rax
+	movb	%bl,%dl
+	shrq	$4,%rbx
+	movq	%r9,%r10
+	shrq	$4,%r9
+	movq	%r8,48(%rbp)
+	movq	128+0-128(%rsi),%r8
+	shlb	$4,%dl
+	movq	%rax,48-128(%rbp)
+	movq	128+8-128(%rsi),%rax
+	shlq	$60,%r10
+	movb	%dl,7(%rsp)
+	orq	%r10,%rbx
+	movb	%al,%dl
+	shrq	$4,%rax
+	movq	%r8,%r10
+	shrq	$4,%r8
+	movq	%r9,56(%rbp)
+	movq	144+0-128(%rsi),%r9
+	shlb	$4,%dl
+	movq	%rbx,56-128(%rbp)
+	movq	144+8-128(%rsi),%rbx
+	shlq	$60,%r10
+	movb	%dl,8(%rsp)
+	orq	%r10,%rax
+	movb	%bl,%dl
+	shrq	$4,%rbx
+	movq	%r9,%r10
+	shrq	$4,%r9
+	movq	%r8,64(%rbp)
+	movq	160+0-128(%rsi),%r8
+	shlb	$4,%dl
+	movq	%rax,64-128(%rbp)
+	movq	160+8-128(%rsi),%rax
+	shlq	$60,%r10
+	movb	%dl,9(%rsp)
+	orq	%r10,%rbx
+	movb	%al,%dl
+	shrq	$4,%rax
+	movq	%r8,%r10
+	shrq	$4,%r8
+	movq	%r9,72(%rbp)
+	movq	176+0-128(%rsi),%r9
+	shlb	$4,%dl
+	movq	%rbx,72-128(%rbp)
+	movq	176+8-128(%rsi),%rbx
+	shlq	$60,%r10
+	movb	%dl,10(%rsp)
+	orq	%r10,%rax
+	movb	%bl,%dl
+	shrq	$4,%rbx
+	movq	%r9,%r10
+	shrq	$4,%r9
+	movq	%r8,80(%rbp)
+	movq	192+0-128(%rsi),%r8
+	shlb	$4,%dl
+	movq	%rax,80-128(%rbp)
+	movq	192+8-128(%rsi),%rax
+	shlq	$60,%r10
+	movb	%dl,11(%rsp)
+	orq	%r10,%rbx
+	movb	%al,%dl
+	shrq	$4,%rax
+	movq	%r8,%r10
+	shrq	$4,%r8
+	movq	%r9,88(%rbp)
+	movq	208+0-128(%rsi),%r9
+	shlb	$4,%dl
+	movq	%rbx,88-128(%rbp)
+	movq	208+8-128(%rsi),%rbx
+	shlq	$60,%r10
+	movb	%dl,12(%rsp)
+	orq	%r10,%rax
+	movb	%bl,%dl
+	shrq	$4,%rbx
+	movq	%r9,%r10
+	shrq	$4,%r9
+	movq	%r8,96(%rbp)
+	movq	224+0-128(%rsi),%r8
+	shlb	$4,%dl
+	movq	%rax,96-128(%rbp)
+	movq	224+8-128(%rsi),%rax
+	shlq	$60,%r10
+	movb	%dl,13(%rsp)
+	orq	%r10,%rbx
+	movb	%al,%dl
+	shrq	$4,%rax
+	movq	%r8,%r10
+	shrq	$4,%r8
+	movq	%r9,104(%rbp)
+	movq	240+0-128(%rsi),%r9
+	shlb	$4,%dl
+	movq	%rbx,104-128(%rbp)
+	movq	240+8-128(%rsi),%rbx
+	shlq	$60,%r10
+	movb	%dl,14(%rsp)
+	orq	%r10,%rax
+	movb	%bl,%dl
+	shrq	$4,%rbx
+	movq	%r9,%r10
+	shrq	$4,%r9
+	movq	%r8,112(%rbp)
+	shlb	$4,%dl
+	movq	%rax,112-128(%rbp)
+	shlq	$60,%r10
+	movb	%dl,15(%rsp)
+	orq	%r10,%rbx
+	movq	%r9,120(%rbp)
+	movq	%rbx,120-128(%rbp)
+	addq	$-128,%rsi
+	movq	8(%rdi),%r8
+	movq	0(%rdi),%r9
+	addq	%r14,%r15
+	leaq	L$rem_8bit(%rip),%r11
+	jmp	L$outer_loop
+.p2align	4
+L$outer_loop:
+	xorq	(%r14),%r9
+	movq	8(%r14),%rdx
+	leaq	16(%r14),%r14
+	xorq	%r8,%rdx
+	movq	%r9,(%rdi)
+	movq	%rdx,8(%rdi)
+	shrq	$32,%rdx
+	xorq	%rax,%rax
+	roll	$8,%edx
+	movb	%dl,%al
+	movzbl	%dl,%ebx
+	shlb	$4,%al
+	shrl	$4,%ebx
+	roll	$8,%edx
+	movq	8(%rsi,%rax,1),%r8
+	movq	(%rsi,%rax,1),%r9
+	movb	%dl,%al
+	movzbl	%dl,%ecx
+	shlb	$4,%al
+	movzbq	(%rsp,%rbx,1),%r12
+	shrl	$4,%ecx
+	xorq	%r8,%r12
+	movq	%r9,%r10
+	shrq	$8,%r8
+	movzbq	%r12b,%r12
+	shrq	$8,%r9
+	xorq	-128(%rbp,%rbx,8),%r8
+	shlq	$56,%r10
+	xorq	(%rbp,%rbx,8),%r9
+	roll	$8,%edx
+	xorq	8(%rsi,%rax,1),%r8
+	xorq	(%rsi,%rax,1),%r9
+	movb	%dl,%al
+	xorq	%r10,%r8
+	movzwq	(%r11,%r12,2),%r12
+	movzbl	%dl,%ebx
+	shlb	$4,%al
+	movzbq	(%rsp,%rcx,1),%r13
+	shrl	$4,%ebx
+	shlq	$48,%r12
+	xorq	%r8,%r13
+	movq	%r9,%r10
+	xorq	%r12,%r9
+	shrq	$8,%r8
+	movzbq	%r13b,%r13
+	shrq	$8,%r9
+	xorq	-128(%rbp,%rcx,8),%r8
+	shlq	$56,%r10
+	xorq	(%rbp,%rcx,8),%r9
+	roll	$8,%edx
+	xorq	8(%rsi,%rax,1),%r8
+	xorq	(%rsi,%rax,1),%r9
+	movb	%dl,%al
+	xorq	%r10,%r8
+	movzwq	(%r11,%r13,2),%r13
+	movzbl	%dl,%ecx
+	shlb	$4,%al
+	movzbq	(%rsp,%rbx,1),%r12
+	shrl	$4,%ecx
+	shlq	$48,%r13
+	xorq	%r8,%r12
+	movq	%r9,%r10
+	xorq	%r13,%r9
+	shrq	$8,%r8
+	movzbq	%r12b,%r12
+	movl	8(%rdi),%edx
+	shrq	$8,%r9
+	xorq	-128(%rbp,%rbx,8),%r8
+	shlq	$56,%r10
+	xorq	(%rbp,%rbx,8),%r9
+	roll	$8,%edx
+	xorq	8(%rsi,%rax,1),%r8
+	xorq	(%rsi,%rax,1),%r9
+	movb	%dl,%al
+	xorq	%r10,%r8
+	movzwq	(%r11,%r12,2),%r12
+	movzbl	%dl,%ebx
+	shlb	$4,%al
+	movzbq	(%rsp,%rcx,1),%r13
+	shrl	$4,%ebx
+	shlq	$48,%r12
+	xorq	%r8,%r13
+	movq	%r9,%r10
+	xorq	%r12,%r9
+	shrq	$8,%r8
+	movzbq	%r13b,%r13
+	shrq	$8,%r9
+	xorq	-128(%rbp,%rcx,8),%r8
+	shlq	$56,%r10
+	xorq	(%rbp,%rcx,8),%r9
+	roll	$8,%edx
+	xorq	8(%rsi,%rax,1),%r8
+	xorq	(%rsi,%rax,1),%r9
+	movb	%dl,%al
+	xorq	%r10,%r8
+	movzwq	(%r11,%r13,2),%r13
+	movzbl	%dl,%ecx
+	shlb	$4,%al
+	movzbq	(%rsp,%rbx,1),%r12
+	shrl	$4,%ecx
+	shlq	$48,%r13
+	xorq	%r8,%r12
+	movq	%r9,%r10
+	xorq	%r13,%r9
+	shrq	$8,%r8
+	movzbq	%r12b,%r12
+	shrq	$8,%r9
+	xorq	-128(%rbp,%rbx,8),%r8
+	shlq	$56,%r10
+	xorq	(%rbp,%rbx,8),%r9
+	roll	$8,%edx
+	xorq	8(%rsi,%rax,1),%r8
+	xorq	(%rsi,%rax,1),%r9
+	movb	%dl,%al
+	xorq	%r10,%r8
+	movzwq	(%r11,%r12,2),%r12
+	movzbl	%dl,%ebx
+	shlb	$4,%al
+	movzbq	(%rsp,%rcx,1),%r13
+	shrl	$4,%ebx
+	shlq	$48,%r12
+	xorq	%r8,%r13
+	movq	%r9,%r10
+	xorq	%r12,%r9
+	shrq	$8,%r8
+	movzbq	%r13b,%r13
+	shrq	$8,%r9
+	xorq	-128(%rbp,%rcx,8),%r8
+	shlq	$56,%r10
+	xorq	(%rbp,%rcx,8),%r9
+	roll	$8,%edx
+	xorq	8(%rsi,%rax,1),%r8
+	xorq	(%rsi,%rax,1),%r9
+	movb	%dl,%al
+	xorq	%r10,%r8
+	movzwq	(%r11,%r13,2),%r13
+	movzbl	%dl,%ecx
+	shlb	$4,%al
+	movzbq	(%rsp,%rbx,1),%r12
+	shrl	$4,%ecx
+	shlq	$48,%r13
+	xorq	%r8,%r12
+	movq	%r9,%r10
+	xorq	%r13,%r9
+	shrq	$8,%r8
+	movzbq	%r12b,%r12
+	movl	4(%rdi),%edx
+	shrq	$8,%r9
+	xorq	-128(%rbp,%rbx,8),%r8
+	shlq	$56,%r10
+	xorq	(%rbp,%rbx,8),%r9
+	roll	$8,%edx
+	xorq	8(%rsi,%rax,1),%r8
+	xorq	(%rsi,%rax,1),%r9
+	movb	%dl,%al
+	xorq	%r10,%r8
+	movzwq	(%r11,%r12,2),%r12
+	movzbl	%dl,%ebx
+	shlb	$4,%al
+	movzbq	(%rsp,%rcx,1),%r13
+	shrl	$4,%ebx
+	shlq	$48,%r12
+	xorq	%r8,%r13
+	movq	%r9,%r10
+	xorq	%r12,%r9
+	shrq	$8,%r8
+	movzbq	%r13b,%r13
+	shrq	$8,%r9
+	xorq	-128(%rbp,%rcx,8),%r8
+	shlq	$56,%r10
+	xorq	(%rbp,%rcx,8),%r9
+	roll	$8,%edx
+	xorq	8(%rsi,%rax,1),%r8
+	xorq	(%rsi,%rax,1),%r9
+	movb	%dl,%al
+	xorq	%r10,%r8
+	movzwq	(%r11,%r13,2),%r13
+	movzbl	%dl,%ecx
+	shlb	$4,%al
+	movzbq	(%rsp,%rbx,1),%r12
+	shrl	$4,%ecx
+	shlq	$48,%r13
+	xorq	%r8,%r12
+	movq	%r9,%r10
+	xorq	%r13,%r9
+	shrq	$8,%r8
+	movzbq	%r12b,%r12
+	shrq	$8,%r9
+	xorq	-128(%rbp,%rbx,8),%r8
+	shlq	$56,%r10
+	xorq	(%rbp,%rbx,8),%r9
+	roll	$8,%edx
+	xorq	8(%rsi,%rax,1),%r8
+	xorq	(%rsi,%rax,1),%r9
+	movb	%dl,%al
+	xorq	%r10,%r8
+	movzwq	(%r11,%r12,2),%r12
+	movzbl	%dl,%ebx
+	shlb	$4,%al
+	movzbq	(%rsp,%rcx,1),%r13
+	shrl	$4,%ebx
+	shlq	$48,%r12
+	xorq	%r8,%r13
+	movq	%r9,%r10
+	xorq	%r12,%r9
+	shrq	$8,%r8
+	movzbq	%r13b,%r13
+	shrq	$8,%r9
+	xorq	-128(%rbp,%rcx,8),%r8
+	shlq	$56,%r10
+	xorq	(%rbp,%rcx,8),%r9
+	roll	$8,%edx
+	xorq	8(%rsi,%rax,1),%r8
+	xorq	(%rsi,%rax,1),%r9
+	movb	%dl,%al
+	xorq	%r10,%r8
+	movzwq	(%r11,%r13,2),%r13
+	movzbl	%dl,%ecx
+	shlb	$4,%al
+	movzbq	(%rsp,%rbx,1),%r12
+	shrl	$4,%ecx
+	shlq	$48,%r13
+	xorq	%r8,%r12
+	movq	%r9,%r10
+	xorq	%r13,%r9
+	shrq	$8,%r8
+	movzbq	%r12b,%r12
+	movl	0(%rdi),%edx
+	shrq	$8,%r9
+	xorq	-128(%rbp,%rbx,8),%r8
+	shlq	$56,%r10
+	xorq	(%rbp,%rbx,8),%r9
+	roll	$8,%edx
+	xorq	8(%rsi,%rax,1),%r8
+	xorq	(%rsi,%rax,1),%r9
+	movb	%dl,%al
+	xorq	%r10,%r8
+	movzwq	(%r11,%r12,2),%r12
+	movzbl	%dl,%ebx
+	shlb	$4,%al
+	movzbq	(%rsp,%rcx,1),%r13
+	shrl	$4,%ebx
+	shlq	$48,%r12
+	xorq	%r8,%r13
+	movq	%r9,%r10
+	xorq	%r12,%r9
+	shrq	$8,%r8
+	movzbq	%r13b,%r13
+	shrq	$8,%r9
+	xorq	-128(%rbp,%rcx,8),%r8
+	shlq	$56,%r10
+	xorq	(%rbp,%rcx,8),%r9
+	roll	$8,%edx
+	xorq	8(%rsi,%rax,1),%r8
+	xorq	(%rsi,%rax,1),%r9
+	movb	%dl,%al
+	xorq	%r10,%r8
+	movzwq	(%r11,%r13,2),%r13
+	movzbl	%dl,%ecx
+	shlb	$4,%al
+	movzbq	(%rsp,%rbx,1),%r12
+	shrl	$4,%ecx
+	shlq	$48,%r13
+	xorq	%r8,%r12
+	movq	%r9,%r10
+	xorq	%r13,%r9
+	shrq	$8,%r8
+	movzbq	%r12b,%r12
+	shrq	$8,%r9
+	xorq	-128(%rbp,%rbx,8),%r8
+	shlq	$56,%r10
+	xorq	(%rbp,%rbx,8),%r9
+	roll	$8,%edx
+	xorq	8(%rsi,%rax,1),%r8
+	xorq	(%rsi,%rax,1),%r9
+	movb	%dl,%al
+	xorq	%r10,%r8
+	movzwq	(%r11,%r12,2),%r12
+	movzbl	%dl,%ebx
+	shlb	$4,%al
+	movzbq	(%rsp,%rcx,1),%r13
+	shrl	$4,%ebx
+	shlq	$48,%r12
+	xorq	%r8,%r13
+	movq	%r9,%r10
+	xorq	%r12,%r9
+	shrq	$8,%r8
+	movzbq	%r13b,%r13
+	shrq	$8,%r9
+	xorq	-128(%rbp,%rcx,8),%r8
+	shlq	$56,%r10
+	xorq	(%rbp,%rcx,8),%r9
+	roll	$8,%edx
+	xorq	8(%rsi,%rax,1),%r8
+	xorq	(%rsi,%rax,1),%r9
+	movb	%dl,%al
+	xorq	%r10,%r8
+	movzwq	(%r11,%r13,2),%r13
+	movzbl	%dl,%ecx
+	shlb	$4,%al
+	movzbq	(%rsp,%rbx,1),%r12
+	andl	$240,%ecx
+	shlq	$48,%r13
+	xorq	%r8,%r12
+	movq	%r9,%r10
+	xorq	%r13,%r9
+	shrq	$8,%r8
+	movzbq	%r12b,%r12
+	movl	-4(%rdi),%edx
+	shrq	$8,%r9
+	xorq	-128(%rbp,%rbx,8),%r8
+	shlq	$56,%r10
+	xorq	(%rbp,%rbx,8),%r9
+	movzwq	(%r11,%r12,2),%r12
+	xorq	8(%rsi,%rax,1),%r8
+	xorq	(%rsi,%rax,1),%r9
+	shlq	$48,%r12
+	xorq	%r10,%r8
+	xorq	%r12,%r9
+	movzbq	%r8b,%r13
+	shrq	$4,%r8
+	movq	%r9,%r10
+	shlb	$4,%r13b
+	shrq	$4,%r9
+	xorq	8(%rsi,%rcx,1),%r8
+	movzwq	(%r11,%r13,2),%r13
+	shlq	$60,%r10
+	xorq	(%rsi,%rcx,1),%r9
+	xorq	%r10,%r8
+	shlq	$48,%r13
+	bswapq	%r8
+	xorq	%r13,%r9
+	bswapq	%r9
+	cmpq	%r15,%r14
+	jb	L$outer_loop
+	movq	%r8,8(%rdi)
+	movq	%r9,(%rdi)
+
+	leaq	280(%rsp),%rsi
+	movq	0(%rsi),%r15
+	movq	8(%rsi),%r14
+	movq	16(%rsi),%r13
+	movq	24(%rsi),%r12
+	movq	32(%rsi),%rbp
+	movq	40(%rsi),%rbx
+	leaq	48(%rsi),%rsp
+L$ghash_epilogue:
+	.byte	0xf3,0xc3
+
+.globl	_gcm_init_clmul
+
+.p2align	4
+_gcm_init_clmul:
+	movdqu	(%rsi),%xmm2
+	pshufd	$78,%xmm2,%xmm2
+
+
+	pshufd	$255,%xmm2,%xmm4
+	movdqa	%xmm2,%xmm3
+	psllq	$1,%xmm2
+	pxor	%xmm5,%xmm5
+	psrlq	$63,%xmm3
+	pcmpgtd	%xmm4,%xmm5
+	pslldq	$8,%xmm3
+	por	%xmm3,%xmm2
+
+
+	pand	L$0x1c2_polynomial(%rip),%xmm5
+	pxor	%xmm5,%xmm2
+
+
+	movdqa	%xmm2,%xmm0
+	movdqa	%xmm0,%xmm1
+	pshufd	$78,%xmm0,%xmm3
+	pshufd	$78,%xmm2,%xmm4
+	pxor	%xmm0,%xmm3
+	pxor	%xmm2,%xmm4
+.byte	102,15,58,68,194,0
+.byte	102,15,58,68,202,17
+.byte	102,15,58,68,220,0
+	pxor	%xmm0,%xmm3
+	pxor	%xmm1,%xmm3
+
+	movdqa	%xmm3,%xmm4
+	psrldq	$8,%xmm3
+	pslldq	$8,%xmm4
+	pxor	%xmm3,%xmm1
+	pxor	%xmm4,%xmm0
+
+	movdqa	%xmm0,%xmm3
+	psllq	$1,%xmm0
+	pxor	%xmm3,%xmm0
+	psllq	$5,%xmm0
+	pxor	%xmm3,%xmm0
+	psllq	$57,%xmm0
+	movdqa	%xmm0,%xmm4
+	pslldq	$8,%xmm0
+	psrldq	$8,%xmm4
+	pxor	%xmm3,%xmm0
+	pxor	%xmm4,%xmm1
+
+
+	movdqa	%xmm0,%xmm4
+	psrlq	$5,%xmm0
+	pxor	%xmm4,%xmm0
+	psrlq	$1,%xmm0
+	pxor	%xmm4,%xmm0
+	pxor	%xmm1,%xmm4
+	psrlq	$1,%xmm0
+	pxor	%xmm4,%xmm0
+	movdqu	%xmm2,(%rdi)
+	movdqu	%xmm0,16(%rdi)
+	.byte	0xf3,0xc3
+
+.globl	_gcm_gmult_clmul
+
+.p2align	4
+_gcm_gmult_clmul:
+	movdqu	(%rdi),%xmm0
+	movdqa	L$bswap_mask(%rip),%xmm5
+	movdqu	(%rsi),%xmm2
+.byte	102,15,56,0,197
+	movdqa	%xmm0,%xmm1
+	pshufd	$78,%xmm0,%xmm3
+	pshufd	$78,%xmm2,%xmm4
+	pxor	%xmm0,%xmm3
+	pxor	%xmm2,%xmm4
+.byte	102,15,58,68,194,0
+.byte	102,15,58,68,202,17
+.byte	102,15,58,68,220,0
+	pxor	%xmm0,%xmm3
+	pxor	%xmm1,%xmm3
+
+	movdqa	%xmm3,%xmm4
+	psrldq	$8,%xmm3
+	pslldq	$8,%xmm4
+	pxor	%xmm3,%xmm1
+	pxor	%xmm4,%xmm0
+
+	movdqa	%xmm0,%xmm3
+	psllq	$1,%xmm0
+	pxor	%xmm3,%xmm0
+	psllq	$5,%xmm0
+	pxor	%xmm3,%xmm0
+	psllq	$57,%xmm0
+	movdqa	%xmm0,%xmm4
+	pslldq	$8,%xmm0
+	psrldq	$8,%xmm4
+	pxor	%xmm3,%xmm0
+	pxor	%xmm4,%xmm1
+
+
+	movdqa	%xmm0,%xmm4
+	psrlq	$5,%xmm0
+	pxor	%xmm4,%xmm0
+	psrlq	$1,%xmm0
+	pxor	%xmm4,%xmm0
+	pxor	%xmm1,%xmm4
+	psrlq	$1,%xmm0
+	pxor	%xmm4,%xmm0
+.byte	102,15,56,0,197
+	movdqu	%xmm0,(%rdi)
+	.byte	0xf3,0xc3
+
+.globl	_gcm_ghash_clmul
+
+.p2align	4
+_gcm_ghash_clmul:
+	movdqa	L$bswap_mask(%rip),%xmm5
+
+	movdqu	(%rdi),%xmm0
+	movdqu	(%rsi),%xmm2
+.byte	102,15,56,0,197
+
+	subq	$16,%rcx
+	jz	L$odd_tail
+
+	movdqu	16(%rsi),%xmm8
+
+
+
+
+
+	movdqu	(%rdx),%xmm3
+	movdqu	16(%rdx),%xmm6
+.byte	102,15,56,0,221
+.byte	102,15,56,0,245
+	pxor	%xmm3,%xmm0
+	movdqa	%xmm6,%xmm7
+	pshufd	$78,%xmm6,%xmm3
+	pshufd	$78,%xmm2,%xmm4
+	pxor	%xmm6,%xmm3
+	pxor	%xmm2,%xmm4
+.byte	102,15,58,68,242,0
+.byte	102,15,58,68,250,17
+.byte	102,15,58,68,220,0
+	pxor	%xmm6,%xmm3
+	pxor	%xmm7,%xmm3
+
+	movdqa	%xmm3,%xmm4
+	psrldq	$8,%xmm3
+	pslldq	$8,%xmm4
+	pxor	%xmm3,%xmm7
+	pxor	%xmm4,%xmm6
+	movdqa	%xmm0,%xmm1
+	pshufd	$78,%xmm0,%xmm3
+	pshufd	$78,%xmm8,%xmm4
+	pxor	%xmm0,%xmm3
+	pxor	%xmm8,%xmm4
+
+	leaq	32(%rdx),%rdx
+	subq	$32,%rcx
+	jbe	L$even_tail
+
+L$mod_loop:
+.byte	102,65,15,58,68,192,0
+.byte	102,65,15,58,68,200,17
+.byte	102,15,58,68,220,0
+	pxor	%xmm0,%xmm3
+	pxor	%xmm1,%xmm3
+
+	movdqa	%xmm3,%xmm4
+	psrldq	$8,%xmm3
+	pslldq	$8,%xmm4
+	pxor	%xmm3,%xmm1
+	pxor	%xmm4,%xmm0
+	movdqu	(%rdx),%xmm3
+	pxor	%xmm6,%xmm0
+	pxor	%xmm7,%xmm1
+
+	movdqu	16(%rdx),%xmm6
+.byte	102,15,56,0,221
+.byte	102,15,56,0,245
+
+	movdqa	%xmm6,%xmm7
+	pshufd	$78,%xmm6,%xmm9
+	pshufd	$78,%xmm2,%xmm10
+	pxor	%xmm6,%xmm9
+	pxor	%xmm2,%xmm10
+	pxor	%xmm3,%xmm1
+
+	movdqa	%xmm0,%xmm3
+	psllq	$1,%xmm0
+	pxor	%xmm3,%xmm0
+	psllq	$5,%xmm0
+	pxor	%xmm3,%xmm0
+.byte	102,15,58,68,242,0
+	psllq	$57,%xmm0
+	movdqa	%xmm0,%xmm4
+	pslldq	$8,%xmm0
+	psrldq	$8,%xmm4
+	pxor	%xmm3,%xmm0
+	pxor	%xmm4,%xmm1
+
+.byte	102,15,58,68,250,17
+	movdqa	%xmm0,%xmm4
+	psrlq	$5,%xmm0
+	pxor	%xmm4,%xmm0
+	psrlq	$1,%xmm0
+	pxor	%xmm4,%xmm0
+	pxor	%xmm1,%xmm4
+	psrlq	$1,%xmm0
+	pxor	%xmm4,%xmm0
+
+.byte	102,69,15,58,68,202,0
+	movdqa	%xmm0,%xmm1
+	pshufd	$78,%xmm0,%xmm3
+	pshufd	$78,%xmm8,%xmm4
+	pxor	%xmm0,%xmm3
+	pxor	%xmm8,%xmm4
+
+	pxor	%xmm6,%xmm9
+	pxor	%xmm7,%xmm9
+	movdqa	%xmm9,%xmm10
+	psrldq	$8,%xmm9
+	pslldq	$8,%xmm10
+	pxor	%xmm9,%xmm7
+	pxor	%xmm10,%xmm6
+
+	leaq	32(%rdx),%rdx
+	subq	$32,%rcx
+	ja	L$mod_loop
+
+L$even_tail:
+.byte	102,65,15,58,68,192,0
+.byte	102,65,15,58,68,200,17
+.byte	102,15,58,68,220,0
+	pxor	%xmm0,%xmm3
+	pxor	%xmm1,%xmm3
+
+	movdqa	%xmm3,%xmm4
+	psrldq	$8,%xmm3
+	pslldq	$8,%xmm4
+	pxor	%xmm3,%xmm1
+	pxor	%xmm4,%xmm0
+	pxor	%xmm6,%xmm0
+	pxor	%xmm7,%xmm1
+
+	movdqa	%xmm0,%xmm3
+	psllq	$1,%xmm0
+	pxor	%xmm3,%xmm0
+	psllq	$5,%xmm0
+	pxor	%xmm3,%xmm0
+	psllq	$57,%xmm0
+	movdqa	%xmm0,%xmm4
+	pslldq	$8,%xmm0
+	psrldq	$8,%xmm4
+	pxor	%xmm3,%xmm0
+	pxor	%xmm4,%xmm1
+
+
+	movdqa	%xmm0,%xmm4
+	psrlq	$5,%xmm0
+	pxor	%xmm4,%xmm0
+	psrlq	$1,%xmm0
+	pxor	%xmm4,%xmm0
+	pxor	%xmm1,%xmm4
+	psrlq	$1,%xmm0
+	pxor	%xmm4,%xmm0
+	testq	%rcx,%rcx
+	jnz	L$done
+
+L$odd_tail:
+	movdqu	(%rdx),%xmm3
+.byte	102,15,56,0,221
+	pxor	%xmm3,%xmm0
+	movdqa	%xmm0,%xmm1
+	pshufd	$78,%xmm0,%xmm3
+	pshufd	$78,%xmm2,%xmm4
+	pxor	%xmm0,%xmm3
+	pxor	%xmm2,%xmm4
+.byte	102,15,58,68,194,0
+.byte	102,15,58,68,202,17
+.byte	102,15,58,68,220,0
+	pxor	%xmm0,%xmm3
+	pxor	%xmm1,%xmm3
+
+	movdqa	%xmm3,%xmm4
+	psrldq	$8,%xmm3
+	pslldq	$8,%xmm4
+	pxor	%xmm3,%xmm1
+	pxor	%xmm4,%xmm0
+
+	movdqa	%xmm0,%xmm3
+	psllq	$1,%xmm0
+	pxor	%xmm3,%xmm0
+	psllq	$5,%xmm0
+	pxor	%xmm3,%xmm0
+	psllq	$57,%xmm0
+	movdqa	%xmm0,%xmm4
+	pslldq	$8,%xmm0
+	psrldq	$8,%xmm4
+	pxor	%xmm3,%xmm0
+	pxor	%xmm4,%xmm1
+
+
+	movdqa	%xmm0,%xmm4
+	psrlq	$5,%xmm0
+	pxor	%xmm4,%xmm0
+	psrlq	$1,%xmm0
+	pxor	%xmm4,%xmm0
+	pxor	%xmm1,%xmm4
+	psrlq	$1,%xmm0
+	pxor	%xmm4,%xmm0
+L$done:
+.byte	102,15,56,0,197
+	movdqu	%xmm0,(%rdi)
+	.byte	0xf3,0xc3
+L$SEH_end_gcm_ghash_clmul:
+
+.p2align	6
+L$bswap_mask:
+.byte	15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0
+L$0x1c2_polynomial:
+.byte	1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0xc2
+.p2align	6
+
+L$rem_4bit:
+.long	0,0,0,471859200,0,943718400,0,610271232
+.long	0,1887436800,0,1822425088,0,1220542464,0,1423966208
+.long	0,3774873600,0,4246732800,0,3644850176,0,3311403008
+.long	0,2441084928,0,2376073216,0,2847932416,0,3051356160
+
+L$rem_8bit:
+.value	0x0000,0x01C2,0x0384,0x0246,0x0708,0x06CA,0x048C,0x054E
+.value	0x0E10,0x0FD2,0x0D94,0x0C56,0x0918,0x08DA,0x0A9C,0x0B5E
+.value	0x1C20,0x1DE2,0x1FA4,0x1E66,0x1B28,0x1AEA,0x18AC,0x196E
+.value	0x1230,0x13F2,0x11B4,0x1076,0x1538,0x14FA,0x16BC,0x177E
+.value	0x3840,0x3982,0x3BC4,0x3A06,0x3F48,0x3E8A,0x3CCC,0x3D0E
+.value	0x3650,0x3792,0x35D4,0x3416,0x3158,0x309A,0x32DC,0x331E
+.value	0x2460,0x25A2,0x27E4,0x2626,0x2368,0x22AA,0x20EC,0x212E
+.value	0x2A70,0x2BB2,0x29F4,0x2836,0x2D78,0x2CBA,0x2EFC,0x2F3E
+.value	0x7080,0x7142,0x7304,0x72C6,0x7788,0x764A,0x740C,0x75CE
+.value	0x7E90,0x7F52,0x7D14,0x7CD6,0x7998,0x785A,0x7A1C,0x7BDE
+.value	0x6CA0,0x6D62,0x6F24,0x6EE6,0x6BA8,0x6A6A,0x682C,0x69EE
+.value	0x62B0,0x6372,0x6134,0x60F6,0x65B8,0x647A,0x663C,0x67FE
+.value	0x48C0,0x4902,0x4B44,0x4A86,0x4FC8,0x4E0A,0x4C4C,0x4D8E
+.value	0x46D0,0x4712,0x4554,0x4496,0x41D8,0x401A,0x425C,0x439E
+.value	0x54E0,0x5522,0x5764,0x56A6,0x53E8,0x522A,0x506C,0x51AE
+.value	0x5AF0,0x5B32,0x5974,0x58B6,0x5DF8,0x5C3A,0x5E7C,0x5FBE
+.value	0xE100,0xE0C2,0xE284,0xE346,0xE608,0xE7CA,0xE58C,0xE44E
+.value	0xEF10,0xEED2,0xEC94,0xED56,0xE818,0xE9DA,0xEB9C,0xEA5E
+.value	0xFD20,0xFCE2,0xFEA4,0xFF66,0xFA28,0xFBEA,0xF9AC,0xF86E
+.value	0xF330,0xF2F2,0xF0B4,0xF176,0xF438,0xF5FA,0xF7BC,0xF67E
+.value	0xD940,0xD882,0xDAC4,0xDB06,0xDE48,0xDF8A,0xDDCC,0xDC0E
+.value	0xD750,0xD692,0xD4D4,0xD516,0xD058,0xD19A,0xD3DC,0xD21E
+.value	0xC560,0xC4A2,0xC6E4,0xC726,0xC268,0xC3AA,0xC1EC,0xC02E
+.value	0xCB70,0xCAB2,0xC8F4,0xC936,0xCC78,0xCDBA,0xCFFC,0xCE3E
+.value	0x9180,0x9042,0x9204,0x93C6,0x9688,0x974A,0x950C,0x94CE
+.value	0x9F90,0x9E52,0x9C14,0x9DD6,0x9898,0x995A,0x9B1C,0x9ADE
+.value	0x8DA0,0x8C62,0x8E24,0x8FE6,0x8AA8,0x8B6A,0x892C,0x88EE
+.value	0x83B0,0x8272,0x8034,0x81F6,0x84B8,0x857A,0x873C,0x86FE
+.value	0xA9C0,0xA802,0xAA44,0xAB86,0xAEC8,0xAF0A,0xAD4C,0xAC8E
+.value	0xA7D0,0xA612,0xA454,0xA596,0xA0D8,0xA11A,0xA35C,0xA29E
+.value	0xB5E0,0xB422,0xB664,0xB7A6,0xB2E8,0xB32A,0xB16C,0xB0AE
+.value	0xBBF0,0xBA32,0xB874,0xB9B6,0xBCF8,0xBD3A,0xBF7C,0xBEBE
+
+.byte	71,72,65,83,72,32,102,111,114,32,120,56,54,95,54,52,44,32,67,82,89,80,84,79,71,65,77,83,32,98,121,32,60,97,112,112,114,111,64,111,112,101,110,115,115,108,46,111,114,103,62,0
+.p2align	6
diff --git a/lib/accelerated/x86/asm-macosx/appro-aes-x86-64-macosx.s b/lib/accelerated/x86/asm-macosx/appro-aes-x86-64-macosx.s
new file mode 100644
index 0000000000..de7ca030c1
--- /dev/null
+++ b/lib/accelerated/x86/asm-macosx/appro-aes-x86-64-macosx.s
@@ -0,0 +1,2572 @@
+# Copyright (c) 2011, Andy Polyakov by <appro@openssl.org>
+# All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions
+# are met:
+# 
+#     * Redistributions of source code must retain copyright notices,
+#      this list of conditions and the following disclaimer.
