diff --git a/.gitignore b/.gitignore
index ead24b74..29f57adc 100644
--- a/.gitignore
+++ b/.gitignore
@@ -10,5 +10,4 @@ CMakeCache.txt
 cmake_install.cmake
 Makefile
 cmake-build-release/
-src/crypto/asm/*.inc
-src/crypto/asm/win/*.inc
+src/crypto/asm/*.inc
\ No newline at end of file
diff --git a/cmake/asm.cmake b/cmake/asm.cmake
index 15a265b5..74b75fc1 100644
--- a/cmake/asm.cmake
+++ b/cmake/asm.cmake
@@ -1,3 +1,5 @@
+message("Generating ASM files")
+
 # CN v1 original
 set(ALGO "original")
 set(ITERATIONS "524288") #0x80000
diff --git a/src/crypto/asm/win/cnv1_main_loop_fast_sandybridge.inc b/src/crypto/asm/win/cnv1_main_loop_fast_sandybridge.inc
new file mode 100644
index 00000000..e0fc8f14
--- /dev/null
+++ b/src/crypto/asm/win/cnv1_main_loop_fast_sandybridge.inc
@@ -0,0 +1,70 @@
+	mov	QWORD PTR [rsp+8], rbx
+	mov	QWORD PTR [rsp+16], rbp
+	mov	QWORD PTR [rsp+24], rsi
+	mov	QWORD PTR [rsp+32], rdi
+	push	r14
+	push	r15
+	mov	rax, QWORD PTR [rcx+48]
+	mov	ebp, 262144
+	xor	rax, QWORD PTR [rcx+16]
+	mov	rdx, QWORD PTR [rcx+56]
+	xor	rdx, QWORD PTR [rcx+24]
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	movq	xmm3, rax
+	mov	rax, QWORD PTR [rcx+256]
+	mov	rdi, QWORD PTR [rcx+40]
+	movq	xmm0, rdx
+	xor	rdi, QWORD PTR [rcx+8]
+	mov	rdx, r8
+	mov	r15, QWORD PTR [rcx+264]
+	and	edx, 2097136
+	mov	r14, QWORD PTR [rax+35]
+	xor	r14, QWORD PTR [rcx+192]
+	mov	rsi, QWORD PTR [rcx+224]
+	punpcklqdq xmm3, xmm0
+	movdqu	xmm2, XMMWORD PTR [rdx+rsi]
+
+	ALIGN 64
+cnv1_main_loop_fast_sandybridge:
+	movq	xmm0, rdi
+	movq	xmm1, r8
+	punpcklqdq xmm1, xmm0
+	aesenc	xmm2, xmm1
+	movq r10, xmm2
+	mov r9d, r10d
+	and r9d, 2097136
+	add r9, rsi
+	movdqa	xmm0, xmm2
+	pxor	xmm0, xmm3
+	movdqa	xmm3, xmm2
+	movdqu	XMMWORD PTR [rdx+rsi], xmm0
+	psrldq	xmm0, 11
+	movq	rax, xmm0
+	movzx	eax, al
+	movzx	eax, BYTE PTR [rax+r15]
+	mov	BYTE PTR [rsi+rdx+11], al
+	mov	rbx, QWORD PTR [r9]
+	mov	r11, QWORD PTR [r9+8]
+	mov	rax, rbx
+	mul	r10
+	add	r8, rdx
+	mov	QWORD PTR [r9], r8
+	add	rdi, rax
+	mov	rax, r14
+	xor	rax, rdi
+	mov	QWORD PTR [r9+8], rax
+	xor	r8, rbx
+	mov	rdx, r8
+	and	edx, 2097136
+	movdqu	xmm2, XMMWORD PTR [rdx+rsi]
+	xor	rdi, r11
+	dec	ebp
+	jne	cnv1_main_loop_fast_sandybridge
+
+	mov	rbx, QWORD PTR [rsp+24]
+	mov	rbp, QWORD PTR [rsp+32]
+	mov	rsi, QWORD PTR [rsp+40]
+	mov	rdi, QWORD PTR [rsp+48]
+	pop	r15
+	pop	r14
diff --git a/src/crypto/asm/win/cnv1_main_loop_fast_soft_aes_sandybridge.inc b/src/crypto/asm/win/cnv1_main_loop_fast_soft_aes_sandybridge.inc
new file mode 100644
index 00000000..9e09a83d
--- /dev/null
+++ b/src/crypto/asm/win/cnv1_main_loop_fast_soft_aes_sandybridge.inc
@@ -0,0 +1,162 @@
+	push	rbx
+	push	rbp
+	push	rsi
+	push	rdi
+	push	r12
+	push	r13
+	push	r14
+	push	r15
+	sub	rsp, 72
+
+	movaps XMMWORD PTR [rsp], xmm6
+	movaps XMMWORD PTR [rsp+16], xmm7
+	movaps XMMWORD PTR [rsp+32], xmm8
+	movaps XMMWORD PTR [rsp+48], xmm9
+
+	mov	rax, QWORD PTR [rcx+48]
+	xor	rax, QWORD PTR [rcx+16]
+	mov	rdx, QWORD PTR [rcx+56]
+	xor	rdx, QWORD PTR [rcx+24]
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	movq	xmm4, rax
+	mov	rax, QWORD PTR [rcx+256]
+	mov	r13, QWORD PTR [rcx+40]
+	movq	xmm0, rdx
+	xor	r13, QWORD PTR [rcx+8]
+	mov	rdx, r8
+	mov	rdi, QWORD PTR [rcx+224]
+	and	edx, 2097136
+	mov	rax, QWORD PTR [rax+35]
+	xor	rax, QWORD PTR [rcx+192]
+	movq	xmm5, rax
+	movq	xmm8, rdi
+	punpcklqdq xmm4, xmm0
+	mov	QWORD PTR [rsp+64], rdx
+
+	movq xmm6, rcx
+	mov rax, QWORD PTR [rcx+264]
+	movq xmm7, rax
+
+	mov eax, 262144
+
+	ALIGN 64
+cnv1_main_loop_fast_soft_aes_sandybridge:
+	movq xmm9, rax
+	mov	r12, QWORD PTR [rcx+272]
+	mov	esi, DWORD PTR [rdx+rdi]
+	mov	r10d, DWORD PTR [rdx+rdi+4]
+	mov	ebp, DWORD PTR [rdx+rdi+12]
+	mov	r14d, DWORD PTR [rdx+rdi+8]
+	mov	rdx, QWORD PTR [rsp+64]
+	movzx	ecx, sil
+	shr	esi, 8
+	mov	r15d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r10b
+	shr	r10d, 8
+	mov	edi, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r14b
+	shr	r14d, 8
+	mov	ebx, DWORD PTR [r12+rcx*4]
+	movzx	ecx, bpl
+	shr	ebp, 8
+	mov	r9d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r10b
+	shr	r10d, 8
+	xor	r15d, DWORD PTR [r12+rcx*4+1024]
+	movzx	ecx, r14b
+	shr	r14d, 8
+	mov	eax, r14d
+	shr	eax, 8
+	xor	edi, DWORD PTR [r12+rcx*4+1024]
+	add	eax, 256
+	movzx	ecx, bpl
+	shr	ebp, 8
+	xor	ebx, DWORD PTR [r12+rcx*4+1024]
+	movzx	ecx, sil
+	shr	esi, 8
+	xor	r9d, DWORD PTR [r12+rcx*4+1024]
+	add	r12, 2048
+	movzx	ecx, r10b
+	shr	r10d, 8
+	add	r10d, 256
+	mov	r11d, DWORD PTR [r12+rax*4]
+	xor	r11d, DWORD PTR [r12+rcx*4]
+	xor	r11d, r9d
+	movzx	ecx, sil
+	mov	r10d, DWORD PTR [r12+r10*4]
+	shr	esi, 8
+	add	esi, 256
+	xor	r10d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, bpl
+	xor	r10d, ebx
+	shr	ebp, 8
+	add	ebp, 256
+	movd	xmm1, r11d
+	mov	r9d, DWORD PTR [r12+rcx*4]
+	xor	r9d, DWORD PTR [r12+rsi*4]
+	mov	eax, DWORD PTR [r12+rbp*4]
+	xor	r9d, edi
+	movq rdi, xmm8
+	movzx	ecx, r14b
+	movd	xmm0, r10d
+	movd	xmm2, r9d
+	punpckldq xmm2, xmm1
+	movq	xmm1, r8
+	xor	eax, DWORD PTR [r12+rcx*4]
+	xor	eax, r15d
+	movd	xmm3, eax
+	movq	rax, xmm7
+	punpckldq xmm3, xmm0
+	movq	xmm0, r13
+	punpcklqdq xmm1, xmm0
+	punpckldq xmm3, xmm2
+	pxor	xmm3, xmm1
+	movq	r9, xmm3
+	mov	r10d, r9d
+	and	r10d, 2097136
+	movdqa	xmm0, xmm3
+	pxor	xmm0, xmm4
+	movdqu	XMMWORD PTR [rdx+rdi], xmm0
+	psrldq	xmm0, 11
+	movq	rcx, xmm0
+	movzx	ecx, cl
+	mov	cl, BYTE PTR [rcx+rax]
+	mov	BYTE PTR [rdi+rdx+11], cl
+	mov	rbx, QWORD PTR [r10+rdi]
+	mov rcx, r9
+	lea	r9, QWORD PTR [r10+rdi]
+	mov	r11, QWORD PTR [r9+8]
+	mov	rax, rbx
+	movdqa	xmm4, xmm3
+	mul	rcx
+	movq	rcx, xmm6
+	add	r8, rdx
+	add	r13, rax
+	movq	rax, xmm5
+	xor	rax, r13
+	mov	QWORD PTR [r9], r8
+	xor	r8, rbx
+	mov	QWORD PTR [r9+8], rax
+	movq rax, xmm9
+	mov	rdx, r8
+	xor	r13, r11
+	and	edx, 2097136
+	mov	QWORD PTR [rsp+64], rdx
+	sub eax, 1
+	jne	cnv1_main_loop_fast_soft_aes_sandybridge
+
+	movaps xmm6, XMMWORD PTR [rsp]
+	movaps xmm7, XMMWORD PTR [rsp+16]
+	movaps xmm8, XMMWORD PTR [rsp+32]
+	movaps xmm9, XMMWORD PTR [rsp+48]
+
+	add	rsp, 72
+	pop	r15
+	pop	r14
+	pop	r13
+	pop	r12
+	pop	rdi
+	pop	rsi
+	pop	rbp
+	pop	rbx
diff --git a/src/crypto/asm/win/cnv1_main_loop_lite_sandybridge.inc b/src/crypto/asm/win/cnv1_main_loop_lite_sandybridge.inc
new file mode 100644
index 00000000..4f369c7a
--- /dev/null
+++ b/src/crypto/asm/win/cnv1_main_loop_lite_sandybridge.inc
@@ -0,0 +1,70 @@
+	mov	QWORD PTR [rsp+8], rbx
+	mov	QWORD PTR [rsp+16], rbp
+	mov	QWORD PTR [rsp+24], rsi
+	mov	QWORD PTR [rsp+32], rdi
+	push	r14
+	push	r15
+	mov	rax, QWORD PTR [rcx+48]
+	mov	ebp, 262144
+	xor	rax, QWORD PTR [rcx+16]
+	mov	rdx, QWORD PTR [rcx+56]
+	xor	rdx, QWORD PTR [rcx+24]
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	movq	xmm3, rax
+	mov	rax, QWORD PTR [rcx+256]
+	mov	rdi, QWORD PTR [rcx+40]
+	movq	xmm0, rdx
+	xor	rdi, QWORD PTR [rcx+8]
+	mov	rdx, r8
+	mov	r15, QWORD PTR [rcx+264]
+	and	edx, 1048560
+	mov	r14, QWORD PTR [rax+35]
+	xor	r14, QWORD PTR [rcx+192]
+	mov	rsi, QWORD PTR [rcx+224]
+	punpcklqdq xmm3, xmm0
+	movdqu	xmm2, XMMWORD PTR [rdx+rsi]
+
+	ALIGN 64
+cnv1_main_loop_lite_sandybridge:
+	movq	xmm0, rdi
+	movq	xmm1, r8
+	punpcklqdq xmm1, xmm0
+	aesenc	xmm2, xmm1
+	movq r10, xmm2
+	mov r9d, r10d
+	and r9d, 1048560
+	add r9, rsi
+	movdqa	xmm0, xmm2
+	pxor	xmm0, xmm3
+	movdqa	xmm3, xmm2
+	movdqu	XMMWORD PTR [rdx+rsi], xmm0
+	psrldq	xmm0, 11
+	movq	rax, xmm0
+	movzx	eax, al
+	movzx	eax, BYTE PTR [rax+r15]
+	mov	BYTE PTR [rsi+rdx+11], al
+	mov	rbx, QWORD PTR [r9]
+	mov	r11, QWORD PTR [r9+8]
+	mov	rax, rbx
+	mul	r10
+	add	r8, rdx
+	mov	QWORD PTR [r9], r8
+	add	rdi, rax
+	mov	rax, r14
+	xor	rax, rdi
+	mov	QWORD PTR [r9+8], rax
+	xor	r8, rbx
+	mov	rdx, r8
+	and	edx, 1048560
+	movdqu	xmm2, XMMWORD PTR [rdx+rsi]
+	xor	rdi, r11
+	dec	ebp
+	jne	cnv1_main_loop_lite_sandybridge
+
+	mov	rbx, QWORD PTR [rsp+24]
+	mov	rbp, QWORD PTR [rsp+32]
+	mov	rsi, QWORD PTR [rsp+40]
+	mov	rdi, QWORD PTR [rsp+48]
+	pop	r15
+	pop	r14
diff --git a/src/crypto/asm/win/cnv1_main_loop_lite_soft_aes_sandybridge.inc b/src/crypto/asm/win/cnv1_main_loop_lite_soft_aes_sandybridge.inc
new file mode 100644
index 00000000..7dc9ac8a
--- /dev/null
+++ b/src/crypto/asm/win/cnv1_main_loop_lite_soft_aes_sandybridge.inc
@@ -0,0 +1,162 @@
+	push	rbx
+	push	rbp
+	push	rsi
+	push	rdi
+	push	r12
+	push	r13
+	push	r14
+	push	r15
+	sub	rsp, 72
+
+	movaps XMMWORD PTR [rsp], xmm6
+	movaps XMMWORD PTR [rsp+16], xmm7
+	movaps XMMWORD PTR [rsp+32], xmm8
+	movaps XMMWORD PTR [rsp+48], xmm9
+
+	mov	rax, QWORD PTR [rcx+48]
+	xor	rax, QWORD PTR [rcx+16]
+	mov	rdx, QWORD PTR [rcx+56]
+	xor	rdx, QWORD PTR [rcx+24]
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	movq	xmm4, rax
+	mov	rax, QWORD PTR [rcx+256]
+	mov	r13, QWORD PTR [rcx+40]
+	movq	xmm0, rdx
+	xor	r13, QWORD PTR [rcx+8]
+	mov	rdx, r8
+	mov	rdi, QWORD PTR [rcx+224]
+	and	edx, 1048560
+	mov	rax, QWORD PTR [rax+35]
+	xor	rax, QWORD PTR [rcx+192]
+	movq	xmm5, rax
+	movq	xmm8, rdi
+	punpcklqdq xmm4, xmm0
+	mov	QWORD PTR [rsp+64], rdx
+
+	movq xmm6, rcx
+	mov rax, QWORD PTR [rcx+264]
+	movq xmm7, rax
+
+	mov eax, 262144
+
+	ALIGN 64
+cnv1_main_loop_lite_soft_aes_sandybridge:
+	movq xmm9, rax
+	mov	r12, QWORD PTR [rcx+272]
+	mov	esi, DWORD PTR [rdx+rdi]
+	mov	r10d, DWORD PTR [rdx+rdi+4]
+	mov	ebp, DWORD PTR [rdx+rdi+12]
+	mov	r14d, DWORD PTR [rdx+rdi+8]
+	mov	rdx, QWORD PTR [rsp+64]
+	movzx	ecx, sil
+	shr	esi, 8
+	mov	r15d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r10b
+	shr	r10d, 8
+	mov	edi, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r14b
+	shr	r14d, 8
+	mov	ebx, DWORD PTR [r12+rcx*4]
+	movzx	ecx, bpl
+	shr	ebp, 8
+	mov	r9d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r10b
+	shr	r10d, 8
+	xor	r15d, DWORD PTR [r12+rcx*4+1024]
+	movzx	ecx, r14b
+	shr	r14d, 8
+	mov	eax, r14d
+	shr	eax, 8
+	xor	edi, DWORD PTR [r12+rcx*4+1024]
+	add	eax, 256
+	movzx	ecx, bpl
+	shr	ebp, 8
+	xor	ebx, DWORD PTR [r12+rcx*4+1024]
+	movzx	ecx, sil
+	shr	esi, 8
+	xor	r9d, DWORD PTR [r12+rcx*4+1024]
+	add	r12, 2048
+	movzx	ecx, r10b
+	shr	r10d, 8
+	add	r10d, 256
+	mov	r11d, DWORD PTR [r12+rax*4]
+	xor	r11d, DWORD PTR [r12+rcx*4]
+	xor	r11d, r9d
+	movzx	ecx, sil
+	mov	r10d, DWORD PTR [r12+r10*4]
+	shr	esi, 8
+	add	esi, 256
+	xor	r10d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, bpl
+	xor	r10d, ebx
+	shr	ebp, 8
+	add	ebp, 256
+	movd	xmm1, r11d
+	mov	r9d, DWORD PTR [r12+rcx*4]
+	xor	r9d, DWORD PTR [r12+rsi*4]
+	mov	eax, DWORD PTR [r12+rbp*4]
+	xor	r9d, edi
+	movq rdi, xmm8
+	movzx	ecx, r14b
+	movd	xmm0, r10d
+	movd	xmm2, r9d
+	punpckldq xmm2, xmm1
+	movq	xmm1, r8
+	xor	eax, DWORD PTR [r12+rcx*4]
+	xor	eax, r15d
+	movd	xmm3, eax
+	movq	rax, xmm7
+	punpckldq xmm3, xmm0
+	movq	xmm0, r13
+	punpcklqdq xmm1, xmm0
+	punpckldq xmm3, xmm2
+	pxor	xmm3, xmm1
+	movq	r9, xmm3
+	mov	r10d, r9d
+	and	r10d, 1048560
+	movdqa	xmm0, xmm3
+	pxor	xmm0, xmm4
+	movdqu	XMMWORD PTR [rdx+rdi], xmm0
+	psrldq	xmm0, 11
+	movq	rcx, xmm0
+	movzx	ecx, cl
+	mov	cl, BYTE PTR [rcx+rax]
+	mov	BYTE PTR [rdi+rdx+11], cl
+	mov	rbx, QWORD PTR [r10+rdi]
+	mov rcx, r9
+	lea	r9, QWORD PTR [r10+rdi]
+	mov	r11, QWORD PTR [r9+8]
+	mov	rax, rbx
+	movdqa	xmm4, xmm3
+	mul	rcx
+	movq	rcx, xmm6
+	add	r8, rdx
+	add	r13, rax
+	movq	rax, xmm5
+	xor	rax, r13
+	mov	QWORD PTR [r9], r8
+	xor	r8, rbx
+	mov	QWORD PTR [r9+8], rax
+	movq rax, xmm9
+	mov	rdx, r8
+	xor	r13, r11
+	and	edx, 1048560
+	mov	QWORD PTR [rsp+64], rdx
+	sub eax, 1
+	jne	cnv1_main_loop_lite_soft_aes_sandybridge
+
+	movaps xmm6, XMMWORD PTR [rsp]
+	movaps xmm7, XMMWORD PTR [rsp+16]
+	movaps xmm8, XMMWORD PTR [rsp+32]
+	movaps xmm9, XMMWORD PTR [rsp+48]
+
+	add	rsp, 72
+	pop	r15
+	pop	r14
+	pop	r13
+	pop	r12
+	pop	rdi
+	pop	rsi
+	pop	rbp
+	pop	rbx
diff --git a/src/crypto/asm/win/cnv1_main_loop_sandybridge.inc b/src/crypto/asm/win/cnv1_main_loop_sandybridge.inc
new file mode 100644
index 00000000..e48183da
--- /dev/null
+++ b/src/crypto/asm/win/cnv1_main_loop_sandybridge.inc
@@ -0,0 +1,70 @@
+	mov	QWORD PTR [rsp+8], rbx
+	mov	QWORD PTR [rsp+16], rbp
+	mov	QWORD PTR [rsp+24], rsi
+	mov	QWORD PTR [rsp+32], rdi
+	push	r14
+	push	r15
+	mov	rax, QWORD PTR [rcx+48]
+	mov	ebp, 524288
+	xor	rax, QWORD PTR [rcx+16]
+	mov	rdx, QWORD PTR [rcx+56]
+	xor	rdx, QWORD PTR [rcx+24]
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	movq	xmm3, rax
+	mov	rax, QWORD PTR [rcx+256]
+	mov	rdi, QWORD PTR [rcx+40]
+	movq	xmm0, rdx
+	xor	rdi, QWORD PTR [rcx+8]
+	mov	rdx, r8
+	mov	r15, QWORD PTR [rcx+264]
+	and	edx, 2097136
+	mov	r14, QWORD PTR [rax+35]
+	xor	r14, QWORD PTR [rcx+192]
+	mov	rsi, QWORD PTR [rcx+224]
+	punpcklqdq xmm3, xmm0
+	movdqu	xmm2, XMMWORD PTR [rdx+rsi]
+
+	ALIGN 64
+cnv1_main_loop_original_sandybridge:
+	movq	xmm0, rdi
+	movq	xmm1, r8
+	punpcklqdq xmm1, xmm0
+	aesenc	xmm2, xmm1
+	movq r10, xmm2
+	mov r9d, r10d
+	and r9d, 2097136
+	add r9, rsi
+	movdqa	xmm0, xmm2
+	pxor	xmm0, xmm3
+	movdqa	xmm3, xmm2
+	movdqu	XMMWORD PTR [rdx+rsi], xmm0
+	psrldq	xmm0, 11
+	movq	rax, xmm0
+	movzx	eax, al
+	movzx	eax, BYTE PTR [rax+r15]
+	mov	BYTE PTR [rsi+rdx+11], al
+	mov	rbx, QWORD PTR [r9]
+	mov	r11, QWORD PTR [r9+8]
+	mov	rax, rbx
+	mul	r10
+	add	r8, rdx
+	mov	QWORD PTR [r9], r8
+	add	rdi, rax
+	mov	rax, r14
+	xor	rax, rdi
+	mov	QWORD PTR [r9+8], rax
+	xor	r8, rbx
+	mov	rdx, r8
+	and	edx, 2097136
+	movdqu	xmm2, XMMWORD PTR [rdx+rsi]
+	xor	rdi, r11
+	dec	ebp
+	jne	cnv1_main_loop_original_sandybridge
+
+	mov	rbx, QWORD PTR [rsp+24]
+	mov	rbp, QWORD PTR [rsp+32]
+	mov	rsi, QWORD PTR [rsp+40]
+	mov	rdi, QWORD PTR [rsp+48]
+	pop	r15
+	pop	r14
diff --git a/src/crypto/asm/win/cnv1_main_loop_soft_aes_sandybridge.inc b/src/crypto/asm/win/cnv1_main_loop_soft_aes_sandybridge.inc
new file mode 100644
index 00000000..5fd5aecb
--- /dev/null
+++ b/src/crypto/asm/win/cnv1_main_loop_soft_aes_sandybridge.inc
@@ -0,0 +1,162 @@
+	push	rbx
+	push	rbp
+	push	rsi
+	push	rdi
+	push	r12
+	push	r13
+	push	r14
+	push	r15
+	sub	rsp, 72
+
+	movaps XMMWORD PTR [rsp], xmm6
+	movaps XMMWORD PTR [rsp+16], xmm7
+	movaps XMMWORD PTR [rsp+32], xmm8
+	movaps XMMWORD PTR [rsp+48], xmm9
+
+	mov	rax, QWORD PTR [rcx+48]
+	xor	rax, QWORD PTR [rcx+16]
+	mov	rdx, QWORD PTR [rcx+56]
+	xor	rdx, QWORD PTR [rcx+24]
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	movq	xmm4, rax
+	mov	rax, QWORD PTR [rcx+256]
+	mov	r13, QWORD PTR [rcx+40]
+	movq	xmm0, rdx
+	xor	r13, QWORD PTR [rcx+8]
+	mov	rdx, r8
+	mov	rdi, QWORD PTR [rcx+224]
+	and	edx, 2097136
+	mov	rax, QWORD PTR [rax+35]
+	xor	rax, QWORD PTR [rcx+192]
+	movq	xmm5, rax
+	movq	xmm8, rdi
+	punpcklqdq xmm4, xmm0
+	mov	QWORD PTR [rsp+64], rdx
+
+	movq xmm6, rcx
+	mov rax, QWORD PTR [rcx+264]
+	movq xmm7, rax
+
+	mov eax, 524288
+
+	ALIGN 64
+cnv1_main_loop_original_soft_aes_sandybridge:
+	movq xmm9, rax
+	mov	r12, QWORD PTR [rcx+272]
+	mov	esi, DWORD PTR [rdx+rdi]
+	mov	r10d, DWORD PTR [rdx+rdi+4]
+	mov	ebp, DWORD PTR [rdx+rdi+12]
+	mov	r14d, DWORD PTR [rdx+rdi+8]
+	mov	rdx, QWORD PTR [rsp+64]
+	movzx	ecx, sil
+	shr	esi, 8
+	mov	r15d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r10b
+	shr	r10d, 8
+	mov	edi, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r14b
+	shr	r14d, 8
+	mov	ebx, DWORD PTR [r12+rcx*4]
+	movzx	ecx, bpl
+	shr	ebp, 8
+	mov	r9d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r10b
+	shr	r10d, 8
+	xor	r15d, DWORD PTR [r12+rcx*4+1024]
+	movzx	ecx, r14b
+	shr	r14d, 8
+	mov	eax, r14d
+	shr	eax, 8
+	xor	edi, DWORD PTR [r12+rcx*4+1024]
+	add	eax, 256
+	movzx	ecx, bpl
+	shr	ebp, 8
+	xor	ebx, DWORD PTR [r12+rcx*4+1024]
+	movzx	ecx, sil
+	shr	esi, 8
+	xor	r9d, DWORD PTR [r12+rcx*4+1024]
+	add	r12, 2048
+	movzx	ecx, r10b
+	shr	r10d, 8
+	add	r10d, 256
+	mov	r11d, DWORD PTR [r12+rax*4]
+	xor	r11d, DWORD PTR [r12+rcx*4]
+	xor	r11d, r9d
+	movzx	ecx, sil
+	mov	r10d, DWORD PTR [r12+r10*4]
+	shr	esi, 8
+	add	esi, 256
+	xor	r10d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, bpl
+	xor	r10d, ebx
+	shr	ebp, 8
+	add	ebp, 256
+	movd	xmm1, r11d
+	mov	r9d, DWORD PTR [r12+rcx*4]
+	xor	r9d, DWORD PTR [r12+rsi*4]
+	mov	eax, DWORD PTR [r12+rbp*4]
+	xor	r9d, edi
+	movq rdi, xmm8
+	movzx	ecx, r14b
+	movd	xmm0, r10d
+	movd	xmm2, r9d
+	punpckldq xmm2, xmm1
+	movq	xmm1, r8
+	xor	eax, DWORD PTR [r12+rcx*4]
+	xor	eax, r15d
+	movd	xmm3, eax
+	movq	rax, xmm7
+	punpckldq xmm3, xmm0
+	movq	xmm0, r13
+	punpcklqdq xmm1, xmm0
+	punpckldq xmm3, xmm2
+	pxor	xmm3, xmm1
+	movq	r9, xmm3
+	mov	r10d, r9d
+	and	r10d, 2097136
+	movdqa	xmm0, xmm3
+	pxor	xmm0, xmm4
+	movdqu	XMMWORD PTR [rdx+rdi], xmm0
+	psrldq	xmm0, 11
+	movq	rcx, xmm0
+	movzx	ecx, cl
+	mov	cl, BYTE PTR [rcx+rax]
+	mov	BYTE PTR [rdi+rdx+11], cl
+	mov	rbx, QWORD PTR [r10+rdi]
+	mov rcx, r9
+	lea	r9, QWORD PTR [r10+rdi]
+	mov	r11, QWORD PTR [r9+8]
+	mov	rax, rbx
+	movdqa	xmm4, xmm3
+	mul	rcx
+	movq	rcx, xmm6
+	add	r8, rdx
+	add	r13, rax
+	movq	rax, xmm5
+	xor	rax, r13
+	mov	QWORD PTR [r9], r8
+	xor	r8, rbx
+	mov	QWORD PTR [r9+8], rax
+	movq rax, xmm9
+	mov	rdx, r8
+	xor	r13, r11
+	and	edx, 2097136
+	mov	QWORD PTR [rsp+64], rdx
+	sub eax, 1
+	jne	cnv1_main_loop_original_soft_aes_sandybridge
+
+	movaps xmm6, XMMWORD PTR [rsp]
+	movaps xmm7, XMMWORD PTR [rsp+16]
+	movaps xmm8, XMMWORD PTR [rsp+32]
+	movaps xmm9, XMMWORD PTR [rsp+48]
+
+	add	rsp, 72
+	pop	r15
+	pop	r14
+	pop	r13
+	pop	r12
+	pop	rdi
+	pop	rsi
+	pop	rbp
+	pop	rbx
diff --git a/src/crypto/asm/win/cnv1_main_loop_upx_sandybridge.inc b/src/crypto/asm/win/cnv1_main_loop_upx_sandybridge.inc
new file mode 100644
index 00000000..b8beee2a
--- /dev/null
+++ b/src/crypto/asm/win/cnv1_main_loop_upx_sandybridge.inc
@@ -0,0 +1,70 @@
+	mov	QWORD PTR [rsp+8], rbx
+	mov	QWORD PTR [rsp+16], rbp
+	mov	QWORD PTR [rsp+24], rsi
+	mov	QWORD PTR [rsp+32], rdi
+	push	r14
+	push	r15
+	mov	rax, QWORD PTR [rcx+48]
+	mov	ebp, 131072
+	xor	rax, QWORD PTR [rcx+16]
+	mov	rdx, QWORD PTR [rcx+56]
+	xor	rdx, QWORD PTR [rcx+24]
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	movq	xmm3, rax
+	mov	rax, QWORD PTR [rcx+256]
+	mov	rdi, QWORD PTR [rcx+40]
+	movq	xmm0, rdx
+	xor	rdi, QWORD PTR [rcx+8]
+	mov	rdx, r8
+	mov	r15, QWORD PTR [rcx+264]
+	and	edx, 1048560
+	mov	r14, QWORD PTR [rax+35]
+	xor	r14, QWORD PTR [rcx+192]
+	mov	rsi, QWORD PTR [rcx+224]
+	punpcklqdq xmm3, xmm0
+	movdqu	xmm2, XMMWORD PTR [rdx+rsi]
+
+	ALIGN 64
+cnv1_main_loop_upx_sandybridge:
+	movq	xmm0, rdi
+	movq	xmm1, r8
+	punpcklqdq xmm1, xmm0
+	aesenc	xmm2, xmm1
+	movq r10, xmm2
+	mov r9d, r10d
+	and r9d, 1048560
+	add r9, rsi
+	movdqa	xmm0, xmm2
+	pxor	xmm0, xmm3
+	movdqa	xmm3, xmm2
+	movdqu	XMMWORD PTR [rdx+rsi], xmm0
+	psrldq	xmm0, 11
+	movq	rax, xmm0
+	movzx	eax, al
+	movzx	eax, BYTE PTR [rax+r15]
+	mov	BYTE PTR [rsi+rdx+11], al
+	mov	rbx, QWORD PTR [r9]
+	mov	r11, QWORD PTR [r9+8]
+	mov	rax, rbx
+	mul	r10
+	add	r8, rdx
+	mov	QWORD PTR [r9], r8
+	add	rdi, rax
+	mov	rax, r14
+	xor	rax, rdi
+	mov	QWORD PTR [r9+8], rax
+	xor	r8, rbx
+	mov	rdx, r8
+	and	edx, 1048560
+	movdqu	xmm2, XMMWORD PTR [rdx+rsi]
+	xor	rdi, r11
+	dec	ebp
+	jne	cnv1_main_loop_upx_sandybridge
+
+	mov	rbx, QWORD PTR [rsp+24]
+	mov	rbp, QWORD PTR [rsp+32]
+	mov	rsi, QWORD PTR [rsp+40]
+	mov	rdi, QWORD PTR [rsp+48]
+	pop	r15
+	pop	r14
diff --git a/src/crypto/asm/win/cnv1_main_loop_upx_soft_aes_sandybridge.inc b/src/crypto/asm/win/cnv1_main_loop_upx_soft_aes_sandybridge.inc
new file mode 100644
index 00000000..13351f22
--- /dev/null
+++ b/src/crypto/asm/win/cnv1_main_loop_upx_soft_aes_sandybridge.inc
@@ -0,0 +1,162 @@
+	push	rbx
+	push	rbp
+	push	rsi
+	push	rdi
+	push	r12
+	push	r13
+	push	r14
+	push	r15
+	sub	rsp, 72
+
+	movaps XMMWORD PTR [rsp], xmm6
+	movaps XMMWORD PTR [rsp+16], xmm7
+	movaps XMMWORD PTR [rsp+32], xmm8
+	movaps XMMWORD PTR [rsp+48], xmm9
+
+	mov	rax, QWORD PTR [rcx+48]
+	xor	rax, QWORD PTR [rcx+16]
+	mov	rdx, QWORD PTR [rcx+56]
+	xor	rdx, QWORD PTR [rcx+24]
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	movq	xmm4, rax
+	mov	rax, QWORD PTR [rcx+256]
+	mov	r13, QWORD PTR [rcx+40]
+	movq	xmm0, rdx
+	xor	r13, QWORD PTR [rcx+8]
+	mov	rdx, r8
+	mov	rdi, QWORD PTR [rcx+224]
+	and	edx, 1048560
+	mov	rax, QWORD PTR [rax+35]
+	xor	rax, QWORD PTR [rcx+192]
+	movq	xmm5, rax
+	movq	xmm8, rdi
+	punpcklqdq xmm4, xmm0
+	mov	QWORD PTR [rsp+64], rdx
+
+	movq xmm6, rcx
+	mov rax, QWORD PTR [rcx+264]
+	movq xmm7, rax
+
+	mov eax, 131072
+
+	ALIGN 64
+cnv1_main_loop_upx_soft_aes_sandybridge:
+	movq xmm9, rax
+	mov	r12, QWORD PTR [rcx+272]
+	mov	esi, DWORD PTR [rdx+rdi]
+	mov	r10d, DWORD PTR [rdx+rdi+4]
+	mov	ebp, DWORD PTR [rdx+rdi+12]
+	mov	r14d, DWORD PTR [rdx+rdi+8]
+	mov	rdx, QWORD PTR [rsp+64]
+	movzx	ecx, sil
+	shr	esi, 8
+	mov	r15d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r10b
+	shr	r10d, 8
+	mov	edi, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r14b
+	shr	r14d, 8
+	mov	ebx, DWORD PTR [r12+rcx*4]
+	movzx	ecx, bpl
+	shr	ebp, 8