+#
+#     * 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.
+#
+#     * Neither the name of the Andy Polyakov nor the names of its
+#      copyright holder and contributors may be used to endorse or
+#      promote products derived from this software without specific
+#      prior written permission.
+#
+# ALTERNATIVELY, provided that this notice is retained in full, this
+# product may be distributed under the terms of the GNU General Public
+# License (GPL), in which case the provisions of the GPL apply INSTEAD OF
+# those given above.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDER AND CONTRIBUTORS
+# "AS IS" AND ANY EXPRESS 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 COPYRIGHT
+# OWNER OR 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.
+
+.text	
+.globl	_aesni_encrypt
+
+.p2align	4
+_aesni_encrypt:
+	movups	(%rdi),%xmm2
+	movl	240(%rdx),%eax
+	movups	(%rdx),%xmm0
+	movups	16(%rdx),%xmm1
+	leaq	32(%rdx),%rdx
+	xorps	%xmm0,%xmm2
+L$oop_enc1_1:
+.byte	102,15,56,220,209
+	decl	%eax
+	movups	(%rdx),%xmm1
+	leaq	16(%rdx),%rdx
+	jnz	L$oop_enc1_1	
+.byte	102,15,56,221,209
+	movups	%xmm2,(%rsi)
+	.byte	0xf3,0xc3
+
+
+.globl	_aesni_decrypt
+
+.p2align	4
+_aesni_decrypt:
+	movups	(%rdi),%xmm2
+	movl	240(%rdx),%eax
+	movups	(%rdx),%xmm0
+	movups	16(%rdx),%xmm1
+	leaq	32(%rdx),%rdx
+	xorps	%xmm0,%xmm2
+L$oop_dec1_2:
+.byte	102,15,56,222,209
+	decl	%eax
+	movups	(%rdx),%xmm1
+	leaq	16(%rdx),%rdx
+	jnz	L$oop_dec1_2	
+.byte	102,15,56,223,209
+	movups	%xmm2,(%rsi)
+	.byte	0xf3,0xc3
+
+
+.p2align	4
+_aesni_encrypt3:
+	movups	(%rcx),%xmm0
+	shrl	$1,%eax
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+	xorps	%xmm0,%xmm3
+	xorps	%xmm0,%xmm4
+	movups	(%rcx),%xmm0
+
+L$enc_loop3:
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+	decl	%eax
+.byte	102,15,56,220,225
+	movups	16(%rcx),%xmm1
+.byte	102,15,56,220,208
+.byte	102,15,56,220,216
+	leaq	32(%rcx),%rcx
+.byte	102,15,56,220,224
+	movups	(%rcx),%xmm0
+	jnz	L$enc_loop3
+
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+.byte	102,15,56,220,225
+.byte	102,15,56,221,208
+.byte	102,15,56,221,216
+.byte	102,15,56,221,224
+	.byte	0xf3,0xc3
+
+
+.p2align	4
+_aesni_decrypt3:
+	movups	(%rcx),%xmm0
+	shrl	$1,%eax
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+	xorps	%xmm0,%xmm3
+	xorps	%xmm0,%xmm4
+	movups	(%rcx),%xmm0
+
+L$dec_loop3:
+.byte	102,15,56,222,209
+.byte	102,15,56,222,217
+	decl	%eax
+.byte	102,15,56,222,225
+	movups	16(%rcx),%xmm1
+.byte	102,15,56,222,208
+.byte	102,15,56,222,216
+	leaq	32(%rcx),%rcx
+.byte	102,15,56,222,224
+	movups	(%rcx),%xmm0
+	jnz	L$dec_loop3
+
+.byte	102,15,56,222,209
+.byte	102,15,56,222,217
+.byte	102,15,56,222,225
+.byte	102,15,56,223,208
+.byte	102,15,56,223,216
+.byte	102,15,56,223,224
+	.byte	0xf3,0xc3
+
+
+.p2align	4
+_aesni_encrypt4:
+	movups	(%rcx),%xmm0
+	shrl	$1,%eax
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+	xorps	%xmm0,%xmm3
+	xorps	%xmm0,%xmm4
+	xorps	%xmm0,%xmm5
+	movups	(%rcx),%xmm0
+
+L$enc_loop4:
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+	decl	%eax
+.byte	102,15,56,220,225
+.byte	102,15,56,220,233
+	movups	16(%rcx),%xmm1
+.byte	102,15,56,220,208
+.byte	102,15,56,220,216
+	leaq	32(%rcx),%rcx
+.byte	102,15,56,220,224
+.byte	102,15,56,220,232
+	movups	(%rcx),%xmm0
+	jnz	L$enc_loop4
+
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+.byte	102,15,56,220,225
+.byte	102,15,56,220,233
+.byte	102,15,56,221,208
+.byte	102,15,56,221,216
+.byte	102,15,56,221,224
+.byte	102,15,56,221,232
+	.byte	0xf3,0xc3
+
+
+.p2align	4
+_aesni_decrypt4:
+	movups	(%rcx),%xmm0
+	shrl	$1,%eax
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+	xorps	%xmm0,%xmm3
+	xorps	%xmm0,%xmm4
+	xorps	%xmm0,%xmm5
+	movups	(%rcx),%xmm0
+
+L$dec_loop4:
+.byte	102,15,56,222,209
+.byte	102,15,56,222,217
+	decl	%eax
+.byte	102,15,56,222,225
+.byte	102,15,56,222,233
+	movups	16(%rcx),%xmm1
+.byte	102,15,56,222,208
+.byte	102,15,56,222,216
+	leaq	32(%rcx),%rcx
+.byte	102,15,56,222,224
+.byte	102,15,56,222,232
+	movups	(%rcx),%xmm0
+	jnz	L$dec_loop4
+
+.byte	102,15,56,222,209
+.byte	102,15,56,222,217
+.byte	102,15,56,222,225
+.byte	102,15,56,222,233
+.byte	102,15,56,223,208
+.byte	102,15,56,223,216
+.byte	102,15,56,223,224
+.byte	102,15,56,223,232
+	.byte	0xf3,0xc3
+
+
+.p2align	4
+_aesni_encrypt6:
+	movups	(%rcx),%xmm0
+	shrl	$1,%eax
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+	pxor	%xmm0,%xmm3
+.byte	102,15,56,220,209
+	pxor	%xmm0,%xmm4
+.byte	102,15,56,220,217
+	pxor	%xmm0,%xmm5
+.byte	102,15,56,220,225
+	pxor	%xmm0,%xmm6
+.byte	102,15,56,220,233
+	pxor	%xmm0,%xmm7
+	decl	%eax
+.byte	102,15,56,220,241
+	movups	(%rcx),%xmm0
+.byte	102,15,56,220,249
+	jmp	L$enc_loop6_enter
+.p2align	4
+L$enc_loop6:
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+	decl	%eax
+.byte	102,15,56,220,225
+.byte	102,15,56,220,233
+.byte	102,15,56,220,241
+.byte	102,15,56,220,249
+L$enc_loop6_enter:
+	movups	16(%rcx),%xmm1
+.byte	102,15,56,220,208
+.byte	102,15,56,220,216
+	leaq	32(%rcx),%rcx
+.byte	102,15,56,220,224
+.byte	102,15,56,220,232
+.byte	102,15,56,220,240
+.byte	102,15,56,220,248
+	movups	(%rcx),%xmm0
+	jnz	L$enc_loop6
+
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+.byte	102,15,56,220,225
+.byte	102,15,56,220,233
+.byte	102,15,56,220,241
+.byte	102,15,56,220,249
+.byte	102,15,56,221,208
+.byte	102,15,56,221,216
+.byte	102,15,56,221,224
+.byte	102,15,56,221,232
+.byte	102,15,56,221,240
+.byte	102,15,56,221,248
+	.byte	0xf3,0xc3
+
+
+.p2align	4
+_aesni_decrypt6:
+	movups	(%rcx),%xmm0
+	shrl	$1,%eax
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+	pxor	%xmm0,%xmm3
+.byte	102,15,56,222,209
+	pxor	%xmm0,%xmm4
+.byte	102,15,56,222,217
+	pxor	%xmm0,%xmm5
+.byte	102,15,56,222,225
+	pxor	%xmm0,%xmm6
+.byte	102,15,56,222,233
+	pxor	%xmm0,%xmm7
+	decl	%eax
+.byte	102,15,56,222,241
+	movups	(%rcx),%xmm0
+.byte	102,15,56,222,249
+	jmp	L$dec_loop6_enter
+.p2align	4
+L$dec_loop6:
+.byte	102,15,56,222,209
+.byte	102,15,56,222,217
+	decl	%eax
+.byte	102,15,56,222,225
+.byte	102,15,56,222,233
+.byte	102,15,56,222,241
+.byte	102,15,56,222,249
+L$dec_loop6_enter:
+	movups	16(%rcx),%xmm1
+.byte	102,15,56,222,208
+.byte	102,15,56,222,216
+	leaq	32(%rcx),%rcx
+.byte	102,15,56,222,224
+.byte	102,15,56,222,232
+.byte	102,15,56,222,240
+.byte	102,15,56,222,248
+	movups	(%rcx),%xmm0
+	jnz	L$dec_loop6
+
+.byte	102,15,56,222,209
+.byte	102,15,56,222,217
+.byte	102,15,56,222,225
+.byte	102,15,56,222,233
+.byte	102,15,56,222,241
+.byte	102,15,56,222,249
+.byte	102,15,56,223,208
+.byte	102,15,56,223,216
+.byte	102,15,56,223,224
+.byte	102,15,56,223,232
+.byte	102,15,56,223,240
+.byte	102,15,56,223,248
+	.byte	0xf3,0xc3
+
+
+.p2align	4
+_aesni_encrypt8:
+	movups	(%rcx),%xmm0
+	shrl	$1,%eax
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+	xorps	%xmm0,%xmm3
+.byte	102,15,56,220,209
+	pxor	%xmm0,%xmm4
+.byte	102,15,56,220,217
+	pxor	%xmm0,%xmm5
+.byte	102,15,56,220,225
+	pxor	%xmm0,%xmm6
+.byte	102,15,56,220,233
+	pxor	%xmm0,%xmm7
+	decl	%eax
+.byte	102,15,56,220,241
+	pxor	%xmm0,%xmm8
+.byte	102,15,56,220,249
+	pxor	%xmm0,%xmm9
+	movups	(%rcx),%xmm0
+.byte	102,68,15,56,220,193
+.byte	102,68,15,56,220,201
+	movups	16(%rcx),%xmm1
+	jmp	L$enc_loop8_enter
+.p2align	4
+L$enc_loop8:
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+	decl	%eax
+.byte	102,15,56,220,225
+.byte	102,15,56,220,233
+.byte	102,15,56,220,241
+.byte	102,15,56,220,249
+.byte	102,68,15,56,220,193
+.byte	102,68,15,56,220,201
+	movups	16(%rcx),%xmm1
+L$enc_loop8_enter:
+.byte	102,15,56,220,208
+.byte	102,15,56,220,216
+	leaq	32(%rcx),%rcx
+.byte	102,15,56,220,224
+.byte	102,15,56,220,232
+.byte	102,15,56,220,240
+.byte	102,15,56,220,248
+.byte	102,68,15,56,220,192
+.byte	102,68,15,56,220,200
+	movups	(%rcx),%xmm0
+	jnz	L$enc_loop8
+
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+.byte	102,15,56,220,225
+.byte	102,15,56,220,233
+.byte	102,15,56,220,241
+.byte	102,15,56,220,249
+.byte	102,68,15,56,220,193
+.byte	102,68,15,56,220,201
+.byte	102,15,56,221,208
+.byte	102,15,56,221,216
+.byte	102,15,56,221,224
+.byte	102,15,56,221,232
+.byte	102,15,56,221,240
+.byte	102,15,56,221,248
+.byte	102,68,15,56,221,192
+.byte	102,68,15,56,221,200
+	.byte	0xf3,0xc3
+
+
+.p2align	4
+_aesni_decrypt8:
+	movups	(%rcx),%xmm0
+	shrl	$1,%eax
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+	xorps	%xmm0,%xmm3
+.byte	102,15,56,222,209
+	pxor	%xmm0,%xmm4
+.byte	102,15,56,222,217
+	pxor	%xmm0,%xmm5
+.byte	102,15,56,222,225
+	pxor	%xmm0,%xmm6
+.byte	102,15,56,222,233
+	pxor	%xmm0,%xmm7
+	decl	%eax
+.byte	102,15,56,222,241
+	pxor	%xmm0,%xmm8
+.byte	102,15,56,222,249
+	pxor	%xmm0,%xmm9
+	movups	(%rcx),%xmm0
+.byte	102,68,15,56,222,193
+.byte	102,68,15,56,222,201
+	movups	16(%rcx),%xmm1
+	jmp	L$dec_loop8_enter
+.p2align	4
+L$dec_loop8:
+.byte	102,15,56,222,209
+.byte	102,15,56,222,217
+	decl	%eax
+.byte	102,15,56,222,225
+.byte	102,15,56,222,233
+.byte	102,15,56,222,241
+.byte	102,15,56,222,249
+.byte	102,68,15,56,222,193
+.byte	102,68,15,56,222,201
+	movups	16(%rcx),%xmm1
+L$dec_loop8_enter:
+.byte	102,15,56,222,208
+.byte	102,15,56,222,216
+	leaq	32(%rcx),%rcx
+.byte	102,15,56,222,224
+.byte	102,15,56,222,232
+.byte	102,15,56,222,240
+.byte	102,15,56,222,248
+.byte	102,68,15,56,222,192
+.byte	102,68,15,56,222,200
+	movups	(%rcx),%xmm0
+	jnz	L$dec_loop8
+
+.byte	102,15,56,222,209
+.byte	102,15,56,222,217
+.byte	102,15,56,222,225
+.byte	102,15,56,222,233
+.byte	102,15,56,222,241
+.byte	102,15,56,222,249
+.byte	102,68,15,56,222,193
+.byte	102,68,15,56,222,201
+.byte	102,15,56,223,208
+.byte	102,15,56,223,216
+.byte	102,15,56,223,224
+.byte	102,15,56,223,232
+.byte	102,15,56,223,240
+.byte	102,15,56,223,248
+.byte	102,68,15,56,223,192
+.byte	102,68,15,56,223,200
+	.byte	0xf3,0xc3
+
+.globl	_aesni_ecb_encrypt
+
+.p2align	4
+_aesni_ecb_encrypt:
+	andq	$-16,%rdx
+	jz	L$ecb_ret
+
+	movl	240(%rcx),%eax
+	movups	(%rcx),%xmm0
+	movq	%rcx,%r11
+	movl	%eax,%r10d
+	testl	%r8d,%r8d
+	jz	L$ecb_decrypt
+
+	cmpq	$128,%rdx
+	jb	L$ecb_enc_tail
+
+	movdqu	(%rdi),%xmm2
+	movdqu	16(%rdi),%xmm3
+	movdqu	32(%rdi),%xmm4
+	movdqu	48(%rdi),%xmm5
+	movdqu	64(%rdi),%xmm6
+	movdqu	80(%rdi),%xmm7
+	movdqu	96(%rdi),%xmm8
+	movdqu	112(%rdi),%xmm9
+	leaq	128(%rdi),%rdi
+	subq	$128,%rdx
+	jmp	L$ecb_enc_loop8_enter
+.p2align	4
+L$ecb_enc_loop8:
+	movups	%xmm2,(%rsi)
+	movq	%r11,%rcx
+	movdqu	(%rdi),%xmm2
+	movl	%r10d,%eax
+	movups	%xmm3,16(%rsi)
+	movdqu	16(%rdi),%xmm3
+	movups	%xmm4,32(%rsi)
+	movdqu	32(%rdi),%xmm4
+	movups	%xmm5,48(%rsi)
+	movdqu	48(%rdi),%xmm5
+	movups	%xmm6,64(%rsi)
+	movdqu	64(%rdi),%xmm6
+	movups	%xmm7,80(%rsi)
+	movdqu	80(%rdi),%xmm7
+	movups	%xmm8,96(%rsi)
+	movdqu	96(%rdi),%xmm8
+	movups	%xmm9,112(%rsi)
+	leaq	128(%rsi),%rsi
+	movdqu	112(%rdi),%xmm9
+	leaq	128(%rdi),%rdi
+L$ecb_enc_loop8_enter:
+
+	call	_aesni_encrypt8
+
+	subq	$128,%rdx
+	jnc	L$ecb_enc_loop8
+
+	movups	%xmm2,(%rsi)
+	movq	%r11,%rcx
+	movups	%xmm3,16(%rsi)
+	movl	%r10d,%eax
+	movups	%xmm4,32(%rsi)
+	movups	%xmm5,48(%rsi)
+	movups	%xmm6,64(%rsi)
+	movups	%xmm7,80(%rsi)
+	movups	%xmm8,96(%rsi)
+	movups	%xmm9,112(%rsi)
+	leaq	128(%rsi),%rsi
+	addq	$128,%rdx
+	jz	L$ecb_ret
+
+L$ecb_enc_tail:
+	movups	(%rdi),%xmm2
+	cmpq	$32,%rdx
+	jb	L$ecb_enc_one
+	movups	16(%rdi),%xmm3
+	je	L$ecb_enc_two
+	movups	32(%rdi),%xmm4
+	cmpq	$64,%rdx
+	jb	L$ecb_enc_three
+	movups	48(%rdi),%xmm5
+	je	L$ecb_enc_four
+	movups	64(%rdi),%xmm6
+	cmpq	$96,%rdx
+	jb	L$ecb_enc_five
+	movups	80(%rdi),%xmm7
+	je	L$ecb_enc_six
+	movdqu	96(%rdi),%xmm8
+	call	_aesni_encrypt8
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	movups	%xmm5,48(%rsi)
+	movups	%xmm6,64(%rsi)
+	movups	%xmm7,80(%rsi)
+	movups	%xmm8,96(%rsi)
+	jmp	L$ecb_ret
+.p2align	4
+L$ecb_enc_one:
+	movups	(%rcx),%xmm0
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+L$oop_enc1_3:
+.byte	102,15,56,220,209
+	decl	%eax
+	movups	(%rcx),%xmm1
+	leaq	16(%rcx),%rcx
+	jnz	L$oop_enc1_3	
+.byte	102,15,56,221,209
+	movups	%xmm2,(%rsi)
+	jmp	L$ecb_ret
+.p2align	4
+L$ecb_enc_two:
+	xorps	%xmm4,%xmm4
+	call	_aesni_encrypt3
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	jmp	L$ecb_ret
+.p2align	4
+L$ecb_enc_three:
+	call	_aesni_encrypt3
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	jmp	L$ecb_ret
+.p2align	4
+L$ecb_enc_four:
+	call	_aesni_encrypt4
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	movups	%xmm5,48(%rsi)
+	jmp	L$ecb_ret
+.p2align	4
+L$ecb_enc_five:
+	xorps	%xmm7,%xmm7
+	call	_aesni_encrypt6
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	movups	%xmm5,48(%rsi)
+	movups	%xmm6,64(%rsi)
+	jmp	L$ecb_ret
+.p2align	4
+L$ecb_enc_six:
+	call	_aesni_encrypt6
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	movups	%xmm5,48(%rsi)
+	movups	%xmm6,64(%rsi)
+	movups	%xmm7,80(%rsi)
+	jmp	L$ecb_ret
+
+.p2align	4
+L$ecb_decrypt:
+	cmpq	$128,%rdx
+	jb	L$ecb_dec_tail
+
+	movdqu	(%rdi),%xmm2
+	movdqu	16(%rdi),%xmm3
+	movdqu	32(%rdi),%xmm4
+	movdqu	48(%rdi),%xmm5
+	movdqu	64(%rdi),%xmm6
+	movdqu	80(%rdi),%xmm7
+	movdqu	96(%rdi),%xmm8
+	movdqu	112(%rdi),%xmm9
+	leaq	128(%rdi),%rdi
+	subq	$128,%rdx
+	jmp	L$ecb_dec_loop8_enter
+.p2align	4
+L$ecb_dec_loop8:
+	movups	%xmm2,(%rsi)
+	movq	%r11,%rcx
+	movdqu	(%rdi),%xmm2
+	movl	%r10d,%eax
+	movups	%xmm3,16(%rsi)
+	movdqu	16(%rdi),%xmm3
+	movups	%xmm4,32(%rsi)
+	movdqu	32(%rdi),%xmm4
+	movups	%xmm5,48(%rsi)
+	movdqu	48(%rdi),%xmm5
+	movups	%xmm6,64(%rsi)
+	movdqu	64(%rdi),%xmm6
+	movups	%xmm7,80(%rsi)
+	movdqu	80(%rdi),%xmm7
+	movups	%xmm8,96(%rsi)
+	movdqu	96(%rdi),%xmm8
+	movups	%xmm9,112(%rsi)
+	leaq	128(%rsi),%rsi
+	movdqu	112(%rdi),%xmm9
+	leaq	128(%rdi),%rdi
+L$ecb_dec_loop8_enter:
+
+	call	_aesni_decrypt8
+
+	movups	(%r11),%xmm0
+	subq	$128,%rdx
+	jnc	L$ecb_dec_loop8
+
+	movups	%xmm2,(%rsi)
+	movq	%r11,%rcx
+	movups	%xmm3,16(%rsi)
+	movl	%r10d,%eax
+	movups	%xmm4,32(%rsi)
+	movups	%xmm5,48(%rsi)
+	movups	%xmm6,64(%rsi)
+	movups	%xmm7,80(%rsi)
+	movups	%xmm8,96(%rsi)
+	movups	%xmm9,112(%rsi)
+	leaq	128(%rsi),%rsi
+	addq	$128,%rdx
+	jz	L$ecb_ret
+
+L$ecb_dec_tail:
+	movups	(%rdi),%xmm2
+	cmpq	$32,%rdx
+	jb	L$ecb_dec_one
+	movups	16(%rdi),%xmm3
+	je	L$ecb_dec_two
+	movups	32(%rdi),%xmm4
+	cmpq	$64,%rdx
+	jb	L$ecb_dec_three
+	movups	48(%rdi),%xmm5
+	je	L$ecb_dec_four
+	movups	64(%rdi),%xmm6
+	cmpq	$96,%rdx
+	jb	L$ecb_dec_five
+	movups	80(%rdi),%xmm7
+	je	L$ecb_dec_six
+	movups	96(%rdi),%xmm8
+	movups	(%rcx),%xmm0
+	call	_aesni_decrypt8
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	movups	%xmm5,48(%rsi)
+	movups	%xmm6,64(%rsi)
+	movups	%xmm7,80(%rsi)
+	movups	%xmm8,96(%rsi)
+	jmp	L$ecb_ret
+.p2align	4
+L$ecb_dec_one:
+	movups	(%rcx),%xmm0
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+L$oop_dec1_4:
+.byte	102,15,56,222,209
+	decl	%eax
+	movups	(%rcx),%xmm1
+	leaq	16(%rcx),%rcx
+	jnz	L$oop_dec1_4	
+.byte	102,15,56,223,209
+	movups	%xmm2,(%rsi)
+	jmp	L$ecb_ret
+.p2align	4
+L$ecb_dec_two:
+	xorps	%xmm4,%xmm4
+	call	_aesni_decrypt3
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	jmp	L$ecb_ret
+.p2align	4
+L$ecb_dec_three:
+	call	_aesni_decrypt3
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	jmp	L$ecb_ret
+.p2align	4
+L$ecb_dec_four:
+	call	_aesni_decrypt4
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	movups	%xmm5,48(%rsi)
+	jmp	L$ecb_ret
+.p2align	4
+L$ecb_dec_five:
+	xorps	%xmm7,%xmm7
+	call	_aesni_decrypt6
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	movups	%xmm5,48(%rsi)
+	movups	%xmm6,64(%rsi)
+	jmp	L$ecb_ret
+.p2align	4
+L$ecb_dec_six:
+	call	_aesni_decrypt6
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	movups	%xmm5,48(%rsi)
+	movups	%xmm6,64(%rsi)
+	movups	%xmm7,80(%rsi)
+
+L$ecb_ret:
+	.byte	0xf3,0xc3
+
+.globl	_aesni_ccm64_encrypt_blocks
+
+.p2align	4
+_aesni_ccm64_encrypt_blocks:
+	movl	240(%rcx),%eax
+	movdqu	(%r8),%xmm9
+	movdqa	L$increment64(%rip),%xmm6
+	movdqa	L$bswap_mask(%rip),%xmm7
+
+	shrl	$1,%eax
+	leaq	0(%rcx),%r11
+	movdqu	(%r9),%xmm3
+	movdqa	%xmm9,%xmm2
+	movl	%eax,%r10d
+.byte	102,68,15,56,0,207
+	jmp	L$ccm64_enc_outer
+.p2align	4
+L$ccm64_enc_outer:
+	movups	(%r11),%xmm0
+	movl	%r10d,%eax
+	movups	(%rdi),%xmm8
+
+	xorps	%xmm0,%xmm2
+	movups	16(%r11),%xmm1
+	xorps	%xmm8,%xmm0
+	leaq	32(%r11),%rcx
+	xorps	%xmm0,%xmm3
+	movups	(%rcx),%xmm0
+
+L$ccm64_enc2_loop:
+.byte	102,15,56,220,209
+	decl	%eax
+.byte	102,15,56,220,217
+	movups	16(%rcx),%xmm1
+.byte	102,15,56,220,208
+	leaq	32(%rcx),%rcx
+.byte	102,15,56,220,216
+	movups	0(%rcx),%xmm0
+	jnz	L$ccm64_enc2_loop
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+	paddq	%xmm6,%xmm9
+.byte	102,15,56,221,208
+.byte	102,15,56,221,216
+
+	decq	%rdx
+	leaq	16(%rdi),%rdi
+	xorps	%xmm2,%xmm8
+	movdqa	%xmm9,%xmm2
+	movups	%xmm8,(%rsi)
+	leaq	16(%rsi),%rsi
+.byte	102,15,56,0,215
+	jnz	L$ccm64_enc_outer
+
+	movups	%xmm3,(%r9)
+	.byte	0xf3,0xc3
+
+.globl	_aesni_ccm64_decrypt_blocks
+
+.p2align	4
+_aesni_ccm64_decrypt_blocks:
+	movl	240(%rcx),%eax
+	movups	(%r8),%xmm9
+	movdqu	(%r9),%xmm3
+	movdqa	L$increment64(%rip),%xmm6
+	movdqa	L$bswap_mask(%rip),%xmm7
+
+	movaps	%xmm9,%xmm2
+	movl	%eax,%r10d
+	movq	%rcx,%r11
+.byte	102,68,15,56,0,207
+	movups	(%rcx),%xmm0
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+L$oop_enc1_5:
+.byte	102,15,56,220,209
+	decl	%eax
+	movups	(%rcx),%xmm1
+	leaq	16(%rcx),%rcx
+	jnz	L$oop_enc1_5	
+.byte	102,15,56,221,209
+	movups	(%rdi),%xmm8
+	paddq	%xmm6,%xmm9
+	leaq	16(%rdi),%rdi
+	jmp	L$ccm64_dec_outer
+.p2align	4
+L$ccm64_dec_outer:
+	xorps	%xmm2,%xmm8
+	movdqa	%xmm9,%xmm2
+	movl	%r10d,%eax
+	movups	%xmm8,(%rsi)
+	leaq	16(%rsi),%rsi
+.byte	102,15,56,0,215
+
+	subq	$1,%rdx
+	jz	L$ccm64_dec_break
+
+	movups	(%r11),%xmm0
+	shrl	$1,%eax
+	movups	16(%r11),%xmm1
+	xorps	%xmm0,%xmm8
+	leaq	32(%r11),%rcx
+	xorps	%xmm0,%xmm2
+	xorps	%xmm8,%xmm3
+	movups	(%rcx),%xmm0
+
+L$ccm64_dec2_loop:
+.byte	102,15,56,220,209
+	decl	%eax
+.byte	102,15,56,220,217
+	movups	16(%rcx),%xmm1
+.byte	102,15,56,220,208
+	leaq	32(%rcx),%rcx
+.byte	102,15,56,220,216
+	movups	0(%rcx),%xmm0
+	jnz	L$ccm64_dec2_loop
+	movups	(%rdi),%xmm8
+	paddq	%xmm6,%xmm9
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+	leaq	16(%rdi),%rdi
+.byte	102,15,56,221,208
+.byte	102,15,56,221,216
+	jmp	L$ccm64_dec_outer
+
+.p2align	4
+L$ccm64_dec_break:
+
+	movups	(%r11),%xmm0
+	movups	16(%r11),%xmm1
+	xorps	%xmm0,%xmm8
+	leaq	32(%r11),%r11
+	xorps	%xmm8,%xmm3
+L$oop_enc1_6:
+.byte	102,15,56,220,217
+	decl	%eax
+	movups	(%r11),%xmm1
+	leaq	16(%r11),%r11
+	jnz	L$oop_enc1_6	
+.byte	102,15,56,221,217
+	movups	%xmm3,(%r9)
+	.byte	0xf3,0xc3
+
+.globl	_aesni_ctr32_encrypt_blocks
+
+.p2align	4
+_aesni_ctr32_encrypt_blocks:
+	cmpq	$1,%rdx
+	je	L$ctr32_one_shortcut
+
+	movdqu	(%r8),%xmm14
+	movdqa	L$bswap_mask(%rip),%xmm15
+	xorl	%eax,%eax
+.byte	102,69,15,58,22,242,3
+.byte	102,68,15,58,34,240,3
+
+	movl	240(%rcx),%eax
+	bswapl	%r10d
+	pxor	%xmm12,%xmm12
+	pxor	%xmm13,%xmm13
+.byte	102,69,15,58,34,226,0
+	leaq	3(%r10),%r11
+.byte	102,69,15,58,34,235,0
+	incl	%r10d
+.byte	102,69,15,58,34,226,1
+	incq	%r11
+.byte	102,69,15,58,34,235,1
+	incl	%r10d
+.byte	102,69,15,58,34,226,2
+	incq	%r11
+.byte	102,69,15,58,34,235,2
+	movdqa	%xmm12,-40(%rsp)
+.byte	102,69,15,56,0,231
+	movdqa	%xmm13,-24(%rsp)
+.byte	102,69,15,56,0,239
+
+	pshufd	$192,%xmm12,%xmm2
+	pshufd	$128,%xmm12,%xmm3
+	pshufd	$64,%xmm12,%xmm4
+	cmpq	$6,%rdx
+	jb	L$ctr32_tail
+	shrl	$1,%eax
+	movq	%rcx,%r11
+	movl	%eax,%r10d
+	subq	$6,%rdx
+	jmp	L$ctr32_loop6
+
+.p2align	4
+L$ctr32_loop6:
+	pshufd	$192,%xmm13,%xmm5
+	por	%xmm14,%xmm2
+	movups	(%r11),%xmm0
+	pshufd	$128,%xmm13,%xmm6
+	por	%xmm14,%xmm3
+	movups	16(%r11),%xmm1
+	pshufd	$64,%xmm13,%xmm7
+	por	%xmm14,%xmm4
+	por	%xmm14,%xmm5
+	xorps	%xmm0,%xmm2
+	por	%xmm14,%xmm6
+	por	%xmm14,%xmm7
+
+
+
+
+	pxor	%xmm0,%xmm3
+.byte	102,15,56,220,209
+	leaq	32(%r11),%rcx
+	pxor	%xmm0,%xmm4
+.byte	102,15,56,220,217
+	movdqa	L$increment32(%rip),%xmm13
+	pxor	%xmm0,%xmm5
+.byte	102,15,56,220,225
+	movdqa	-40(%rsp),%xmm12
+	pxor	%xmm0,%xmm6
+.byte	102,15,56,220,233
+	pxor	%xmm0,%xmm7
+	movups	(%rcx),%xmm0
+	decl	%eax
+.byte	102,15,56,220,241
+.byte	102,15,56,220,249
+	jmp	L$ctr32_enc_loop6_enter
+.p2align	4
+L$ctr32_enc_loop6:
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+	decl	%eax
+.byte	102,15,56,220,225
+.byte	102,15,56,220,233
+.byte	102,15,56,220,241
+.byte	102,15,56,220,249
+L$ctr32_enc_loop6_enter:
+	movups	16(%rcx),%xmm1
+.byte	102,15,56,220,208
+.byte	102,15,56,220,216
+	leaq	32(%rcx),%rcx
+.byte	102,15,56,220,224
+.byte	102,15,56,220,232
+.byte	102,15,56,220,240
+.byte	102,15,56,220,248
+	movups	(%rcx),%xmm0
+	jnz	L$ctr32_enc_loop6
+
+.byte	102,15,56,220,209
+	paddd	%xmm13,%xmm12
+.byte	102,15,56,220,217
+	paddd	-24(%rsp),%xmm13
+.byte	102,15,56,220,225
+	movdqa	%xmm12,-40(%rsp)
+.byte	102,15,56,220,233
+	movdqa	%xmm13,-24(%rsp)
+.byte	102,15,56,220,241
+.byte	102,69,15,56,0,231
+.byte	102,15,56,220,249
+.byte	102,69,15,56,0,239
+
+.byte	102,15,56,221,208
+	movups	(%rdi),%xmm8
+.byte	102,15,56,221,216
+	movups	16(%rdi),%xmm9
+.byte	102,15,56,221,224
+	movups	32(%rdi),%xmm10
+.byte	102,15,56,221,232
+	movups	48(%rdi),%xmm11
+.byte	102,15,56,221,240
+	movups	64(%rdi),%xmm1
+.byte	102,15,56,221,248
+	movups	80(%rdi),%xmm0
+	leaq	96(%rdi),%rdi
+
+	xorps	%xmm2,%xmm8
+	pshufd	$192,%xmm12,%xmm2
+	xorps	%xmm3,%xmm9
+	pshufd	$128,%xmm12,%xmm3
+	movups	%xmm8,(%rsi)
+	xorps	%xmm4,%xmm10
+	pshufd	$64,%xmm12,%xmm4
+	movups	%xmm9,16(%rsi)
+	xorps	%xmm5,%xmm11
+	movups	%xmm10,32(%rsi)
+	xorps	%xmm6,%xmm1
+	movups	%xmm11,48(%rsi)
+	xorps	%xmm7,%xmm0
+	movups	%xmm1,64(%rsi)
+	movups	%xmm0,80(%rsi)
+	leaq	96(%rsi),%rsi
+	movl	%r10d,%eax
+	subq	$6,%rdx
+	jnc	L$ctr32_loop6
+
+	addq	$6,%rdx
+	jz	L$ctr32_done
+	movq	%r11,%rcx
+	leal	1(%rax,%rax,1),%eax
+
+L$ctr32_tail:
+	por	%xmm14,%xmm2
+	movups	(%rdi),%xmm8
+	cmpq	$2,%rdx
+	jb	L$ctr32_one
+
+	por	%xmm14,%xmm3
+	movups	16(%rdi),%xmm9
+	je	L$ctr32_two
+
+	pshufd	$192,%xmm13,%xmm5
+	por	%xmm14,%xmm4
+	movups	32(%rdi),%xmm10
+	cmpq	$4,%rdx
+	jb	L$ctr32_three
+
+	pshufd	$128,%xmm13,%xmm6
+	por	%xmm14,%xmm5
+	movups	48(%rdi),%xmm11
+	je	L$ctr32_four
+
+	por	%xmm14,%xmm6
+	xorps	%xmm7,%xmm7
+
+	call	_aesni_encrypt6
+
+	movups	64(%rdi),%xmm1
+	xorps	%xmm2,%xmm8
+	xorps	%xmm3,%xmm9
+	movups	%xmm8,(%rsi)
+	xorps	%xmm4,%xmm10
+	movups	%xmm9,16(%rsi)
+	xorps	%xmm5,%xmm11
+	movups	%xmm10,32(%rsi)
+	xorps	%xmm6,%xmm1
+	movups	%xmm11,48(%rsi)
+	movups	%xmm1,64(%rsi)
+	jmp	L$ctr32_done
+
+.p2align	4
+L$ctr32_one_shortcut:
+	movups	(%r8),%xmm2
+	movups	(%rdi),%xmm8
+	movl	240(%rcx),%eax
+L$ctr32_one:
+	movups	(%rcx),%xmm0
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+L$oop_enc1_7:
+.byte	102,15,56,220,209
+	decl	%eax
+	movups	(%rcx),%xmm1
+	leaq	16(%rcx),%rcx
+	jnz	L$oop_enc1_7	
+.byte	102,15,56,221,209
+	xorps	%xmm2,%xmm8
+	movups	%xmm8,(%rsi)
+	jmp	L$ctr32_done
+
+.p2align	4
+L$ctr32_two:
+	xorps	%xmm4,%xmm4
+	call	_aesni_encrypt3
+	xorps	%xmm2,%xmm8
+	xorps	%xmm3,%xmm9
+	movups	%xmm8,(%rsi)
+	movups	%xmm9,16(%rsi)
+	jmp	L$ctr32_done
+
+.p2align	4
+L$ctr32_three:
+	call	_aesni_encrypt3
+	xorps	%xmm2,%xmm8
+	xorps	%xmm3,%xmm9
+	movups	%xmm8,(%rsi)
+	xorps	%xmm4,%xmm10
+	movups	%xmm9,16(%rsi)
+	movups	%xmm10,32(%rsi)
+	jmp	L$ctr32_done
+
+.p2align	4
+L$ctr32_four:
+	call	_aesni_encrypt4
+	xorps	%xmm2,%xmm8
+	xorps	%xmm3,%xmm9
+	movups	%xmm8,(%rsi)
+	xorps	%xmm4,%xmm10
+	movups	%xmm9,16(%rsi)
+	xorps	%xmm5,%xmm11
+	movups	%xmm10,32(%rsi)
+	movups	%xmm11,48(%rsi)
+
+L$ctr32_done:
+	.byte	0xf3,0xc3
+
+.globl	_aesni_xts_encrypt
+
+.p2align	4
+_aesni_xts_encrypt:
+	leaq	-104(%rsp),%rsp
+	movups	(%r9),%xmm15
+	movl	240(%r8),%eax
+	movl	240(%rcx),%r10d
+	movups	(%r8),%xmm0
+	movups	16(%r8),%xmm1
+	leaq	32(%r8),%r8
+	xorps	%xmm0,%xmm15
+L$oop_enc1_8:
+.byte	102,68,15,56,220,249
+	decl	%eax
+	movups	(%r8),%xmm1
+	leaq	16(%r8),%r8
+	jnz	L$oop_enc1_8	
+.byte	102,68,15,56,221,249
+	movq	%rcx,%r11
+	movl	%r10d,%eax
+	movq	%rdx,%r9
+	andq	$-16,%rdx
+
+	movdqa	L$xts_magic(%rip),%xmm8
+	pxor	%xmm14,%xmm14
+	pcmpgtd	%xmm15,%xmm14
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	movdqa	%xmm15,%xmm10
+	paddq	%xmm15,%xmm15
+	pand	%xmm8,%xmm9
+	pcmpgtd	%xmm15,%xmm14
+	pxor	%xmm9,%xmm15
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	movdqa	%xmm15,%xmm11
+	paddq	%xmm15,%xmm15
+	pand	%xmm8,%xmm9
+	pcmpgtd	%xmm15,%xmm14
+	pxor	%xmm9,%xmm15
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	movdqa	%xmm15,%xmm12
+	paddq	%xmm15,%xmm15
+	pand	%xmm8,%xmm9
+	pcmpgtd	%xmm15,%xmm14
+	pxor	%xmm9,%xmm15
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	movdqa	%xmm15,%xmm13
+	paddq	%xmm15,%xmm15
+	pand	%xmm8,%xmm9
+	pcmpgtd	%xmm15,%xmm14
+	pxor	%xmm9,%xmm15
+	subq	$96,%rdx
+	jc	L$xts_enc_short
+
+	shrl	$1,%eax
+	subl	$1,%eax
+	movl	%eax,%r10d
+	jmp	L$xts_enc_grandloop
+
+.p2align	4
+L$xts_enc_grandloop:
+	pshufd	$19,%xmm14,%xmm9
+	movdqa	%xmm15,%xmm14
+	paddq	%xmm15,%xmm15
+	movdqu	0(%rdi),%xmm2
+	pand	%xmm8,%xmm9
+	movdqu	16(%rdi),%xmm3
+	pxor	%xmm9,%xmm15
+
+	movdqu	32(%rdi),%xmm4
+	pxor	%xmm10,%xmm2
+	movdqu	48(%rdi),%xmm5
+	pxor	%xmm11,%xmm3
+	movdqu	64(%rdi),%xmm6
+	pxor	%xmm12,%xmm4
+	movdqu	80(%rdi),%xmm7
+	leaq	96(%rdi),%rdi
+	pxor	%xmm13,%xmm5
+	movups	(%r11),%xmm0
+	pxor	%xmm14,%xmm6
+	pxor	%xmm15,%xmm7
+
+
+
+	movups	16(%r11),%xmm1
+	pxor	%xmm0,%xmm2
+	pxor	%xmm0,%xmm3
+	movdqa	%xmm10,0(%rsp)
+.byte	102,15,56,220,209
+	leaq	32(%r11),%rcx
+	pxor	%xmm0,%xmm4
+	movdqa	%xmm11,16(%rsp)
+.byte	102,15,56,220,217
+	pxor	%xmm0,%xmm5
+	movdqa	%xmm12,32(%rsp)
+.byte	102,15,56,220,225
+	pxor	%xmm0,%xmm6
+	movdqa	%xmm13,48(%rsp)
+.byte	102,15,56,220,233
+	pxor	%xmm0,%xmm7
+	movups	(%rcx),%xmm0
+	decl	%eax
+	movdqa	%xmm14,64(%rsp)
+.byte	102,15,56,220,241
+	movdqa	%xmm15,80(%rsp)
+.byte	102,15,56,220,249
+	pxor	%xmm14,%xmm14
+	pcmpgtd	%xmm15,%xmm14
+	jmp	L$xts_enc_loop6_enter
+
+.p2align	4
+L$xts_enc_loop6:
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+	decl	%eax
+.byte	102,15,56,220,225
+.byte	102,15,56,220,233
+.byte	102,15,56,220,241
+.byte	102,15,56,220,249
+L$xts_enc_loop6_enter:
+	movups	16(%rcx),%xmm1
+.byte	102,15,56,220,208
+.byte	102,15,56,220,216
+	leaq	32(%rcx),%rcx
+.byte	102,15,56,220,224
+.byte	102,15,56,220,232
+.byte	102,15,56,220,240
+.byte	102,15,56,220,248
+	movups	(%rcx),%xmm0
+	jnz	L$xts_enc_loop6
+
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	paddq	%xmm15,%xmm15
+.byte	102,15,56,220,209
+	pand	%xmm8,%xmm9
+.byte	102,15,56,220,217
+	pcmpgtd	%xmm15,%xmm14
+.byte	102,15,56,220,225
+	pxor	%xmm9,%xmm15
+.byte	102,15,56,220,233
+.byte	102,15,56,220,241
+.byte	102,15,56,220,249
+	movups	16(%rcx),%xmm1
+
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	movdqa	%xmm15,%xmm10
+	paddq	%xmm15,%xmm15
+.byte	102,15,56,220,208
+	pand	%xmm8,%xmm9
+.byte	102,15,56,220,216
+	pcmpgtd	%xmm15,%xmm14
+.byte	102,15,56,220,224
+	pxor	%xmm9,%xmm15
+.byte	102,15,56,220,232
+.byte	102,15,56,220,240
+.byte	102,15,56,220,248
+	movups	32(%rcx),%xmm0
+
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	movdqa	%xmm15,%xmm11
+	paddq	%xmm15,%xmm15
+.byte	102,15,56,220,209
+	pand	%xmm8,%xmm9
+.byte	102,15,56,220,217
+	pcmpgtd	%xmm15,%xmm14
+.byte	102,15,56,220,225
+	pxor	%xmm9,%xmm15
+.byte	102,15,56,220,233
+.byte	102,15,56,220,241
+.byte	102,15,56,220,249
+
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	movdqa	%xmm15,%xmm12
+	paddq	%xmm15,%xmm15
+.byte	102,15,56,221,208
+	pand	%xmm8,%xmm9
+.byte	102,15,56,221,216
+	pcmpgtd	%xmm15,%xmm14
+.byte	102,15,56,221,224
+	pxor	%xmm9,%xmm15
+.byte	102,15,56,221,232
+.byte	102,15,56,221,240
+.byte	102,15,56,221,248
+
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	movdqa	%xmm15,%xmm13
+	paddq	%xmm15,%xmm15
+	xorps	0(%rsp),%xmm2
+	pand	%xmm8,%xmm9
+	xorps	16(%rsp),%xmm3
+	pcmpgtd	%xmm15,%xmm14
+	pxor	%xmm9,%xmm15
+
+	xorps	32(%rsp),%xmm4
+	movups	%xmm2,0(%rsi)
+	xorps	48(%rsp),%xmm5
+	movups	%xmm3,16(%rsi)
+	xorps	64(%rsp),%xmm6
+	movups	%xmm4,32(%rsi)
+	xorps	80(%rsp),%xmm7
+	movups	%xmm5,48(%rsi)
+	movl	%r10d,%eax
+	movups	%xmm6,64(%rsi)
+	movups	%xmm7,80(%rsi)
+	leaq	96(%rsi),%rsi
+	subq	$96,%rdx
+	jnc	L$xts_enc_grandloop
+
+	leal	3(%rax,%rax,1),%eax
+	movq	%r11,%rcx
+	movl	%eax,%r10d
+
+L$xts_enc_short:
+	addq	$96,%rdx
+	jz	L$xts_enc_done
+
+	cmpq	$32,%rdx
+	jb	L$xts_enc_one
+	je	L$xts_enc_two
+
+	cmpq	$64,%rdx
+	jb	L$xts_enc_three
+	je	L$xts_enc_four
+
+	pshufd	$19,%xmm14,%xmm9
+	movdqa	%xmm15,%xmm14
+	paddq	%xmm15,%xmm15
+	movdqu	(%rdi),%xmm2
+	pand	%xmm8,%xmm9
+	movdqu	16(%rdi),%xmm3
+	pxor	%xmm9,%xmm15
+
+	movdqu	32(%rdi),%xmm4
+	pxor	%xmm10,%xmm2
+	movdqu	48(%rdi),%xmm5
+	pxor	%xmm11,%xmm3
+	movdqu	64(%rdi),%xmm6
+	leaq	80(%rdi),%rdi
+	pxor	%xmm12,%xmm4
+	pxor	%xmm13,%xmm5
+	pxor	%xmm14,%xmm6
+
+	call	_aesni_encrypt6
+
+	xorps	%xmm10,%xmm2
+	movdqa	%xmm15,%xmm10
+	xorps	%xmm11,%xmm3
+	xorps	%xmm12,%xmm4
+	movdqu	%xmm2,(%rsi)
+	xorps	%xmm13,%xmm5
+	movdqu	%xmm3,16(%rsi)
+	xorps	%xmm14,%xmm6
+	movdqu	%xmm4,32(%rsi)
+	movdqu	%xmm5,48(%rsi)
+	movdqu	%xmm6,64(%rsi)
+	leaq	80(%rsi),%rsi
+	jmp	L$xts_enc_done
+
+.p2align	4
+L$xts_enc_one:
+	movups	(%rdi),%xmm2
+	leaq	16(%rdi),%rdi
+	xorps	%xmm10,%xmm2
+	movups	(%rcx),%xmm0
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+L$oop_enc1_9:
+.byte	102,15,56,220,209
+	decl	%eax
+	movups	(%rcx),%xmm1
+	leaq	16(%rcx),%rcx
+	jnz	L$oop_enc1_9	
+.byte	102,15,56,221,209
+	xorps	%xmm10,%xmm2
+	movdqa	%xmm11,%xmm10
+	movups	%xmm2,(%rsi)
+	leaq	16(%rsi),%rsi
+	jmp	L$xts_enc_done
+
+.p2align	4
+L$xts_enc_two:
+	movups	(%rdi),%xmm2
+	movups	16(%rdi),%xmm3
+	leaq	32(%rdi),%rdi
+	xorps	%xmm10,%xmm2
+	xorps	%xmm11,%xmm3
+
+	call	_aesni_encrypt3
+
+	xorps	%xmm10,%xmm2
+	movdqa	%xmm12,%xmm10
+	xorps	%xmm11,%xmm3
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	leaq	32(%rsi),%rsi
+	jmp	L$xts_enc_done
+
+.p2align	4
+L$xts_enc_three:
+	movups	(%rdi),%xmm2
+	movups	16(%rdi),%xmm3
+	movups	32(%rdi),%xmm4
+	leaq	48(%rdi),%rdi
+	xorps	%xmm10,%xmm2
+	xorps	%xmm11,%xmm3
+	xorps	%xmm12,%xmm4
+
+	call	_aesni_encrypt3
+
+	xorps	%xmm10,%xmm2
+	movdqa	%xmm13,%xmm10
+	xorps	%xmm11,%xmm3
+	xorps	%xmm12,%xmm4
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	leaq	48(%rsi),%rsi
+	jmp	L$xts_enc_done
+
+.p2align	4
+L$xts_enc_four:
+	movups	(%rdi),%xmm2
+	movups	16(%rdi),%xmm3
+	movups	32(%rdi),%xmm4
+	xorps	%xmm10,%xmm2
+	movups	48(%rdi),%xmm5
+	leaq	64(%rdi),%rdi
+	xorps	%xmm11,%xmm3
+	xorps	%xmm12,%xmm4
+	xorps	%xmm13,%xmm5
+
+	call	_aesni_encrypt4
+
+	xorps	%xmm10,%xmm2
+	movdqa	%xmm15,%xmm10
+	xorps	%xmm11,%xmm3
+	xorps	%xmm12,%xmm4
+	movups	%xmm2,(%rsi)
+	xorps	%xmm13,%xmm5
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	movups	%xmm5,48(%rsi)
+	leaq	64(%rsi),%rsi
+	jmp	L$xts_enc_done
+
+.p2align	4
+L$xts_enc_done:
+	andq	$15,%r9
+	jz	L$xts_enc_ret
+	movq	%r9,%rdx
+
+L$xts_enc_steal:
+	movzbl	(%rdi),%eax
+	movzbl	-16(%rsi),%ecx
+	leaq	1(%rdi),%rdi
+	movb	%al,-16(%rsi)
+	movb	%cl,0(%rsi)
+	leaq	1(%rsi),%rsi
+	subq	$1,%rdx
+	jnz	L$xts_enc_steal
+
+	subq	%r9,%rsi
+	movq	%r11,%rcx
+	movl	%r10d,%eax
+
+	movups	-16(%rsi),%xmm2
+	xorps	%xmm10,%xmm2
+	movups	(%rcx),%xmm0
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+L$oop_enc1_10:
+.byte	102,15,56,220,209
+	decl	%eax
+	movups	(%rcx),%xmm1
+	leaq	16(%rcx),%rcx
+	jnz	L$oop_enc1_10	
+.byte	102,15,56,221,209
+	xorps	%xmm10,%xmm2
+	movups	%xmm2,-16(%rsi)
+
+L$xts_enc_ret:
+	leaq	104(%rsp),%rsp
+L$xts_enc_epilogue:
+	.byte	0xf3,0xc3
+
+.globl	_aesni_xts_decrypt
+
+.p2align	4
+_aesni_xts_decrypt:
+	leaq	-104(%rsp),%rsp
+	movups	(%r9),%xmm15
+	movl	240(%r8),%eax
+	movl	240(%rcx),%r10d
+	movups	(%r8),%xmm0
+	movups	16(%r8),%xmm1
+	leaq	32(%r8),%r8
+	xorps	%xmm0,%xmm15
+L$oop_enc1_11:
+.byte	102,68,15,56,220,249
+	decl	%eax
+	movups	(%r8),%xmm1
+	leaq	16(%r8),%r8
+	jnz	L$oop_enc1_11	
+.byte	102,68,15,56,221,249
+	xorl	%eax,%eax
+	testq	$15,%rdx
+	setnz	%al
+	shlq	$4,%rax
+	subq	%rax,%rdx
+
+	movq	%rcx,%r11
+	movl	%r10d,%eax
+	movq	%rdx,%r9
+	andq	$-16,%rdx
+
+	movdqa	L$xts_magic(%rip),%xmm8
+	pxor	%xmm14,%xmm14
+	pcmpgtd	%xmm15,%xmm14
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	movdqa	%xmm15,%xmm10
+	paddq	%xmm15,%xmm15
+	pand	%xmm8,%xmm9
+	pcmpgtd	%xmm15,%xmm14
+	pxor	%xmm9,%xmm15
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	movdqa	%xmm15,%xmm11
+	paddq	%xmm15,%xmm15
+	pand	%xmm8,%xmm9
+	pcmpgtd	%xmm15,%xmm14
+	pxor	%xmm9,%xmm15
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	movdqa	%xmm15,%xmm12
+	paddq	%xmm15,%xmm15
+	pand	%xmm8,%xmm9
+	pcmpgtd	%xmm15,%xmm14
+	pxor	%xmm9,%xmm15
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	movdqa	%xmm15,%xmm13
+	paddq	%xmm15,%xmm15
+	pand	%xmm8,%xmm9
+	pcmpgtd	%xmm15,%xmm14
+	pxor	%xmm9,%xmm15
+	subq	$96,%rdx
+	jc	L$xts_dec_short
+
+	shrl	$1,%eax
+	subl	$1,%eax
+	movl	%eax,%r10d
+	jmp	L$xts_dec_grandloop
+
+.p2align	4
+L$xts_dec_grandloop:
+	pshufd	$19,%xmm14,%xmm9
+	movdqa	%xmm15,%xmm14
+	paddq	%xmm15,%xmm15
+	movdqu	0(%rdi),%xmm2
+	pand	%xmm8,%xmm9
+	movdqu	16(%rdi),%xmm3
+	pxor	%xmm9,%xmm15
+
+	movdqu	32(%rdi),%xmm4
+	pxor	%xmm10,%xmm2
+	movdqu	48(%rdi),%xmm5
+	pxor	%xmm11,%xmm3
+	movdqu	64(%rdi),%xmm6
+	pxor	%xmm12,%xmm4
+	movdqu	80(%rdi),%xmm7
+	leaq	96(%rdi),%rdi
+	pxor	%xmm13,%xmm5
+	movups	(%r11),%xmm0
+	pxor	%xmm14,%xmm6
+	pxor	%xmm15,%xmm7
+
+
+
+	movups	16(%r11),%xmm1
+	pxor	%xmm0,%xmm2
+	pxor	%xmm0,%xmm3
+	movdqa	%xmm10,0(%rsp)
+.byte	102,15,56,222,209
+	leaq	32(%r11),%rcx
+	pxor	%xmm0,%xmm4
+	movdqa	%xmm11,16(%rsp)
+.byte	102,15,56,222,217
+	pxor	%xmm0,%xmm5
+	movdqa	%xmm12,32(%rsp)
+.byte	102,15,56,222,225
+	pxor	%xmm0,%xmm6
+	movdqa	%xmm13,48(%rsp)
+.byte	102,15,56,222,233
+	pxor	%xmm0,%xmm7
+	movups	(%rcx),%xmm0
+	decl	%eax
+	movdqa	%xmm14,64(%rsp)
+.byte	102,15,56,222,241
+	movdqa	%xmm15,80(%rsp)
+.byte	102,15,56,222,249
+	pxor	%xmm14,%xmm14
+	pcmpgtd	%xmm15,%xmm14
+	jmp	L$xts_dec_loop6_enter
+
+.p2align	4
+L$xts_dec_loop6:
+.byte	102,15,56,222,209
+.byte	102,15,56,222,217
+	decl	%eax
+.byte	102,15,56,222,225
+.byte	102,15,56,222,233
+.byte	102,15,56,222,241
+.byte	102,15,56,222,249
+L$xts_dec_loop6_enter:
+	movups	16(%rcx),%xmm1
+.byte	102,15,56,222,208
+.byte	102,15,56,222,216
+	leaq	32(%rcx),%rcx
+.byte	102,15,56,222,224
+.byte	102,15,56,222,232
+.byte	102,15,56,222,240
+.byte	102,15,56,222,248
+	movups	(%rcx),%xmm0
+	jnz	L$xts_dec_loop6
+
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	paddq	%xmm15,%xmm15
+.byte	102,15,56,222,209
+	pand	%xmm8,%xmm9
+.byte	102,15,56,222,217
+	pcmpgtd	%xmm15,%xmm14
+.byte	102,15,56,222,225
+	pxor	%xmm9,%xmm15
+.byte	102,15,56,222,233
+.byte	102,15,56,222,241
+.byte	102,15,56,222,249
+	movups	16(%rcx),%xmm1
+
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	movdqa	%xmm15,%xmm10
+	paddq	%xmm15,%xmm15
+.byte	102,15,56,222,208
+	pand	%xmm8,%xmm9
+.byte	102,15,56,222,216
+	pcmpgtd	%xmm15,%xmm14
+.byte	102,15,56,222,224
+	pxor	%xmm9,%xmm15
+.byte	102,15,56,222,232
+.byte	102,15,56,222,240
+.byte	102,15,56,222,248
+	movups	32(%rcx),%xmm0
+