+	mov	r9d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r10b
+	shr	r10d, 8
+	xor	r15d, DWORD PTR [r12+rcx*4+1024]
+	movzx	ecx, r14b
+	shr	r14d, 8
+	mov	eax, r14d
+	shr	eax, 8
+	xor	edi, DWORD PTR [r12+rcx*4+1024]
+	add	eax, 256
+	movzx	ecx, bpl
+	shr	ebp, 8
+	xor	ebx, DWORD PTR [r12+rcx*4+1024]
+	movzx	ecx, sil
+	shr	esi, 8
+	xor	r9d, DWORD PTR [r12+rcx*4+1024]
+	add	r12, 2048
+	movzx	ecx, r10b
+	shr	r10d, 8
+	add	r10d, 256
+	mov	r11d, DWORD PTR [r12+rax*4]
+	xor	r11d, DWORD PTR [r12+rcx*4]
+	xor	r11d, r9d
+	movzx	ecx, sil
+	mov	r10d, DWORD PTR [r12+r10*4]
+	shr	esi, 8
+	add	esi, 256
+	xor	r10d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, bpl
+	xor	r10d, ebx
+	shr	ebp, 8
+	add	ebp, 256
+	movd	xmm1, r11d
+	mov	r9d, DWORD PTR [r12+rcx*4]
+	xor	r9d, DWORD PTR [r12+rsi*4]
+	mov	eax, DWORD PTR [r12+rbp*4]
+	xor	r9d, edi
+	movq rdi, xmm8
+	movzx	ecx, r14b
+	movd	xmm0, r10d
+	movd	xmm2, r9d
+	punpckldq xmm2, xmm1
+	movq	xmm1, r8
+	xor	eax, DWORD PTR [r12+rcx*4]
+	xor	eax, r15d
+	movd	xmm3, eax
+	movq	rax, xmm7
+	punpckldq xmm3, xmm0
+	movq	xmm0, r13
+	punpcklqdq xmm1, xmm0
+	punpckldq xmm3, xmm2
+	pxor	xmm3, xmm1
+	movq	r9, xmm3
+	mov	r10d, r9d
+	and	r10d, 1048560
+	movdqa	xmm0, xmm3
+	pxor	xmm0, xmm4
+	movdqu	XMMWORD PTR [rdx+rdi], xmm0
+	psrldq	xmm0, 11
+	movq	rcx, xmm0
+	movzx	ecx, cl
+	mov	cl, BYTE PTR [rcx+rax]
+	mov	BYTE PTR [rdi+rdx+11], cl
+	mov	rbx, QWORD PTR [r10+rdi]
+	mov rcx, r9
+	lea	r9, QWORD PTR [r10+rdi]
+	mov	r11, QWORD PTR [r9+8]
+	mov	rax, rbx
+	movdqa	xmm4, xmm3
+	mul	rcx
+	movq	rcx, xmm6
+	add	r8, rdx
+	add	r13, rax
+	movq	rax, xmm5
+	xor	rax, r13
+	mov	QWORD PTR [r9], r8
+	xor	r8, rbx
+	mov	QWORD PTR [r9+8], rax
+	movq rax, xmm9
+	mov	rdx, r8
+	xor	r13, r11
+	and	edx, 1048560
+	mov	QWORD PTR [rsp+64], rdx
+	sub eax, 1
+	jne	cnv1_main_loop_upx_soft_aes_sandybridge
+
+	movaps xmm6, XMMWORD PTR [rsp]
+	movaps xmm7, XMMWORD PTR [rsp+16]
+	movaps xmm8, XMMWORD PTR [rsp+32]
+	movaps xmm9, XMMWORD PTR [rsp+48]
+
+	add	rsp, 72
+	pop	r15
+	pop	r14
+	pop	r13
+	pop	r12
+	pop	rdi
+	pop	rsi
+	pop	rbp
+	pop	rbx
diff --git a/src/crypto/asm/win/cnv2_double_main_loop_fastv2_sandybridge.inc b/src/crypto/asm/win/cnv2_double_main_loop_fastv2_sandybridge.inc
new file mode 100644
index 00000000..8295ef06
--- /dev/null
+++ b/src/crypto/asm/win/cnv2_double_main_loop_fastv2_sandybridge.inc
@@ -0,0 +1,410 @@
+	mov	rax, rsp
+	push	rbx
+	push	rbp
+	push	rsi
+	push	rdi
+	push	r12
+	push	r13
+	push	r14
+	push	r15
+	sub	rsp, 184
+
+	stmxcsr DWORD PTR [rsp+272]
+	mov DWORD PTR [rsp+276], 24448
+	ldmxcsr DWORD PTR [rsp+276]
+
+	mov	r13, QWORD PTR [rcx+224]
+	mov	r9, rdx
+	mov	r10, QWORD PTR [rcx+32]
+	mov	r8, rcx
+	xor	r10, QWORD PTR [rcx]
+	mov	r14d, 262144
+	mov	r11, QWORD PTR [rcx+40]
+	xor	r11, QWORD PTR [rcx+8]
+	mov	rsi, QWORD PTR [rdx+224]
+	mov	rdx, QWORD PTR [rcx+56]
+	xor	rdx, QWORD PTR [rcx+24]
+	mov	rdi, QWORD PTR [r9+32]
+	xor	rdi, QWORD PTR [r9]
+	mov	rbp, QWORD PTR [r9+40]
+	xor	rbp, QWORD PTR [r9+8]
+	movq	xmm0, rdx
+	movaps	XMMWORD PTR [rax-88], xmm6
+	movaps	XMMWORD PTR [rax-104], xmm7
+	movaps	XMMWORD PTR [rax-120], xmm8
+	movaps	XMMWORD PTR [rsp+112], xmm9
+	movaps	XMMWORD PTR [rsp+96], xmm10
+	movaps	XMMWORD PTR [rsp+80], xmm11
+	movaps	XMMWORD PTR [rsp+64], xmm12
+	movaps	XMMWORD PTR [rsp+48], xmm13
+	movaps	XMMWORD PTR [rsp+32], xmm14
+	movaps	XMMWORD PTR [rsp+16], xmm15
+	mov	rdx, r10
+	movq	xmm4, QWORD PTR [r8+96]
+	and	edx, 2097136
+	mov	rax, QWORD PTR [rcx+48]
+	xorps	xmm13, xmm13
+	xor	rax, QWORD PTR [rcx+16]
+	mov	rcx, QWORD PTR [rcx+88]
+	xor	rcx, QWORD PTR [r8+72]
+	movq	xmm5, QWORD PTR [r8+104]
+	movq	xmm7, rax
+
+	mov eax, 1
+	shl rax, 52
+	movq xmm14, rax
+	punpcklqdq xmm14, xmm14
+
+	mov eax, 1023
+	shl rax, 52
+	movq xmm12, rax
+	punpcklqdq xmm12, xmm12
+
+	mov	rax, QWORD PTR [r8+80]
+	xor	rax, QWORD PTR [r8+64]
+	punpcklqdq xmm7, xmm0
+	movq	xmm0, rcx
+	mov	rcx, QWORD PTR [r9+56]
+	xor	rcx, QWORD PTR [r9+24]
+	movq	xmm3, rax
+	mov	rax, QWORD PTR [r9+48]
+	xor	rax, QWORD PTR [r9+16]
+	punpcklqdq xmm3, xmm0
+	movq	xmm0, rcx
+	mov	QWORD PTR [rsp], r13
+	mov	rcx, QWORD PTR [r9+88]
+	xor	rcx, QWORD PTR [r9+72]
+	movq	xmm6, rax
+	mov	rax, QWORD PTR [r9+80]
+	xor	rax, QWORD PTR [r9+64]
+	punpcklqdq xmm6, xmm0
+	movq	xmm0, rcx
+	mov	QWORD PTR [rsp+256], r10
+	mov	rcx, rdi
+	mov	QWORD PTR [rsp+264], r11
+	movq	xmm8, rax
+	and	ecx, 2097136
+	punpcklqdq xmm8, xmm0
+	movq	xmm0, QWORD PTR [r9+96]
+	punpcklqdq xmm4, xmm0
+	movq	xmm0, QWORD PTR [r9+104]
+	lea	r8, QWORD PTR [rcx+rsi]
+	movdqu	xmm11, XMMWORD PTR [r8]
+	punpcklqdq xmm5, xmm0
+	lea	r9, QWORD PTR [rdx+r13]
+	movdqu	xmm15, XMMWORD PTR [r9]
+
+    ALIGN 64
+cnv2_double_main_loop_fastv2_sandybridge:
+	movdqu	xmm9, xmm15
+	mov eax, edx
+	mov ebx, edx
+	xor eax, 16
+	xor ebx, 32
+	xor edx, 48
+
+	movq	xmm0, r11
+	movq	xmm2, r10
+	punpcklqdq xmm2, xmm0
+	aesenc	xmm9, xmm2
+
+	movdqu	xmm0, XMMWORD PTR [rax+r13]
+	movdqu	xmm1, XMMWORD PTR [rbx+r13]
+	paddq	xmm0, xmm7
+	paddq	xmm1, xmm2
+	movdqu	XMMWORD PTR [rbx+r13], xmm0
+	movdqu	xmm0, XMMWORD PTR [rdx+r13]
+	movdqu	XMMWORD PTR [rdx+r13], xmm1
+	paddq	xmm0, xmm3
+	movdqu	XMMWORD PTR [rax+r13], xmm0
+
+	movq	r11, xmm9
+	mov	edx, r11d
+	and	edx, 2097136
+	movdqa	xmm0, xmm9
+	pxor	xmm0, xmm7
+	movdqu	XMMWORD PTR [r9], xmm0
+
+	lea	rbx, QWORD PTR [rdx+r13]
+	mov	r10, QWORD PTR [rdx+r13]
+
+	movdqu	xmm10, xmm11
+	movq	xmm0, rbp
+	movq	xmm11, rdi
+	punpcklqdq xmm11, xmm0
+	aesenc	xmm10, xmm11
+
+	mov eax, ecx
+	mov r12d, ecx
+	xor eax, 16
+	xor r12d, 32
+	xor ecx, 48
+
+	movdqu	xmm0, XMMWORD PTR [rax+rsi]
+	paddq	xmm0, xmm6
+	movdqu	xmm1, XMMWORD PTR [r12+rsi]
+	movdqu	XMMWORD PTR [r12+rsi], xmm0
+	paddq	xmm1, xmm11
+	movdqu	xmm0, XMMWORD PTR [rcx+rsi]
+	movdqu	XMMWORD PTR [rcx+rsi], xmm1
+	paddq	xmm0, xmm8
+	movdqu	XMMWORD PTR [rax+rsi], xmm0
+
+	movq	rcx, xmm10
+	and	ecx, 2097136
+
+	movdqa	xmm0, xmm10
+	pxor	xmm0, xmm6
+	movdqu	XMMWORD PTR [r8], xmm0
+	mov r12, QWORD PTR [rcx+rsi]
+
+	mov	r9, QWORD PTR [rbx+8]
+
+	xor edx, 16
+	mov r8d, edx
+	mov r15d, edx
+
+	movq	rdx, xmm5
+	shl	rdx, 32
+	movq	rax, xmm4
+	xor	rdx, rax
+	xor	r10, rdx
+	mov	rax, r10
+	mul	r11
+	mov r11d, r8d
+	xor r11d, 48
+	movq xmm0, rdx
+	xor rdx, [r11+r13]
+	movq xmm1, rax
+	xor rax, [r11+r13+8]
+	punpcklqdq xmm0, xmm1
+
+	pxor xmm0, XMMWORD PTR [r8+r13]
+	xor	r8d, 32
+	movdqu	xmm1, XMMWORD PTR [r11+r13]
+	paddq	xmm0, xmm7
+	paddq	xmm1, xmm2
+	movdqu	XMMWORD PTR [r11+r13], xmm0
+	movdqu	xmm0, XMMWORD PTR [r8+r13]
+	movdqu	XMMWORD PTR [r8+r13], xmm1
+	paddq	xmm0, xmm3
+	movdqu	XMMWORD PTR [r15+r13], xmm0
+
+	mov	r11, QWORD PTR [rsp+256]
+	add	r11, rdx
+	mov	rdx, QWORD PTR [rsp+264]
+	add	rdx, rax
+	mov	QWORD PTR [rbx], r11
+	xor	r11, r10
+	mov	QWORD PTR [rbx+8], rdx
+	xor	rdx, r9
+	mov	QWORD PTR [rsp+256], r11
+	and	r11d, 2097136
+	mov	QWORD PTR [rsp+264], rdx
+	mov	QWORD PTR [rsp+8], r11
+	lea	r15, QWORD PTR [r11+r13]
+	movdqu xmm15, XMMWORD PTR [r11+r13]
+	lea	r13, QWORD PTR [rsi+rcx]
+	movdqa	xmm0, xmm5
+	psrldq	xmm0, 8
+	movaps	xmm2, xmm13
+	movq	r10, xmm0
+	psllq	xmm5, 1
+	shl	r10, 32
+	movdqa	xmm0, xmm9
+	psrldq	xmm0, 8
+	movdqa	xmm1, xmm10
+	movq	r11, xmm0
+	psrldq	xmm1, 8
+	movq	r8, xmm1
+	psrldq	xmm4, 8
+	movaps	xmm0, xmm13
+	movq	rax, xmm4
+	xor	r10, rax
+	movaps	xmm1, xmm13
+	xor	r10, r12
+	lea	rax, QWORD PTR [r11+1]
+	shr	rax, 1
+	movdqa	xmm3, xmm9
+	punpcklqdq xmm3, xmm10
+	paddq	xmm5, xmm3
+	movq	rdx, xmm5
+	psrldq	xmm5, 8
+	cvtsi2sd xmm2, rax
+	or	edx, -2147483647
+	lea	rax, QWORD PTR [r8+1]
+	shr	rax, 1
+	movq	r9, xmm5
+	cvtsi2sd xmm0, rax
+	or	r9d, -2147483647
+	cvtsi2sd xmm1, rdx
+	unpcklpd xmm2, xmm0
+	movaps	xmm0, xmm13
+	cvtsi2sd xmm0, r9
+	unpcklpd xmm1, xmm0
+	divpd	xmm2, xmm1
+	paddq	xmm2, xmm14
+	cvttsd2si rax, xmm2
+	psrldq	xmm2, 8
+	mov	rbx, rax
+	imul	rax, rdx
+	sub	r11, rax
+	js	div_fix_1_fastv2_sandybridge
+div_fix_1_ret_fastv2_sandybridge:
+
+	cvttsd2si rdx, xmm2
+	mov	rax, rdx
+	imul	rax, r9
+	movd	xmm2, r11d
+	movd	xmm4, ebx
+	sub	r8, rax
+	js	div_fix_2_fastv2_sandybridge
+div_fix_2_ret_fastv2_sandybridge:
+
+	movd	xmm1, r8d
+	movd	xmm0, edx
+	punpckldq xmm2, xmm1
+	punpckldq xmm4, xmm0
+	punpckldq xmm4, xmm2
+	paddq	xmm3, xmm4
+	movdqa	xmm0, xmm3
+	psrlq	xmm0, 12
+	paddq	xmm0, xmm12
+	sqrtpd	xmm1, xmm0
+	movq	r9, xmm1
+	movdqa xmm5, xmm1
+	psrlq xmm5, 19
+	test	r9, 524287
+	je	sqrt_fix_1_fastv2_sandybridge
+sqrt_fix_1_ret_fastv2_sandybridge:
+
+	movq r9, xmm10
+	psrldq	xmm1, 8
+	movq	r8, xmm1
+	test	r8, 524287
+	je	sqrt_fix_2_fastv2_sandybridge
+sqrt_fix_2_ret_fastv2_sandybridge:
+
+	mov r12d, ecx
+	mov r8d, ecx
+	xor r12d, 16
+	xor r8d, 32
+	xor ecx, 48
+	mov	rax, r10
+	mul	r9
+	movq xmm0, rax
+	movq xmm3, rdx
+	punpcklqdq xmm3, xmm0
+
+	movdqu	xmm0, XMMWORD PTR [r12+rsi]
+	pxor xmm0, xmm3
+	movdqu	xmm1, XMMWORD PTR [r8+rsi]
+	xor rdx, [r8+rsi]
+	xor rax, [r8+rsi+8]
+	movdqu	xmm3, XMMWORD PTR [rcx+rsi]
+	paddq	xmm0, xmm6
+	paddq	xmm1, xmm11
+	paddq	xmm3, xmm8
+	movdqu	XMMWORD PTR [r8+rsi], xmm0
+	movdqu	XMMWORD PTR [rcx+rsi], xmm1
+	movdqu	XMMWORD PTR [r12+rsi], xmm3
+
+	add	rdi, rdx
+	mov	QWORD PTR [r13], rdi
+	xor	rdi, r10
+	mov	ecx, edi
+	and	ecx, 2097136
+	lea	r8, QWORD PTR [rcx+rsi]
+
+	mov rdx, QWORD PTR [r13+8]	
+	add	rbp, rax
+	mov	QWORD PTR [r13+8], rbp
+	movdqu xmm11, XMMWORD PTR [rcx+rsi]
+	xor	rbp, rdx
+	mov	r13, QWORD PTR [rsp]
+	movdqa	xmm3, xmm7
+	mov	rdx, QWORD PTR [rsp+8]
+	movdqa	xmm8, xmm6
+	mov	r10, QWORD PTR [rsp+256]
+	movdqa	xmm7, xmm9
+	mov	r11, QWORD PTR [rsp+264]
+	movdqa	xmm6, xmm10
+	mov	r9, r15
+	dec r14d
+	jne	cnv2_double_main_loop_fastv2_sandybridge
+
+	ldmxcsr DWORD PTR [rsp+272]
+	movaps	xmm13, XMMWORD PTR [rsp+48]
+	lea	r11, QWORD PTR [rsp+184]
+	movaps	xmm6, XMMWORD PTR [r11-24]
+	movaps	xmm7, XMMWORD PTR [r11-40]
+	movaps	xmm8, XMMWORD PTR [r11-56]
+	movaps	xmm9, XMMWORD PTR [r11-72]
+	movaps	xmm10, XMMWORD PTR [r11-88]
+	movaps	xmm11, XMMWORD PTR [r11-104]
+	movaps	xmm12, XMMWORD PTR [r11-120]
+	movaps	xmm14, XMMWORD PTR [rsp+32]
+	movaps	xmm15, XMMWORD PTR [rsp+16]
+	mov	rsp, r11
+	pop	r15
+	pop	r14
+	pop	r13
+	pop	r12
+	pop	rdi
+	pop	rsi
+	pop	rbp
+	pop	rbx
+	jmp cnv2_double_main_loop_fastv2_sandybridge_endp
+
+div_fix_1_fastv2_sandybridge:
+	dec	rbx
+	add	r11, rdx
+	jmp	div_fix_1_ret_fastv2_sandybridge
+
+div_fix_2_fastv2_sandybridge:
+	dec	rdx
+	add	r8, r9
+	jmp	div_fix_2_ret_fastv2_sandybridge
+
+sqrt_fix_1_fastv2_sandybridge:
+	movq	r8, xmm3
+	movdqa xmm0, xmm5