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	movdqa	%xmm15,%xmm11
+	paddq	%xmm15,%xmm15
+.byte	102,15,56,222,209
+	pand	%xmm8,%xmm9
+.byte	102,15,56,222,217
+	pcmpgtd	%xmm15,%xmm14
+.byte	102,15,56,222,225
+	pxor	%xmm9,%xmm15
+.byte	102,15,56,222,233
+.byte	102,15,56,222,241
+.byte	102,15,56,222,249
+
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	movdqa	%xmm15,%xmm12
+	paddq	%xmm15,%xmm15
+.byte	102,15,56,223,208
+	pand	%xmm8,%xmm9
+.byte	102,15,56,223,216
+	pcmpgtd	%xmm15,%xmm14
+.byte	102,15,56,223,224
+	pxor	%xmm9,%xmm15
+.byte	102,15,56,223,232
+.byte	102,15,56,223,240
+.byte	102,15,56,223,248
+
+	pshufd	$19,%xmm14,%xmm9
+	pxor	%xmm14,%xmm14
+	movdqa	%xmm15,%xmm13
+	paddq	%xmm15,%xmm15
+	xorps	0(%rsp),%xmm2
+	pand	%xmm8,%xmm9
+	xorps	16(%rsp),%xmm3
+	pcmpgtd	%xmm15,%xmm14
+	pxor	%xmm9,%xmm15
+
+	xorps	32(%rsp),%xmm4
+	movups	%xmm2,0(%rsi)
+	xorps	48(%rsp),%xmm5
+	movups	%xmm3,16(%rsi)
+	xorps	64(%rsp),%xmm6
+	movups	%xmm4,32(%rsi)
+	xorps	80(%rsp),%xmm7
+	movups	%xmm5,48(%rsi)
+	movl	%r10d,%eax
+	movups	%xmm6,64(%rsi)
+	movups	%xmm7,80(%rsi)
+	leaq	96(%rsi),%rsi
+	subq	$96,%rdx
+	jnc	L$xts_dec_grandloop
+
+	leal	3(%rax,%rax,1),%eax
+	movq	%r11,%rcx
+	movl	%eax,%r10d
+
+L$xts_dec_short:
+	addq	$96,%rdx
+	jz	L$xts_dec_done
+
+	cmpq	$32,%rdx
+	jb	L$xts_dec_one
+	je	L$xts_dec_two
+
+	cmpq	$64,%rdx
+	jb	L$xts_dec_three
+	je	L$xts_dec_four
+
+	pshufd	$19,%xmm14,%xmm9
+	movdqa	%xmm15,%xmm14
+	paddq	%xmm15,%xmm15
+	movdqu	(%rdi),%xmm2
+	pand	%xmm8,%xmm9
+	movdqu	16(%rdi),%xmm3
+	pxor	%xmm9,%xmm15
+
+	movdqu	32(%rdi),%xmm4
+	pxor	%xmm10,%xmm2
+	movdqu	48(%rdi),%xmm5
+	pxor	%xmm11,%xmm3
+	movdqu	64(%rdi),%xmm6
+	leaq	80(%rdi),%rdi
+	pxor	%xmm12,%xmm4
+	pxor	%xmm13,%xmm5
+	pxor	%xmm14,%xmm6
+
+	call	_aesni_decrypt6
+
+	xorps	%xmm10,%xmm2
+	xorps	%xmm11,%xmm3
+	xorps	%xmm12,%xmm4
+	movdqu	%xmm2,(%rsi)
+	xorps	%xmm13,%xmm5
+	movdqu	%xmm3,16(%rsi)
+	xorps	%xmm14,%xmm6
+	movdqu	%xmm4,32(%rsi)
+	pxor	%xmm14,%xmm14
+	movdqu	%xmm5,48(%rsi)
+	pcmpgtd	%xmm15,%xmm14
+	movdqu	%xmm6,64(%rsi)
+	leaq	80(%rsi),%rsi
+	pshufd	$19,%xmm14,%xmm11
+	andq	$15,%r9
+	jz	L$xts_dec_ret
+
+	movdqa	%xmm15,%xmm10
+	paddq	%xmm15,%xmm15
+	pand	%xmm8,%xmm11
+	pxor	%xmm15,%xmm11
+	jmp	L$xts_dec_done2
+
+.p2align	4
+L$xts_dec_one:
+	movups	(%rdi),%xmm2
+	leaq	16(%rdi),%rdi
+	xorps	%xmm10,%xmm2
+	movups	(%rcx),%xmm0
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+L$oop_dec1_12:
+.byte	102,15,56,222,209
+	decl	%eax
+	movups	(%rcx),%xmm1
+	leaq	16(%rcx),%rcx
+	jnz	L$oop_dec1_12	
+.byte	102,15,56,223,209
+	xorps	%xmm10,%xmm2
+	movdqa	%xmm11,%xmm10
+	movups	%xmm2,(%rsi)
+	movdqa	%xmm12,%xmm11
+	leaq	16(%rsi),%rsi
+	jmp	L$xts_dec_done
+
+.p2align	4
+L$xts_dec_two:
+	movups	(%rdi),%xmm2
+	movups	16(%rdi),%xmm3
+	leaq	32(%rdi),%rdi
+	xorps	%xmm10,%xmm2
+	xorps	%xmm11,%xmm3
+
+	call	_aesni_decrypt3
+
+	xorps	%xmm10,%xmm2
+	movdqa	%xmm12,%xmm10
+	xorps	%xmm11,%xmm3
+	movdqa	%xmm13,%xmm11
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	leaq	32(%rsi),%rsi
+	jmp	L$xts_dec_done
+
+.p2align	4
+L$xts_dec_three:
+	movups	(%rdi),%xmm2
+	movups	16(%rdi),%xmm3
+	movups	32(%rdi),%xmm4
+	leaq	48(%rdi),%rdi
+	xorps	%xmm10,%xmm2
+	xorps	%xmm11,%xmm3
+	xorps	%xmm12,%xmm4
+
+	call	_aesni_decrypt3
+
+	xorps	%xmm10,%xmm2
+	movdqa	%xmm13,%xmm10
+	xorps	%xmm11,%xmm3
+	movdqa	%xmm15,%xmm11
+	xorps	%xmm12,%xmm4
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	leaq	48(%rsi),%rsi
+	jmp	L$xts_dec_done
+
+.p2align	4
+L$xts_dec_four:
+	pshufd	$19,%xmm14,%xmm9
+	movdqa	%xmm15,%xmm14
+	paddq	%xmm15,%xmm15
+	movups	(%rdi),%xmm2
+	pand	%xmm8,%xmm9
+	movups	16(%rdi),%xmm3
+	pxor	%xmm9,%xmm15
+
+	movups	32(%rdi),%xmm4
+	xorps	%xmm10,%xmm2
+	movups	48(%rdi),%xmm5
+	leaq	64(%rdi),%rdi
+	xorps	%xmm11,%xmm3
+	xorps	%xmm12,%xmm4
+	xorps	%xmm13,%xmm5
+
+	call	_aesni_decrypt4
+
+	xorps	%xmm10,%xmm2
+	movdqa	%xmm14,%xmm10
+	xorps	%xmm11,%xmm3
+	movdqa	%xmm15,%xmm11
+	xorps	%xmm12,%xmm4
+	movups	%xmm2,(%rsi)
+	xorps	%xmm13,%xmm5
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	movups	%xmm5,48(%rsi)
+	leaq	64(%rsi),%rsi
+	jmp	L$xts_dec_done
+
+.p2align	4
+L$xts_dec_done:
+	andq	$15,%r9
+	jz	L$xts_dec_ret
+L$xts_dec_done2:
+	movq	%r9,%rdx
+	movq	%r11,%rcx
+	movl	%r10d,%eax
+
+	movups	(%rdi),%xmm2
+	xorps	%xmm11,%xmm2
+	movups	(%rcx),%xmm0
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+L$oop_dec1_13:
+.byte	102,15,56,222,209
+	decl	%eax
+	movups	(%rcx),%xmm1
+	leaq	16(%rcx),%rcx
+	jnz	L$oop_dec1_13	
+.byte	102,15,56,223,209
+	xorps	%xmm11,%xmm2
+	movups	%xmm2,(%rsi)
+
+L$xts_dec_steal:
+	movzbl	16(%rdi),%eax
+	movzbl	(%rsi),%ecx
+	leaq	1(%rdi),%rdi
+	movb	%al,(%rsi)
+	movb	%cl,16(%rsi)
+	leaq	1(%rsi),%rsi
+	subq	$1,%rdx
+	jnz	L$xts_dec_steal
+
+	subq	%r9,%rsi
+	movq	%r11,%rcx
+	movl	%r10d,%eax
+
+	movups	(%rsi),%xmm2
+	xorps	%xmm10,%xmm2
+	movups	(%rcx),%xmm0
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+L$oop_dec1_14:
+.byte	102,15,56,222,209
+	decl	%eax
+	movups	(%rcx),%xmm1
+	leaq	16(%rcx),%rcx
+	jnz	L$oop_dec1_14	
+.byte	102,15,56,223,209
+	xorps	%xmm10,%xmm2
+	movups	%xmm2,(%rsi)
+
+L$xts_dec_ret:
+	leaq	104(%rsp),%rsp
+L$xts_dec_epilogue:
+	.byte	0xf3,0xc3
+
+.globl	_aesni_cbc_encrypt
+
+.p2align	4
+_aesni_cbc_encrypt:
+	testq	%rdx,%rdx
+	jz	L$cbc_ret
+
+	movl	240(%rcx),%r10d
+	movq	%rcx,%r11
+	testl	%r9d,%r9d
+	jz	L$cbc_decrypt
+
+	movups	(%r8),%xmm2
+	movl	%r10d,%eax
+	cmpq	$16,%rdx
+	jb	L$cbc_enc_tail
+	subq	$16,%rdx
+	jmp	L$cbc_enc_loop
+.p2align	4
+L$cbc_enc_loop:
+	movups	(%rdi),%xmm3
+	leaq	16(%rdi),%rdi
+
+	movups	(%rcx),%xmm0
+	movups	16(%rcx),%xmm1
+	xorps	%xmm0,%xmm3
+	leaq	32(%rcx),%rcx
+	xorps	%xmm3,%xmm2
+L$oop_enc1_15:
+.byte	102,15,56,220,209
+	decl	%eax
+	movups	(%rcx),%xmm1
+	leaq	16(%rcx),%rcx
+	jnz	L$oop_enc1_15	
+.byte	102,15,56,221,209
+	movl	%r10d,%eax
+	movq	%r11,%rcx
+	movups	%xmm2,0(%rsi)
+	leaq	16(%rsi),%rsi
+	subq	$16,%rdx
+	jnc	L$cbc_enc_loop
+	addq	$16,%rdx
+	jnz	L$cbc_enc_tail
+	movups	%xmm2,(%r8)
+	jmp	L$cbc_ret
+
+L$cbc_enc_tail:
+	movq	%rdx,%rcx
+	xchgq	%rdi,%rsi
+.long	0x9066A4F3	
+	movl	$16,%ecx
+	subq	%rdx,%rcx
+	xorl	%eax,%eax
+.long	0x9066AAF3	
+	leaq	-16(%rdi),%rdi
+	movl	%r10d,%eax
+	movq	%rdi,%rsi
+	movq	%r11,%rcx
+	xorq	%rdx,%rdx
+	jmp	L$cbc_enc_loop	
+
+.p2align	4
+L$cbc_decrypt:
+	movups	(%r8),%xmm9
+	movl	%r10d,%eax
+	cmpq	$112,%rdx
+	jbe	L$cbc_dec_tail
+	shrl	$1,%r10d
+	subq	$112,%rdx
+	movl	%r10d,%eax
+	movaps	%xmm9,-24(%rsp)
+	jmp	L$cbc_dec_loop8_enter
+.p2align	4
+L$cbc_dec_loop8:
+	movaps	%xmm0,-24(%rsp)
+	movups	%xmm9,(%rsi)
+	leaq	16(%rsi),%rsi
+L$cbc_dec_loop8_enter:
+	movups	(%rcx),%xmm0
+	movups	(%rdi),%xmm2
+	movups	16(%rdi),%xmm3
+	movups	16(%rcx),%xmm1
+
+	leaq	32(%rcx),%rcx
+	movdqu	32(%rdi),%xmm4
+	xorps	%xmm0,%xmm2
+	movdqu	48(%rdi),%xmm5
+	xorps	%xmm0,%xmm3
+	movdqu	64(%rdi),%xmm6
+.byte	102,15,56,222,209
+	pxor	%xmm0,%xmm4
+	movdqu	80(%rdi),%xmm7
+.byte	102,15,56,222,217
+	pxor	%xmm0,%xmm5
+	movdqu	96(%rdi),%xmm8
+.byte	102,15,56,222,225
+	pxor	%xmm0,%xmm6
+	movdqu	112(%rdi),%xmm9
+.byte	102,15,56,222,233
+	pxor	%xmm0,%xmm7
+	decl	%eax
+.byte	102,15,56,222,241
+	pxor	%xmm0,%xmm8
+.byte	102,15,56,222,249
+	pxor	%xmm0,%xmm9
+	movups	(%rcx),%xmm0
+.byte	102,68,15,56,222,193
+.byte	102,68,15,56,222,201
+	movups	16(%rcx),%xmm1
+
+	call	L$dec_loop8_enter
+
+	movups	(%rdi),%xmm1
+	movups	16(%rdi),%xmm0
+	xorps	-24(%rsp),%xmm2
+	xorps	%xmm1,%xmm3
+	movups	32(%rdi),%xmm1
+	xorps	%xmm0,%xmm4
+	movups	48(%rdi),%xmm0
+	xorps	%xmm1,%xmm5
+	movups	64(%rdi),%xmm1
+	xorps	%xmm0,%xmm6
+	movups	80(%rdi),%xmm0
+	xorps	%xmm1,%xmm7
+	movups	96(%rdi),%xmm1
+	xorps	%xmm0,%xmm8
+	movups	112(%rdi),%xmm0
+	xorps	%xmm1,%xmm9
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	movups	%xmm5,48(%rsi)
+	movl	%r10d,%eax
+	movups	%xmm6,64(%rsi)
+	movq	%r11,%rcx
+	movups	%xmm7,80(%rsi)
+	leaq	128(%rdi),%rdi
+	movups	%xmm8,96(%rsi)
+	leaq	112(%rsi),%rsi
+	subq	$128,%rdx
+	ja	L$cbc_dec_loop8
+
+	movaps	%xmm9,%xmm2
+	movaps	%xmm0,%xmm9
+	addq	$112,%rdx
+	jle	L$cbc_dec_tail_collected
+	movups	%xmm2,(%rsi)
+	leal	1(%r10,%r10,1),%eax
+	leaq	16(%rsi),%rsi
+L$cbc_dec_tail:
+	movups	(%rdi),%xmm2
+	movaps	%xmm2,%xmm8
+	cmpq	$16,%rdx
+	jbe	L$cbc_dec_one
+
+	movups	16(%rdi),%xmm3
+	movaps	%xmm3,%xmm7
+	cmpq	$32,%rdx
+	jbe	L$cbc_dec_two
+
+	movups	32(%rdi),%xmm4
+	movaps	%xmm4,%xmm6
+	cmpq	$48,%rdx
+	jbe	L$cbc_dec_three
+
+	movups	48(%rdi),%xmm5
+	cmpq	$64,%rdx
+	jbe	L$cbc_dec_four
+
+	movups	64(%rdi),%xmm6
+	cmpq	$80,%rdx
+	jbe	L$cbc_dec_five
+
+	movups	80(%rdi),%xmm7
+	cmpq	$96,%rdx
+	jbe	L$cbc_dec_six
+
+	movups	96(%rdi),%xmm8
+	movaps	%xmm9,-24(%rsp)
+	call	_aesni_decrypt8
+	movups	(%rdi),%xmm1
+	movups	16(%rdi),%xmm0
+	xorps	-24(%rsp),%xmm2
+	xorps	%xmm1,%xmm3
+	movups	32(%rdi),%xmm1
+	xorps	%xmm0,%xmm4
+	movups	48(%rdi),%xmm0
+	xorps	%xmm1,%xmm5
+	movups	64(%rdi),%xmm1
+	xorps	%xmm0,%xmm6
+	movups	80(%rdi),%xmm0
+	xorps	%xmm1,%xmm7
+	movups	96(%rdi),%xmm9
+	xorps	%xmm0,%xmm8
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	movups	%xmm5,48(%rsi)
+	movups	%xmm6,64(%rsi)
+	movups	%xmm7,80(%rsi)
+	leaq	96(%rsi),%rsi
+	movaps	%xmm8,%xmm2
+	subq	$112,%rdx
+	jmp	L$cbc_dec_tail_collected
+.p2align	4
+L$cbc_dec_one:
+	movups	(%rcx),%xmm0
+	movups	16(%rcx),%xmm1
+	leaq	32(%rcx),%rcx
+	xorps	%xmm0,%xmm2
+L$oop_dec1_16:
+.byte	102,15,56,222,209
+	decl	%eax
+	movups	(%rcx),%xmm1
+	leaq	16(%rcx),%rcx
+	jnz	L$oop_dec1_16	
+.byte	102,15,56,223,209
+	xorps	%xmm9,%xmm2
+	movaps	%xmm8,%xmm9
+	subq	$16,%rdx
+	jmp	L$cbc_dec_tail_collected
+.p2align	4
+L$cbc_dec_two:
+	xorps	%xmm4,%xmm4
+	call	_aesni_decrypt3
+	xorps	%xmm9,%xmm2
+	xorps	%xmm8,%xmm3
+	movups	%xmm2,(%rsi)
+	movaps	%xmm7,%xmm9
+	movaps	%xmm3,%xmm2
+	leaq	16(%rsi),%rsi
+	subq	$32,%rdx
+	jmp	L$cbc_dec_tail_collected
+.p2align	4
+L$cbc_dec_three:
+	call	_aesni_decrypt3
+	xorps	%xmm9,%xmm2
+	xorps	%xmm8,%xmm3
+	movups	%xmm2,(%rsi)
+	xorps	%xmm7,%xmm4
+	movups	%xmm3,16(%rsi)
+	movaps	%xmm6,%xmm9
+	movaps	%xmm4,%xmm2
+	leaq	32(%rsi),%rsi
+	subq	$48,%rdx
+	jmp	L$cbc_dec_tail_collected
+.p2align	4
+L$cbc_dec_four:
+	call	_aesni_decrypt4
+	xorps	%xmm9,%xmm2
+	movups	48(%rdi),%xmm9
+	xorps	%xmm8,%xmm3
+	movups	%xmm2,(%rsi)
+	xorps	%xmm7,%xmm4
+	movups	%xmm3,16(%rsi)
+	xorps	%xmm6,%xmm5
+	movups	%xmm4,32(%rsi)
+	movaps	%xmm5,%xmm2
+	leaq	48(%rsi),%rsi
+	subq	$64,%rdx
+	jmp	L$cbc_dec_tail_collected
+.p2align	4
+L$cbc_dec_five:
+	xorps	%xmm7,%xmm7
+	call	_aesni_decrypt6
+	movups	16(%rdi),%xmm1
+	movups	32(%rdi),%xmm0
+	xorps	%xmm9,%xmm2
+	xorps	%xmm8,%xmm3
+	xorps	%xmm1,%xmm4
+	movups	48(%rdi),%xmm1
+	xorps	%xmm0,%xmm5
+	movups	64(%rdi),%xmm9
+	xorps	%xmm1,%xmm6
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	movups	%xmm5,48(%rsi)
+	leaq	64(%rsi),%rsi
+	movaps	%xmm6,%xmm2
+	subq	$80,%rdx
+	jmp	L$cbc_dec_tail_collected
+.p2align	4
+L$cbc_dec_six:
+	call	_aesni_decrypt6
+	movups	16(%rdi),%xmm1
+	movups	32(%rdi),%xmm0
+	xorps	%xmm9,%xmm2
+	xorps	%xmm8,%xmm3
+	xorps	%xmm1,%xmm4
+	movups	48(%rdi),%xmm1
+	xorps	%xmm0,%xmm5
+	movups	64(%rdi),%xmm0
+	xorps	%xmm1,%xmm6
+	movups	80(%rdi),%xmm9
+	xorps	%xmm0,%xmm7
+	movups	%xmm2,(%rsi)
+	movups	%xmm3,16(%rsi)
+	movups	%xmm4,32(%rsi)
+	movups	%xmm5,48(%rsi)
+	movups	%xmm6,64(%rsi)
+	leaq	80(%rsi),%rsi
+	movaps	%xmm7,%xmm2
+	subq	$96,%rdx
+	jmp	L$cbc_dec_tail_collected
+.p2align	4
+L$cbc_dec_tail_collected:
+	andq	$15,%rdx
+	movups	%xmm9,(%r8)
+	jnz	L$cbc_dec_tail_partial
+	movups	%xmm2,(%rsi)
+	jmp	L$cbc_dec_ret
+.p2align	4
+L$cbc_dec_tail_partial:
+	movaps	%xmm2,-24(%rsp)
+	movq	$16,%rcx
+	movq	%rsi,%rdi
+	subq	%rdx,%rcx
+	leaq	-24(%rsp),%rsi
+.long	0x9066A4F3	
+
+L$cbc_dec_ret:
+L$cbc_ret:
+	.byte	0xf3,0xc3
+
+.globl	_aesni_set_decrypt_key
+
+.p2align	4
+_aesni_set_decrypt_key:
+.byte	0x48,0x83,0xEC,0x08	
+	call	__aesni_set_encrypt_key
+	shll	$4,%esi
+	testl	%eax,%eax
+	jnz	L$dec_key_ret
+	leaq	16(%rdx,%rsi,1),%rdi
+
+	movups	(%rdx),%xmm0
+	movups	(%rdi),%xmm1
+	movups	%xmm0,(%rdi)
+	movups	%xmm1,(%rdx)
+	leaq	16(%rdx),%rdx
+	leaq	-16(%rdi),%rdi
+
+L$dec_key_inverse:
+	movups	(%rdx),%xmm0
+	movups	(%rdi),%xmm1
+.byte	102,15,56,219,192
+.byte	102,15,56,219,201
+	leaq	16(%rdx),%rdx
+	leaq	-16(%rdi),%rdi
+	movups	%xmm0,16(%rdi)
+	movups	%xmm1,-16(%rdx)
+	cmpq	%rdx,%rdi
+	ja	L$dec_key_inverse
+
+	movups	(%rdx),%xmm0
+.byte	102,15,56,219,192
+	movups	%xmm0,(%rdi)
+L$dec_key_ret:
+	addq	$8,%rsp
+	.byte	0xf3,0xc3
+L$SEH_end_set_decrypt_key:
+
+.globl	_aesni_set_encrypt_key
+
+.p2align	4
+_aesni_set_encrypt_key:
+__aesni_set_encrypt_key:
+.byte	0x48,0x83,0xEC,0x08	
+	movq	$-1,%rax
+	testq	%rdi,%rdi
+	jz	L$enc_key_ret
+	testq	%rdx,%rdx
+	jz	L$enc_key_ret
+
+	movups	(%rdi),%xmm0
+	xorps	%xmm4,%xmm4
+	leaq	16(%rdx),%rax
+	cmpl	$256,%esi
+	je	L$14rounds
+	cmpl	$192,%esi
+	je	L$12rounds
+	cmpl	$128,%esi
+	jne	L$bad_keybits
+
+L$10rounds:
+	movl	$9,%esi
+	movups	%xmm0,(%rdx)
+.byte	102,15,58,223,200,1
+	call	L$key_expansion_128_cold
+.byte	102,15,58,223,200,2
+	call	L$key_expansion_128
+.byte	102,15,58,223,200,4
+	call	L$key_expansion_128
+.byte	102,15,58,223,200,8
+	call	L$key_expansion_128
+.byte	102,15,58,223,200,16
+	call	L$key_expansion_128
+.byte	102,15,58,223,200,32
+	call	L$key_expansion_128
+.byte	102,15,58,223,200,64
+	call	L$key_expansion_128
+.byte	102,15,58,223,200,128
+	call	L$key_expansion_128
+.byte	102,15,58,223,200,27
+	call	L$key_expansion_128
+.byte	102,15,58,223,200,54
+	call	L$key_expansion_128
+	movups	%xmm0,(%rax)
+	movl	%esi,80(%rax)
+	xorl	%eax,%eax
+	jmp	L$enc_key_ret
+
+.p2align	4
+L$12rounds:
+	movq	16(%rdi),%xmm2
+	movl	$11,%esi
+	movups	%xmm0,(%rdx)
+.byte	102,15,58,223,202,1
+	call	L$key_expansion_192a_cold
+.byte	102,15,58,223,202,2
+	call	L$key_expansion_192b
+.byte	102,15,58,223,202,4
+	call	L$key_expansion_192a
+.byte	102,15,58,223,202,8
+	call	L$key_expansion_192b
+.byte	102,15,58,223,202,16
+	call	L$key_expansion_192a
+.byte	102,15,58,223,202,32
+	call	L$key_expansion_192b
+.byte	102,15,58,223,202,64
+	call	L$key_expansion_192a
+.byte	102,15,58,223,202,128
+	call	L$key_expansion_192b
+	movups	%xmm0,(%rax)
+	movl	%esi,48(%rax)
+	xorq	%rax,%rax
+	jmp	L$enc_key_ret
+
+.p2align	4
+L$14rounds:
+	movups	16(%rdi),%xmm2
+	movl	$13,%esi
+	leaq	16(%rax),%rax
+	movups	%xmm0,(%rdx)
+	movups	%xmm2,16(%rdx)
+.byte	102,15,58,223,202,1
+	call	L$key_expansion_256a_cold
+.byte	102,15,58,223,200,1
+	call	L$key_expansion_256b
+.byte	102,15,58,223,202,2
+	call	L$key_expansion_256a
+.byte	102,15,58,223,200,2
+	call	L$key_expansion_256b
+.byte	102,15,58,223,202,4
+	call	L$key_expansion_256a
+.byte	102,15,58,223,200,4
+	call	L$key_expansion_256b
+.byte	102,15,58,223,202,8
+	call	L$key_expansion_256a
+.byte	102,15,58,223,200,8
+	call	L$key_expansion_256b
+.byte	102,15,58,223,202,16
+	call	L$key_expansion_256a
+.byte	102,15,58,223,200,16
+	call	L$key_expansion_256b
+.byte	102,15,58,223,202,32
+	call	L$key_expansion_256a
+.byte	102,15,58,223,200,32
+	call	L$key_expansion_256b
+.byte	102,15,58,223,202,64
+	call	L$key_expansion_256a
+	movups	%xmm0,(%rax)
+	movl	%esi,16(%rax)
+	xorq	%rax,%rax
+	jmp	L$enc_key_ret
+
+.p2align	4
+L$bad_keybits:
+	movq	$-2,%rax
+L$enc_key_ret:
+	addq	$8,%rsp
+	.byte	0xf3,0xc3
+L$SEH_end_set_encrypt_key:
+
+.p2align	4
+L$key_expansion_128:
+	movups	%xmm0,(%rax)
+	leaq	16(%rax),%rax
+L$key_expansion_128_cold:
+	shufps	$16,%xmm0,%xmm4
+	xorps	%xmm4,%xmm0
+	shufps	$140,%xmm0,%xmm4
+	xorps	%xmm4,%xmm0
+	shufps	$255,%xmm1,%xmm1
+	xorps	%xmm1,%xmm0
+	.byte	0xf3,0xc3
+
+.p2align	4
+L$key_expansion_192a:
+	movups	%xmm0,(%rax)
+	leaq	16(%rax),%rax
+L$key_expansion_192a_cold:
+	movaps	%xmm2,%xmm5
+L$key_expansion_192b_warm:
+	shufps	$16,%xmm0,%xmm4
+	movdqa	%xmm2,%xmm3
+	xorps	%xmm4,%xmm0
+	shufps	$140,%xmm0,%xmm4
+	pslldq	$4,%xmm3
+	xorps	%xmm4,%xmm0
+	pshufd	$85,%xmm1,%xmm1
+	pxor	%xmm3,%xmm2
+	pxor	%xmm1,%xmm0
+	pshufd	$255,%xmm0,%xmm3
+	pxor	%xmm3,%xmm2
+	.byte	0xf3,0xc3
+
+.p2align	4
+L$key_expansion_192b:
+	movaps	%xmm0,%xmm3
+	shufps	$68,%xmm0,%xmm5
+	movups	%xmm5,(%rax)
+	shufps	$78,%xmm2,%xmm3
+	movups	%xmm3,16(%rax)
+	leaq	32(%rax),%rax
+	jmp	L$key_expansion_192b_warm
+
+.p2align	4
+L$key_expansion_256a:
+	movups	%xmm2,(%rax)
+	leaq	16(%rax),%rax
+L$key_expansion_256a_cold:
+	shufps	$16,%xmm0,%xmm4
+	xorps	%xmm4,%xmm0
+	shufps	$140,%xmm0,%xmm4
+	xorps	%xmm4,%xmm0
+	shufps	$255,%xmm1,%xmm1
+	xorps	%xmm1,%xmm0
+	.byte	0xf3,0xc3
+
+.p2align	4
+L$key_expansion_256b:
+	movups	%xmm0,(%rax)
+	leaq	16(%rax),%rax
+
+	shufps	$16,%xmm2,%xmm4
+	xorps	%xmm4,%xmm2
+	shufps	$140,%xmm2,%xmm4
+	xorps	%xmm4,%xmm2
+	shufps	$170,%xmm1,%xmm1
+	xorps	%xmm1,%xmm2
+	.byte	0xf3,0xc3
+
+
+.p2align	6
+L$bswap_mask:
+.byte	15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0
+L$increment32:
+.long	6,6,6,0
+L$increment64:
+.long	1,0,0,0
+L$xts_magic:
+.long	0x87,0,1,0
+
+.byte	65,69,83,32,102,111,114,32,73,110,116,101,108,32,65,69,83,45,78,73,44,32,67,82,89,80,84,79,71,65,77,83,32,98,121,32,60,97,112,112,114,111,64,111,112,101,110,115,115,108,46,111,114,103,62,0
+.p2align	6
diff --git a/lib/accelerated/x86/asm-macosx/appro-aes-x86-macosx.s b/lib/accelerated/x86/asm-macosx/appro-aes-x86-macosx.s
new file mode 100644
index 0000000000..ca05d9e496
--- /dev/null
+++ b/lib/accelerated/x86/asm-macosx/appro-aes-x86-macosx.s
@@ -0,0 +1,2144 @@
+# Copyright (c) 2011, Andy Polyakov by <appro@openssl.org>
+# All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions
+# are met:
+# 
+#     * Redistributions of source code must retain copyright notices,
+#      this list of conditions and the following disclaimer.
+#
+#     * 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.
+#
+#     * Neither the name of the Andy Polyakov nor the names of its
+#      copyright holder and contributors may be used to endorse or
+#      promote products derived from this software without specific
+#      prior written permission.
+#
+# ALTERNATIVELY, provided that this notice is retained in full, this
+# product may be distributed under the terms of the GNU General Public
+# License (GPL), in which case the provisions of the GPL apply INSTEAD OF
+# those given above.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDER AND CONTRIBUTORS
+# "AS IS" AND ANY EXPRESS 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 COPYRIGHT
+# OWNER OR 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.
+
+.file	"devel/perlasm/aesni-x86.s"
+.text
+.globl	_aesni_encrypt
+.align	4
+_aesni_encrypt:
+L_aesni_encrypt_begin:
+	movl	4(%esp),%eax
+	movl	12(%esp),%edx
+	movups	(%eax),%xmm2
+	movl	240(%edx),%ecx
+	movl	8(%esp),%eax
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+L000enc1_loop_1:
+.byte	102,15,56,220,209
+	decl	%ecx
+	movups	(%edx),%xmm1
+	leal	16(%edx),%edx
+	jnz	L000enc1_loop_1
+.byte	102,15,56,221,209
+	movups	%xmm2,(%eax)
+	ret
+.globl	_aesni_decrypt
+.align	4
+_aesni_decrypt:
+L_aesni_decrypt_begin:
+	movl	4(%esp),%eax
+	movl	12(%esp),%edx
+	movups	(%eax),%xmm2
+	movl	240(%edx),%ecx
+	movl	8(%esp),%eax
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+L001dec1_loop_2:
+.byte	102,15,56,222,209
+	decl	%ecx
+	movups	(%edx),%xmm1
+	leal	16(%edx),%edx
+	jnz	L001dec1_loop_2
+.byte	102,15,56,223,209
+	movups	%xmm2,(%eax)
+	ret
+.align	4
+__aesni_encrypt3:
+	movups	(%edx),%xmm0
+	shrl	$1,%ecx
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+	pxor	%xmm0,%xmm3
+	pxor	%xmm0,%xmm4
+	movups	(%edx),%xmm0
+L002enc3_loop:
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+	decl	%ecx
+.byte	102,15,56,220,225
+	movups	16(%edx),%xmm1
+.byte	102,15,56,220,208
+.byte	102,15,56,220,216
+	leal	32(%edx),%edx
+.byte	102,15,56,220,224
+	movups	(%edx),%xmm0
+	jnz	L002enc3_loop
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+.byte	102,15,56,220,225
+.byte	102,15,56,221,208
+.byte	102,15,56,221,216
+.byte	102,15,56,221,224
+	ret
+.align	4
+__aesni_decrypt3:
+	movups	(%edx),%xmm0
+	shrl	$1,%ecx
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+	pxor	%xmm0,%xmm3
+	pxor	%xmm0,%xmm4
+	movups	(%edx),%xmm0
+L003dec3_loop:
+.byte	102,15,56,222,209
+.byte	102,15,56,222,217
+	decl	%ecx
+.byte	102,15,56,222,225
+	movups	16(%edx),%xmm1
+.byte	102,15,56,222,208
+.byte	102,15,56,222,216
+	leal	32(%edx),%edx
+.byte	102,15,56,222,224
+	movups	(%edx),%xmm0
+	jnz	L003dec3_loop
+.byte	102,15,56,222,209
+.byte	102,15,56,222,217
+.byte	102,15,56,222,225
+.byte	102,15,56,223,208
+.byte	102,15,56,223,216
+.byte	102,15,56,223,224
+	ret
+.align	4
+__aesni_encrypt4:
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	shrl	$1,%ecx
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+	pxor	%xmm0,%xmm3
+	pxor	%xmm0,%xmm4
+	pxor	%xmm0,%xmm5
+	movups	(%edx),%xmm0
+L004enc4_loop:
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+	decl	%ecx
+.byte	102,15,56,220,225
+.byte	102,15,56,220,233
+	movups	16(%edx),%xmm1
+.byte	102,15,56,220,208
+.byte	102,15,56,220,216
+	leal	32(%edx),%edx
+.byte	102,15,56,220,224
+.byte	102,15,56,220,232
+	movups	(%edx),%xmm0
+	jnz	L004enc4_loop
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+.byte	102,15,56,220,225
+.byte	102,15,56,220,233
+.byte	102,15,56,221,208
+.byte	102,15,56,221,216
+.byte	102,15,56,221,224
+.byte	102,15,56,221,232
+	ret
+.align	4
+__aesni_decrypt4:
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	shrl	$1,%ecx
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+	pxor	%xmm0,%xmm3
+	pxor	%xmm0,%xmm4
+	pxor	%xmm0,%xmm5
+	movups	(%edx),%xmm0
+L005dec4_loop:
+.byte	102,15,56,222,209
+.byte	102,15,56,222,217
+	decl	%ecx
+.byte	102,15,56,222,225
+.byte	102,15,56,222,233
+	movups	16(%edx),%xmm1
+.byte	102,15,56,222,208
+.byte	102,15,56,222,216
+	leal	32(%edx),%edx
+.byte	102,15,56,222,224
+.byte	102,15,56,222,232
+	movups	(%edx),%xmm0
+	jnz	L005dec4_loop
+.byte	102,15,56,222,209
+.byte	102,15,56,222,217
+.byte	102,15,56,222,225
+.byte	102,15,56,222,233
+.byte	102,15,56,223,208
+.byte	102,15,56,223,216
+.byte	102,15,56,223,224
+.byte	102,15,56,223,232
+	ret
+.align	4
+__aesni_encrypt6:
+	movups	(%edx),%xmm0
+	shrl	$1,%ecx
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+	pxor	%xmm0,%xmm3
+.byte	102,15,56,220,209
+	pxor	%xmm0,%xmm4
+.byte	102,15,56,220,217
+	pxor	%xmm0,%xmm5
+	decl	%ecx
+.byte	102,15,56,220,225
+	pxor	%xmm0,%xmm6
+.byte	102,15,56,220,233
+	pxor	%xmm0,%xmm7
+.byte	102,15,56,220,241
+	movups	(%edx),%xmm0
+.byte	102,15,56,220,249
+	jmp	L_aesni_encrypt6_enter
+.align	4,0x90
+L006enc6_loop:
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+	decl	%ecx
+.byte	102,15,56,220,225
+.byte	102,15,56,220,233
+.byte	102,15,56,220,241
+.byte	102,15,56,220,249
+.align	4,0x90
+L_aesni_encrypt6_enter:
+	movups	16(%edx),%xmm1
+.byte	102,15,56,220,208
+.byte	102,15,56,220,216
+	leal	32(%edx),%edx
+.byte	102,15,56,220,224
+.byte	102,15,56,220,232
+.byte	102,15,56,220,240
+.byte	102,15,56,220,248
+	movups	(%edx),%xmm0
+	jnz	L006enc6_loop
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+.byte	102,15,56,220,225
+.byte	102,15,56,220,233
+.byte	102,15,56,220,241
+.byte	102,15,56,220,249
+.byte	102,15,56,221,208
+.byte	102,15,56,221,216
+.byte	102,15,56,221,224
+.byte	102,15,56,221,232
+.byte	102,15,56,221,240
+.byte	102,15,56,221,248
+	ret
+.align	4
+__aesni_decrypt6:
+	movups	(%edx),%xmm0
+	shrl	$1,%ecx
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+	pxor	%xmm0,%xmm3
+.byte	102,15,56,222,209
+	pxor	%xmm0,%xmm4
+.byte	102,15,56,222,217
+	pxor	%xmm0,%xmm5
+	decl	%ecx
+.byte	102,15,56,222,225
+	pxor	%xmm0,%xmm6
+.byte	102,15,56,222,233
+	pxor	%xmm0,%xmm7
+.byte	102,15,56,222,241
+	movups	(%edx),%xmm0
+.byte	102,15,56,222,249
+	jmp	L_aesni_decrypt6_enter
+.align	4,0x90
+L007dec6_loop:
+.byte	102,15,56,222,209
+.byte	102,15,56,222,217
+	decl	%ecx
+.byte	102,15,56,222,225
+.byte	102,15,56,222,233
+.byte	102,15,56,222,241
+.byte	102,15,56,222,249
+.align	4,0x90
+L_aesni_decrypt6_enter:
+	movups	16(%edx),%xmm1
+.byte	102,15,56,222,208
+.byte	102,15,56,222,216
+	leal	32(%edx),%edx
+.byte	102,15,56,222,224
+.byte	102,15,56,222,232
+.byte	102,15,56,222,240
+.byte	102,15,56,222,248
+	movups	(%edx),%xmm0
+	jnz	L007dec6_loop
+.byte	102,15,56,222,209
+.byte	102,15,56,222,217
+.byte	102,15,56,222,225
+.byte	102,15,56,222,233
+.byte	102,15,56,222,241
+.byte	102,15,56,222,249
+.byte	102,15,56,223,208
+.byte	102,15,56,223,216
+.byte	102,15,56,223,224
+.byte	102,15,56,223,232
+.byte	102,15,56,223,240
+.byte	102,15,56,223,248
+	ret
+.globl	_aesni_ecb_encrypt
+.align	4
+_aesni_ecb_encrypt:
+L_aesni_ecb_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
+	movl	36(%esp),%ebx
+	andl	$-16,%eax
+	jz	L008ecb_ret
+	movl	240(%edx),%ecx
+	testl	%ebx,%ebx
+	jz	L009ecb_decrypt
+	movl	%edx,%ebp
+	movl	%ecx,%ebx
+	cmpl	$96,%eax
+	jb	L010ecb_enc_tail
+	movdqu	(%esi),%xmm2
+	movdqu	16(%esi),%xmm3
+	movdqu	32(%esi),%xmm4
+	movdqu	48(%esi),%xmm5
+	movdqu	64(%esi),%xmm6
+	movdqu	80(%esi),%xmm7
+	leal	96(%esi),%esi
+	subl	$96,%eax
+	jmp	L011ecb_enc_loop6_enter
+.align	4,0x90
+L012ecb_enc_loop6:
+	movups	%xmm2,(%edi)
+	movdqu	(%esi),%xmm2
+	movups	%xmm3,16(%edi)
+	movdqu	16(%esi),%xmm3
+	movups	%xmm4,32(%edi)
+	movdqu	32(%esi),%xmm4
+	movups	%xmm5,48(%edi)
+	movdqu	48(%esi),%xmm5
+	movups	%xmm6,64(%edi)
+	movdqu	64(%esi),%xmm6
+	movups	%xmm7,80(%edi)
+	leal	96(%edi),%edi
+	movdqu	80(%esi),%xmm7
+	leal	96(%esi),%esi
+L011ecb_enc_loop6_enter:
+	call	__aesni_encrypt6
+	movl	%ebp,%edx
+	movl	%ebx,%ecx
+	subl	$96,%eax
+	jnc	L012ecb_enc_loop6
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	movups	%xmm4,32(%edi)
+	movups	%xmm5,48(%edi)
+	movups	%xmm6,64(%edi)
+	movups	%xmm7,80(%edi)
+	leal	96(%edi),%edi
+	addl	$96,%eax
+	jz	L008ecb_ret
+L010ecb_enc_tail:
+	movups	(%esi),%xmm2
+	cmpl	$32,%eax
+	jb	L013ecb_enc_one
+	movups	16(%esi),%xmm3
+	je	L014ecb_enc_two
+	movups	32(%esi),%xmm4
+	cmpl	$64,%eax
+	jb	L015ecb_enc_three
+	movups	48(%esi),%xmm5
+	je	L016ecb_enc_four
+	movups	64(%esi),%xmm6
+	xorps	%xmm7,%xmm7
+	call	__aesni_encrypt6
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	movups	%xmm4,32(%edi)
+	movups	%xmm5,48(%edi)
+	movups	%xmm6,64(%edi)
+	jmp	L008ecb_ret
+.align	4,0x90
+L013ecb_enc_one:
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+L017enc1_loop_3:
+.byte	102,15,56,220,209
+	decl	%ecx
+	movups	(%edx),%xmm1
+	leal	16(%edx),%edx
+	jnz	L017enc1_loop_3
+.byte	102,15,56,221,209
+	movups	%xmm2,(%edi)
+	jmp	L008ecb_ret
+.align	4,0x90
+L014ecb_enc_two:
+	xorps	%xmm4,%xmm4
+	call	__aesni_encrypt3
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	jmp	L008ecb_ret
+.align	4,0x90
+L015ecb_enc_three:
+	call	__aesni_encrypt3
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	movups	%xmm4,32(%edi)
+	jmp	L008ecb_ret
+.align	4,0x90
+L016ecb_enc_four:
+	call	__aesni_encrypt4
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	movups	%xmm4,32(%edi)
+	movups	%xmm5,48(%edi)
+	jmp	L008ecb_ret
+.align	4,0x90
+L009ecb_decrypt:
+	movl	%edx,%ebp
+	movl	%ecx,%ebx
+	cmpl	$96,%eax
+	jb	L018ecb_dec_tail
+	movdqu	(%esi),%xmm2
+	movdqu	16(%esi),%xmm3
+	movdqu	32(%esi),%xmm4
+	movdqu	48(%esi),%xmm5
+	movdqu	64(%esi),%xmm6
+	movdqu	80(%esi),%xmm7
+	leal	96(%esi),%esi
+	subl	$96,%eax
+	jmp	L019ecb_dec_loop6_enter
+.align	4,0x90
+L020ecb_dec_loop6:
+	movups	%xmm2,(%edi)
+	movdqu	(%esi),%xmm2
+	movups	%xmm3,16(%edi)
+	movdqu	16(%esi),%xmm3
+	movups	%xmm4,32(%edi)
+	movdqu	32(%esi),%xmm4
+	movups	%xmm5,48(%edi)
+	movdqu	48(%esi),%xmm5
+	movups	%xmm6,64(%edi)
+	movdqu	64(%esi),%xmm6
+	movups	%xmm7,80(%edi)
+	leal	96(%edi),%edi
+	movdqu	80(%esi),%xmm7
+	leal	96(%esi),%esi
+L019ecb_dec_loop6_enter:
+	call	__aesni_decrypt6
+	movl	%ebp,%edx
+	movl	%ebx,%ecx
+	subl	$96,%eax
+	jnc	L020ecb_dec_loop6
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	movups	%xmm4,32(%edi)
+	movups	%xmm5,48(%edi)
+	movups	%xmm6,64(%edi)
+	movups	%xmm7,80(%edi)
+	leal	96(%edi),%edi
+	addl	$96,%eax
+	jz	L008ecb_ret
+L018ecb_dec_tail:
+	movups	(%esi),%xmm2
+	cmpl	$32,%eax
+	jb	L021ecb_dec_one
+	movups	16(%esi),%xmm3
+	je	L022ecb_dec_two
+	movups	32(%esi),%xmm4
+	cmpl	$64,%eax
+	jb	L023ecb_dec_three
+	movups	48(%esi),%xmm5
+	je	L024ecb_dec_four
+	movups	64(%esi),%xmm6
+	xorps	%xmm7,%xmm7
+	call	__aesni_decrypt6
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	movups	%xmm4,32(%edi)
+	movups	%xmm5,48(%edi)
+	movups	%xmm6,64(%edi)
+	jmp	L008ecb_ret
+.align	4,0x90
+L021ecb_dec_one:
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+L025dec1_loop_4:
+.byte	102,15,56,222,209
+	decl	%ecx
+	movups	(%edx),%xmm1
+	leal	16(%edx),%edx
+	jnz	L025dec1_loop_4
+.byte	102,15,56,223,209
+	movups	%xmm2,(%edi)
+	jmp	L008ecb_ret
+.align	4,0x90
+L022ecb_dec_two:
+	xorps	%xmm4,%xmm4
+	call	__aesni_decrypt3
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	jmp	L008ecb_ret
+.align	4,0x90
+L023ecb_dec_three:
+	call	__aesni_decrypt3
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	movups	%xmm4,32(%edi)
+	jmp	L008ecb_ret
+.align	4,0x90
+L024ecb_dec_four:
+	call	__aesni_decrypt4
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	movups	%xmm4,32(%edi)
+	movups	%xmm5,48(%edi)
+L008ecb_ret:
+	popl	%edi
+	popl	%esi
+	popl	%ebx
+	popl	%ebp
+	ret
+.globl	_aesni_ccm64_encrypt_blocks
+.align	4
+_aesni_ccm64_encrypt_blocks:
+L_aesni_ccm64_encrypt_blocks_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
+	movl	36(%esp),%ebx
+	movl	40(%esp),%ecx
+	movl	%esp,%ebp
+	subl	$60,%esp
+	andl	$-16,%esp
+	movl	%ebp,48(%esp)
+	movdqu	(%ebx),%xmm7
+	movdqu	(%ecx),%xmm3
+	movl	240(%edx),%ecx
+	movl	$202182159,(%esp)
+	movl	$134810123,4(%esp)
+	movl	$67438087,8(%esp)
+	movl	$66051,12(%esp)
+	movl	$1,%ebx
+	xorl	%ebp,%ebp
+	movl	%ebx,16(%esp)
+	movl	%ebp,20(%esp)
+	movl	%ebp,24(%esp)
+	movl	%ebp,28(%esp)
+	shrl	$1,%ecx
+	leal	(%edx),%ebp
+	movdqa	(%esp),%xmm5
+	movdqa	%xmm7,%xmm2
+	movl	%ecx,%ebx
+.byte	102,15,56,0,253
+L026ccm64_enc_outer:
+	movups	(%ebp),%xmm0
+	movl	%ebx,%ecx
+	movups	(%esi),%xmm6
+	xorps	%xmm0,%xmm2
+	movups	16(%ebp),%xmm1
+	xorps	%xmm6,%xmm0
+	leal	32(%ebp),%edx
+	xorps	%xmm0,%xmm3
+	movups	(%edx),%xmm0
+L027ccm64_enc2_loop:
+.byte	102,15,56,220,209
+	decl	%ecx
+.byte	102,15,56,220,217
+	movups	16(%edx),%xmm1
+.byte	102,15,56,220,208
+	leal	32(%edx),%edx
+.byte	102,15,56,220,216
+	movups	(%edx),%xmm0
+	jnz	L027ccm64_enc2_loop
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+	paddq	16(%esp),%xmm7
+.byte	102,15,56,221,208
+.byte	102,15,56,221,216
+	decl	%eax
+	leal	16(%esi),%esi
+	xorps	%xmm2,%xmm6
+	movdqa	%xmm7,%xmm2
+	movups	%xmm6,(%edi)
+	leal	16(%edi),%edi
+.byte	102,15,56,0,213
+	jnz	L026ccm64_enc_outer
+	movl	48(%esp),%esp
+	movl	40(%esp),%edi
+	movups	%xmm3,(%edi)
+	popl	%edi
+	popl	%esi
+	popl	%ebx
+	popl	%ebp
+	ret
+.globl	_aesni_ccm64_decrypt_blocks
+.align	4
+_aesni_ccm64_decrypt_blocks:
+L_aesni_ccm64_decrypt_blocks_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
+	movl	36(%esp),%ebx
+	movl	40(%esp),%ecx
+	movl	%esp,%ebp
+	subl	$60,%esp
+	andl	$-16,%esp
+	movl	%ebp,48(%esp)
+	movdqu	(%ebx),%xmm7
+	movdqu	(%ecx),%xmm3
+	movl	240(%edx),%ecx
+	movl	$202182159,(%esp)
+	movl	$134810123,4(%esp)
+	movl	$67438087,8(%esp)
+	movl	$66051,12(%esp)
+	movl	$1,%ebx
+	xorl	%ebp,%ebp
+	movl	%ebx,16(%esp)
+	movl	%ebp,20(%esp)
+	movl	%ebp,24(%esp)
+	movl	%ebp,28(%esp)
+	movdqa	(%esp),%xmm5
+	movdqa	%xmm7,%xmm2
+	movl	%edx,%ebp
+	movl	%ecx,%ebx
+.byte	102,15,56,0,253
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+L028enc1_loop_5:
+.byte	102,15,56,220,209
+	decl	%ecx
+	movups	(%edx),%xmm1
+	leal	16(%edx),%edx
+	jnz	L028enc1_loop_5
+.byte	102,15,56,221,209
+	movups	(%esi),%xmm6
+	paddq	16(%esp),%xmm7
+	leal	16(%esi),%esi
+	jmp	L029ccm64_dec_outer
+.align	4,0x90
+L029ccm64_dec_outer:
+	xorps	%xmm2,%xmm6
+	movdqa	%xmm7,%xmm2
+	movl	%ebx,%ecx
+	movups	%xmm6,(%edi)
+	leal	16(%edi),%edi
+.byte	102,15,56,0,213
+	subl	$1,%eax
+	jz	L030ccm64_dec_break
+	movups	(%ebp),%xmm0
+	shrl	$1,%ecx
+	movups	16(%ebp),%xmm1
+	xorps	%xmm0,%xmm6
+	leal	32(%ebp),%edx
+	xorps	%xmm0,%xmm2
+	xorps	%xmm6,%xmm3
+	movups	(%edx),%xmm0
+L031ccm64_dec2_loop:
+.byte	102,15,56,220,209
+	decl	%ecx
+.byte	102,15,56,220,217
+	movups	16(%edx),%xmm1
+.byte	102,15,56,220,208
+	leal	32(%edx),%edx
+.byte	102,15,56,220,216
+	movups	(%edx),%xmm0
+	jnz	L031ccm64_dec2_loop
+	movups	(%esi),%xmm6
+	paddq	16(%esp),%xmm7
+.byte	102,15,56,220,209
+.byte	102,15,56,220,217
+	leal	16(%esi),%esi
+.byte	102,15,56,221,208
+.byte	102,15,56,221,216
+	jmp	L029ccm64_dec_outer
+.align	4,0x90
+L030ccm64_dec_break:
+	movl	%ebp,%edx
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	xorps	%xmm0,%xmm6
+	leal	32(%edx),%edx
+	xorps	%xmm6,%xmm3
+L032enc1_loop_6:
+.byte	102,15,56,220,217
+	decl	%ecx
+	movups	(%edx),%xmm1
+	leal	16(%edx),%edx
+	jnz	L032enc1_loop_6
+.byte	102,15,56,221,217
+	movl	48(%esp),%esp
+	movl	40(%esp),%edi
+	movups	%xmm3,(%edi)
+	popl	%edi
+	popl	%esi
+	popl	%ebx
+	popl	%ebp
+	ret
+.globl	_aesni_ctr32_encrypt_blocks
+.align	4
+_aesni_ctr32_encrypt_blocks:
+L_aesni_ctr32_encrypt_blocks_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
+	movl	36(%esp),%ebx
+	movl	%esp,%ebp
+	subl	$88,%esp
+	andl	$-16,%esp
+	movl	%ebp,80(%esp)
+	cmpl	$1,%eax
+	je	L033ctr32_one_shortcut
+	movdqu	(%ebx),%xmm7
+	movl	$202182159,(%esp)
+	movl	$134810123,4(%esp)
+	movl	$67438087,8(%esp)
+	movl	$66051,12(%esp)
+	movl	$6,%ecx
+	xorl	%ebp,%ebp
+	movl	%ecx,16(%esp)
+	movl	%ecx,20(%esp)
+	movl	%ecx,24(%esp)
+	movl	%ebp,28(%esp)
+.byte	102,15,58,22,251,3
+.byte	102,15,58,34,253,3
+	movl	240(%edx),%ecx
+	bswap	%ebx
+	pxor	%xmm1,%xmm1
+	pxor	%xmm0,%xmm0
+	movdqa	(%esp),%xmm2
+.byte	102,15,58,34,203,0
+	leal	3(%ebx),%ebp
+.byte	102,15,58,34,197,0
+	incl	%ebx
+.byte	102,15,58,34,203,1
+	incl	%ebp
+.byte	102,15,58,34,197,1
+	incl	%ebx
+.byte	102,15,58,34,203,2
+	incl	%ebp
+.byte	102,15,58,34,197,2
+	movdqa	%xmm1,48(%esp)
+.byte	102,15,56,0,202
+	movdqa	%xmm0,64(%esp)
+.byte	102,15,56,0,194
+	pshufd	$192,%xmm1,%xmm2
+	pshufd	$128,%xmm1,%xmm3
+	cmpl	$6,%eax
+	jb	L034ctr32_tail
+	movdqa	%xmm7,32(%esp)
+	shrl	$1,%ecx
+	movl	%edx,%ebp
+	movl	%ecx,%ebx