+	psrldq xmm0, 8
+	dec	r9
+	mov r11d, -1022
+	shl r11, 32
+	mov	rax, r9
+	shr	r9, 19
+	shr	rax, 20
+	mov	rdx, r9
+	sub	rdx, rax
+	lea	rdx, [rdx+r11+1]
+	add	rax, r11
+	imul	rdx, rax
+	sub	rdx, r8
+	adc	r9, 0
+	movq xmm5, r9
+	punpcklqdq xmm5, xmm0
+	jmp	sqrt_fix_1_ret_fastv2_sandybridge
+
+sqrt_fix_2_fastv2_sandybridge:
+	psrldq	xmm3, 8
+	movq	r11, xmm3
+	dec	r8
+	mov ebx, -1022
+	shl rbx, 32
+	mov	rax, r8
+	shr	r8, 19
+	shr	rax, 20
+	mov	rdx, r8
+	sub	rdx, rax
+	lea	rdx, [rdx+rbx+1]
+	add	rax, rbx
+	imul	rdx, rax
+	sub	rdx, r11
+	adc	r8, 0
+	movq xmm0, r8
+	punpcklqdq xmm5, xmm0
+	jmp	sqrt_fix_2_ret_fastv2_sandybridge
+
+cnv2_double_main_loop_fastv2_sandybridge_endp:
diff --git a/src/crypto/asm/win/cnv2_double_main_loop_sandybridge.inc b/src/crypto/asm/win/cnv2_double_main_loop_sandybridge.inc
new file mode 100644
index 00000000..da2c1b9e
--- /dev/null
+++ b/src/crypto/asm/win/cnv2_double_main_loop_sandybridge.inc
@@ -0,0 +1,410 @@
+	mov	rax, rsp
+	push	rbx
+	push	rbp
+	push	rsi
+	push	rdi
+	push	r12
+	push	r13
+	push	r14
+	push	r15
+	sub	rsp, 184
+
+	stmxcsr DWORD PTR [rsp+272]
+	mov DWORD PTR [rsp+276], 24448
+	ldmxcsr DWORD PTR [rsp+276]
+
+	mov	r13, QWORD PTR [rcx+224]
+	mov	r9, rdx
+	mov	r10, QWORD PTR [rcx+32]
+	mov	r8, rcx
+	xor	r10, QWORD PTR [rcx]
+	mov	r14d, 524288
+	mov	r11, QWORD PTR [rcx+40]
+	xor	r11, QWORD PTR [rcx+8]
+	mov	rsi, QWORD PTR [rdx+224]
+	mov	rdx, QWORD PTR [rcx+56]
+	xor	rdx, QWORD PTR [rcx+24]
+	mov	rdi, QWORD PTR [r9+32]
+	xor	rdi, QWORD PTR [r9]
+	mov	rbp, QWORD PTR [r9+40]
+	xor	rbp, QWORD PTR [r9+8]
+	movq	xmm0, rdx
+	movaps	XMMWORD PTR [rax-88], xmm6
+	movaps	XMMWORD PTR [rax-104], xmm7
+	movaps	XMMWORD PTR [rax-120], xmm8
+	movaps	XMMWORD PTR [rsp+112], xmm9
+	movaps	XMMWORD PTR [rsp+96], xmm10
+	movaps	XMMWORD PTR [rsp+80], xmm11
+	movaps	XMMWORD PTR [rsp+64], xmm12
+	movaps	XMMWORD PTR [rsp+48], xmm13
+	movaps	XMMWORD PTR [rsp+32], xmm14
+	movaps	XMMWORD PTR [rsp+16], xmm15
+	mov	rdx, r10
+	movq	xmm4, QWORD PTR [r8+96]
+	and	edx, 2097136
+	mov	rax, QWORD PTR [rcx+48]
+	xorps	xmm13, xmm13
+	xor	rax, QWORD PTR [rcx+16]
+	mov	rcx, QWORD PTR [rcx+88]
+	xor	rcx, QWORD PTR [r8+72]
+	movq	xmm5, QWORD PTR [r8+104]
+	movq	xmm7, rax
+
+	mov eax, 1
+	shl rax, 52
+	movq xmm14, rax
+	punpcklqdq xmm14, xmm14
+
+	mov eax, 1023
+	shl rax, 52
+	movq xmm12, rax
+	punpcklqdq xmm12, xmm12
+
+	mov	rax, QWORD PTR [r8+80]
+	xor	rax, QWORD PTR [r8+64]
+	punpcklqdq xmm7, xmm0
+	movq	xmm0, rcx
+	mov	rcx, QWORD PTR [r9+56]
+	xor	rcx, QWORD PTR [r9+24]
+	movq	xmm3, rax
+	mov	rax, QWORD PTR [r9+48]
+	xor	rax, QWORD PTR [r9+16]
+	punpcklqdq xmm3, xmm0
+	movq	xmm0, rcx
+	mov	QWORD PTR [rsp], r13
+	mov	rcx, QWORD PTR [r9+88]
+	xor	rcx, QWORD PTR [r9+72]
+	movq	xmm6, rax
+	mov	rax, QWORD PTR [r9+80]
+	xor	rax, QWORD PTR [r9+64]
+	punpcklqdq xmm6, xmm0
+	movq	xmm0, rcx
+	mov	QWORD PTR [rsp+256], r10
+	mov	rcx, rdi
+	mov	QWORD PTR [rsp+264], r11
+	movq	xmm8, rax
+	and	ecx, 2097136
+	punpcklqdq xmm8, xmm0
+	movq	xmm0, QWORD PTR [r9+96]
+	punpcklqdq xmm4, xmm0
+	movq	xmm0, QWORD PTR [r9+104]
+	lea	r8, QWORD PTR [rcx+rsi]
+	movdqu	xmm11, XMMWORD PTR [r8]
+	punpcklqdq xmm5, xmm0
+	lea	r9, QWORD PTR [rdx+r13]
+	movdqu	xmm15, XMMWORD PTR [r9]
+
+    ALIGN 64
+cnv2_double_main_loop_originalv2_sandybridge:
+	movdqu	xmm9, xmm15
+	mov eax, edx
+	mov ebx, edx
+	xor eax, 16
+	xor ebx, 32
+	xor edx, 48
+
+	movq	xmm0, r11
+	movq	xmm2, r10
+	punpcklqdq xmm2, xmm0
+	aesenc	xmm9, xmm2
+
+	movdqu	xmm0, XMMWORD PTR [rax+r13]
+	movdqu	xmm1, XMMWORD PTR [rbx+r13]
+	paddq	xmm0, xmm7
+	paddq	xmm1, xmm2
+	movdqu	XMMWORD PTR [rbx+r13], xmm0
+	movdqu	xmm0, XMMWORD PTR [rdx+r13]
+	movdqu	XMMWORD PTR [rdx+r13], xmm1
+	paddq	xmm0, xmm3
+	movdqu	XMMWORD PTR [rax+r13], xmm0
+
+	movq	r11, xmm9
+	mov	edx, r11d
+	and	edx, 2097136
+	movdqa	xmm0, xmm9
+	pxor	xmm0, xmm7
+	movdqu	XMMWORD PTR [r9], xmm0
+
+	lea	rbx, QWORD PTR [rdx+r13]
+	mov	r10, QWORD PTR [rdx+r13]
+
+	movdqu	xmm10, xmm11
+	movq	xmm0, rbp
+	movq	xmm11, rdi
+	punpcklqdq xmm11, xmm0
+	aesenc	xmm10, xmm11
+
+	mov eax, ecx
+	mov r12d, ecx
+	xor eax, 16
+	xor r12d, 32
+	xor ecx, 48
+
+	movdqu	xmm0, XMMWORD PTR [rax+rsi]
+	paddq	xmm0, xmm6
+	movdqu	xmm1, XMMWORD PTR [r12+rsi]
+	movdqu	XMMWORD PTR [r12+rsi], xmm0
+	paddq	xmm1, xmm11
+	movdqu	xmm0, XMMWORD PTR [rcx+rsi]
+	movdqu	XMMWORD PTR [rcx+rsi], xmm1
+	paddq	xmm0, xmm8
+	movdqu	XMMWORD PTR [rax+rsi], xmm0
+
+	movq	rcx, xmm10
+	and	ecx, 2097136
+
+	movdqa	xmm0, xmm10
+	pxor	xmm0, xmm6
+	movdqu	XMMWORD PTR [r8], xmm0
+	mov r12, QWORD PTR [rcx+rsi]
+
+	mov	r9, QWORD PTR [rbx+8]
+
+	xor edx, 16
+	mov r8d, edx
+	mov r15d, edx
+
+	movq	rdx, xmm5
+	shl	rdx, 32
+	movq	rax, xmm4
+	xor	rdx, rax
+	xor	r10, rdx
+	mov	rax, r10
+	mul	r11
+	mov r11d, r8d
+	xor r11d, 48
+	movq xmm0, rdx
+	xor rdx, [r11+r13]
+	movq xmm1, rax
+	xor rax, [r11+r13+8]
+	punpcklqdq xmm0, xmm1
+
+	pxor xmm0, XMMWORD PTR [r8+r13]
+	xor	r8d, 32
+	movdqu	xmm1, XMMWORD PTR [r11+r13]
+	paddq	xmm0, xmm7
+	paddq	xmm1, xmm2
+	movdqu	XMMWORD PTR [r11+r13], xmm0
+	movdqu	xmm0, XMMWORD PTR [r8+r13]
+	movdqu	XMMWORD PTR [r8+r13], xmm1
+	paddq	xmm0, xmm3
+	movdqu	XMMWORD PTR [r15+r13], xmm0
+
+	mov	r11, QWORD PTR [rsp+256]
+	add	r11, rdx
+	mov	rdx, QWORD PTR [rsp+264]
+	add	rdx, rax
+	mov	QWORD PTR [rbx], r11
+	xor	r11, r10
+	mov	QWORD PTR [rbx+8], rdx
+	xor	rdx, r9
+	mov	QWORD PTR [rsp+256], r11
+	and	r11d, 2097136
+	mov	QWORD PTR [rsp+264], rdx
+	mov	QWORD PTR [rsp+8], r11
+	lea	r15, QWORD PTR [r11+r13]
+	movdqu xmm15, XMMWORD PTR [r11+r13]
+	lea	r13, QWORD PTR [rsi+rcx]
+	movdqa	xmm0, xmm5
+	psrldq	xmm0, 8
+	movaps	xmm2, xmm13
+	movq	r10, xmm0
+	psllq	xmm5, 1
+	shl	r10, 32
+	movdqa	xmm0, xmm9
+	psrldq	xmm0, 8
+	movdqa	xmm1, xmm10
+	movq	r11, xmm0
+	psrldq	xmm1, 8
+	movq	r8, xmm1
+	psrldq	xmm4, 8
+	movaps	xmm0, xmm13
+	movq	rax, xmm4
+	xor	r10, rax
+	movaps	xmm1, xmm13
+	xor	r10, r12
+	lea	rax, QWORD PTR [r11+1]
+	shr	rax, 1
+	movdqa	xmm3, xmm9
+	punpcklqdq xmm3, xmm10
+	paddq	xmm5, xmm3
+	movq	rdx, xmm5
+	psrldq	xmm5, 8
+	cvtsi2sd xmm2, rax
+	or	edx, -2147483647
+	lea	rax, QWORD PTR [r8+1]
+	shr	rax, 1
+	movq	r9, xmm5
+	cvtsi2sd xmm0, rax
+	or	r9d, -2147483647
+	cvtsi2sd xmm1, rdx
+	unpcklpd xmm2, xmm0
+	movaps	xmm0, xmm13
+	cvtsi2sd xmm0, r9
+	unpcklpd xmm1, xmm0
+	divpd	xmm2, xmm1
+	paddq	xmm2, xmm14
+	cvttsd2si rax, xmm2
+	psrldq	xmm2, 8
+	mov	rbx, rax
+	imul	rax, rdx
+	sub	r11, rax
+	js	div_fix_1_originalv2_sandybridge
+div_fix_1_ret_originalv2_sandybridge:
+
+	cvttsd2si rdx, xmm2
+	mov	rax, rdx
+	imul	rax, r9
+	movd	xmm2, r11d
+	movd	xmm4, ebx
+	sub	r8, rax
+	js	div_fix_2_originalv2_sandybridge
+div_fix_2_ret_originalv2_sandybridge:
+
+	movd	xmm1, r8d
+	movd	xmm0, edx
+	punpckldq xmm2, xmm1
+	punpckldq xmm4, xmm0
+	punpckldq xmm4, xmm2
+	paddq	xmm3, xmm4
+	movdqa	xmm0, xmm3
+	psrlq	xmm0, 12
+	paddq	xmm0, xmm12
+	sqrtpd	xmm1, xmm0
+	movq	r9, xmm1
+	movdqa xmm5, xmm1
+	psrlq xmm5, 19
+	test	r9, 524287
+	je	sqrt_fix_1_originalv2_sandybridge
+sqrt_fix_1_ret_originalv2_sandybridge:
+
+	movq r9, xmm10
+	psrldq	xmm1, 8
+	movq	r8, xmm1
+	test	r8, 524287
+	je	sqrt_fix_2_originalv2_sandybridge
+sqrt_fix_2_ret_originalv2_sandybridge:
+
+	mov r12d, ecx
+	mov r8d, ecx
+	xor r12d, 16
+	xor r8d, 32
+	xor ecx, 48
+	mov	rax, r10
+	mul	r9
+	movq xmm0, rax
+	movq xmm3, rdx
+	punpcklqdq xmm3, xmm0
+
+	movdqu	xmm0, XMMWORD PTR [r12+rsi]
+	pxor xmm0, xmm3
+	movdqu	xmm1, XMMWORD PTR [r8+rsi]
+	xor rdx, [r8+rsi]
+	xor rax, [r8+rsi+8]
+	movdqu	xmm3, XMMWORD PTR [rcx+rsi]
+	paddq	xmm0, xmm6
+	paddq	xmm1, xmm11
+	paddq	xmm3, xmm8
+	movdqu	XMMWORD PTR [r8+rsi], xmm0
+	movdqu	XMMWORD PTR [rcx+rsi], xmm1
+	movdqu	XMMWORD PTR [r12+rsi], xmm3
+
+	add	rdi, rdx
+	mov	QWORD PTR [r13], rdi
+	xor	rdi, r10
+	mov	ecx, edi
+	and	ecx, 2097136
+	lea	r8, QWORD PTR [rcx+rsi]
+
+	mov rdx, QWORD PTR [r13+8]	
+	add	rbp, rax
+	mov	QWORD PTR [r13+8], rbp
+	movdqu xmm11, XMMWORD PTR [rcx+rsi]
+	xor	rbp, rdx
+	mov	r13, QWORD PTR [rsp]
+	movdqa	xmm3, xmm7
+	mov	rdx, QWORD PTR [rsp+8]
+	movdqa	xmm8, xmm6
+	mov	r10, QWORD PTR [rsp+256]
+	movdqa	xmm7, xmm9
+	mov	r11, QWORD PTR [rsp+264]
+	movdqa	xmm6, xmm10
+	mov	r9, r15
+	dec r14d
+	jne	cnv2_double_main_loop_originalv2_sandybridge
+
+	ldmxcsr DWORD PTR [rsp+272]
+	movaps	xmm13, XMMWORD PTR [rsp+48]
+	lea	r11, QWORD PTR [rsp+184]
+	movaps	xmm6, XMMWORD PTR [r11-24]
+	movaps	xmm7, XMMWORD PTR [r11-40]
+	movaps	xmm8, XMMWORD PTR [r11-56]
+	movaps	xmm9, XMMWORD PTR [r11-72]
+	movaps	xmm10, XMMWORD PTR [r11-88]
+	movaps	xmm11, XMMWORD PTR [r11-104]
+	movaps	xmm12, XMMWORD PTR [r11-120]
+	movaps	xmm14, XMMWORD PTR [rsp+32]
+	movaps	xmm15, XMMWORD PTR [rsp+16]
+	mov	rsp, r11
+	pop	r15
+	pop	r14
+	pop	r13
+	pop	r12
+	pop	rdi
+	pop	rsi
+	pop	rbp
+	pop	rbx
+	jmp cnv2_double_main_loop_originalv2_sandybridge_endp
+
+div_fix_1_originalv2_sandybridge:
+	dec	rbx
+	add	r11, rdx
+	jmp	div_fix_1_ret_originalv2_sandybridge
+
+div_fix_2_originalv2_sandybridge:
+	dec	rdx
+	add	r8, r9
+	jmp	div_fix_2_ret_originalv2_sandybridge
+
+sqrt_fix_1_originalv2_sandybridge:
+	movq	r8, xmm3
+	movdqa xmm0, xmm5
+	psrldq xmm0, 8
+	dec	r9
+	mov r11d, -1022
+	shl r11, 32
+	mov	rax, r9
+	shr	r9, 19
+	shr	rax, 20
+	mov	rdx, r9
+	sub	rdx, rax
+	lea	rdx, [rdx+r11+1]
+	add	rax, r11
+	imul	rdx, rax
+	sub	rdx, r8
+	adc	r9, 0
+	movq xmm5, r9
+	punpcklqdq xmm5, xmm0
+	jmp	sqrt_fix_1_ret_originalv2_sandybridge
+
+sqrt_fix_2_originalv2_sandybridge:
+	psrldq	xmm3, 8
+	movq	r11, xmm3
+	dec	r8
+	mov ebx, -1022
+	shl rbx, 32
+	mov	rax, r8
+	shr	r8, 19
+	shr	rax, 20
+	mov	rdx, r8
+	sub	rdx, rax
+	lea	rdx, [rdx+rbx+1]
+	add	rax, rbx
+	imul	rdx, rax
+	sub	rdx, r11
+	adc	r8, 0
+	movq xmm0, r8
+	punpcklqdq xmm5, xmm0
+	jmp	sqrt_fix_2_ret_originalv2_sandybridge
+
+cnv2_double_main_loop_originalv2_sandybridge_endp:
diff --git a/src/crypto/asm/win/cnv2_double_main_loop_ultralite_sandybridge.inc b/src/crypto/asm/win/cnv2_double_main_loop_ultralite_sandybridge.inc
new file mode 100644
index 00000000..e7c0eb60
--- /dev/null
+++ b/src/crypto/asm/win/cnv2_double_main_loop_ultralite_sandybridge.inc
@@ -0,0 +1,410 @@
+	mov	rax, rsp
+	push	rbx
+	push	rbp
+	push	rsi
+	push	rdi
+	push	r12
+	push	r13
+	push	r14
+	push	r15
+	sub	rsp, 184
+
+	stmxcsr DWORD PTR [rsp+272]
+	mov DWORD PTR [rsp+276], 24448
+	ldmxcsr DWORD PTR [rsp+276]
+
+	mov	r13, QWORD PTR [rcx+224]
+	mov	r9, rdx
+	mov	r10, QWORD PTR [rcx+32]
+	mov	r8, rcx
+	xor	r10, QWORD PTR [rcx]
+	mov	r14d, 65536