+	subl	$6,%eax
+	jmp	L035ctr32_loop6
+.align	4,0x90
+L035ctr32_loop6:
+	pshufd	$64,%xmm1,%xmm4
+	movdqa	32(%esp),%xmm1
+	pshufd	$192,%xmm0,%xmm5
+	por	%xmm1,%xmm2
+	pshufd	$128,%xmm0,%xmm6
+	por	%xmm1,%xmm3
+	pshufd	$64,%xmm0,%xmm7
+	por	%xmm1,%xmm4
+	por	%xmm1,%xmm5
+	por	%xmm1,%xmm6
+	por	%xmm1,%xmm7
+	movups	(%ebp),%xmm0
+	movups	16(%ebp),%xmm1
+	leal	32(%ebp),%edx
+	decl	%ecx
+	pxor	%xmm0,%xmm2
+	pxor	%xmm0,%xmm3
+.byte	102,15,56,220,209
+	pxor	%xmm0,%xmm4
+.byte	102,15,56,220,217
+	pxor	%xmm0,%xmm5
+.byte	102,15,56,220,225
+	pxor	%xmm0,%xmm6
+.byte	102,15,56,220,233
+	pxor	%xmm0,%xmm7
+.byte	102,15,56,220,241
+	movups	(%edx),%xmm0
+.byte	102,15,56,220,249
+	call	L_aesni_encrypt6_enter
+	movups	(%esi),%xmm1
+	movups	16(%esi),%xmm0
+	xorps	%xmm1,%xmm2
+	movups	32(%esi),%xmm1
+	xorps	%xmm0,%xmm3
+	movups	%xmm2,(%edi)
+	movdqa	16(%esp),%xmm0
+	xorps	%xmm1,%xmm4
+	movdqa	48(%esp),%xmm1
+	movups	%xmm3,16(%edi)
+	movups	%xmm4,32(%edi)
+	paddd	%xmm0,%xmm1
+	paddd	64(%esp),%xmm0
+	movdqa	(%esp),%xmm2
+	movups	48(%esi),%xmm3
+	movups	64(%esi),%xmm4
+	xorps	%xmm3,%xmm5
+	movups	80(%esi),%xmm3
+	leal	96(%esi),%esi
+	movdqa	%xmm1,48(%esp)
+.byte	102,15,56,0,202
+	xorps	%xmm4,%xmm6
+	movups	%xmm5,48(%edi)
+	xorps	%xmm3,%xmm7
+	movdqa	%xmm0,64(%esp)
+.byte	102,15,56,0,194
+	movups	%xmm6,64(%edi)
+	pshufd	$192,%xmm1,%xmm2
+	movups	%xmm7,80(%edi)
+	leal	96(%edi),%edi
+	movl	%ebx,%ecx
+	pshufd	$128,%xmm1,%xmm3
+	subl	$6,%eax
+	jnc	L035ctr32_loop6
+	addl	$6,%eax
+	jz	L036ctr32_ret
+	movl	%ebp,%edx
+	leal	1(,%ecx,2),%ecx
+	movdqa	32(%esp),%xmm7
+L034ctr32_tail:
+	por	%xmm7,%xmm2
+	cmpl	$2,%eax
+	jb	L037ctr32_one
+	pshufd	$64,%xmm1,%xmm4
+	por	%xmm7,%xmm3
+	je	L038ctr32_two
+	pshufd	$192,%xmm0,%xmm5
+	por	%xmm7,%xmm4
+	cmpl	$4,%eax
+	jb	L039ctr32_three
+	pshufd	$128,%xmm0,%xmm6
+	por	%xmm7,%xmm5
+	je	L040ctr32_four
+	por	%xmm7,%xmm6
+	call	__aesni_encrypt6
+	movups	(%esi),%xmm1
+	movups	16(%esi),%xmm0
+	xorps	%xmm1,%xmm2
+	movups	32(%esi),%xmm1
+	xorps	%xmm0,%xmm3
+	movups	48(%esi),%xmm0
+	xorps	%xmm1,%xmm4
+	movups	64(%esi),%xmm1
+	xorps	%xmm0,%xmm5
+	movups	%xmm2,(%edi)
+	xorps	%xmm1,%xmm6
+	movups	%xmm3,16(%edi)
+	movups	%xmm4,32(%edi)
+	movups	%xmm5,48(%edi)
+	movups	%xmm6,64(%edi)
+	jmp	L036ctr32_ret
+.align	4,0x90
+L033ctr32_one_shortcut:
+	movups	(%ebx),%xmm2
+	movl	240(%edx),%ecx
+L037ctr32_one:
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+L041enc1_loop_7:
+.byte	102,15,56,220,209
+	decl	%ecx
+	movups	(%edx),%xmm1
+	leal	16(%edx),%edx
+	jnz	L041enc1_loop_7
+.byte	102,15,56,221,209
+	movups	(%esi),%xmm6
+	xorps	%xmm2,%xmm6
+	movups	%xmm6,(%edi)
+	jmp	L036ctr32_ret
+.align	4,0x90
+L038ctr32_two:
+	call	__aesni_encrypt3
+	movups	(%esi),%xmm5
+	movups	16(%esi),%xmm6
+	xorps	%xmm5,%xmm2
+	xorps	%xmm6,%xmm3
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	jmp	L036ctr32_ret
+.align	4,0x90
+L039ctr32_three:
+	call	__aesni_encrypt3
+	movups	(%esi),%xmm5
+	movups	16(%esi),%xmm6
+	xorps	%xmm5,%xmm2
+	movups	32(%esi),%xmm7
+	xorps	%xmm6,%xmm3
+	movups	%xmm2,(%edi)
+	xorps	%xmm7,%xmm4
+	movups	%xmm3,16(%edi)
+	movups	%xmm4,32(%edi)
+	jmp	L036ctr32_ret
+.align	4,0x90
+L040ctr32_four:
+	call	__aesni_encrypt4
+	movups	(%esi),%xmm6
+	movups	16(%esi),%xmm7
+	movups	32(%esi),%xmm1
+	xorps	%xmm6,%xmm2
+	movups	48(%esi),%xmm0
+	xorps	%xmm7,%xmm3
+	movups	%xmm2,(%edi)
+	xorps	%xmm1,%xmm4
+	movups	%xmm3,16(%edi)
+	xorps	%xmm0,%xmm5
+	movups	%xmm4,32(%edi)
+	movups	%xmm5,48(%edi)
+L036ctr32_ret:
+	movl	80(%esp),%esp
+	popl	%edi
+	popl	%esi
+	popl	%ebx
+	popl	%ebp
+	ret
+.globl	_aesni_xts_encrypt
+.align	4
+_aesni_xts_encrypt:
+L_aesni_xts_encrypt_begin:
+	pushl	%ebp
+	pushl	%ebx
+	pushl	%esi
+	pushl	%edi
+	movl	36(%esp),%edx
+	movl	40(%esp),%esi
+	movl	240(%edx),%ecx
+	movups	(%esi),%xmm2
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+L042enc1_loop_8:
+.byte	102,15,56,220,209
+	decl	%ecx
+	movups	(%edx),%xmm1
+	leal	16(%edx),%edx
+	jnz	L042enc1_loop_8
+.byte	102,15,56,221,209
+	movl	20(%esp),%esi
+	movl	24(%esp),%edi
+	movl	28(%esp),%eax
+	movl	32(%esp),%edx
+	movl	%esp,%ebp
+	subl	$120,%esp
+	movl	240(%edx),%ecx
+	andl	$-16,%esp
+	movl	$135,96(%esp)
+	movl	$0,100(%esp)
+	movl	$1,104(%esp)
+	movl	$0,108(%esp)
+	movl	%eax,112(%esp)
+	movl	%ebp,116(%esp)
+	movdqa	%xmm2,%xmm1
+	pxor	%xmm0,%xmm0
+	movdqa	96(%esp),%xmm3
+	pcmpgtd	%xmm1,%xmm0
+	andl	$-16,%eax
+	movl	%edx,%ebp
+	movl	%ecx,%ebx
+	subl	$96,%eax
+	jc	L043xts_enc_short
+	shrl	$1,%ecx
+	movl	%ecx,%ebx
+	jmp	L044xts_enc_loop6
+.align	4,0x90
+L044xts_enc_loop6:
+	pshufd	$19,%xmm0,%xmm2
+	pxor	%xmm0,%xmm0
+	movdqa	%xmm1,(%esp)
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	pxor	%xmm2,%xmm1
+	pshufd	$19,%xmm0,%xmm2
+	pxor	%xmm0,%xmm0
+	movdqa	%xmm1,16(%esp)
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	pxor	%xmm2,%xmm1
+	pshufd	$19,%xmm0,%xmm2
+	pxor	%xmm0,%xmm0
+	movdqa	%xmm1,32(%esp)
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	pxor	%xmm2,%xmm1
+	pshufd	$19,%xmm0,%xmm2
+	pxor	%xmm0,%xmm0
+	movdqa	%xmm1,48(%esp)
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	pxor	%xmm2,%xmm1
+	pshufd	$19,%xmm0,%xmm7
+	movdqa	%xmm1,64(%esp)
+	paddq	%xmm1,%xmm1
+	movups	(%ebp),%xmm0
+	pand	%xmm3,%xmm7
+	movups	(%esi),%xmm2
+	pxor	%xmm1,%xmm7
+	movdqu	16(%esi),%xmm3
+	xorps	%xmm0,%xmm2
+	movdqu	32(%esi),%xmm4
+	pxor	%xmm0,%xmm3
+	movdqu	48(%esi),%xmm5
+	pxor	%xmm0,%xmm4
+	movdqu	64(%esi),%xmm6
+	pxor	%xmm0,%xmm5
+	movdqu	80(%esi),%xmm1
+	pxor	%xmm0,%xmm6
+	leal	96(%esi),%esi
+	pxor	(%esp),%xmm2
+	movdqa	%xmm7,80(%esp)
+	pxor	%xmm1,%xmm7
+	movups	16(%ebp),%xmm1
+	leal	32(%ebp),%edx
+	pxor	16(%esp),%xmm3
+.byte	102,15,56,220,209
+	pxor	32(%esp),%xmm4
+.byte	102,15,56,220,217
+	pxor	48(%esp),%xmm5
+	decl	%ecx
+.byte	102,15,56,220,225
+	pxor	64(%esp),%xmm6
+.byte	102,15,56,220,233
+	pxor	%xmm0,%xmm7
+.byte	102,15,56,220,241
+	movups	(%edx),%xmm0
+.byte	102,15,56,220,249
+	call	L_aesni_encrypt6_enter
+	movdqa	80(%esp),%xmm1
+	pxor	%xmm0,%xmm0
+	xorps	(%esp),%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	xorps	16(%esp),%xmm3
+	movups	%xmm2,(%edi)
+	xorps	32(%esp),%xmm4
+	movups	%xmm3,16(%edi)
+	xorps	48(%esp),%xmm5
+	movups	%xmm4,32(%edi)
+	xorps	64(%esp),%xmm6
+	movups	%xmm5,48(%edi)
+	xorps	%xmm1,%xmm7
+	movups	%xmm6,64(%edi)
+	pshufd	$19,%xmm0,%xmm2
+	movups	%xmm7,80(%edi)
+	leal	96(%edi),%edi
+	movdqa	96(%esp),%xmm3
+	pxor	%xmm0,%xmm0
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	movl	%ebx,%ecx
+	pxor	%xmm2,%xmm1
+	subl	$96,%eax
+	jnc	L044xts_enc_loop6
+	leal	1(,%ecx,2),%ecx
+	movl	%ebp,%edx
+	movl	%ecx,%ebx
+L043xts_enc_short:
+	addl	$96,%eax
+	jz	L045xts_enc_done6x
+	movdqa	%xmm1,%xmm5
+	cmpl	$32,%eax
+	jb	L046xts_enc_one
+	pshufd	$19,%xmm0,%xmm2
+	pxor	%xmm0,%xmm0
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	pxor	%xmm2,%xmm1
+	je	L047xts_enc_two
+	pshufd	$19,%xmm0,%xmm2
+	pxor	%xmm0,%xmm0
+	movdqa	%xmm1,%xmm6
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	pxor	%xmm2,%xmm1
+	cmpl	$64,%eax
+	jb	L048xts_enc_three
+	pshufd	$19,%xmm0,%xmm2
+	pxor	%xmm0,%xmm0
+	movdqa	%xmm1,%xmm7
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	pxor	%xmm2,%xmm1
+	movdqa	%xmm5,(%esp)
+	movdqa	%xmm6,16(%esp)
+	je	L049xts_enc_four
+	movdqa	%xmm7,32(%esp)
+	pshufd	$19,%xmm0,%xmm7
+	movdqa	%xmm1,48(%esp)
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm7
+	pxor	%xmm1,%xmm7
+	movdqu	(%esi),%xmm2
+	movdqu	16(%esi),%xmm3
+	movdqu	32(%esi),%xmm4
+	pxor	(%esp),%xmm2
+	movdqu	48(%esi),%xmm5
+	pxor	16(%esp),%xmm3
+	movdqu	64(%esi),%xmm6
+	pxor	32(%esp),%xmm4
+	leal	80(%esi),%esi
+	pxor	48(%esp),%xmm5
+	movdqa	%xmm7,64(%esp)
+	pxor	%xmm7,%xmm6
+	call	__aesni_encrypt6
+	movaps	64(%esp),%xmm1
+	xorps	(%esp),%xmm2
+	xorps	16(%esp),%xmm3
+	xorps	32(%esp),%xmm4
+	movups	%xmm2,(%edi)
+	xorps	48(%esp),%xmm5
+	movups	%xmm3,16(%edi)
+	xorps	%xmm1,%xmm6
+	movups	%xmm4,32(%edi)
+	movups	%xmm5,48(%edi)
+	movups	%xmm6,64(%edi)
+	leal	80(%edi),%edi
+	jmp	L050xts_enc_done
+.align	4,0x90
+L046xts_enc_one:
+	movups	(%esi),%xmm2
+	leal	16(%esi),%esi
+	xorps	%xmm5,%xmm2
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+L051enc1_loop_9:
+.byte	102,15,56,220,209
+	decl	%ecx
+	movups	(%edx),%xmm1
+	leal	16(%edx),%edx
+	jnz	L051enc1_loop_9
+.byte	102,15,56,221,209
+	xorps	%xmm5,%xmm2
+	movups	%xmm2,(%edi)
+	leal	16(%edi),%edi
+	movdqa	%xmm5,%xmm1
+	jmp	L050xts_enc_done
+.align	4,0x90
+L047xts_enc_two:
+	movaps	%xmm1,%xmm6
+	movups	(%esi),%xmm2
+	movups	16(%esi),%xmm3
+	leal	32(%esi),%esi
+	xorps	%xmm5,%xmm2
+	xorps	%xmm6,%xmm3
+	xorps	%xmm4,%xmm4
+	call	__aesni_encrypt3
+	xorps	%xmm5,%xmm2
+	xorps	%xmm6,%xmm3
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	leal	32(%edi),%edi
+	movdqa	%xmm6,%xmm1
+	jmp	L050xts_enc_done
+.align	4,0x90
+L048xts_enc_three:
+	movaps	%xmm1,%xmm7
+	movups	(%esi),%xmm2
+	movups	16(%esi),%xmm3
+	movups	32(%esi),%xmm4
+	leal	48(%esi),%esi
+	xorps	%xmm5,%xmm2
+	xorps	%xmm6,%xmm3
+	xorps	%xmm7,%xmm4
+	call	__aesni_encrypt3
+	xorps	%xmm5,%xmm2
+	xorps	%xmm6,%xmm3
+	xorps	%xmm7,%xmm4
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	movups	%xmm4,32(%edi)
+	leal	48(%edi),%edi
+	movdqa	%xmm7,%xmm1
+	jmp	L050xts_enc_done
+.align	4,0x90
+L049xts_enc_four:
+	movaps	%xmm1,%xmm6
+	movups	(%esi),%xmm2
+	movups	16(%esi),%xmm3
+	movups	32(%esi),%xmm4
+	xorps	(%esp),%xmm2
+	movups	48(%esi),%xmm5
+	leal	64(%esi),%esi
+	xorps	16(%esp),%xmm3
+	xorps	%xmm7,%xmm4
+	xorps	%xmm6,%xmm5
+	call	__aesni_encrypt4
+	xorps	(%esp),%xmm2
+	xorps	16(%esp),%xmm3
+	xorps	%xmm7,%xmm4
+	movups	%xmm2,(%edi)
+	xorps	%xmm6,%xmm5
+	movups	%xmm3,16(%edi)
+	movups	%xmm4,32(%edi)
+	movups	%xmm5,48(%edi)
+	leal	64(%edi),%edi
+	movdqa	%xmm6,%xmm1
+	jmp	L050xts_enc_done
+.align	4,0x90
+L045xts_enc_done6x:
+	movl	112(%esp),%eax
+	andl	$15,%eax
+	jz	L052xts_enc_ret
+	movdqa	%xmm1,%xmm5
+	movl	%eax,112(%esp)
+	jmp	L053xts_enc_steal
+.align	4,0x90
+L050xts_enc_done:
+	movl	112(%esp),%eax
+	pxor	%xmm0,%xmm0
+	andl	$15,%eax
+	jz	L052xts_enc_ret
+	pcmpgtd	%xmm1,%xmm0
+	movl	%eax,112(%esp)
+	pshufd	$19,%xmm0,%xmm5
+	paddq	%xmm1,%xmm1
+	pand	96(%esp),%xmm5
+	pxor	%xmm1,%xmm5
+L053xts_enc_steal:
+	movzbl	(%esi),%ecx
+	movzbl	-16(%edi),%edx
+	leal	1(%esi),%esi
+	movb	%cl,-16(%edi)
+	movb	%dl,(%edi)
+	leal	1(%edi),%edi
+	subl	$1,%eax
+	jnz	L053xts_enc_steal
+	subl	112(%esp),%edi
+	movl	%ebp,%edx
+	movl	%ebx,%ecx
+	movups	-16(%edi),%xmm2
+	xorps	%xmm5,%xmm2
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+L054enc1_loop_10:
+.byte	102,15,56,220,209
+	decl	%ecx
+	movups	(%edx),%xmm1
+	leal	16(%edx),%edx
+	jnz	L054enc1_loop_10
+.byte	102,15,56,221,209
+	xorps	%xmm5,%xmm2
+	movups	%xmm2,-16(%edi)
+L052xts_enc_ret:
+	movl	116(%esp),%esp
+	popl	%edi
+	popl	%esi
+	popl	%ebx
+	popl	%ebp
+	ret
+.globl	_aesni_xts_decrypt
+.align	4
+_aesni_xts_decrypt:
+L_aesni_xts_decrypt_begin:
+	pushl	%ebp
+	pushl	%ebx
+	pushl	%esi
+	pushl	%edi
+	movl	36(%esp),%edx
+	movl	40(%esp),%esi
+	movl	240(%edx),%ecx
+	movups	(%esi),%xmm2
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+L055enc1_loop_11:
+.byte	102,15,56,220,209
+	decl	%ecx
+	movups	(%edx),%xmm1
+	leal	16(%edx),%edx
+	jnz	L055enc1_loop_11
+.byte	102,15,56,221,209
+	movl	20(%esp),%esi
+	movl	24(%esp),%edi
+	movl	28(%esp),%eax
+	movl	32(%esp),%edx
+	movl	%esp,%ebp
+	subl	$120,%esp
+	andl	$-16,%esp
+	xorl	%ebx,%ebx
+	testl	$15,%eax
+	setnz	%bl
+	shll	$4,%ebx
+	subl	%ebx,%eax
+	movl	$135,96(%esp)
+	movl	$0,100(%esp)
+	movl	$1,104(%esp)
+	movl	$0,108(%esp)
+	movl	%eax,112(%esp)
+	movl	%ebp,116(%esp)
+	movl	240(%edx),%ecx
+	movl	%edx,%ebp
+	movl	%ecx,%ebx
+	movdqa	%xmm2,%xmm1
+	pxor	%xmm0,%xmm0
+	movdqa	96(%esp),%xmm3
+	pcmpgtd	%xmm1,%xmm0
+	andl	$-16,%eax
+	subl	$96,%eax
+	jc	L056xts_dec_short
+	shrl	$1,%ecx
+	movl	%ecx,%ebx
+	jmp	L057xts_dec_loop6
+.align	4,0x90
+L057xts_dec_loop6:
+	pshufd	$19,%xmm0,%xmm2
+	pxor	%xmm0,%xmm0
+	movdqa	%xmm1,(%esp)
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	pxor	%xmm2,%xmm1
+	pshufd	$19,%xmm0,%xmm2
+	pxor	%xmm0,%xmm0
+	movdqa	%xmm1,16(%esp)
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	pxor	%xmm2,%xmm1
+	pshufd	$19,%xmm0,%xmm2
+	pxor	%xmm0,%xmm0
+	movdqa	%xmm1,32(%esp)
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	pxor	%xmm2,%xmm1
+	pshufd	$19,%xmm0,%xmm2
+	pxor	%xmm0,%xmm0
+	movdqa	%xmm1,48(%esp)
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	pxor	%xmm2,%xmm1
+	pshufd	$19,%xmm0,%xmm7
+	movdqa	%xmm1,64(%esp)
+	paddq	%xmm1,%xmm1
+	movups	(%ebp),%xmm0
+	pand	%xmm3,%xmm7
+	movups	(%esi),%xmm2
+	pxor	%xmm1,%xmm7
+	movdqu	16(%esi),%xmm3
+	xorps	%xmm0,%xmm2
+	movdqu	32(%esi),%xmm4
+	pxor	%xmm0,%xmm3
+	movdqu	48(%esi),%xmm5
+	pxor	%xmm0,%xmm4
+	movdqu	64(%esi),%xmm6
+	pxor	%xmm0,%xmm5
+	movdqu	80(%esi),%xmm1
+	pxor	%xmm0,%xmm6
+	leal	96(%esi),%esi
+	pxor	(%esp),%xmm2
+	movdqa	%xmm7,80(%esp)
+	pxor	%xmm1,%xmm7
+	movups	16(%ebp),%xmm1
+	leal	32(%ebp),%edx
+	pxor	16(%esp),%xmm3
+.byte	102,15,56,222,209
+	pxor	32(%esp),%xmm4
+.byte	102,15,56,222,217
+	pxor	48(%esp),%xmm5
+	decl	%ecx
+.byte	102,15,56,222,225
+	pxor	64(%esp),%xmm6
+.byte	102,15,56,222,233
+	pxor	%xmm0,%xmm7
+.byte	102,15,56,222,241
+	movups	(%edx),%xmm0
+.byte	102,15,56,222,249
+	call	L_aesni_decrypt6_enter
+	movdqa	80(%esp),%xmm1
+	pxor	%xmm0,%xmm0
+	xorps	(%esp),%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	xorps	16(%esp),%xmm3
+	movups	%xmm2,(%edi)
+	xorps	32(%esp),%xmm4
+	movups	%xmm3,16(%edi)
+	xorps	48(%esp),%xmm5
+	movups	%xmm4,32(%edi)
+	xorps	64(%esp),%xmm6
+	movups	%xmm5,48(%edi)
+	xorps	%xmm1,%xmm7
+	movups	%xmm6,64(%edi)
+	pshufd	$19,%xmm0,%xmm2
+	movups	%xmm7,80(%edi)
+	leal	96(%edi),%edi
+	movdqa	96(%esp),%xmm3
+	pxor	%xmm0,%xmm0
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	movl	%ebx,%ecx
+	pxor	%xmm2,%xmm1
+	subl	$96,%eax
+	jnc	L057xts_dec_loop6
+	leal	1(,%ecx,2),%ecx
+	movl	%ebp,%edx
+	movl	%ecx,%ebx
+L056xts_dec_short:
+	addl	$96,%eax
+	jz	L058xts_dec_done6x
+	movdqa	%xmm1,%xmm5
+	cmpl	$32,%eax
+	jb	L059xts_dec_one
+	pshufd	$19,%xmm0,%xmm2
+	pxor	%xmm0,%xmm0
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	pxor	%xmm2,%xmm1
+	je	L060xts_dec_two
+	pshufd	$19,%xmm0,%xmm2
+	pxor	%xmm0,%xmm0
+	movdqa	%xmm1,%xmm6
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	pxor	%xmm2,%xmm1
+	cmpl	$64,%eax
+	jb	L061xts_dec_three
+	pshufd	$19,%xmm0,%xmm2
+	pxor	%xmm0,%xmm0
+	movdqa	%xmm1,%xmm7
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	pxor	%xmm2,%xmm1
+	movdqa	%xmm5,(%esp)
+	movdqa	%xmm6,16(%esp)
+	je	L062xts_dec_four
+	movdqa	%xmm7,32(%esp)
+	pshufd	$19,%xmm0,%xmm7
+	movdqa	%xmm1,48(%esp)
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm7
+	pxor	%xmm1,%xmm7
+	movdqu	(%esi),%xmm2
+	movdqu	16(%esi),%xmm3
+	movdqu	32(%esi),%xmm4
+	pxor	(%esp),%xmm2
+	movdqu	48(%esi),%xmm5
+	pxor	16(%esp),%xmm3
+	movdqu	64(%esi),%xmm6
+	pxor	32(%esp),%xmm4
+	leal	80(%esi),%esi
+	pxor	48(%esp),%xmm5
+	movdqa	%xmm7,64(%esp)
+	pxor	%xmm7,%xmm6
+	call	__aesni_decrypt6
+	movaps	64(%esp),%xmm1
+	xorps	(%esp),%xmm2
+	xorps	16(%esp),%xmm3
+	xorps	32(%esp),%xmm4
+	movups	%xmm2,(%edi)
+	xorps	48(%esp),%xmm5
+	movups	%xmm3,16(%edi)
+	xorps	%xmm1,%xmm6
+	movups	%xmm4,32(%edi)
+	movups	%xmm5,48(%edi)
+	movups	%xmm6,64(%edi)
+	leal	80(%edi),%edi
+	jmp	L063xts_dec_done
+.align	4,0x90
+L059xts_dec_one:
+	movups	(%esi),%xmm2
+	leal	16(%esi),%esi
+	xorps	%xmm5,%xmm2
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+L064dec1_loop_12:
+.byte	102,15,56,222,209
+	decl	%ecx
+	movups	(%edx),%xmm1
+	leal	16(%edx),%edx
+	jnz	L064dec1_loop_12
+.byte	102,15,56,223,209
+	xorps	%xmm5,%xmm2
+	movups	%xmm2,(%edi)
+	leal	16(%edi),%edi
+	movdqa	%xmm5,%xmm1
+	jmp	L063xts_dec_done
+.align	4,0x90
+L060xts_dec_two:
+	movaps	%xmm1,%xmm6
+	movups	(%esi),%xmm2
+	movups	16(%esi),%xmm3
+	leal	32(%esi),%esi
+	xorps	%xmm5,%xmm2
+	xorps	%xmm6,%xmm3
+	call	__aesni_decrypt3
+	xorps	%xmm5,%xmm2
+	xorps	%xmm6,%xmm3
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	leal	32(%edi),%edi
+	movdqa	%xmm6,%xmm1
+	jmp	L063xts_dec_done
+.align	4,0x90
+L061xts_dec_three:
+	movaps	%xmm1,%xmm7
+	movups	(%esi),%xmm2
+	movups	16(%esi),%xmm3
+	movups	32(%esi),%xmm4
+	leal	48(%esi),%esi
+	xorps	%xmm5,%xmm2
+	xorps	%xmm6,%xmm3
+	xorps	%xmm7,%xmm4
+	call	__aesni_decrypt3
+	xorps	%xmm5,%xmm2
+	xorps	%xmm6,%xmm3
+	xorps	%xmm7,%xmm4
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	movups	%xmm4,32(%edi)
+	leal	48(%edi),%edi
+	movdqa	%xmm7,%xmm1
+	jmp	L063xts_dec_done
+.align	4,0x90
+L062xts_dec_four:
+	movaps	%xmm1,%xmm6
+	movups	(%esi),%xmm2
+	movups	16(%esi),%xmm3
+	movups	32(%esi),%xmm4
+	xorps	(%esp),%xmm2
+	movups	48(%esi),%xmm5
+	leal	64(%esi),%esi
+	xorps	16(%esp),%xmm3
+	xorps	%xmm7,%xmm4
+	xorps	%xmm6,%xmm5
+	call	__aesni_decrypt4
+	xorps	(%esp),%xmm2
+	xorps	16(%esp),%xmm3
+	xorps	%xmm7,%xmm4
+	movups	%xmm2,(%edi)
+	xorps	%xmm6,%xmm5
+	movups	%xmm3,16(%edi)
+	movups	%xmm4,32(%edi)
+	movups	%xmm5,48(%edi)
+	leal	64(%edi),%edi
+	movdqa	%xmm6,%xmm1
+	jmp	L063xts_dec_done
+.align	4,0x90
+L058xts_dec_done6x:
+	movl	112(%esp),%eax
+	andl	$15,%eax
+	jz	L065xts_dec_ret
+	movl	%eax,112(%esp)
+	jmp	L066xts_dec_only_one_more
+.align	4,0x90
+L063xts_dec_done:
+	movl	112(%esp),%eax
+	pxor	%xmm0,%xmm0
+	andl	$15,%eax
+	jz	L065xts_dec_ret
+	pcmpgtd	%xmm1,%xmm0
+	movl	%eax,112(%esp)
+	pshufd	$19,%xmm0,%xmm2
+	pxor	%xmm0,%xmm0
+	movdqa	96(%esp),%xmm3
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm2
+	pcmpgtd	%xmm1,%xmm0
+	pxor	%xmm2,%xmm1
+L066xts_dec_only_one_more:
+	pshufd	$19,%xmm0,%xmm5
+	movdqa	%xmm1,%xmm6
+	paddq	%xmm1,%xmm1
+	pand	%xmm3,%xmm5
+	pxor	%xmm1,%xmm5
+	movl	%ebp,%edx
+	movl	%ebx,%ecx
+	movups	(%esi),%xmm2
+	xorps	%xmm5,%xmm2
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+L067dec1_loop_13:
+.byte	102,15,56,222,209
+	decl	%ecx
+	movups	(%edx),%xmm1
+	leal	16(%edx),%edx
+	jnz	L067dec1_loop_13
+.byte	102,15,56,223,209
+	xorps	%xmm5,%xmm2
+	movups	%xmm2,(%edi)
+L068xts_dec_steal:
+	movzbl	16(%esi),%ecx
+	movzbl	(%edi),%edx
+	leal	1(%esi),%esi
+	movb	%cl,(%edi)
+	movb	%dl,16(%edi)
+	leal	1(%edi),%edi
+	subl	$1,%eax
+	jnz	L068xts_dec_steal
+	subl	112(%esp),%edi
+	movl	%ebp,%edx
+	movl	%ebx,%ecx
+	movups	(%edi),%xmm2
+	xorps	%xmm6,%xmm2
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+L069dec1_loop_14:
+.byte	102,15,56,222,209
+	decl	%ecx
+	movups	(%edx),%xmm1
+	leal	16(%edx),%edx
+	jnz	L069dec1_loop_14
+.byte	102,15,56,223,209
+	xorps	%xmm6,%xmm2
+	movups	%xmm2,(%edi)
+L065xts_dec_ret:
+	movl	116(%esp),%esp
+	popl	%edi
+	popl	%esi
+	popl	%ebx
+	popl	%ebp
+	ret
+.globl	_aesni_cbc_encrypt
+.align	4
+_aesni_cbc_encrypt:
+L_aesni_cbc_encrypt_begin:
+	pushl	%ebp
+	pushl	%ebx
+	pushl	%esi
+	pushl	%edi
+	movl	20(%esp),%esi
+	movl	%esp,%ebx
+	movl	24(%esp),%edi
+	subl	$24,%ebx
+	movl	28(%esp),%eax
+	andl	$-16,%ebx
+	movl	32(%esp),%edx
+	movl	36(%esp),%ebp
+	testl	%eax,%eax
+	jz	L070cbc_abort
+	cmpl	$0,40(%esp)
+	xchgl	%esp,%ebx
+	movups	(%ebp),%xmm7
+	movl	240(%edx),%ecx
+	movl	%edx,%ebp
+	movl	%ebx,16(%esp)
+	movl	%ecx,%ebx
+	je	L071cbc_decrypt
+	movaps	%xmm7,%xmm2
+	cmpl	$16,%eax
+	jb	L072cbc_enc_tail
+	subl	$16,%eax
+	jmp	L073cbc_enc_loop
+.align	4,0x90
+L073cbc_enc_loop:
+	movups	(%esi),%xmm7
+	leal	16(%esi),%esi
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	xorps	%xmm0,%xmm7
+	leal	32(%edx),%edx
+	xorps	%xmm7,%xmm2
+L074enc1_loop_15:
+.byte	102,15,56,220,209
+	decl	%ecx
+	movups	(%edx),%xmm1
+	leal	16(%edx),%edx
+	jnz	L074enc1_loop_15
+.byte	102,15,56,221,209
+	movl	%ebx,%ecx
+	movl	%ebp,%edx
+	movups	%xmm2,(%edi)
+	leal	16(%edi),%edi
+	subl	$16,%eax
+	jnc	L073cbc_enc_loop
+	addl	$16,%eax
+	jnz	L072cbc_enc_tail
+	movaps	%xmm2,%xmm7
+	jmp	L075cbc_ret
+L072cbc_enc_tail:
+	movl	%eax,%ecx
+.long	2767451785
+	movl	$16,%ecx
+	subl	%eax,%ecx
+	xorl	%eax,%eax
+.long	2868115081
+	leal	-16(%edi),%edi
+	movl	%ebx,%ecx
+	movl	%edi,%esi
+	movl	%ebp,%edx
+	jmp	L073cbc_enc_loop
+.align	4,0x90
+L071cbc_decrypt:
+	cmpl	$80,%eax
+	jbe	L076cbc_dec_tail
+	movaps	%xmm7,(%esp)
+	subl	$80,%eax
+	jmp	L077cbc_dec_loop6_enter
+.align	4,0x90
+L078cbc_dec_loop6:
+	movaps	%xmm0,(%esp)
+	movups	%xmm7,(%edi)
+	leal	16(%edi),%edi
+L077cbc_dec_loop6_enter:
+	movdqu	(%esi),%xmm2
+	movdqu	16(%esi),%xmm3
+	movdqu	32(%esi),%xmm4
+	movdqu	48(%esi),%xmm5
+	movdqu	64(%esi),%xmm6
+	movdqu	80(%esi),%xmm7
+	call	__aesni_decrypt6
+	movups	(%esi),%xmm1
+	movups	16(%esi),%xmm0
+	xorps	(%esp),%xmm2
+	xorps	%xmm1,%xmm3
+	movups	32(%esi),%xmm1
+	xorps	%xmm0,%xmm4
+	movups	48(%esi),%xmm0
+	xorps	%xmm1,%xmm5
+	movups	64(%esi),%xmm1
+	xorps	%xmm0,%xmm6
+	movups	80(%esi),%xmm0
+	xorps	%xmm1,%xmm7
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	leal	96(%esi),%esi
+	movups	%xmm4,32(%edi)
+	movl	%ebx,%ecx
+	movups	%xmm5,48(%edi)
+	movl	%ebp,%edx
+	movups	%xmm6,64(%edi)
+	leal	80(%edi),%edi
+	subl	$96,%eax
+	ja	L078cbc_dec_loop6
+	movaps	%xmm7,%xmm2
+	movaps	%xmm0,%xmm7
+	addl	$80,%eax
+	jle	L079cbc_dec_tail_collected
+	movups	%xmm2,(%edi)
+	leal	16(%edi),%edi
+L076cbc_dec_tail:
+	movups	(%esi),%xmm2
+	movaps	%xmm2,%xmm6
+	cmpl	$16,%eax
+	jbe	L080cbc_dec_one
+	movups	16(%esi),%xmm3
+	movaps	%xmm3,%xmm5
+	cmpl	$32,%eax
+	jbe	L081cbc_dec_two
+	movups	32(%esi),%xmm4
+	cmpl	$48,%eax
+	jbe	L082cbc_dec_three
+	movups	48(%esi),%xmm5
+	cmpl	$64,%eax
+	jbe	L083cbc_dec_four
+	movups	64(%esi),%xmm6
+	movaps	%xmm7,(%esp)
+	movups	(%esi),%xmm2
+	xorps	%xmm7,%xmm7
+	call	__aesni_decrypt6
+	movups	(%esi),%xmm1
+	movups	16(%esi),%xmm0
+	xorps	(%esp),%xmm2
+	xorps	%xmm1,%xmm3
+	movups	32(%esi),%xmm1
+	xorps	%xmm0,%xmm4
+	movups	48(%esi),%xmm0
+	xorps	%xmm1,%xmm5
+	movups	64(%esi),%xmm7
+	xorps	%xmm0,%xmm6
+	movups	%xmm2,(%edi)
+	movups	%xmm3,16(%edi)
+	movups	%xmm4,32(%edi)
+	movups	%xmm5,48(%edi)
+	leal	64(%edi),%edi
+	movaps	%xmm6,%xmm2
+	subl	$80,%eax
+	jmp	L079cbc_dec_tail_collected
+.align	4,0x90
+L080cbc_dec_one:
+	movups	(%edx),%xmm0
+	movups	16(%edx),%xmm1
+	leal	32(%edx),%edx
+	xorps	%xmm0,%xmm2
+L084dec1_loop_16:
+.byte	102,15,56,222,209
+	decl	%ecx
+	movups	(%edx),%xmm1
+	leal	16(%edx),%edx
+	jnz	L084dec1_loop_16
+.byte	102,15,56,223,209
+	xorps	%xmm7,%xmm2
+	movaps	%xmm6,%xmm7
+	subl	$16,%eax
+	jmp	L079cbc_dec_tail_collected
+.align	4,0x90
+L081cbc_dec_two:
+	xorps	%xmm4,%xmm4
+	call	__aesni_decrypt3
+	xorps	%xmm7,%xmm2
+	xorps	%xmm6,%xmm3
+	movups	%xmm2,(%edi)
+	movaps	%xmm3,%xmm2
+	leal	16(%edi),%edi
+	movaps	%xmm5,%xmm7
+	subl	$32,%eax
+	jmp	L079cbc_dec_tail_collected
+.align	4,0x90
+L082cbc_dec_three:
+	call	__aesni_decrypt3
+	xorps	%xmm7,%xmm2
+	xorps	%xmm6,%xmm3
+	xorps	%xmm5,%xmm4
+	movups	%xmm2,(%edi)
+	movaps	%xmm4,%xmm2
+	movups	%xmm3,16(%edi)
+	leal	32(%edi),%edi
+	movups	32(%esi),%xmm7
+	subl	$48,%eax
+	jmp	L079cbc_dec_tail_collected
+.align	4,0x90
+L083cbc_dec_four:
+	call	__aesni_decrypt4
+	movups	16(%esi),%xmm1
+	movups	32(%esi),%xmm0
+	xorps	%xmm7,%xmm2
+	movups	48(%esi),%xmm7
+	xorps	%xmm6,%xmm3
+	movups	%xmm2,(%edi)
+	xorps	%xmm1,%xmm4
+	movups	%xmm3,16(%edi)
+	xorps	%xmm0,%xmm5
+	movups	%xmm4,32(%edi)
+	leal	48(%edi),%edi
+	movaps	%xmm5,%xmm2
+	subl	$64,%eax
+L079cbc_dec_tail_collected:
+	andl	$15,%eax
+	jnz	L085cbc_dec_tail_partial
+	movups	%xmm2,(%edi)
+	jmp	L075cbc_ret
+.align	4,0x90
+L085cbc_dec_tail_partial:
+	movaps	%xmm2,(%esp)
+	movl	$16,%ecx
+	movl	%esp,%esi
+	subl	%eax,%ecx
+.long	2767451785
+L075cbc_ret:
+	movl	16(%esp),%esp
+	movl	36(%esp),%ebp
+	movups	%xmm7,(%ebp)
+L070cbc_abort:
+	popl	%edi
+	popl	%esi
+	popl	%ebx
+	popl	%ebp
+	ret
+.align	4
+__aesni_set_encrypt_key:
+	testl	%eax,%eax
+	jz	L086bad_pointer
+	testl	%edx,%edx
+	jz	L086bad_pointer
+	movups	(%eax),%xmm0
+	xorps	%xmm4,%xmm4
+	leal	16(%edx),%edx
+	cmpl	$256,%ecx
+	je	L08714rounds
+	cmpl	$192,%ecx
+	je	L08812rounds
+	cmpl	$128,%ecx
+	jne	L089bad_keybits
+.align	4,0x90
+L09010rounds:
+	movl	$9,%ecx
+	movups	%xmm0,-16(%edx)
+.byte	102,15,58,223,200,1
+	call	L091key_128_cold
+.byte	102,15,58,223,200,2
+	call	L092key_128
+.byte	102,15,58,223,200,4
+	call	L092key_128
+.byte	102,15,58,223,200,8
+	call	L092key_128
+.byte	102,15,58,223,200,16
+	call	L092key_128
+.byte	102,15,58,223,200,32
+	call	L092key_128
+.byte	102,15,58,223,200,64
+	call	L092key_128
+.byte	102,15,58,223,200,128
+	call	L092key_128
+.byte	102,15,58,223,200,27
+	call	L092key_128
+.byte	102,15,58,223,200,54
+	call	L092key_128
+	movups	%xmm0,(%edx)
+	movl	%ecx,80(%edx)
+	xorl	%eax,%eax
+	ret
+.align	4,0x90
+L092key_128:
+	movups	%xmm0,(%edx)
+	leal	16(%edx),%edx
+L091key_128_cold:
+	shufps	$16,%xmm0,%xmm4
+	xorps	%xmm4,%xmm0
+	shufps	$140,%xmm0,%xmm4
+	xorps	%xmm4,%xmm0
+	shufps	$255,%xmm1,%xmm1
+	xorps	%xmm1,%xmm0
+	ret
+.align	4,0x90
+L08812rounds:
+	movq	16(%eax),%xmm2
+	movl	$11,%ecx
+	movups	%xmm0,-16(%edx)
+.byte	102,15,58,223,202,1
+	call	L093key_192a_cold
+.byte	102,15,58,223,202,2
+	call	L094key_192b
+.byte	102,15,58,223,202,4
+	call	L095key_192a
+.byte	102,15,58,223,202,8
+	call	L094key_192b
+.byte	102,15,58,223,202,16
+	call	L095key_192a
+.byte	102,15,58,223,202,32
+	call	L094key_192b
+.byte	102,15,58,223,202,64
+	call	L095key_192a
+.byte	102,15,58,223,202,128
+	call	L094key_192b
+	movups	%xmm0,(%edx)
+	movl	%ecx,48(%edx)
+	xorl	%eax,%eax
+	ret
+.align	4,0x90
+L095key_192a:
+	movups	%xmm0,(%edx)
+	leal	16(%edx),%edx
+.align	4,0x90
+L093key_192a_cold:
+	movaps	%xmm2,%xmm5
+L096key_192b_warm:
+	shufps	$16,%xmm0,%xmm4
+	movdqa	%xmm2,%xmm3
+	xorps	%xmm4,%xmm0
+	shufps	$140,%xmm0,%xmm4
+	pslldq	$4,%xmm3
+	xorps	%xmm4,%xmm0
+	pshufd	$85,%xmm1,%xmm1
+	pxor	%xmm3,%xmm2
+	pxor	%xmm1,%xmm0
+	pshufd	$255,%xmm0,%xmm3
+	pxor	%xmm3,%xmm2
+	ret
+.align	4,0x90
+L094key_192b:
+	movaps	%xmm0,%xmm3
+	shufps	$68,%xmm0,%xmm5
+	movups	%xmm5,(%edx)
+	shufps	$78,%xmm2,%xmm3
+	movups	%xmm3,16(%edx)
+	leal	32(%edx),%edx
+	jmp	L096key_192b_warm
+.align	4,0x90
+L08714rounds:
+	movups	16(%eax),%xmm2
+	movl	$13,%ecx
+	leal	16(%edx),%edx
+	movups	%xmm0,-32(%edx)
+	movups	%xmm2,-16(%edx)
+.byte	102,15,58,223,202,1
+	call	L097key_256a_cold
+.byte	102,15,58,223,200,1
+	call	L098key_256b
+.byte	102,15,58,223,202,2
+	call	L099key_256a
+.byte	102,15,58,223,200,2
+	call	L098key_256b
+.byte	102,15,58,223,202,4
+	call	L099key_256a
+.byte	102,15,58,223,200,4
+	call	L098key_256b
+.byte	102,15,58,223,202,8
+	call	L099key_256a
+.byte	102,15,58,223,200,8
+	call	L098key_256b
+.byte	102,15,58,223,202,16
+	call	L099key_256a
+.byte	102,15,58,223,200,16
+	call	L098key_256b
+.byte	102,15,58,223,202,32
+	call	L099key_256a
+.byte	102,15,58,223,200,32
+	call	L098key_256b
+.byte	102,15,58,223,202,64
+	call	L099key_256a
+	movups	%xmm0,(%edx)
+	movl	%ecx,16(%edx)
+	xorl	%eax,%eax
+	ret
+.align	4,0x90
+L099key_256a:
+	movups	%xmm2,(%edx)
+	leal	16(%edx),%edx
+L097key_256a_cold:
+	shufps	$16,%xmm0,%xmm4
+	xorps	%xmm4,%xmm0
+	shufps	$140,%xmm0,%xmm4
+	xorps	%xmm4,%xmm0
+	shufps	$255,%xmm1,%xmm1
+	xorps	%xmm1,%xmm0
+	ret
+.align	4,0x90
+L098key_256b:
+	movups	%xmm0,(%edx)
+	leal	16(%edx),%edx
+	shufps	$16,%xmm2,%xmm4
+	xorps	%xmm4,%xmm2
+	shufps	$140,%xmm2,%xmm4
+	xorps	%xmm4,%xmm2
+	shufps	$170,%xmm1,%xmm1
+	xorps	%xmm1,%xmm2
+	ret
+.align	2,0x90
+L086bad_pointer:
+	movl	$-1,%eax
+	ret
+.align	2,0x90
+L089bad_keybits:
+	movl	$-2,%eax
+	ret
+.globl	_aesni_set_encrypt_key
+.align	4
+_aesni_set_encrypt_key:
+L_aesni_set_encrypt_key_begin:
+	movl	4(%esp),%eax
+	movl	8(%esp),%ecx
+	movl	12(%esp),%edx
+	call	__aesni_set_encrypt_key
+	ret
+.globl	_aesni_set_decrypt_key
+.align	4
+_aesni_set_decrypt_key:
+L_aesni_set_decrypt_key_begin:
+	movl	4(%esp),%eax
+	movl	8(%esp),%ecx
+	movl	12(%esp),%edx
+	call	__aesni_set_encrypt_key
+	movl	12(%esp),%edx
+	shll	$4,%ecx
+	testl	%eax,%eax
+	jnz	L100dec_key_ret
+	leal	16(%edx,%ecx,1),%eax
+	movups	(%edx),%xmm0
+	movups	(%eax),%xmm1
+	movups	%xmm0,(%eax)
+	movups	%xmm1,(%edx)
+	leal	16(%edx),%edx
+	leal	-16(%eax),%eax
+L101dec_key_inverse:
+	movups	(%edx),%xmm0
+	movups	(%eax),%xmm1
+.byte	102,15,56,219,192
+.byte	102,15,56,219,201
+	leal	16(%edx),%edx
+	leal	-16(%eax),%eax
+	movups	%xmm0,16(%eax)
+	movups	%xmm1,-16(%edx)
+	cmpl	%edx,%eax
+	ja	L101dec_key_inverse
+	movups	(%edx),%xmm0
+.byte	102,15,56,219,192
+	movups	%xmm0,(%edx)
+	xorl	%eax,%eax
+L100dec_key_ret:
+	ret
+.byte	65,69,83,32,102,111,114,32,73,110,116,101,108,32,65,69
+.byte	83,45,78,73,44,32,67,82,89,80,84,79,71,65,77,83
+.byte	32,98,121,32,60,97,112,112,114,111,64,111,112,101,110,115
+.byte	115,108,46,111,114,103,62,0
diff --git a/lib/accelerated/x86/asm-macosx/cpuid-x86-64-macosx.s b/lib/accelerated/x86/asm-macosx/cpuid-x86-64-macosx.s
new file mode 100644
index 0000000000..b4a4e941a4
--- /dev/null
+++ b/lib/accelerated/x86/asm-macosx/cpuid-x86-64-macosx.s
@@ -0,0 +1,54 @@
+#
+# Copyright (C) 2011 Free Software Foundation, Inc.
+#
+# Author: Nikos Mavrogiannopoulos
+#
+# This file is part of GnuTLS.
+#
+# The GnuTLS is free software; you can redistribute it and/or
+# modify it under the terms of the GNU Lesser General Public License
+# as published by the Free Software Foundation; either version 3 of
+# the License, or (at your option) any later version.
+#
+# This library is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# Lesser General Public License for more details.
+#
+# You should have received a copy of the GNU Lesser General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+#
+.text	
+.globl	_gnutls_cpuid
+
+.p2align	4
+_gnutls_cpuid:
+	pushq	%rbp
+	movq	%rsp,%rbp
+	pushq	%rbx
+	movl	%edi,-12(%rbp)
+	movq	%rsi,-24(%rbp)
+	movq	%rdx,-32(%rbp)
+	movq	%rcx,-40(%rbp)
+	movq	%r8,-48(%rbp)
+	movl	-12(%rbp),%eax
+	movl	%eax,-60(%rbp)
+	movl	-60(%rbp),%eax
+	cpuid
+	movl	%edx,-56(%rbp)
+	movl	%ecx,%esi
+	movl	%eax,-52(%rbp)
+	movq	-24(%rbp),%rax
+	movl	-52(%rbp),%edx
+	movl	%edx,(%rax)
+	movq	-32(%rbp),%rax
+	movl	%ebx,(%rax)
+	movq	-40(%rbp),%rax
+	movl	%esi,(%rax)
+	movq	-48(%rbp),%rax
+	movl	-56(%rbp),%ecx
+	movl	%ecx,(%rax)
+	popq	%rbx
+	leave
+	.byte	0xf3,0xc3
+
diff --git a/lib/accelerated/x86/asm-macosx/cpuid-x86-macosx.s b/lib/accelerated/x86/asm-macosx/cpuid-x86-macosx.s
new file mode 100644
index 0000000000..a69a5c7fad
--- /dev/null
+++ b/lib/accelerated/x86/asm-macosx/cpuid-x86-macosx.s
@@ -0,0 +1,66 @@
+#
+# Copyright (C) 2011 Free Software Foundation, Inc.
+#
+# Author: Nikos Mavrogiannopoulos
+#
+# This file is part of GnuTLS.
+#
+# The GnuTLS is free software; you can redistribute it and/or
+# modify it under the terms of the GNU Lesser General Public License
+# as published by the Free Software Foundation; either version 3 of
+# the License, or (at your option) any later version.
+#
+# This library is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# Lesser General Public License for more details.
+#
+# You should have received a copy of the GNU Lesser General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+#
+.file	"devel/perlasm/cpuid-x86.s"
+.text
+.globl	_gnutls_cpuid
+.align	4
+_gnutls_cpuid:
+L_gnutls_cpuid_begin:
+	pushl	%ebp
+	movl	%esp,%ebp
+	subl	$12,%esp
+	movl	%ebx,(%esp)
+	movl	8(%ebp),%eax
+	movl	%esi,4(%esp)
+	movl	%edi,8(%esp)
+	pushl	%ebx
+	.byte	0x0f,0xa2
+	movl	%ebx,%edi
+	popl	%ebx
+	movl	%edx,%esi
+	movl	12(%ebp),%edx
+	movl	%eax,(%edx)
+	movl	16(%ebp),%eax
+	movl	%edi,(%eax)
+	movl	20(%ebp),%eax
+	movl	%ecx,(%eax)
+	movl	24(%ebp),%eax
+	movl	%esi,(%eax)
+	movl	(%esp),%ebx
+	movl	4(%esp),%esi
+	movl	8(%esp),%edi
+	movl	%ebp,%esp
+	popl	%ebp
+	ret
+.globl	_gnutls_have_cpuid
+.align	4
+_gnutls_have_cpuid:
+L_gnutls_have_cpuid_begin:
+	pushfl
+	popl	%eax
+	orl	$2097152,%eax
+	pushl	%eax
+	popfl
+	pushfl
+	popl	%eax
+	andl	$2097152,%eax
+	ret
+.byte	67,80,85,73,68,32,102,111,114,32,120,56,54,0
diff --git a/lib/accelerated/x86/asm-macosx/padlock-x86-64-macosx.s b/lib/accelerated/x86/asm-macosx/padlock-x86-64-macosx.s
new file mode 100644
index 0000000000..a19e824ad3
--- /dev/null
+++ b/lib/accelerated/x86/asm-macosx/padlock-x86-64-macosx.s
@@ -0,0 +1,515 @@
+# Copyright (c) 2011, Andy Polyakov by <appro@openssl.org>
+# All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions
+# are met:
+# 
+#     * Redistributions of source code must retain copyright notices,
+#      this list of conditions and the following disclaimer.
+#
+#     * 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.
+#
+#     * Neither the name of the Andy Polyakov nor the names of its
+#      copyright holder and contributors may be used to endorse or
+#      promote products derived from this software without specific
+#      prior written permission.
+#
+# ALTERNATIVELY, provided that this notice is retained in full, this
+# product may be distributed under the terms of the GNU General Public
+# License (GPL), in which case the provisions of the GPL apply INSTEAD OF
+# those given above.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDER AND CONTRIBUTORS
+# "AS IS" AND ANY EXPRESS 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 COPYRIGHT
+# OWNER OR 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.