+	mov	r11, QWORD PTR [rcx+40]
+	xor	r11, QWORD PTR [rcx+8]
+	mov	rsi, QWORD PTR [rdx+224]
+	mov	rdx, QWORD PTR [rcx+56]
+	xor	rdx, QWORD PTR [rcx+24]
+	mov	rdi, QWORD PTR [r9+32]
+	xor	rdi, QWORD PTR [r9]
+	mov	rbp, QWORD PTR [r9+40]
+	xor	rbp, QWORD PTR [r9+8]
+	movq	xmm0, rdx
+	movaps	XMMWORD PTR [rax-88], xmm6
+	movaps	XMMWORD PTR [rax-104], xmm7
+	movaps	XMMWORD PTR [rax-120], xmm8
+	movaps	XMMWORD PTR [rsp+112], xmm9
+	movaps	XMMWORD PTR [rsp+96], xmm10
+	movaps	XMMWORD PTR [rsp+80], xmm11
+	movaps	XMMWORD PTR [rsp+64], xmm12
+	movaps	XMMWORD PTR [rsp+48], xmm13
+	movaps	XMMWORD PTR [rsp+32], xmm14
+	movaps	XMMWORD PTR [rsp+16], xmm15
+	mov	rdx, r10
+	movq	xmm4, QWORD PTR [r8+96]
+	and	edx, 131056
+	mov	rax, QWORD PTR [rcx+48]
+	xorps	xmm13, xmm13
+	xor	rax, QWORD PTR [rcx+16]
+	mov	rcx, QWORD PTR [rcx+88]
+	xor	rcx, QWORD PTR [r8+72]
+	movq	xmm5, QWORD PTR [r8+104]
+	movq	xmm7, rax
+
+	mov eax, 1
+	shl rax, 52
+	movq xmm14, rax
+	punpcklqdq xmm14, xmm14
+
+	mov eax, 1023
+	shl rax, 52
+	movq xmm12, rax
+	punpcklqdq xmm12, xmm12
+
+	mov	rax, QWORD PTR [r8+80]
+	xor	rax, QWORD PTR [r8+64]
+	punpcklqdq xmm7, xmm0
+	movq	xmm0, rcx
+	mov	rcx, QWORD PTR [r9+56]
+	xor	rcx, QWORD PTR [r9+24]
+	movq	xmm3, rax
+	mov	rax, QWORD PTR [r9+48]
+	xor	rax, QWORD PTR [r9+16]
+	punpcklqdq xmm3, xmm0
+	movq	xmm0, rcx
+	mov	QWORD PTR [rsp], r13
+	mov	rcx, QWORD PTR [r9+88]
+	xor	rcx, QWORD PTR [r9+72]
+	movq	xmm6, rax
+	mov	rax, QWORD PTR [r9+80]
+	xor	rax, QWORD PTR [r9+64]
+	punpcklqdq xmm6, xmm0
+	movq	xmm0, rcx
+	mov	QWORD PTR [rsp+256], r10
+	mov	rcx, rdi
+	mov	QWORD PTR [rsp+264], r11
+	movq	xmm8, rax
+	and	ecx, 131056
+	punpcklqdq xmm8, xmm0
+	movq	xmm0, QWORD PTR [r9+96]
+	punpcklqdq xmm4, xmm0
+	movq	xmm0, QWORD PTR [r9+104]
+	lea	r8, QWORD PTR [rcx+rsi]
+	movdqu	xmm11, XMMWORD PTR [r8]
+	punpcklqdq xmm5, xmm0
+	lea	r9, QWORD PTR [rdx+r13]
+	movdqu	xmm15, XMMWORD PTR [r9]
+
+    ALIGN 64
+cnv2_double_main_loop_ultralite_sandybridge:
+	movdqu	xmm9, xmm15
+	mov eax, edx
+	mov ebx, edx
+	xor eax, 16
+	xor ebx, 32
+	xor edx, 48
+
+	movq	xmm0, r11
+	movq	xmm2, r10
+	punpcklqdq xmm2, xmm0
+	aesenc	xmm9, xmm2
+
+	movdqu	xmm0, XMMWORD PTR [rax+r13]
+	movdqu	xmm1, XMMWORD PTR [rbx+r13]
+	paddq	xmm0, xmm7
+	paddq	xmm1, xmm2
+	movdqu	XMMWORD PTR [rbx+r13], xmm0
+	movdqu	xmm0, XMMWORD PTR [rdx+r13]
+	movdqu	XMMWORD PTR [rdx+r13], xmm1
+	paddq	xmm0, xmm3
+	movdqu	XMMWORD PTR [rax+r13], xmm0
+
+	movq	r11, xmm9
+	mov	edx, r11d
+	and	edx, 131056
+	movdqa	xmm0, xmm9
+	pxor	xmm0, xmm7
+	movdqu	XMMWORD PTR [r9], xmm0
+
+	lea	rbx, QWORD PTR [rdx+r13]
+	mov	r10, QWORD PTR [rdx+r13]
+
+	movdqu	xmm10, xmm11
+	movq	xmm0, rbp
+	movq	xmm11, rdi
+	punpcklqdq xmm11, xmm0
+	aesenc	xmm10, xmm11
+
+	mov eax, ecx
+	mov r12d, ecx
+	xor eax, 16
+	xor r12d, 32
+	xor ecx, 48
+
+	movdqu	xmm0, XMMWORD PTR [rax+rsi]
+	paddq	xmm0, xmm6
+	movdqu	xmm1, XMMWORD PTR [r12+rsi]
+	movdqu	XMMWORD PTR [r12+rsi], xmm0
+	paddq	xmm1, xmm11
+	movdqu	xmm0, XMMWORD PTR [rcx+rsi]
+	movdqu	XMMWORD PTR [rcx+rsi], xmm1
+	paddq	xmm0, xmm8
+	movdqu	XMMWORD PTR [rax+rsi], xmm0
+
+	movq	rcx, xmm10
+	and	ecx, 131056
+
+	movdqa	xmm0, xmm10
+	pxor	xmm0, xmm6
+	movdqu	XMMWORD PTR [r8], xmm0
+	mov r12, QWORD PTR [rcx+rsi]
+
+	mov	r9, QWORD PTR [rbx+8]
+
+	xor edx, 16
+	mov r8d, edx
+	mov r15d, edx
+
+	movq	rdx, xmm5
+	shl	rdx, 32
+	movq	rax, xmm4
+	xor	rdx, rax
+	xor	r10, rdx
+	mov	rax, r10
+	mul	r11
+	mov r11d, r8d
+	xor r11d, 48
+	movq xmm0, rdx
+	xor rdx, [r11+r13]
+	movq xmm1, rax
+	xor rax, [r11+r13+8]
+	punpcklqdq xmm0, xmm1
+
+	pxor xmm0, XMMWORD PTR [r8+r13]
+	xor	r8d, 32
+	movdqu	xmm1, XMMWORD PTR [r11+r13]
+	paddq	xmm0, xmm7
+	paddq	xmm1, xmm2
+	movdqu	XMMWORD PTR [r11+r13], xmm0
+	movdqu	xmm0, XMMWORD PTR [r8+r13]
+	movdqu	XMMWORD PTR [r8+r13], xmm1
+	paddq	xmm0, xmm3
+	movdqu	XMMWORD PTR [r15+r13], xmm0
+
+	mov	r11, QWORD PTR [rsp+256]
+	add	r11, rdx
+	mov	rdx, QWORD PTR [rsp+264]
+	add	rdx, rax
+	mov	QWORD PTR [rbx], r11
+	xor	r11, r10
+	mov	QWORD PTR [rbx+8], rdx
+	xor	rdx, r9
+	mov	QWORD PTR [rsp+256], r11
+	and	r11d, 131056
+	mov	QWORD PTR [rsp+264], rdx
+	mov	QWORD PTR [rsp+8], r11
+	lea	r15, QWORD PTR [r11+r13]
+	movdqu xmm15, XMMWORD PTR [r11+r13]
+	lea	r13, QWORD PTR [rsi+rcx]
+	movdqa	xmm0, xmm5
+	psrldq	xmm0, 8
+	movaps	xmm2, xmm13
+	movq	r10, xmm0
+	psllq	xmm5, 1
+	shl	r10, 32
+	movdqa	xmm0, xmm9
+	psrldq	xmm0, 8
+	movdqa	xmm1, xmm10
+	movq	r11, xmm0
+	psrldq	xmm1, 8
+	movq	r8, xmm1
+	psrldq	xmm4, 8
+	movaps	xmm0, xmm13
+	movq	rax, xmm4
+	xor	r10, rax
+	movaps	xmm1, xmm13
+	xor	r10, r12
+	lea	rax, QWORD PTR [r11+1]
+	shr	rax, 1
+	movdqa	xmm3, xmm9
+	punpcklqdq xmm3, xmm10
+	paddq	xmm5, xmm3
+	movq	rdx, xmm5
+	psrldq	xmm5, 8
+	cvtsi2sd xmm2, rax
+	or	edx, -2147483647
+	lea	rax, QWORD PTR [r8+1]
+	shr	rax, 1
+	movq	r9, xmm5
+	cvtsi2sd xmm0, rax
+	or	r9d, -2147483647
+	cvtsi2sd xmm1, rdx
+	unpcklpd xmm2, xmm0
+	movaps	xmm0, xmm13
+	cvtsi2sd xmm0, r9
+	unpcklpd xmm1, xmm0
+	divpd	xmm2, xmm1
+	paddq	xmm2, xmm14
+	cvttsd2si rax, xmm2
+	psrldq	xmm2, 8
+	mov	rbx, rax
+	imul	rax, rdx
+	sub	r11, rax
+	js	div_fix_1_ultralite_sandybridge
+div_fix_1_ret_ultralite_sandybridge:
+
+	cvttsd2si rdx, xmm2
+	mov	rax, rdx
+	imul	rax, r9
+	movd	xmm2, r11d
+	movd	xmm4, ebx
+	sub	r8, rax
+	js	div_fix_2_ultralite_sandybridge
+div_fix_2_ret_ultralite_sandybridge:
+
+	movd	xmm1, r8d
+	movd	xmm0, edx
+	punpckldq xmm2, xmm1
+	punpckldq xmm4, xmm0
+	punpckldq xmm4, xmm2
+	paddq	xmm3, xmm4
+	movdqa	xmm0, xmm3
+	psrlq	xmm0, 12
+	paddq	xmm0, xmm12
+	sqrtpd	xmm1, xmm0
+	movq	r9, xmm1
+	movdqa xmm5, xmm1
+	psrlq xmm5, 19
+	test	r9, 524287
+	je	sqrt_fix_1_ultralite_sandybridge
+sqrt_fix_1_ret_ultralite_sandybridge:
+
+	movq r9, xmm10
+	psrldq	xmm1, 8
+	movq	r8, xmm1
+	test	r8, 524287
+	je	sqrt_fix_2_ultralite_sandybridge
+sqrt_fix_2_ret_ultralite_sandybridge:
+
+	mov r12d, ecx
+	mov r8d, ecx
+	xor r12d, 16
+	xor r8d, 32
+	xor ecx, 48
+	mov	rax, r10
+	mul	r9
+	movq xmm0, rax
+	movq xmm3, rdx
+	punpcklqdq xmm3, xmm0
+
+	movdqu	xmm0, XMMWORD PTR [r12+rsi]
+	pxor xmm0, xmm3
+	movdqu	xmm1, XMMWORD PTR [r8+rsi]
+	xor rdx, [r8+rsi]
+	xor rax, [r8+rsi+8]
+	movdqu	xmm3, XMMWORD PTR [rcx+rsi]
+	paddq	xmm0, xmm6
+	paddq	xmm1, xmm11
+	paddq	xmm3, xmm8
+	movdqu	XMMWORD PTR [r8+rsi], xmm0
+	movdqu	XMMWORD PTR [rcx+rsi], xmm1
+	movdqu	XMMWORD PTR [r12+rsi], xmm3
+
+	add	rdi, rdx
+	mov	QWORD PTR [r13], rdi
+	xor	rdi, r10
+	mov	ecx, edi
+	and	ecx, 131056
+	lea	r8, QWORD PTR [rcx+rsi]
+
+	mov rdx, QWORD PTR [r13+8]	
+	add	rbp, rax
+	mov	QWORD PTR [r13+8], rbp
+	movdqu xmm11, XMMWORD PTR [rcx+rsi]
+	xor	rbp, rdx
+	mov	r13, QWORD PTR [rsp]
+	movdqa	xmm3, xmm7
+	mov	rdx, QWORD PTR [rsp+8]
+	movdqa	xmm8, xmm6
+	mov	r10, QWORD PTR [rsp+256]
+	movdqa	xmm7, xmm9
+	mov	r11, QWORD PTR [rsp+264]
+	movdqa	xmm6, xmm10
+	mov	r9, r15
+	dec r14d
+	jne	cnv2_double_main_loop_ultralite_sandybridge
+
+	ldmxcsr DWORD PTR [rsp+272]
+	movaps	xmm13, XMMWORD PTR [rsp+48]
+	lea	r11, QWORD PTR [rsp+184]
+	movaps	xmm6, XMMWORD PTR [r11-24]
+	movaps	xmm7, XMMWORD PTR [r11-40]
+	movaps	xmm8, XMMWORD PTR [r11-56]
+	movaps	xmm9, XMMWORD PTR [r11-72]
+	movaps	xmm10, XMMWORD PTR [r11-88]
+	movaps	xmm11, XMMWORD PTR [r11-104]
+	movaps	xmm12, XMMWORD PTR [r11-120]
+	movaps	xmm14, XMMWORD PTR [rsp+32]
+	movaps	xmm15, XMMWORD PTR [rsp+16]
+	mov	rsp, r11
+	pop	r15
+	pop	r14
+	pop	r13
+	pop	r12
+	pop	rdi
+	pop	rsi
+	pop	rbp
+	pop	rbx
+	jmp cnv2_double_main_loop_ultralite_sandybridge_endp
+
+div_fix_1_ultralite_sandybridge:
+	dec	rbx
+	add	r11, rdx
+	jmp	div_fix_1_ret_ultralite_sandybridge
+
+div_fix_2_ultralite_sandybridge:
+	dec	rdx
+	add	r8, r9
+	jmp	div_fix_2_ret_ultralite_sandybridge
+
+sqrt_fix_1_ultralite_sandybridge:
+	movq	r8, xmm3
+	movdqa xmm0, xmm5
+	psrldq xmm0, 8
+	dec	r9
+	mov r11d, -1022
+	shl r11, 32
+	mov	rax, r9
+	shr	r9, 19
+	shr	rax, 20
+	mov	rdx, r9
+	sub	rdx, rax
+	lea	rdx, [rdx+r11+1]
+	add	rax, r11
+	imul	rdx, rax
+	sub	rdx, r8
+	adc	r9, 0
+	movq xmm5, r9
+	punpcklqdq xmm5, xmm0
+	jmp	sqrt_fix_1_ret_ultralite_sandybridge
+
+sqrt_fix_2_ultralite_sandybridge:
+	psrldq	xmm3, 8
+	movq	r11, xmm3
+	dec	r8
+	mov ebx, -1022
+	shl rbx, 32
+	mov	rax, r8
+	shr	r8, 19
+	shr	rax, 20
+	mov	rdx, r8
+	sub	rdx, rax
+	lea	rdx, [rdx+rbx+1]
+	add	rax, rbx
+	imul	rdx, rax
+	sub	rdx, r11
+	adc	r8, 0
+	movq xmm0, r8
+	punpcklqdq xmm5, xmm0
+	jmp	sqrt_fix_2_ret_ultralite_sandybridge
+
+cnv2_double_main_loop_ultralite_sandybridge_endp:
diff --git a/src/crypto/asm/win/cnv2_main_loop_bulldozer.inc b/src/crypto/asm/win/cnv2_main_loop_bulldozer.inc
new file mode 100644
index 00000000..6bae7a3c
--- /dev/null
+++ b/src/crypto/asm/win/cnv2_main_loop_bulldozer.inc
@@ -0,0 +1,180 @@
+	mov	QWORD PTR [rsp+16], rbx
+	mov	QWORD PTR [rsp+24], rbp
+	mov	QWORD PTR [rsp+32], rsi
+	push	rdi
+	push	r12
+	push	r13
+	push	r14
+	push	r15
+	sub	rsp, 64
+
+	stmxcsr DWORD PTR [rsp]
+	mov DWORD PTR [rsp+4], 24448
+	ldmxcsr DWORD PTR [rsp+4]
+
+	mov	rax, QWORD PTR [rcx+48]
+	mov	r9, rcx
+	xor	rax, QWORD PTR [rcx+16]
+	mov	ebp, 524288
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	mov	r11, QWORD PTR [rcx+40]
+	mov	r10, r8
+	mov	rdx, QWORD PTR [rcx+56]
+	movd	xmm3, rax
+	xor	rdx, QWORD PTR [rcx+24]
+	xor	r11, QWORD PTR [rcx+8]
+	mov	rbx, QWORD PTR [rcx+224]
+	mov	rax, QWORD PTR [r9+80]
+	xor	rax, QWORD PTR [r9+64]
+	movd	xmm0, rdx
+	mov	rcx, QWORD PTR [rcx+88]
+	xor	rcx, QWORD PTR [r9+72]
+	mov	rdi, QWORD PTR [r9+104]
+	and	r10d, 2097136
+	movaps	XMMWORD PTR [rsp+48], xmm6
+	movd	xmm4, rax
+	movaps	XMMWORD PTR [rsp+32], xmm7
+	movaps	XMMWORD PTR [rsp+16], xmm8
+	xorps	xmm8, xmm8
+	mov ax, 1023
+	shl rax, 52
+	movd xmm7, rax
+	mov	r15, QWORD PTR [r9+96]
+	punpcklqdq xmm3, xmm0
+	movd	xmm0, rcx
+	punpcklqdq xmm4, xmm0
+
+	ALIGN 16
+cnv2_main_loop_originalv2_bulldozer:
+	movdqa	xmm5, XMMWORD PTR [r10+rbx]
+	movd xmm6, r8
+	pinsrq xmm6, r11, 1
+	lea	rdx, QWORD PTR [r10+rbx]
+	lea	r9, QWORD PTR [rdi+rdi]
+	shl	rdi, 32
+
+	mov	ecx, r10d
+	mov	eax, r10d
+	xor	ecx, 16
+	xor	eax, 32
+	xor	r10d, 48
+	aesenc	xmm5, xmm6
+	movdqa	xmm2, XMMWORD PTR [rcx+rbx]
+	movdqa	xmm1, XMMWORD PTR [rax+rbx]
+	movdqa	xmm0, XMMWORD PTR [r10+rbx]
+	paddq	xmm2, xmm3
+	paddq	xmm1, xmm6
+	paddq	xmm0, xmm4
+	movdqa	XMMWORD PTR [rcx+rbx], xmm0
+	movdqa	XMMWORD PTR [rax+rbx], xmm2
+	movdqa	XMMWORD PTR [r10+rbx], xmm1
+
+	movaps	xmm1, xmm8
+	mov	rsi, r15
+	xor	rsi, rdi
+
+	mov edi, 1023
+	shl rdi, 52
+
+	movd	r14, xmm5
+	pextrq rax, xmm5, 1
+
+	movdqa	xmm0, xmm5
+	pxor	xmm0, xmm3
+	mov	r10, r14
+	and	r10d, 2097136
+	movdqa	XMMWORD PTR [rdx], xmm0
+	xor	rsi, QWORD PTR [r10+rbx]
+	lea	r12, QWORD PTR [r10+rbx]
+	mov	r13, QWORD PTR [r10+rbx+8]
+
+	add	r9d, r14d
+	or	r9d, -2147483647
+	xor	edx, edx
+	div	r9
+	mov	eax, eax
+	shl	rdx, 32
+	lea	r15, [rax+rdx]
+	lea	rax, [r14+r15]
+	shr	rax, 12
+	add	rax, rdi
+	movd	xmm0, rax
+	sqrtsd	xmm1, xmm0
+	movd	rdi, xmm1
+	test	rdi, 524287
+	je	sqrt_fixup_originalv2_bulldozer
+	shr	rdi, 19
+
+sqrt_fixup_originalv2_bulldozer_ret:
+	mov	rax, rsi
+	mul	r14
+	movd xmm1, rax
+	movd xmm0, rdx
+	punpcklqdq xmm0, xmm1
+
+	mov	r9d, r10d
+	mov	ecx, r10d
+	xor	r9d, 16
+	xor	ecx, 32
+	xor	r10d, 48
+	movdqa	xmm1, XMMWORD PTR [rcx+rbx]
+	xor rdx, [rcx+rbx]
+	xor rax, [rcx+rbx+8]
+	movdqa	xmm2, XMMWORD PTR [r9+rbx]
+	pxor xmm2, xmm0
+	paddq xmm4, XMMWORD PTR [r10+rbx]
+	paddq	xmm2, xmm3
+	paddq	xmm1, xmm6
+	movdqa	XMMWORD PTR [r9+rbx], xmm4
+	movdqa	XMMWORD PTR [rcx+rbx], xmm2
+	movdqa	XMMWORD PTR [r10+rbx], xmm1
+
+	movdqa	xmm4, xmm3
+	add	r8, rdx
+	add	r11, rax
+	mov	QWORD PTR [r12], r8
+	xor	r8, rsi
+	mov	QWORD PTR [r12+8], r11
+	mov	r10, r8
+	xor	r11, r13
+	and	r10d, 2097136
+	movdqa	xmm3, xmm5
+	dec	ebp
+	jne	cnv2_main_loop_originalv2_bulldozer
+
+	ldmxcsr DWORD PTR [rsp]
+	movaps	xmm6, XMMWORD PTR [rsp+48]
+	lea	r11, QWORD PTR [rsp+64]
+	mov	rbx, QWORD PTR [r11+56]
+	mov	rbp, QWORD PTR [r11+64]
+	mov	rsi, QWORD PTR [r11+72]
+	movaps	xmm8, XMMWORD PTR [r11-48]
+	movaps	xmm7, XMMWORD PTR [rsp+32]
+	mov	rsp, r11
+	pop	r15
+	pop	r14
+	pop	r13
+	pop	r12
+	pop	rdi
+	jmp cnv2_main_loop_originalv2_bulldozer_endp
+
+sqrt_fixup_originalv2_bulldozer:
+	movd r9, xmm5
+	add r9, r15
+	dec	rdi
+	mov edx, -1022
+	shl rdx, 32
+	mov	rax, rdi
+	shr	rdi, 19
+	shr	rax, 20
+	mov	rcx, rdi
+	sub	rcx, rax
+	lea	rcx, [rcx+rdx+1]
+	add	rax, rdx
+	imul	rcx, rax
+	sub	rcx, r9
+	adc	rdi, 0
+	jmp	sqrt_fixup_originalv2_bulldozer_ret
+
+cnv2_main_loop_originalv2_bulldozer_endp:
diff --git a/src/crypto/asm/win/cnv2_main_loop_fastv2_bulldozer.inc b/src/crypto/asm/win/cnv2_main_loop_fastv2_bulldozer.inc
new file mode 100644
index 00000000..3915c632
--- /dev/null
+++ b/src/crypto/asm/win/cnv2_main_loop_fastv2_bulldozer.inc
@@ -0,0 +1,180 @@
+	mov	QWORD PTR [rsp+16], rbx
+	mov	QWORD PTR [rsp+24], rbp
+	mov	QWORD PTR [rsp+32], rsi
+	push	rdi
+	push	r12
+	push	r13
+	push	r14
+	push	r15
+	sub	rsp, 64
+
+	stmxcsr DWORD PTR [rsp]
+	mov DWORD PTR [rsp+4], 24448
+	ldmxcsr DWORD PTR [rsp+4]
+
+	mov	rax, QWORD PTR [rcx+48]
+	mov	r9, rcx
+	xor	rax, QWORD PTR [rcx+16]
+	mov	ebp, 262144
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	mov	r11, QWORD PTR [rcx+40]
+	mov	r10, r8
+	mov	rdx, QWORD PTR [rcx+56]
+	movd	xmm3, rax
+	xor	rdx, QWORD PTR [rcx+24]
+	xor	r11, QWORD PTR [rcx+8]
+	mov	rbx, QWORD PTR [rcx+224]
+	mov	rax, QWORD PTR [r9+80]
+	xor	rax, QWORD PTR [r9+64]
+	movd	xmm0, rdx
+	mov	rcx, QWORD PTR [rcx+88]
+	xor	rcx, QWORD PTR [r9+72]
+	mov	rdi, QWORD PTR [r9+104]
+	and	r10d, 2097136
+	movaps	XMMWORD PTR [rsp+48], xmm6
+	movd	xmm4, rax
+	movaps	XMMWORD PTR [rsp+32], xmm7
+	movaps	XMMWORD PTR [rsp+16], xmm8
+	xorps	xmm8, xmm8
+	mov ax, 1023
+	shl rax, 52
+	movd xmm7, rax
+	mov	r15, QWORD PTR [r9+96]
+	punpcklqdq xmm3, xmm0
+	movd	xmm0, rcx
+	punpcklqdq xmm4, xmm0
+
+	ALIGN 16
+cnv2_main_loop_fastv2_bulldozer:
+	movdqa	xmm5, XMMWORD PTR [r10+rbx]
+	movd xmm6, r8
+	pinsrq xmm6, r11, 1
+	lea	rdx, QWORD PTR [r10+rbx]
+	lea	r9, QWORD PTR [rdi+rdi]
+	shl	rdi, 32
+
+	mov	ecx, r10d
+	mov	eax, r10d
+	xor	ecx, 16
+	xor	eax, 32
+	xor	r10d, 48
+	aesenc	xmm5, xmm6
+	movdqa	xmm2, XMMWORD PTR [rcx+rbx]
+	movdqa	xmm1, XMMWORD PTR [rax+rbx]
+	movdqa	xmm0, XMMWORD PTR [r10+rbx]
+	paddq	xmm2, xmm3
+	paddq	xmm1, xmm6
+	paddq	xmm0, xmm4
+	movdqa	XMMWORD PTR [rcx+rbx], xmm0
+	movdqa	XMMWORD PTR [rax+rbx], xmm2
+	movdqa	XMMWORD PTR [r10+rbx], xmm1
+
+	movaps	xmm1, xmm8
+	mov	rsi, r15
+	xor	rsi, rdi
+
+	mov edi, 1023
+	shl rdi, 52
+
+	movd	r14, xmm5
+	pextrq rax, xmm5, 1
+
+	movdqa	xmm0, xmm5
+	pxor	xmm0, xmm3
+	mov	r10, r14
+	and	r10d, 2097136
+	movdqa	XMMWORD PTR [rdx], xmm0
+	xor	rsi, QWORD PTR [r10+rbx]
+	lea	r12, QWORD PTR [r10+rbx]
+	mov	r13, QWORD PTR [r10+rbx+8]
+
+	add	r9d, r14d
+	or	r9d, -2147483647
+	xor	edx, edx
+	div	r9
+	mov	eax, eax
+	shl	rdx, 32
+	lea	r15, [rax+rdx]
+	lea	rax, [r14+r15]
+	shr	rax, 12
+	add	rax, rdi
+	movd	xmm0, rax
+	sqrtsd	xmm1, xmm0
+	movd	rdi, xmm1
+	test	rdi, 524287
+	je	sqrt_fixup_fastv2_bulldozer
+	shr	rdi, 19
+
+sqrt_fixup_fastv2_bulldozer_ret:
+	mov	rax, rsi
+	mul	r14
+	movd xmm1, rax
+	movd xmm0, rdx
+	punpcklqdq xmm0, xmm1
+
+	mov	r9d, r10d
+	mov	ecx, r10d
+	xor	r9d, 16
+	xor	ecx, 32
+	xor	r10d, 48
+	movdqa	xmm1, XMMWORD PTR [rcx+rbx]
+	xor rdx, [rcx+rbx]
+	xor rax, [rcx+rbx+8]
+	movdqa	xmm2, XMMWORD PTR [r9+rbx]
+	pxor xmm2, xmm0
+	paddq xmm4, XMMWORD PTR [r10+rbx]
+	paddq	xmm2, xmm3
+	paddq	xmm1, xmm6
+	movdqa	XMMWORD PTR [r9+rbx], xmm4
+	movdqa	XMMWORD PTR [rcx+rbx], xmm2
+	movdqa	XMMWORD PTR [r10+rbx], xmm1
+
+	movdqa	xmm4, xmm3
+	add	r8, rdx
+	add	r11, rax
+	mov	QWORD PTR [r12], r8
+	xor	r8, rsi
+	mov	QWORD PTR [r12+8], r11
+	mov	r10, r8
+	xor	r11, r13
+	and	r10d, 2097136
+	movdqa	xmm3, xmm5
+	dec	ebp
+	jne	cnv2_main_loop_fastv2_bulldozer
+
+	ldmxcsr DWORD PTR [rsp]
+	movaps	xmm6, XMMWORD PTR [rsp+48]
+	lea	r11, QWORD PTR [rsp+64]
+	mov	rbx, QWORD PTR [r11+56]
+	mov	rbp, QWORD PTR [r11+64]
+	mov	rsi, QWORD PTR [r11+72]
+	movaps	xmm8, XMMWORD PTR [r11-48]
+	movaps	xmm7, XMMWORD PTR [rsp+32]
+	mov	rsp, r11
+	pop	r15
+	pop	r14
+	pop	r13
+	pop	r12
+	pop	rdi
+	jmp cnv2_main_loop_fastv2_bulldozer_endp
+
+sqrt_fixup_fastv2_bulldozer:
+	movd r9, xmm5
+	add r9, r15
+	dec	rdi
+	mov edx, -1022
+	shl rdx, 32
+	mov	rax, rdi
+	shr	rdi, 19
+	shr	rax, 20
+	mov	rcx, rdi
+	sub	rcx, rax
+	lea	rcx, [rcx+rdx+1]
+	add	rax, rdx
+	imul	rcx, rax
+	sub	rcx, r9
+	adc	rdi, 0
+	jmp	sqrt_fixup_fastv2_bulldozer_ret
+
+cnv2_main_loop_fastv2_bulldozer_endp:
diff --git a/src/crypto/asm/win/cnv2_main_loop_fastv2_ivybridge.inc b/src/crypto/asm/win/cnv2_main_loop_fastv2_ivybridge.inc
new file mode 100644
index 00000000..0d0bc4fe
--- /dev/null
+++ b/src/crypto/asm/win/cnv2_main_loop_fastv2_ivybridge.inc
@@ -0,0 +1,182 @@
+	mov	 QWORD PTR [rsp+24], rbx
+	push	 rbp
+	push	 rsi
+	push	 rdi
+	push	 r12
+	push	 r13
+	push	 r14
+	push	 r15
+	sub	 rsp, 80
+
+	stmxcsr DWORD PTR [rsp]
+	mov DWORD PTR [rsp+4], 24448
+	ldmxcsr DWORD PTR [rsp+4]
+
+	mov	 rax, QWORD PTR [rcx+48]
+	mov	 r9, rcx
+	xor	 rax, QWORD PTR [rcx+16]
+	mov	 esi, 262144
+	mov	 r8, QWORD PTR [rcx+32]
+	mov	 r13d, -2147483647
+	xor	 r8, QWORD PTR [rcx]
+	mov	 r11, QWORD PTR [rcx+40]
+	mov	 r10, r8
+	mov	 rdx, QWORD PTR [rcx+56]
+	movq	 xmm4, rax
+	xor	 rdx, QWORD PTR [rcx+24]
+	xor	 r11, QWORD PTR [rcx+8]
+	mov	 rbx, QWORD PTR [rcx+224]
+	mov	 rax, QWORD PTR [r9+80]
+	xor	 rax, QWORD PTR [r9+64]
+	movq	 xmm0, rdx
+	mov	 rcx, QWORD PTR [rcx+88]
+	xor	 rcx, QWORD PTR [r9+72]
+	movq	 xmm3, QWORD PTR [r9+104]
+	movaps	 XMMWORD PTR [rsp+64], xmm6
+	movaps	 XMMWORD PTR [rsp+48], xmm7
+	movaps	 XMMWORD PTR [rsp+32], xmm8
+	and	 r10d, 2097136
+	movq	 xmm5, rax
+
+	mov ax, 1023
+	shl rax, 52
+	movq xmm8, rax
+	mov r15, QWORD PTR [r9+96]
+	punpcklqdq xmm4, xmm0
+	movq	 xmm0, rcx
+	punpcklqdq xmm5, xmm0
+	movdqu	 xmm6, XMMWORD PTR [r10+rbx]
+
+    ALIGN 64
+cnv2_main_loop_fastv2_ivybridge:
+	lea	 rdx, QWORD PTR [r10+rbx]
+	mov	 ecx, r10d
+	mov	 eax, r10d
+	mov rdi, r15
+	xor	 ecx, 16
+	xor	 eax, 32
+	xor	 r10d, 48
+	movq	 xmm0, r11
+	movq	 xmm7, r8
+	punpcklqdq xmm7, xmm0
+	aesenc	 xmm6, xmm7
+	movq	 rbp, xmm6
+	mov	 r9, rbp
+	and	 r9d, 2097136
+	movdqu	 xmm2, XMMWORD PTR [rcx+rbx]
+	movdqu	 xmm1, XMMWORD PTR [rax+rbx]
+	movdqu	 xmm0, XMMWORD PTR [r10+rbx]
+	paddq	 xmm1, xmm7
+	paddq	 xmm0, xmm5
+	paddq	 xmm2, xmm4
+	movdqu	 XMMWORD PTR [rcx+rbx], xmm0
+	movdqu	 XMMWORD PTR [rax+rbx], xmm2
+	movdqu	 XMMWORD PTR [r10+rbx], xmm1
+	mov r10, r9
+	xor r10d, 32
+	movq	 rcx, xmm3
+	mov	 rax, rcx
+	shl	 rax, 32
+	xor	 rdi, rax
+	movdqa	 xmm0, xmm6
+	pxor	 xmm0, xmm4
+	movdqu	 XMMWORD PTR [rdx], xmm0
+	xor	 rdi, QWORD PTR [r9+rbx]
+	lea	 r14, QWORD PTR [r9+rbx]
+	mov	 r12, QWORD PTR [r14+8]
+	xor	 edx, edx
+	lea	 r9d, DWORD PTR [ecx+ecx]
+	add	 r9d, ebp
+	movdqa	 xmm0, xmm6
+	psrldq	 xmm0, 8
+	or	 r9d, r13d
+	movq	 rax, xmm0
+	div	 r9
+	xorps xmm3, xmm3
+	mov	 eax, eax
+	shl	 rdx, 32
+	add	 rdx, rax
+	lea	 r9, QWORD PTR [rdx+rbp]
+	mov r15, rdx
+	mov	 rax, r9
+	shr	 rax, 12
+	movq	 xmm0, rax
+	paddq	 xmm0, xmm8
+	sqrtsd	 xmm3, xmm0
+	movq	 rdx, xmm3
+	test	 edx, 524287
+	je	 sqrt_fixup_fastv2_ivybridge
+	psrlq	 xmm3, 19
+sqrt_fixup_fastv2_ivybridge_ret:
+
+	mov	 ecx, r10d
+	mov	 rax, rdi
+	mul	 rbp
+	movq xmm2, rdx
+	xor rdx, [rcx+rbx]
+	add	 r8, rdx
+	mov	 QWORD PTR [r14], r8
+	xor	 r8, rdi
+	mov edi, r8d
+	and edi, 2097136
+	movq xmm0, rax
+	xor rax, [rcx+rbx+8]
+	add	 r11, rax
+	mov	 QWORD PTR [r14+8], r11
+	punpcklqdq xmm2, xmm0
+
+	mov	 r9d, r10d
+	xor	 r9d, 48
+	xor	 r10d, 16
+	pxor	 xmm2, XMMWORD PTR [r9+rbx]
+	movdqu	 xmm0, XMMWORD PTR [r10+rbx]
+	paddq	 xmm0, xmm5
+	movdqu	 xmm1, XMMWORD PTR [rcx+rbx]
+	paddq	 xmm2, xmm4
+	paddq	 xmm1, xmm7
+	movdqa	 xmm5, xmm4
+	movdqu	 XMMWORD PTR [r9+rbx], xmm0
+	movdqa	 xmm4, xmm6
+	movdqu	 XMMWORD PTR [rcx+rbx], xmm2
+	movdqu	 XMMWORD PTR [r10+rbx], xmm1
+	movdqu xmm6, [rdi+rbx]
+	mov	 r10d, edi
+	xor	 r11, r12
+	dec rsi
+	jne	 cnv2_main_loop_fastv2_ivybridge
+
+	ldmxcsr DWORD PTR [rsp]
+	mov	 rbx, QWORD PTR [rsp+160]
+	movaps	 xmm6, XMMWORD PTR [rsp+64]
+	movaps	 xmm7, XMMWORD PTR [rsp+48]
+	movaps	 xmm8, XMMWORD PTR [rsp+32]
+	add	 rsp, 80
+	pop	 r15
+	pop	 r14
+	pop	 r13
+	pop	 r12
+	pop	 rdi
+	pop	 rsi
+	pop	 rbp
+	jmp cnv2_main_loop_fastv2_ivybridge_endp
+
+sqrt_fixup_fastv2_ivybridge:
+	dec	 rdx
+	mov r13d, -1022
+	shl r13, 32
+	mov	 rax, rdx
+	shr	 rdx, 19
+	shr	 rax, 20
+	mov	 rcx, rdx
+	sub	 rcx, rax
+	add	 rax, r13
+	not r13
+	sub	 rcx, r13
+	mov	 r13d, -2147483647
+	imul	 rcx, rax
+	sub	 rcx, r9
+	adc	 rdx, 0
+	movq	 xmm3, rdx
+	jmp	 sqrt_fixup_fastv2_ivybridge_ret
+
+cnv2_main_loop_fastv2_ivybridge_endp:
diff --git a/src/crypto/asm/win/cnv2_main_loop_fastv2_ryzen.inc b/src/crypto/asm/win/cnv2_main_loop_fastv2_ryzen.inc
new file mode 100644
index 00000000..33fc7804
--- /dev/null
+++ b/src/crypto/asm/win/cnv2_main_loop_fastv2_ryzen.inc
@@ -0,0 +1,179 @@