+
+.text	
+.globl	_padlock_capability
+
+.p2align	4
+_padlock_capability:
+	movq	%rbx,%r8
+	xorl	%eax,%eax
+	cpuid
+	xorl	%eax,%eax
+	cmpl	$1953391939,%ebx
+	jne	L$noluck
+	cmpl	$1215460705,%edx
+	jne	L$noluck
+	cmpl	$1936487777,%ecx
+	jne	L$noluck
+	movl	$3221225472,%eax
+	cpuid
+	movl	%eax,%edx
+	xorl	%eax,%eax
+	cmpl	$3221225473,%edx
+	jb	L$noluck
+	movl	$3221225473,%eax
+	cpuid
+	movl	%edx,%eax
+	andl	$4294967279,%eax
+	orl	$16,%eax
+L$noluck:
+	movq	%r8,%rbx
+	.byte	0xf3,0xc3
+
+
+.globl	_padlock_key_bswap
+
+.p2align	4
+_padlock_key_bswap:
+	movl	240(%rdi),%edx
+L$bswap_loop:
+	movl	(%rdi),%eax
+	bswapl	%eax
+	movl	%eax,(%rdi)
+	leaq	4(%rdi),%rdi
+	subl	$1,%edx
+	jnz	L$bswap_loop
+	.byte	0xf3,0xc3
+
+
+.globl	_padlock_verify_context
+
+.p2align	4
+_padlock_verify_context:
+	movq	%rdi,%rdx
+	pushf
+	leaq	L$padlock_saved_context(%rip),%rax
+	call	_padlock_verify_ctx
+	leaq	8(%rsp),%rsp
+	.byte	0xf3,0xc3
+
+
+
+.p2align	4
+_padlock_verify_ctx:
+	movq	8(%rsp),%r8
+	btq	$30,%r8
+	jnc	L$verified
+	cmpq	(%rax),%rdx
+	je	L$verified
+	pushf
+	popf
+L$verified:
+	movq	%rdx,(%rax)
+	.byte	0xf3,0xc3
+
+
+.globl	_padlock_reload_key
+
+.p2align	4
+_padlock_reload_key:
+	pushf
+	popf
+	.byte	0xf3,0xc3
+
+
+.globl	_padlock_aes_block
+
+.p2align	4
+_padlock_aes_block:
+	movq	%rbx,%r8
+	movq	$1,%rcx
+	leaq	32(%rdx),%rbx
+	leaq	16(%rdx),%rdx
+.byte	0xf3,0x0f,0xa7,0xc8	
+	movq	%r8,%rbx
+	.byte	0xf3,0xc3
+
+
+.globl	_padlock_xstore
+
+.p2align	4
+_padlock_xstore:
+	movl	%esi,%edx
+.byte	0x0f,0xa7,0xc0		
+	.byte	0xf3,0xc3
+
+
+.globl	_padlock_sha1_oneshot
+
+.p2align	4
+_padlock_sha1_oneshot:
+	movq	%rdx,%rcx
+	movq	%rdi,%rdx
+	movups	(%rdi),%xmm0
+	subq	$128+8,%rsp
+	movl	16(%rdi),%eax
+	movaps	%xmm0,(%rsp)
+	movq	%rsp,%rdi
+	movl	%eax,16(%rsp)
+	xorq	%rax,%rax
+.byte	0xf3,0x0f,0xa6,0xc8	
+	movaps	(%rsp),%xmm0
+	movl	16(%rsp),%eax
+	addq	$128+8,%rsp
+	movups	%xmm0,(%rdx)
+	movl	%eax,16(%rdx)
+	.byte	0xf3,0xc3
+
+
+.globl	_padlock_sha1_blocks
+
+.p2align	4
+_padlock_sha1_blocks:
+	movq	%rdx,%rcx
+	movq	%rdi,%rdx
+	movups	(%rdi),%xmm0
+	subq	$128+8,%rsp
+	movl	16(%rdi),%eax
+	movaps	%xmm0,(%rsp)
+	movq	%rsp,%rdi
+	movl	%eax,16(%rsp)
+	movq	$-1,%rax
+.byte	0xf3,0x0f,0xa6,0xc8	
+	movaps	(%rsp),%xmm0
+	movl	16(%rsp),%eax
+	addq	$128+8,%rsp
+	movups	%xmm0,(%rdx)
+	movl	%eax,16(%rdx)
+	.byte	0xf3,0xc3
+
+
+.globl	_padlock_sha256_oneshot
+
+.p2align	4
+_padlock_sha256_oneshot:
+	movq	%rdx,%rcx
+	movq	%rdi,%rdx
+	movups	(%rdi),%xmm0
+	subq	$128+8,%rsp
+	movups	16(%rdi),%xmm1
+	movaps	%xmm0,(%rsp)
+	movq	%rsp,%rdi
+	movaps	%xmm1,16(%rsp)
+	xorq	%rax,%rax
+.byte	0xf3,0x0f,0xa6,0xd0	
+	movaps	(%rsp),%xmm0
+	movaps	16(%rsp),%xmm1
+	addq	$128+8,%rsp
+	movups	%xmm0,(%rdx)
+	movups	%xmm1,16(%rdx)
+	.byte	0xf3,0xc3
+
+
+.globl	_padlock_sha256_blocks
+
+.p2align	4
+_padlock_sha256_blocks:
+	movq	%rdx,%rcx
+	movq	%rdi,%rdx
+	movups	(%rdi),%xmm0
+	subq	$128+8,%rsp
+	movups	16(%rdi),%xmm1
+	movaps	%xmm0,(%rsp)
+	movq	%rsp,%rdi
+	movaps	%xmm1,16(%rsp)
+	movq	$-1,%rax
+.byte	0xf3,0x0f,0xa6,0xd0	
+	movaps	(%rsp),%xmm0
+	movaps	16(%rsp),%xmm1
+	addq	$128+8,%rsp
+	movups	%xmm0,(%rdx)
+	movups	%xmm1,16(%rdx)
+	.byte	0xf3,0xc3
+
+
+.globl	_padlock_sha512_blocks
+
+.p2align	4
+_padlock_sha512_blocks:
+	movq	%rdx,%rcx
+	movq	%rdi,%rdx
+	movups	(%rdi),%xmm0
+	subq	$128+8,%rsp
+	movups	16(%rdi),%xmm1
+	movups	32(%rdi),%xmm2
+	movups	48(%rdi),%xmm3
+	movaps	%xmm0,(%rsp)
+	movq	%rsp,%rdi
+	movaps	%xmm1,16(%rsp)
+	movaps	%xmm2,32(%rsp)
+	movaps	%xmm3,48(%rsp)
+.byte	0xf3,0x0f,0xa6,0xe0	
+	movaps	(%rsp),%xmm0
+	movaps	16(%rsp),%xmm1
+	movaps	32(%rsp),%xmm2
+	movaps	48(%rsp),%xmm3
+	addq	$128+8,%rsp
+	movups	%xmm0,(%rdx)
+	movups	%xmm1,16(%rdx)
+	movups	%xmm2,32(%rdx)
+	movups	%xmm3,48(%rdx)
+	.byte	0xf3,0xc3
+
+.globl	_padlock_ecb_encrypt
+
+.p2align	4
+_padlock_ecb_encrypt:
+	pushq	%rbp
+	pushq	%rbx
+
+	xorl	%eax,%eax
+	testq	$15,%rdx
+	jnz	L$ecb_abort
+	testq	$15,%rcx
+	jnz	L$ecb_abort
+	leaq	L$padlock_saved_context(%rip),%rax
+	pushf
+	cld
+	call	_padlock_verify_ctx
+	leaq	16(%rdx),%rdx
+	xorl	%eax,%eax
+	xorl	%ebx,%ebx
+	cmpq	$128,%rcx
+	jbe	L$ecb_short
+	testl	$32,(%rdx)
+	jnz	L$ecb_aligned
+	testq	$15,%rdi
+	setz	%al
+	testq	$15,%rsi
+	setz	%bl
+	testl	%ebx,%eax
+	jnz	L$ecb_aligned
+	negq	%rax
+	movq	$512,%rbx
+	notq	%rax
+	leaq	(%rsp),%rbp
+	cmpq	%rbx,%rcx
+	cmovcq	%rcx,%rbx
+	andq	%rbx,%rax
+	movq	%rcx,%rbx
+	negq	%rax
+	andq	$512-1,%rbx
+	leaq	(%rax,%rbp,1),%rsp
+	jmp	L$ecb_loop
+.p2align	4
+L$ecb_loop:
+	cmpq	%rcx,%rbx
+	cmovaq	%rcx,%rbx
+	movq	%rdi,%r8
+	movq	%rsi,%r9
+	movq	%rcx,%r10
+	movq	%rbx,%rcx
+	movq	%rbx,%r11
+	testq	$15,%rdi
+	cmovnzq	%rsp,%rdi
+	testq	$15,%rsi
+	jz	L$ecb_inp_aligned
+	shrq	$3,%rcx
+.byte	0xf3,0x48,0xa5		
+	subq	%rbx,%rdi
+	movq	%rbx,%rcx
+	movq	%rdi,%rsi
+L$ecb_inp_aligned:
+	leaq	-16(%rdx),%rax
+	leaq	16(%rdx),%rbx
+	shrq	$4,%rcx
+.byte	0xf3,0x0f,0xa7,200	
+	movq	%r8,%rdi
+	movq	%r11,%rbx
+	testq	$15,%rdi
+	jz	L$ecb_out_aligned
+	movq	%rbx,%rcx
+	shrq	$3,%rcx
+	leaq	(%rsp),%rsi
+.byte	0xf3,0x48,0xa5		
+	subq	%rbx,%rdi
+L$ecb_out_aligned:
+	movq	%r9,%rsi
+	movq	%r10,%rcx
+	addq	%rbx,%rdi
+	addq	%rbx,%rsi
+	subq	%rbx,%rcx
+	movq	$512,%rbx
+	jnz	L$ecb_loop
+
+	cmpq	%rsp,%rbp
+	je	L$ecb_done
+
+	pxor	%xmm0,%xmm0
+	leaq	(%rsp),%rax
+L$ecb_bzero:
+	movaps	%xmm0,(%rax)
+	leaq	16(%rax),%rax
+	cmpq	%rax,%rbp
+	ja	L$ecb_bzero
+
+L$ecb_done:
+	leaq	(%rbp),%rsp
+	jmp	L$ecb_exit
+.p2align	4
+L$ecb_short:
+	movq	%rsp,%rbp
+	subq	%rcx,%rsp
+	xorq	%rbx,%rbx
+L$ecb_short_copy:
+	movups	(%rsi,%rbx,1),%xmm0
+	leaq	16(%rbx),%rbx
+	cmpq	%rbx,%rcx
+	movaps	%xmm0,-16(%rsp,%rbx,1)
+	ja	L$ecb_short_copy
+	movq	%rsp,%rsi
+	movq	%rcx,%rbx
+	jmp	L$ecb_loop
+.p2align	4
+L$ecb_aligned:
+	leaq	-16(%rdx),%rax
+	leaq	16(%rdx),%rbx
+	shrq	$4,%rcx
+.byte	0xf3,0x0f,0xa7,200	
+L$ecb_exit:
+	movl	$1,%eax
+	leaq	8(%rsp),%rsp
+L$ecb_abort:
+	popq	%rbx
+	popq	%rbp
+	.byte	0xf3,0xc3
+
+.globl	_padlock_cbc_encrypt
+
+.p2align	4
+_padlock_cbc_encrypt:
+	pushq	%rbp
+	pushq	%rbx
+
+	xorl	%eax,%eax
+	testq	$15,%rdx
+	jnz	L$cbc_abort
+	testq	$15,%rcx
+	jnz	L$cbc_abort
+	leaq	L$padlock_saved_context(%rip),%rax
+	pushf
+	cld
+	call	_padlock_verify_ctx
+	leaq	16(%rdx),%rdx
+	xorl	%eax,%eax
+	xorl	%ebx,%ebx
+	cmpq	$64,%rcx
+	jbe	L$cbc_short
+	testl	$32,(%rdx)
+	jnz	L$cbc_aligned
+	testq	$15,%rdi
+	setz	%al
+	testq	$15,%rsi
+	setz	%bl
+	testl	%ebx,%eax
+	jnz	L$cbc_aligned
+	negq	%rax
+	movq	$512,%rbx
+	notq	%rax
+	leaq	(%rsp),%rbp
+	cmpq	%rbx,%rcx
+	cmovcq	%rcx,%rbx
+	andq	%rbx,%rax
+	movq	%rcx,%rbx
+	negq	%rax
+	andq	$512-1,%rbx
+	leaq	(%rax,%rbp,1),%rsp
+	jmp	L$cbc_loop
+.p2align	4
+L$cbc_loop:
+	cmpq	%rcx,%rbx
+	cmovaq	%rcx,%rbx
+	movq	%rdi,%r8
+	movq	%rsi,%r9
+	movq	%rcx,%r10
+	movq	%rbx,%rcx
+	movq	%rbx,%r11
+	testq	$15,%rdi
+	cmovnzq	%rsp,%rdi
+	testq	$15,%rsi
+	jz	L$cbc_inp_aligned
+	shrq	$3,%rcx
+.byte	0xf3,0x48,0xa5		
+	subq	%rbx,%rdi
+	movq	%rbx,%rcx
+	movq	%rdi,%rsi
+L$cbc_inp_aligned:
+	leaq	-16(%rdx),%rax
+	leaq	16(%rdx),%rbx
+	shrq	$4,%rcx
+.byte	0xf3,0x0f,0xa7,208	
+	movdqa	(%rax),%xmm0
+	movdqa	%xmm0,-16(%rdx)
+	movq	%r8,%rdi
+	movq	%r11,%rbx
+	testq	$15,%rdi
+	jz	L$cbc_out_aligned
+	movq	%rbx,%rcx
+	shrq	$3,%rcx
+	leaq	(%rsp),%rsi
+.byte	0xf3,0x48,0xa5		
+	subq	%rbx,%rdi
+L$cbc_out_aligned:
+	movq	%r9,%rsi
+	movq	%r10,%rcx
+	addq	%rbx,%rdi
+	addq	%rbx,%rsi
+	subq	%rbx,%rcx
+	movq	$512,%rbx
+	jnz	L$cbc_loop
+
+	cmpq	%rsp,%rbp
+	je	L$cbc_done
+
+	pxor	%xmm0,%xmm0
+	leaq	(%rsp),%rax
+L$cbc_bzero:
+	movaps	%xmm0,(%rax)
+	leaq	16(%rax),%rax
+	cmpq	%rax,%rbp
+	ja	L$cbc_bzero
+
+L$cbc_done:
+	leaq	(%rbp),%rsp
+	jmp	L$cbc_exit
+.p2align	4
+L$cbc_short:
+	movq	%rsp,%rbp
+	subq	%rcx,%rsp
+	xorq	%rbx,%rbx
+L$cbc_short_copy:
+	movups	(%rsi,%rbx,1),%xmm0
+	leaq	16(%rbx),%rbx
+	cmpq	%rbx,%rcx
+	movaps	%xmm0,-16(%rsp,%rbx,1)
+	ja	L$cbc_short_copy
+	movq	%rsp,%rsi
+	movq	%rcx,%rbx
+	jmp	L$cbc_loop
+.p2align	4
+L$cbc_aligned:
+	leaq	-16(%rdx),%rax
+	leaq	16(%rdx),%rbx
+	shrq	$4,%rcx
+.byte	0xf3,0x0f,0xa7,208	
+	movdqa	(%rax),%xmm0
+	movdqa	%xmm0,-16(%rdx)
+L$cbc_exit:
+	movl	$1,%eax
+	leaq	8(%rsp),%rsp
+L$cbc_abort:
+	popq	%rbx
+	popq	%rbp
+	.byte	0xf3,0xc3
+
+.byte	86,73,65,32,80,97,100,108,111,99,107,32,120,56,54,95,54,52,32,109,111,100,117,108,101,44,32,67,82,89,80,84,79,71,65,77,83,32,98,121,32,60,97,112,112,114,111,64,111,112,101,110,115,115,108,46,111,114,103,62,0
+.p2align	4
+.data	
+.p2align	3
+L$padlock_saved_context:
+.quad	0
diff --git a/lib/accelerated/x86/asm-macosx/padlock-x86-macosx.s b/lib/accelerated/x86/asm-macosx/padlock-x86-macosx.s
new file mode 100644
index 0000000000..b3822f424d
--- /dev/null
+++ b/lib/accelerated/x86/asm-macosx/padlock-x86-macosx.s
@@ -0,0 +1,591 @@
+# Copyright (c) 2011, Andy Polyakov by <appro@openssl.org>
+# All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions
+# are met:
+# 
+#     * Redistributions of source code must retain copyright notices,
+#      this list of conditions and the following disclaimer.
+#
+#     * 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.
+#
+#     * Neither the name of the Andy Polyakov nor the names of its
+#      copyright holder and contributors may be used to endorse or
+#      promote products derived from this software without specific
+#      prior written permission.
+#
+# ALTERNATIVELY, provided that this notice is retained in full, this
+# product may be distributed under the terms of the GNU General Public
+# License (GPL), in which case the provisions of the GPL apply INSTEAD OF
+# those given above.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDER AND CONTRIBUTORS
+# "AS IS" AND ANY EXPRESS 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 COPYRIGHT
+# OWNER OR 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.
+
+.file	"devel/perlasm/e_padlock-x86.s"
+.text
+.globl	_padlock_capability
+.align	4
+_padlock_capability:
+L_padlock_capability_begin:
+	pushl	%ebx
+	pushfl
+	popl	%eax
+	movl	%eax,%ecx
+	xorl	$2097152,%eax
+	pushl	%eax
+	popfl
+	pushfl
+	popl	%eax
+	xorl	%eax,%ecx
+	xorl	%eax,%eax
+	btl	$21,%ecx
+	jnc	L000noluck
+	.byte	0x0f,0xa2
+	xorl	%eax,%eax
+	cmpl	$0x746e6543,%ebx
+	jne	L000noluck
+	cmpl	$0x48727561,%edx
+	jne	L000noluck
+	cmpl	$0x736c7561,%ecx
+	jne	L000noluck
+	movl	$3221225472,%eax
+	.byte	0x0f,0xa2
+	movl	%eax,%edx
+	xorl	%eax,%eax
+	cmpl	$3221225473,%edx
+	jb	L000noluck
+	movl	$1,%eax
+	.byte	0x0f,0xa2
+	orl	$15,%eax
+	xorl	%ebx,%ebx
+	andl	$4095,%eax
+	cmpl	$1791,%eax
+	sete	%bl
+	movl	$3221225473,%eax
+	pushl	%ebx
+	.byte	0x0f,0xa2
+	popl	%ebx
+	movl	%edx,%eax
+	shll	$4,%ebx
+	andl	$4294967279,%eax
+	orl	%ebx,%eax
+L000noluck:
+	popl	%ebx
+	ret
+.globl	_padlock_key_bswap
+.align	4
+_padlock_key_bswap:
+L_padlock_key_bswap_begin:
+	movl	4(%esp),%edx
+	movl	240(%edx),%ecx
+L001bswap_loop:
+	movl	(%edx),%eax
+	bswap	%eax
+	movl	%eax,(%edx)
+	leal	4(%edx),%edx
+	subl	$1,%ecx
+	jnz	L001bswap_loop
+	ret
+.globl	_padlock_verify_context
+.align	4
+_padlock_verify_context:
+L_padlock_verify_context_begin:
+	movl	4(%esp),%edx
+	leal	Lpadlock_saved_context-L002verify_pic_point,%eax
+	pushfl
+	call	__padlock_verify_ctx
+L002verify_pic_point:
+	leal	4(%esp),%esp
+	ret
+.align	4
+__padlock_verify_ctx:
+	addl	(%esp),%eax
+	btl	$30,4(%esp)
+	jnc	L003verified
+	cmpl	(%eax),%edx
+	je	L003verified
+	pushfl
+	popfl
+L003verified:
+	movl	%edx,(%eax)
+	ret
+.globl	_padlock_reload_key
+.align	4
+_padlock_reload_key:
+L_padlock_reload_key_begin:
+	pushfl
+	popfl
+	ret
+.globl	_padlock_aes_block
+.align	4
+_padlock_aes_block:
+L_padlock_aes_block_begin:
+	pushl	%edi
+	pushl	%esi
+	pushl	%ebx
+	movl	16(%esp),%edi
+	movl	20(%esp),%esi
+	movl	24(%esp),%edx
+	movl	$1,%ecx
+	leal	32(%edx),%ebx
+	leal	16(%edx),%edx
+.byte	243,15,167,200
+	popl	%ebx
+	popl	%esi
+	popl	%edi
+	ret
+.globl	_padlock_ecb_encrypt
+.align	4
+_padlock_ecb_encrypt:
+L_padlock_ecb_encrypt_begin:
+	pushl	%ebp
+	pushl	%ebx
+	pushl	%esi
+	pushl	%edi
+	movl	20(%esp),%edi
+	movl	24(%esp),%esi
+	movl	28(%esp),%edx
+	movl	32(%esp),%ecx
+	testl	$15,%edx
+	jnz	L004ecb_abort
+	testl	$15,%ecx
+	jnz	L004ecb_abort
+	leal	Lpadlock_saved_context-L005ecb_pic_point,%eax
+	pushfl
+	cld
+	call	__padlock_verify_ctx
+L005ecb_pic_point:
+	leal	16(%edx),%edx
+	xorl	%eax,%eax
+	xorl	%ebx,%ebx
+	cmpl	$128,%ecx
+	jbe	L006ecb_short
+	testl	$32,(%edx)
+	jnz	L007ecb_aligned
+	testl	$15,%edi
+	setz	%al
+	testl	$15,%esi
+	setz	%bl
+	testl	%ebx,%eax
+	jnz	L007ecb_aligned
+	negl	%eax
+	movl	$512,%ebx
+	notl	%eax
+	leal	-24(%esp),%ebp
+	cmpl	%ebx,%ecx
+	cmovcl	%ecx,%ebx
+	andl	%ebx,%eax
+	movl	%ecx,%ebx
+	negl	%eax
+	andl	$511,%ebx
+	leal	(%eax,%ebp,1),%esp
+	andl	$-16,%esp
+	jmp	L008ecb_loop
+.align	4,0x90
+L008ecb_loop:
+	movl	%edi,(%ebp)
+	movl	%esi,4(%ebp)
+	movl	%ecx,8(%ebp)
+	movl	%ebx,%ecx
+	movl	%ebx,12(%ebp)
+	testl	$15,%edi
+	cmovnzl	%esp,%edi
+	testl	$15,%esi
+	jz	L009ecb_inp_aligned
+	shrl	$2,%ecx
+.byte	243,165
+	subl	%ebx,%edi
+	movl	%ebx,%ecx
+	movl	%edi,%esi
+L009ecb_inp_aligned:
+	leal	-16(%edx),%eax
+	leal	16(%edx),%ebx
+	shrl	$4,%ecx
+.byte	243,15,167,200
+	movl	(%ebp),%edi
+	movl	12(%ebp),%ebx
+	testl	$15,%edi
+	jz	L010ecb_out_aligned
+	movl	%ebx,%ecx
+	shrl	$2,%ecx
+	leal	(%esp),%esi
+.byte	243,165
+	subl	%ebx,%edi
+L010ecb_out_aligned:
+	movl	4(%ebp),%esi
+	movl	8(%ebp),%ecx
+	addl	%ebx,%edi
+	addl	%ebx,%esi
+	subl	%ebx,%ecx
+	movl	$512,%ebx
+	jnz	L008ecb_loop
+	cmpl	%ebp,%esp
+	je	L011ecb_done
+	pxor	%xmm0,%xmm0
+	leal	(%esp),%eax
+L012ecb_bzero:
+	movaps	%xmm0,(%eax)
+	leal	16(%eax),%eax
+	cmpl	%eax,%ebp
+	ja	L012ecb_bzero
+L011ecb_done:
+	leal	24(%ebp),%esp
+	jmp	L013ecb_exit
+.align	4,0x90
+L006ecb_short:
+	xorl	%eax,%eax
+	leal	-24(%esp),%ebp
+	subl	%ecx,%eax
+	leal	(%eax,%ebp,1),%esp
+	andl	$-16,%esp
+	xorl	%ebx,%ebx
+L014ecb_short_copy:
+	movups	(%esi,%ebx,1),%xmm0
+	leal	16(%ebx),%ebx
+	cmpl	%ebx,%ecx
+	movaps	%xmm0,-16(%esp,%ebx,1)
+	ja	L014ecb_short_copy
+	movl	%esp,%esi
+	movl	%ecx,%ebx
+	jmp	L008ecb_loop
+.align	4,0x90
+L007ecb_aligned:
+	leal	-16(%edx),%eax
+	leal	16(%edx),%ebx
+	shrl	$4,%ecx
+.byte	243,15,167,200
+L013ecb_exit:
+	movl	$1,%eax
+	leal	4(%esp),%esp
+L004ecb_abort:
+	popl	%edi
+	popl	%esi
+	popl	%ebx
+	popl	%ebp
+	ret
+.globl	_padlock_cbc_encrypt
+.align	4
+_padlock_cbc_encrypt:
+L_padlock_cbc_encrypt_begin:
+	pushl	%ebp
+	pushl	%ebx
+	pushl	%esi
+	pushl	%edi
+	movl	20(%esp),%edi
+	movl	24(%esp),%esi
+	movl	28(%esp),%edx
+	movl	32(%esp),%ecx
+	testl	$15,%edx
+	jnz	L015cbc_abort
+	testl	$15,%ecx
+	jnz	L015cbc_abort
+	leal	Lpadlock_saved_context-L016cbc_pic_point,%eax
+	pushfl
+	cld
+	call	__padlock_verify_ctx
+L016cbc_pic_point:
+	leal	16(%edx),%edx
+	xorl	%eax,%eax
+	xorl	%ebx,%ebx
+	cmpl	$64,%ecx
+	jbe	L017cbc_short
+	testl	$32,(%edx)
+	jnz	L018cbc_aligned
+	testl	$15,%edi
+	setz	%al
+	testl	$15,%esi
+	setz	%bl
+	testl	%ebx,%eax
+	jnz	L018cbc_aligned
+	negl	%eax
+	movl	$512,%ebx
+	notl	%eax
+	leal	-24(%esp),%ebp
+	cmpl	%ebx,%ecx
+	cmovcl	%ecx,%ebx
+	andl	%ebx,%eax
+	movl	%ecx,%ebx
+	negl	%eax
+	andl	$511,%ebx
+	leal	(%eax,%ebp,1),%esp
+	andl	$-16,%esp
+	jmp	L019cbc_loop
+.align	4,0x90
+L019cbc_loop:
+	movl	%edi,(%ebp)
+	movl	%esi,4(%ebp)
+	movl	%ecx,8(%ebp)
+	movl	%ebx,%ecx
+	movl	%ebx,12(%ebp)
+	testl	$15,%edi
+	cmovnzl	%esp,%edi
+	testl	$15,%esi
+	jz	L020cbc_inp_aligned
+	shrl	$2,%ecx
+.byte	243,165
+	subl	%ebx,%edi
+	movl	%ebx,%ecx
+	movl	%edi,%esi
+L020cbc_inp_aligned:
+	leal	-16(%edx),%eax
+	leal	16(%edx),%ebx
+	shrl	$4,%ecx
+.byte	243,15,167,208
+	movaps	(%eax),%xmm0
+	movaps	%xmm0,-16(%edx)
+	movl	(%ebp),%edi
+	movl	12(%ebp),%ebx
+	testl	$15,%edi
+	jz	L021cbc_out_aligned
+	movl	%ebx,%ecx
+	shrl	$2,%ecx
+	leal	(%esp),%esi
+.byte	243,165
+	subl	%ebx,%edi
+L021cbc_out_aligned:
+	movl	4(%ebp),%esi
+	movl	8(%ebp),%ecx
+	addl	%ebx,%edi
+	addl	%ebx,%esi
+	subl	%ebx,%ecx
+	movl	$512,%ebx
+	jnz	L019cbc_loop
+	cmpl	%ebp,%esp
+	je	L022cbc_done
+	pxor	%xmm0,%xmm0
+	leal	(%esp),%eax
+L023cbc_bzero:
+	movaps	%xmm0,(%eax)
+	leal	16(%eax),%eax
+	cmpl	%eax,%ebp
+	ja	L023cbc_bzero
+L022cbc_done:
+	leal	24(%ebp),%esp
+	jmp	L024cbc_exit
+.align	4,0x90
+L017cbc_short:
+	xorl	%eax,%eax
+	leal	-24(%esp),%ebp
+	subl	%ecx,%eax
+	leal	(%eax,%ebp,1),%esp
+	andl	$-16,%esp
+	xorl	%ebx,%ebx
+L025cbc_short_copy:
+	movups	(%esi,%ebx,1),%xmm0
+	leal	16(%ebx),%ebx
+	cmpl	%ebx,%ecx
+	movaps	%xmm0,-16(%esp,%ebx,1)
+	ja	L025cbc_short_copy
+	movl	%esp,%esi
+	movl	%ecx,%ebx
+	jmp	L019cbc_loop
+.align	4,0x90
+L018cbc_aligned:
+	leal	-16(%edx),%eax
+	leal	16(%edx),%ebx
+	shrl	$4,%ecx
+.byte	243,15,167,208
+	movaps	(%eax),%xmm0
+	movaps	%xmm0,-16(%edx)
+L024cbc_exit:
+	movl	$1,%eax
+	leal	4(%esp),%esp
+L015cbc_abort:
+	popl	%edi
+	popl	%esi
+	popl	%ebx
+	popl	%ebp
+	ret
+.globl	_padlock_xstore
+.align	4
+_padlock_xstore:
+L_padlock_xstore_begin:
+	pushl	%edi
+	movl	8(%esp),%edi
+	movl	12(%esp),%edx
+.byte	15,167,192
+	popl	%edi
+	ret
+.align	4
+__win32_segv_handler:
+	movl	$1,%eax
+	movl	4(%esp),%edx
+	movl	12(%esp),%ecx
+	cmpl	$3221225477,(%edx)
+	jne	L026ret
+	addl	$4,184(%ecx)
+	movl	$0,%eax
+L026ret:
+	ret
+.globl	_padlock_sha1_oneshot
+.align	4
+_padlock_sha1_oneshot:
+L_padlock_sha1_oneshot_begin:
+	pushl	%edi
+	pushl	%esi
+	xorl	%eax,%eax
+	movl	12(%esp),%edi
+	movl	16(%esp),%esi
+	movl	20(%esp),%ecx
+	movl	%esp,%edx
+	addl	$-128,%esp
+	movups	(%edi),%xmm0
+	andl	$-16,%esp
+	movl	16(%edi),%eax
+	movaps	%xmm0,(%esp)
+	movl	%esp,%edi
+	movl	%eax,16(%esp)
+	xorl	%eax,%eax
+.byte	243,15,166,200
+	movaps	(%esp),%xmm0
+	movl	16(%esp),%eax
+	movl	%edx,%esp
+	movl	12(%esp),%edi
+	movups	%xmm0,(%edi)
+	movl	%eax,16(%edi)
+	popl	%esi
+	popl	%edi
+	ret
+.globl	_padlock_sha1_blocks
+.align	4
+_padlock_sha1_blocks:
+L_padlock_sha1_blocks_begin:
+	pushl	%edi
+	pushl	%esi
+	movl	12(%esp),%edi
+	movl	16(%esp),%esi
+	movl	%esp,%edx
+	movl	20(%esp),%ecx
+	addl	$-128,%esp
+	movups	(%edi),%xmm0
+	andl	$-16,%esp
+	movl	16(%edi),%eax
+	movaps	%xmm0,(%esp)
+	movl	%esp,%edi
+	movl	%eax,16(%esp)
+	movl	$-1,%eax
+.byte	243,15,166,200
+	movaps	(%esp),%xmm0
+	movl	16(%esp),%eax
+	movl	%edx,%esp
+	movl	12(%esp),%edi
+	movups	%xmm0,(%edi)
+	movl	%eax,16(%edi)
+	popl	%esi
+	popl	%edi
+	ret
+.globl	_padlock_sha256_oneshot
+.align	4
+_padlock_sha256_oneshot:
+L_padlock_sha256_oneshot_begin:
+	pushl	%edi
+	pushl	%esi
+	xorl	%eax,%eax
+	movl	12(%esp),%edi
+	movl	16(%esp),%esi
+	movl	20(%esp),%ecx
+	movl	%esp,%edx
+	addl	$-128,%esp
+	movups	(%edi),%xmm0
+	andl	$-16,%esp
+	movups	16(%edi),%xmm1
+	movaps	%xmm0,(%esp)
+	movl	%esp,%edi
+	movaps	%xmm1,16(%esp)
+	xorl	%eax,%eax
+.byte	243,15,166,208
+	movaps	(%esp),%xmm0
+	movaps	16(%esp),%xmm1
+	movl	%edx,%esp
+	movl	12(%esp),%edi
+	movups	%xmm0,(%edi)
+	movups	%xmm1,16(%edi)
+	popl	%esi
+	popl	%edi
+	ret
+.globl	_padlock_sha256_blocks
+.align	4
+_padlock_sha256_blocks:
+L_padlock_sha256_blocks_begin:
+	pushl	%edi
+	pushl	%esi
+	movl	12(%esp),%edi
+	movl	16(%esp),%esi
+	movl	20(%esp),%ecx
+	movl	%esp,%edx
+	addl	$-128,%esp
+	movups	(%edi),%xmm0
+	andl	$-16,%esp
+	movups	16(%edi),%xmm1
+	movaps	%xmm0,(%esp)
+	movl	%esp,%edi
+	movaps	%xmm1,16(%esp)
+	movl	$-1,%eax
+.byte	243,15,166,208
+	movaps	(%esp),%xmm0
+	movaps	16(%esp),%xmm1
+	movl	%edx,%esp
+	movl	12(%esp),%edi
+	movups	%xmm0,(%edi)
+	movups	%xmm1,16(%edi)
+	popl	%esi
+	popl	%edi
+	ret
+.globl	_padlock_sha512_blocks
+.align	4
+_padlock_sha512_blocks:
+L_padlock_sha512_blocks_begin:
+	pushl	%edi
+	pushl	%esi
+	movl	12(%esp),%edi
+	movl	16(%esp),%esi
+	movl	20(%esp),%ecx
+	movl	%esp,%edx
+	addl	$-128,%esp
+	movups	(%edi),%xmm0
+	andl	$-16,%esp
+	movups	16(%edi),%xmm1
+	movups	32(%edi),%xmm2
+	movups	48(%edi),%xmm3
+	movaps	%xmm0,(%esp)
+	movl	%esp,%edi
+	movaps	%xmm1,16(%esp)
+	movaps	%xmm2,32(%esp)
+	movaps	%xmm3,48(%esp)
+.byte	243,15,166,224
+	movaps	(%esp),%xmm0
+	movaps	16(%esp),%xmm1
+	movaps	32(%esp),%xmm2
+	movaps	48(%esp),%xmm3
+	movl	%edx,%esp
+	movl	12(%esp),%edi
+	movups	%xmm0,(%edi)
+	movups	%xmm1,16(%edi)
+	movups	%xmm2,32(%edi)
+	movups	%xmm3,48(%edi)
+	popl	%esi
+	popl	%edi
+	ret
+.byte	86,73,65,32,80,97,100,108,111,99,107,32,120,56,54,32
+.byte	109,111,100,117,108,101,44,32,67,82,89,80,84,79,71,65
+.byte	77,83,32,98,121,32,60,97,112,112,114,111,64,111,112,101
+.byte	110,115,115,108,46,111,114,103,62,0
+.align	4,0x90
+.data
+.align	2,0x90
+Lpadlock_saved_context:
+.long	0