+	mov	QWORD PTR [rsp+16], rbx
+	mov	QWORD PTR [rsp+24], rbp
+	mov	QWORD PTR [rsp+32], rsi
+	push	rdi
+	push	r12
+	push	r13
+	push	r14
+	push	r15
+	sub	rsp, 64
+
+	stmxcsr DWORD PTR [rsp]
+	mov DWORD PTR [rsp+4], 24448
+	ldmxcsr DWORD PTR [rsp+4]
+
+	mov	rax, QWORD PTR [rcx+48]
+	mov	r9, rcx
+	xor	rax, QWORD PTR [rcx+16]
+	mov	ebp, 262144
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	mov	r11, QWORD PTR [rcx+40]
+	mov	r10, r8
+	mov	rdx, QWORD PTR [rcx+56]
+	movq	xmm3, rax
+	xor	rdx, QWORD PTR [rcx+24]
+	xor	r11, QWORD PTR [rcx+8]
+	mov	rbx, QWORD PTR [rcx+224]
+	mov	rax, QWORD PTR [r9+80]
+	xor	rax, QWORD PTR [r9+64]
+	movq	xmm0, rdx
+	mov	rcx, QWORD PTR [rcx+88]
+	xor	rcx, QWORD PTR [r9+72]
+	mov	rdi, QWORD PTR [r9+104]
+	and	r10d, 2097136
+	movaps	XMMWORD PTR [rsp+48], xmm6
+	movq	xmm4, rax
+	movaps	XMMWORD PTR [rsp+32], xmm7
+	movaps	XMMWORD PTR [rsp+16], xmm8
+	xorps	xmm8, xmm8
+	mov ax, 1023
+	shl rax, 52
+	movq xmm7, rax
+	mov	r15, QWORD PTR [r9+96]
+	punpcklqdq xmm3, xmm0
+	movq	xmm0, rcx
+	punpcklqdq xmm4, xmm0
+
+    ALIGN 64
+cnv2_main_loop_fastv2_ryzen:
+	movdqa	xmm5, XMMWORD PTR [r10+rbx]
+	movq	xmm0, r11
+	movq	xmm6, r8
+	punpcklqdq xmm6, xmm0
+	lea	rdx, QWORD PTR [r10+rbx]
+	lea	r9, QWORD PTR [rdi+rdi]
+	shl	rdi, 32
+
+	mov	ecx, r10d
+	mov	eax, r10d
+	xor	ecx, 16
+	xor	eax, 32
+	xor	r10d, 48
+	aesenc	xmm5, xmm6
+	movdqa	xmm2, XMMWORD PTR [rcx+rbx]
+	movdqa	xmm1, XMMWORD PTR [rax+rbx]
+	movdqa	xmm0, XMMWORD PTR [r10+rbx]
+	paddq	xmm2, xmm3
+	paddq	xmm1, xmm6
+	paddq	xmm0, xmm4
+	movdqa	XMMWORD PTR [rcx+rbx], xmm0
+	movdqa	XMMWORD PTR [rax+rbx], xmm2
+	movdqa	XMMWORD PTR [r10+rbx], xmm1
+
+	movaps	xmm1, xmm8
+	mov	rsi, r15
+	xor	rsi, rdi
+	movq	r14, xmm5
+	movdqa	xmm0, xmm5
+	pxor	xmm0, xmm3
+	mov	r10, r14
+	and	r10d, 2097136
+	movdqa	XMMWORD PTR [rdx], xmm0
+	xor	rsi, QWORD PTR [r10+rbx]
+	lea	r12, QWORD PTR [r10+rbx]
+	mov	r13, QWORD PTR [r10+rbx+8]
+
+	add	r9d, r14d
+	or	r9d, -2147483647
+	xor	edx, edx
+	movdqa	xmm0, xmm5
+	psrldq	xmm0, 8
+	movq	rax, xmm0
+
+	div	r9
+	movq xmm0, rax
+	movq xmm1, rdx
+	punpckldq xmm0, xmm1
+	movq r15, xmm0
+	paddq xmm0, xmm5
+	movdqa xmm2, xmm0
+	psrlq xmm0, 12
+	paddq	xmm0, xmm7
+	sqrtsd	xmm1, xmm0
+	movq	rdi, xmm1
+	test	rdi, 524287
+	je	sqrt_fixup_fastv2_ryzen
+	shr	rdi, 19
+
+sqrt_fixup_fastv2_ryzen_ret:
+	mov	rax, rsi
+	mul	r14
+	movq xmm1, rax
+	movq xmm0, rdx
+	punpcklqdq xmm0, xmm1
+
+	mov	r9d, r10d
+	mov	ecx, r10d
+	xor	r9d, 16
+	xor	ecx, 32
+	xor	r10d, 48
+	movdqa	xmm1, XMMWORD PTR [rcx+rbx]
+	xor rdx, [rcx+rbx]
+	xor rax, [rcx+rbx+8]
+	movdqa	xmm2, XMMWORD PTR [r9+rbx]
+	pxor xmm2, xmm0
+	paddq xmm4, XMMWORD PTR [r10+rbx]
+	paddq	xmm2, xmm3
+	paddq	xmm1, xmm6
+	movdqa	XMMWORD PTR [r9+rbx], xmm4
+	movdqa	XMMWORD PTR [rcx+rbx], xmm2
+	movdqa	XMMWORD PTR [r10+rbx], xmm1
+
+	movdqa	xmm4, xmm3
+	add	r8, rdx
+	add	r11, rax
+	mov	QWORD PTR [r12], r8
+	xor	r8, rsi
+	mov	QWORD PTR [r12+8], r11
+	mov	r10, r8
+	xor	r11, r13
+	and	r10d, 2097136
+	movdqa	xmm3, xmm5
+	dec	ebp
+	jne	cnv2_main_loop_fastv2_ryzen
+
+	ldmxcsr DWORD PTR [rsp]
+	movaps	xmm6, XMMWORD PTR [rsp+48]
+	lea	r11, QWORD PTR [rsp+64]
+	mov	rbx, QWORD PTR [r11+56]
+	mov	rbp, QWORD PTR [r11+64]
+	mov	rsi, QWORD PTR [r11+72]
+	movaps	xmm8, XMMWORD PTR [r11-48]
+	movaps	xmm7, XMMWORD PTR [rsp+32]
+	mov	rsp, r11
+	pop	r15
+	pop	r14
+	pop	r13
+	pop	r12
+	pop	rdi
+	jmp cnv2_main_loop_fastv2_ryzen_endp
+
+sqrt_fixup_fastv2_ryzen:
+	movq r9, xmm2
+	dec	rdi
+	mov edx, -1022
+	shl rdx, 32
+	mov	rax, rdi
+	shr	rdi, 19
+	shr	rax, 20
+	mov	rcx, rdi
+	sub	rcx, rax
+	lea	rcx, [rcx+rdx+1]
+	add	rax, rdx
+	imul	rcx, rax
+	sub	rcx, r9
+	adc	rdi, 0
+	jmp	sqrt_fixup_fastv2_ryzen_ret
+
+cnv2_main_loop_fastv2_ryzen_endp:
diff --git a/src/crypto/asm/win/cnv2_main_loop_fastv2_soft_aes_sandybridge.inc b/src/crypto/asm/win/cnv2_main_loop_fastv2_soft_aes_sandybridge.inc
new file mode 100644
index 00000000..06a899c0
--- /dev/null
+++ b/src/crypto/asm/win/cnv2_main_loop_fastv2_soft_aes_sandybridge.inc
@@ -0,0 +1,267 @@
+	mov	QWORD PTR [rsp+8], rcx
+	push	rbx
+	push	rbp
+	push	rsi
+	push	rdi
+	push	r12
+	push	r13
+	push	r14
+	push	r15
+	sub	rsp, 152
+
+	stmxcsr DWORD PTR [rsp+4]
+	mov DWORD PTR [rsp], 24448
+	ldmxcsr DWORD PTR [rsp]
+
+	mov	rax, QWORD PTR [rcx+48]
+	mov	r10, rcx
+	xor	rax, QWORD PTR [rcx+16]
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	mov	r9, QWORD PTR [rcx+40]
+	xor	r9, QWORD PTR [rcx+8]
+	movq	xmm4, rax
+	mov	rdx, QWORD PTR [rcx+56]
+	xor	rdx, QWORD PTR [rcx+24]
+	mov	r11, QWORD PTR [rcx+224]
+	mov	rcx, QWORD PTR [rcx+88]
+	xor	rcx, QWORD PTR [r10+72]
+	mov	rax, QWORD PTR [r10+80]
+	movq	xmm0, rdx
+	xor	rax, QWORD PTR [r10+64]
+
+	movaps	XMMWORD PTR [rsp+16], xmm6
+	movaps	XMMWORD PTR [rsp+32], xmm7
+	movaps	XMMWORD PTR [rsp+48], xmm8
+	movaps	XMMWORD PTR [rsp+64], xmm9
+	movaps	XMMWORD PTR [rsp+80], xmm10
+	movaps	XMMWORD PTR [rsp+96], xmm11
+	movaps	XMMWORD PTR [rsp+112], xmm12
+	movaps	XMMWORD PTR [rsp+128], xmm13
+
+	movq	xmm5, rax
+
+	mov ax, 1023
+	shl rax, 52
+	movq xmm8, rax
+
+	mov	rax, r8
+	punpcklqdq xmm4, xmm0
+	and	eax, 2097136
+	movq	xmm10, QWORD PTR [r10+96]
+	movq	xmm0, rcx
+	mov	rcx, QWORD PTR [r10+104]
+	xorps	xmm9, xmm9
+	mov	QWORD PTR [rsp+248], rax
+	movq	xmm12, r11
+	mov	QWORD PTR [rsp+240], r9
+	punpcklqdq xmm5, xmm0
+	movq xmm13, rcx
+	mov r12d, 262144
+
+	ALIGN 64
+cnv2_main_loop_fastv2_soft_aes_sandybridge:
+	movd xmm11, r12d
+	mov	r12, QWORD PTR [r10+272]
+	lea	r13, QWORD PTR [rax+r11]
+	mov	esi, DWORD PTR [r13]
+	movq	xmm0, r9
+	mov	r10d, DWORD PTR [r13+4]
+	movq	xmm7, r8
+	mov	ebp, DWORD PTR [r13+12]
+	mov	r14d, DWORD PTR [r13+8]
+	mov	rdx, QWORD PTR [rsp+248]
+	movzx	ecx, sil
+	shr	esi, 8
+	punpcklqdq xmm7, xmm0
+	mov	r15d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r10b
+	shr	r10d, 8
+	mov	edi, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r14b
+	shr	r14d, 8
+	mov	ebx, DWORD PTR [r12+rcx*4]
+	movzx	ecx, bpl
+	shr	ebp, 8
+	mov	r9d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r10b
+	shr	r10d, 8
+	xor	r15d, DWORD PTR [r12+rcx*4+1024]
+	movzx	ecx, r14b
+	shr	r14d, 8
+	mov	eax, r14d
+	shr	eax, 8
+	xor	edi, DWORD PTR [r12+rcx*4+1024]
+	add	eax, 256
+	movzx	ecx, bpl
+	shr	ebp, 8
+	xor	ebx, DWORD PTR [r12+rcx*4+1024]
+	movzx	ecx, sil
+	shr	esi, 8
+	xor	r9d, DWORD PTR [r12+rcx*4+1024]
+	add	r12, 2048
+	movzx	ecx, r10b
+	shr	r10d, 8
+	add	r10d, 256
+	mov	r11d, DWORD PTR [r12+rax*4]
+	xor	r11d, DWORD PTR [r12+rcx*4]
+	xor	r11d, r9d
+	movzx	ecx, sil
+	mov	r10d, DWORD PTR [r12+r10*4]
+	shr	esi, 8
+	add	esi, 256
+	xor	r10d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, bpl
+	xor	r10d, ebx
+	shr	ebp, 8
+	movd	xmm1, r11d
+	add	ebp, 256
+	movq	r11, xmm12
+	mov	r9d, DWORD PTR [r12+rcx*4]
+	xor	r9d, DWORD PTR [r12+rsi*4]
+	mov	eax, DWORD PTR [r12+rbp*4]
+	xor	r9d, edi
+	movzx	ecx, r14b
+	movd	xmm0, r10d
+	movd	xmm2, r9d
+	xor	eax, DWORD PTR [r12+rcx*4]
+	mov	rcx, rdx
+	xor	eax, r15d
+	punpckldq xmm2, xmm1
+	xor	rcx, 16
+	movd	xmm6, eax
+	mov	rax, rdx
+	punpckldq xmm6, xmm0
+	xor	rax, 32
+	punpckldq xmm6, xmm2
+	xor	rdx, 48
+	movdqu	xmm2, XMMWORD PTR [rcx+r11]
+	pxor	xmm6, xmm7
+	paddq	xmm2, xmm4
+	movdqu	xmm1, XMMWORD PTR [rax+r11]
+	movdqu	xmm0, XMMWORD PTR [rdx+r11]
+	paddq	xmm0, xmm5
+	movdqu	XMMWORD PTR [rcx+r11], xmm0
+	movdqu	XMMWORD PTR [rax+r11], xmm2
+	movq rcx, xmm13
+	paddq	xmm1, xmm7
+	movdqu	XMMWORD PTR [rdx+r11], xmm1
+	movq	rdi, xmm6
+	mov	r10, rdi
+	and	r10d, 2097136
+	xor	edx, edx
+	mov	rax, rcx
+	shl	rax, 32
+	movq	rbx, xmm10
+	xor	rbx, rax
+	lea	r9, QWORD PTR [rcx+rcx]
+	add	r9d, edi
+	movdqa	xmm0, xmm6
+	pxor	xmm0, xmm4
+	mov	ecx, -2147483647
+	movdqu	XMMWORD PTR [r13], xmm0
+	or	r9, rcx
+	movdqa	xmm0, xmm6
+	movaps	xmm1, xmm9
+	psrldq	xmm0, 8
+	movq	rax, xmm0
+	xor	rbx, QWORD PTR [r10+r11]
+	lea	r14, QWORD PTR [r10+r11]
+	mov	rbp, QWORD PTR [r14+8]
+	div	r9
+	shl	rdx, 32
+	mov	eax, eax
+	add	rdx, rax
+	lea	r9, QWORD PTR [rdx+rdi]
+	movq	xmm10, rdx
+	mov	rax, r9
+	shr	rax, 12
+	movq	xmm0, rax
+	paddq	xmm0, xmm8
+	sqrtsd	xmm1, xmm0
+	movq	rdx, xmm1
+	test	rdx, 524287
+	je	sqrt_fixup_fastv2_soft_aes_sandybridge
+	psrlq xmm1, 19
+sqrt_fixup_fastv2_soft_aes_sandybridge_ret:
+
+	mov	r9, r10
+	movdqa	xmm13, xmm1
+	xor	r9, 16
+	mov	rcx, r10
+	xor	rcx, 32
+	xor	r10, 48
+	mov	rax, rbx
+	mul	rdi
+	movdqu	xmm2, XMMWORD PTR [r9+r11]
+	movdqu	xmm1, XMMWORD PTR [rcx+r11]
+	paddq	xmm1, xmm7
+	movq	xmm0, rax
+	movq	xmm3, rdx
+	xor	rax, QWORD PTR [r11+rcx+8]
+	xor	rdx, QWORD PTR [rcx+r11]
+	punpcklqdq xmm3, xmm0
+	add	r8, rdx
+	movdqu	xmm0, XMMWORD PTR [r10+r11]
+	pxor	xmm2, xmm3
+	paddq	xmm0, xmm5
+	paddq	xmm2, xmm4
+	movdqu	XMMWORD PTR [r9+r11], xmm0
+	movdqa	xmm5, xmm4
+	mov	r9, QWORD PTR [rsp+240]
+	movdqa	xmm4, xmm6
+	add	r9, rax
+	movdqu	XMMWORD PTR [rcx+r11], xmm2
+	movdqu	XMMWORD PTR [r10+r11], xmm1
+	mov	r10, QWORD PTR [rsp+224]
+	movd r12d, xmm11
+	mov	QWORD PTR [r14], r8
+	xor	r8, rbx
+	mov	rax, r8
+	mov	QWORD PTR [r14+8], r9
+	and	eax, 2097136
+	xor	r9, rbp
+	mov	QWORD PTR [rsp+240], r9
+	mov	QWORD PTR [rsp+248], rax
+	sub	r12d, 1
+	jne	cnv2_main_loop_fastv2_soft_aes_sandybridge
+
+	ldmxcsr DWORD PTR [rsp+4]
+	movaps	xmm6, XMMWORD PTR [rsp+16]
+	movaps	xmm7, XMMWORD PTR [rsp+32]
+	movaps	xmm8, XMMWORD PTR [rsp+48]
+	movaps	xmm9, XMMWORD PTR [rsp+64]
+	movaps	xmm10, XMMWORD PTR [rsp+80]
+	movaps	xmm11, XMMWORD PTR [rsp+96]
+	movaps	xmm12, XMMWORD PTR [rsp+112]
+	movaps	xmm13, XMMWORD PTR [rsp+128]
+
+	add	rsp, 152
+	pop	r15
+	pop	r14
+	pop	r13
+	pop	r12
+	pop	rdi
+	pop	rsi
+	pop	rbp
+	pop	rbx
+	jmp cnv2_main_loop_fastv2_soft_aes_sandybridge_endp
+
+sqrt_fixup_fastv2_soft_aes_sandybridge:
+	dec	rdx
+	mov r15d, -1022
+	shl r15, 32
+	mov	rax, rdx
+	shr	rdx, 19
+	shr	rax, 20
+	mov	rcx, rdx
+	sub	rcx, rax
+	lea	rcx, [rcx+r15+1]
+	add	rax, r15
+	imul	rcx, rax
+	sub	rcx, r9
+	adc	rdx, 0
+	movq xmm1, rdx
+	jmp	sqrt_fixup_fastv2_soft_aes_sandybridge_ret
+
+cnv2_main_loop_fastv2_soft_aes_sandybridge_endp:
diff --git a/src/crypto/asm/win/cnv2_main_loop_ivybridge.inc b/src/crypto/asm/win/cnv2_main_loop_ivybridge.inc
new file mode 100644
index 00000000..aa2797c3
--- /dev/null
+++ b/src/crypto/asm/win/cnv2_main_loop_ivybridge.inc
@@ -0,0 +1,182 @@
+	mov	 QWORD PTR [rsp+24], rbx
+	push	 rbp
+	push	 rsi
+	push	 rdi
+	push	 r12
+	push	 r13
+	push	 r14
+	push	 r15
+	sub	 rsp, 80
+
+	stmxcsr DWORD PTR [rsp]
+	mov DWORD PTR [rsp+4], 24448
+	ldmxcsr DWORD PTR [rsp+4]
+
+	mov	 rax, QWORD PTR [rcx+48]
+	mov	 r9, rcx
+	xor	 rax, QWORD PTR [rcx+16]
+	mov	 esi, 524288
+	mov	 r8, QWORD PTR [rcx+32]
+	mov	 r13d, -2147483647
+	xor	 r8, QWORD PTR [rcx]
+	mov	 r11, QWORD PTR [rcx+40]
+	mov	 r10, r8
+	mov	 rdx, QWORD PTR [rcx+56]
+	movq	 xmm4, rax
+	xor	 rdx, QWORD PTR [rcx+24]
+	xor	 r11, QWORD PTR [rcx+8]
+	mov	 rbx, QWORD PTR [rcx+224]
+	mov	 rax, QWORD PTR [r9+80]
+	xor	 rax, QWORD PTR [r9+64]
+	movq	 xmm0, rdx
+	mov	 rcx, QWORD PTR [rcx+88]
+	xor	 rcx, QWORD PTR [r9+72]
+	movq	 xmm3, QWORD PTR [r9+104]
+	movaps	 XMMWORD PTR [rsp+64], xmm6
+	movaps	 XMMWORD PTR [rsp+48], xmm7
+	movaps	 XMMWORD PTR [rsp+32], xmm8
+	and	 r10d, 2097136
+	movq	 xmm5, rax
+
+	mov ax, 1023
+	shl rax, 52
+	movq xmm8, rax
+	mov r15, QWORD PTR [r9+96]
+	punpcklqdq xmm4, xmm0
+	movq	 xmm0, rcx
+	punpcklqdq xmm5, xmm0
+	movdqu	 xmm6, XMMWORD PTR [r10+rbx]
+
+    ALIGN 64
+cnv2_main_loop_originalv2_ivybridge:
+	lea	 rdx, QWORD PTR [r10+rbx]
+	mov	 ecx, r10d
+	mov	 eax, r10d
+	mov rdi, r15
+	xor	 ecx, 16
+	xor	 eax, 32
+	xor	 r10d, 48
+	movq	 xmm0, r11
+	movq	 xmm7, r8
+	punpcklqdq xmm7, xmm0
+	aesenc	 xmm6, xmm7
+	movq	 rbp, xmm6
+	mov	 r9, rbp
+	and	 r9d, 2097136
+	movdqu	 xmm2, XMMWORD PTR [rcx+rbx]
+	movdqu	 xmm1, XMMWORD PTR [rax+rbx]
+	movdqu	 xmm0, XMMWORD PTR [r10+rbx]
+	paddq	 xmm1, xmm7
+	paddq	 xmm0, xmm5
+	paddq	 xmm2, xmm4
+	movdqu	 XMMWORD PTR [rcx+rbx], xmm0
+	movdqu	 XMMWORD PTR [rax+rbx], xmm2
+	movdqu	 XMMWORD PTR [r10+rbx], xmm1
+	mov r10, r9
+	xor r10d, 32
+	movq	 rcx, xmm3
+	mov	 rax, rcx
+	shl	 rax, 32
+	xor	 rdi, rax
+	movdqa	 xmm0, xmm6
+	pxor	 xmm0, xmm4
+	movdqu	 XMMWORD PTR [rdx], xmm0
+	xor	 rdi, QWORD PTR [r9+rbx]
+	lea	 r14, QWORD PTR [r9+rbx]
+	mov	 r12, QWORD PTR [r14+8]
+	xor	 edx, edx
+	lea	 r9d, DWORD PTR [ecx+ecx]
+	add	 r9d, ebp
+	movdqa	 xmm0, xmm6
+	psrldq	 xmm0, 8
+	or	 r9d, r13d
+	movq	 rax, xmm0
+	div	 r9
+	xorps xmm3, xmm3
+	mov	 eax, eax
+	shl	 rdx, 32
+	add	 rdx, rax
+	lea	 r9, QWORD PTR [rdx+rbp]
+	mov r15, rdx
+	mov	 rax, r9
+	shr	 rax, 12
+	movq	 xmm0, rax
+	paddq	 xmm0, xmm8
+	sqrtsd	 xmm3, xmm0
+	movq	 rdx, xmm3
+	test	 edx, 524287
+	je	 sqrt_fixup_originalv2_ivybridge
+	psrlq	 xmm3, 19
+sqrt_fixup_originalv2_ivybridge_ret:
+
+	mov	 ecx, r10d
+	mov	 rax, rdi
+	mul	 rbp
+	movq xmm2, rdx
+	xor rdx, [rcx+rbx]
+	add	 r8, rdx
+	mov	 QWORD PTR [r14], r8
+	xor	 r8, rdi
+	mov edi, r8d
+	and edi, 2097136
+	movq xmm0, rax
+	xor rax, [rcx+rbx+8]
+	add	 r11, rax
+	mov	 QWORD PTR [r14+8], r11
+	punpcklqdq xmm2, xmm0
+
+	mov	 r9d, r10d
+	xor	 r9d, 48
+	xor	 r10d, 16
+	pxor	 xmm2, XMMWORD PTR [r9+rbx]
+	movdqu	 xmm0, XMMWORD PTR [r10+rbx]
+	paddq	 xmm0, xmm5
+	movdqu	 xmm1, XMMWORD PTR [rcx+rbx]
+	paddq	 xmm2, xmm4
+	paddq	 xmm1, xmm7
+	movdqa	 xmm5, xmm4
+	movdqu	 XMMWORD PTR [r9+rbx], xmm0
+	movdqa	 xmm4, xmm6
+	movdqu	 XMMWORD PTR [rcx+rbx], xmm2
+	movdqu	 XMMWORD PTR [r10+rbx], xmm1
+	movdqu xmm6, [rdi+rbx]
+	mov	 r10d, edi
+	xor	 r11, r12
+	dec rsi
+	jne	 cnv2_main_loop_originalv2_ivybridge
+
+	ldmxcsr DWORD PTR [rsp]
+	mov	 rbx, QWORD PTR [rsp+160]
+	movaps	 xmm6, XMMWORD PTR [rsp+64]
+	movaps	 xmm7, XMMWORD PTR [rsp+48]
+	movaps	 xmm8, XMMWORD PTR [rsp+32]
+	add	 rsp, 80
+	pop	 r15
+	pop	 r14
+	pop	 r13
+	pop	 r12
+	pop	 rdi
+	pop	 rsi
+	pop	 rbp
+	jmp cnv2_main_loop_originalv2_ivybridge_endp
+
+sqrt_fixup_originalv2_ivybridge:
+	dec	 rdx
+	mov r13d, -1022
+	shl r13, 32
+	mov	 rax, rdx
+	shr	 rdx, 19
+	shr	 rax, 20
+	mov	 rcx, rdx
+	sub	 rcx, rax
+	add	 rax, r13
+	not r13
+	sub	 rcx, r13
+	mov	 r13d, -2147483647
+	imul	 rcx, rax
+	sub	 rcx, r9
+	adc	 rdx, 0
+	movq	 xmm3, rdx
+	jmp	 sqrt_fixup_originalv2_ivybridge_ret
+
+cnv2_main_loop_originalv2_ivybridge_endp:
diff --git a/src/crypto/asm/win/cnv2_main_loop_ryzen.inc b/src/crypto/asm/win/cnv2_main_loop_ryzen.inc
new file mode 100644
index 00000000..5a0ab3b7
--- /dev/null
+++ b/src/crypto/asm/win/cnv2_main_loop_ryzen.inc
@@ -0,0 +1,179 @@
+	mov	QWORD PTR [rsp+16], rbx
+	mov	QWORD PTR [rsp+24], rbp
+	mov	QWORD PTR [rsp+32], rsi
+	push	rdi
+	push	r12
+	push	r13
+	push	r14
+	push	r15
+	sub	rsp, 64
+
+	stmxcsr DWORD PTR [rsp]
+	mov DWORD PTR [rsp+4], 24448
+	ldmxcsr DWORD PTR [rsp+4]
+
+	mov	rax, QWORD PTR [rcx+48]
+	mov	r9, rcx
+	xor	rax, QWORD PTR [rcx+16]
+	mov	ebp, 524288
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	mov	r11, QWORD PTR [rcx+40]
+	mov	r10, r8
+	mov	rdx, QWORD PTR [rcx+56]
+	movq	xmm3, rax
+	xor	rdx, QWORD PTR [rcx+24]
+	xor	r11, QWORD PTR [rcx+8]
+	mov	rbx, QWORD PTR [rcx+224]
+	mov	rax, QWORD PTR [r9+80]
+	xor	rax, QWORD PTR [r9+64]
+	movq	xmm0, rdx
+	mov	rcx, QWORD PTR [rcx+88]
+	xor	rcx, QWORD PTR [r9+72]
+	mov	rdi, QWORD PTR [r9+104]
+	and	r10d, 2097136
+	movaps	XMMWORD PTR [rsp+48], xmm6
+	movq	xmm4, rax
+	movaps	XMMWORD PTR [rsp+32], xmm7
+	movaps	XMMWORD PTR [rsp+16], xmm8
+	xorps	xmm8, xmm8
+	mov ax, 1023
+	shl rax, 52
+	movq xmm7, rax
+	mov	r15, QWORD PTR [r9+96]
+	punpcklqdq xmm3, xmm0
+	movq	xmm0, rcx
+	punpcklqdq xmm4, xmm0
+
+    ALIGN 64
+cnv2_main_loop_originalv2_ryzen:
+	movdqa	xmm5, XMMWORD PTR [r10+rbx]
+	movq	xmm0, r11
+	movq	xmm6, r8
+	punpcklqdq xmm6, xmm0
+	lea	rdx, QWORD PTR [r10+rbx]
+	lea	r9, QWORD PTR [rdi+rdi]
+	shl	rdi, 32
+
+	mov	ecx, r10d
+	mov	eax, r10d
+	xor	ecx, 16
+	xor	eax, 32
+	xor	r10d, 48
+	aesenc	xmm5, xmm6
+	movdqa	xmm2, XMMWORD PTR [rcx+rbx]
+	movdqa	xmm1, XMMWORD PTR [rax+rbx]
+	movdqa	xmm0, XMMWORD PTR [r10+rbx]
+	paddq	xmm2, xmm3
+	paddq	xmm1, xmm6
+	paddq	xmm0, xmm4
+	movdqa	XMMWORD PTR [rcx+rbx], xmm0
+	movdqa	XMMWORD PTR [rax+rbx], xmm2
+	movdqa	XMMWORD PTR [r10+rbx], xmm1
+
+	movaps	xmm1, xmm8
+	mov	rsi, r15
+	xor	rsi, rdi
+	movq	r14, xmm5
+	movdqa	xmm0, xmm5
+	pxor	xmm0, xmm3
+	mov	r10, r14
+	and	r10d, 2097136
+	movdqa	XMMWORD PTR [rdx], xmm0
+	xor	rsi, QWORD PTR [r10+rbx]
+	lea	r12, QWORD PTR [r10+rbx]
+	mov	r13, QWORD PTR [r10+rbx+8]
+
+	add	r9d, r14d
+	or	r9d, -2147483647
+	xor	edx, edx
+	movdqa	xmm0, xmm5
+	psrldq	xmm0, 8
+	movq	rax, xmm0
+
+	div	r9
+	movq xmm0, rax
+	movq xmm1, rdx
+	punpckldq xmm0, xmm1
+	movq r15, xmm0
+	paddq xmm0, xmm5
+	movdqa xmm2, xmm0
+	psrlq xmm0, 12
+	paddq	xmm0, xmm7
+	sqrtsd	xmm1, xmm0
+	movq	rdi, xmm1
+	test	rdi, 524287
+	je	sqrt_fixup_originalv2_ryzen
+	shr	rdi, 19
+
+sqrt_fixup_originalv2_ryzen_ret:
+	mov	rax, rsi
+	mul	r14
+	movq xmm1, rax
+	movq xmm0, rdx
+	punpcklqdq xmm0, xmm1
+
+	mov	r9d, r10d
+	mov	ecx, r10d
+	xor	r9d, 16
+	xor	ecx, 32
+	xor	r10d, 48
+	movdqa	xmm1, XMMWORD PTR [rcx+rbx]
+	xor rdx, [rcx+rbx]
+	xor rax, [rcx+rbx+8]
+	movdqa	xmm2, XMMWORD PTR [r9+rbx]
+	pxor xmm2, xmm0
+	paddq xmm4, XMMWORD PTR [r10+rbx]
+	paddq	xmm2, xmm3
+	paddq	xmm1, xmm6
+	movdqa	XMMWORD PTR [r9+rbx], xmm4
+	movdqa	XMMWORD PTR [rcx+rbx], xmm2
+	movdqa	XMMWORD PTR [r10+rbx], xmm1
+
+	movdqa	xmm4, xmm3
+	add	r8, rdx
+	add	r11, rax
+	mov	QWORD PTR [r12], r8
+	xor	r8, rsi
+	mov	QWORD PTR [r12+8], r11
+	mov	r10, r8
+	xor	r11, r13
+	and	r10d, 2097136
+	movdqa	xmm3, xmm5
+	dec	ebp
+	jne	cnv2_main_loop_originalv2_ryzen
+
+	ldmxcsr DWORD PTR [rsp]
+	movaps	xmm6, XMMWORD PTR [rsp+48]
+	lea	r11, QWORD PTR [rsp+64]
+	mov	rbx, QWORD PTR [r11+56]
+	mov	rbp, QWORD PTR [r11+64]
+	mov	rsi, QWORD PTR [r11+72]
+	movaps	xmm8, XMMWORD PTR [r11-48]
+	movaps	xmm7, XMMWORD PTR [rsp+32]
+	mov	rsp, r11
+	pop	r15
+	pop	r14
+	pop	r13
+	pop	r12
+	pop	rdi
+	jmp cnv2_main_loop_originalv2_ryzen_endp
+
+sqrt_fixup_originalv2_ryzen:
+	movq r9, xmm2
+	dec	rdi
+	mov edx, -1022
+	shl rdx, 32
+	mov	rax, rdi
+	shr	rdi, 19
+	shr	rax, 20
+	mov	rcx, rdi
+	sub	rcx, rax
+	lea	rcx, [rcx+rdx+1]
+	add	rax, rdx
+	imul	rcx, rax
+	sub	rcx, r9
+	adc	rdi, 0
+	jmp	sqrt_fixup_originalv2_ryzen_ret
+
+cnv2_main_loop_originalv2_ryzen_endp:
diff --git a/src/crypto/asm/win/cnv2_main_loop_soft_aes_sandybridge.inc b/src/crypto/asm/win/cnv2_main_loop_soft_aes_sandybridge.inc
new file mode 100644
index 00000000..64650a25
--- /dev/null
+++ b/src/crypto/asm/win/cnv2_main_loop_soft_aes_sandybridge.inc
@@ -0,0 +1,267 @@
+	mov	QWORD PTR [rsp+8], rcx
+	push	rbx
+	push	rbp
+	push	rsi
+	push	rdi
+	push	r12
+	push	r13
+	push	r14
+	push	r15
+	sub	rsp, 152
+
+	stmxcsr DWORD PTR [rsp+4]
+	mov DWORD PTR [rsp], 24448
+	ldmxcsr DWORD PTR [rsp]
+
+	mov	rax, QWORD PTR [rcx+48]
+	mov	r10, rcx
+	xor	rax, QWORD PTR [rcx+16]
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	mov	r9, QWORD PTR [rcx+40]
+	xor	r9, QWORD PTR [rcx+8]
+	movq	xmm4, rax
+	mov	rdx, QWORD PTR [rcx+56]
+	xor	rdx, QWORD PTR [rcx+24]
+	mov	r11, QWORD PTR [rcx+224]
+	mov	rcx, QWORD PTR [rcx+88]
+	xor	rcx, QWORD PTR [r10+72]
+	mov	rax, QWORD PTR [r10+80]
+	movq	xmm0, rdx
+	xor	rax, QWORD PTR [r10+64]
+
+	movaps	XMMWORD PTR [rsp+16], xmm6
+	movaps	XMMWORD PTR [rsp+32], xmm7
+	movaps	XMMWORD PTR [rsp+48], xmm8
+	movaps	XMMWORD PTR [rsp+64], xmm9
+	movaps	XMMWORD PTR [rsp+80], xmm10
+	movaps	XMMWORD PTR [rsp+96], xmm11
+	movaps	XMMWORD PTR [rsp+112], xmm12
+	movaps	XMMWORD PTR [rsp+128], xmm13
+
+	movq	xmm5, rax
+
+	mov ax, 1023
+	shl rax, 52
+	movq xmm8, rax
+
+	mov	rax, r8
+	punpcklqdq xmm4, xmm0
+	and	eax, 2097136
+	movq	xmm10, QWORD PTR [r10+96]
+	movq	xmm0, rcx
+	mov	rcx, QWORD PTR [r10+104]
+	xorps	xmm9, xmm9
+	mov	QWORD PTR [rsp+248], rax
+	movq	xmm12, r11
+	mov	QWORD PTR [rsp+240], r9
+	punpcklqdq xmm5, xmm0
+	movq xmm13, rcx
+	mov r12d, 524288
+
+	ALIGN 64
+cnv2_main_loop_originalv2_soft_aes_sandybridge:
+	movd xmm11, r12d
+	mov	r12, QWORD PTR [r10+272]
+	lea	r13, QWORD PTR [rax+r11]
+	mov	esi, DWORD PTR [r13]
+	movq	xmm0, r9
+	mov	r10d, DWORD PTR [r13+4]
+	movq	xmm7, r8
+	mov	ebp, DWORD PTR [r13+12]
+	mov	r14d, DWORD PTR [r13+8]
+	mov	rdx, QWORD PTR [rsp+248]
+	movzx	ecx, sil
+	shr	esi, 8
+	punpcklqdq xmm7, xmm0
+	mov	r15d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r10b
+	shr	r10d, 8
+	mov	edi, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r14b
+	shr	r14d, 8
+	mov	ebx, DWORD PTR [r12+rcx*4]
+	movzx	ecx, bpl
+	shr	ebp, 8
+	mov	r9d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r10b
+	shr	r10d, 8
+	xor	r15d, DWORD PTR [r12+rcx*4+1024]
+	movzx	ecx, r14b
+	shr	r14d, 8
+	mov	eax, r14d
+	shr	eax, 8
+	xor	edi, DWORD PTR [r12+rcx*4+1024]
+	add	eax, 256
+	movzx	ecx, bpl
+	shr	ebp, 8
+	xor	ebx, DWORD PTR [r12+rcx*4+1024]
+	movzx	ecx, sil
+	shr	esi, 8
+	xor	r9d, DWORD PTR [r12+rcx*4+1024]
+	add	r12, 2048
+	movzx	ecx, r10b
+	shr	r10d, 8
+	add	r10d, 256
+	mov	r11d, DWORD PTR [r12+rax*4]
+	xor	r11d, DWORD PTR [r12+rcx*4]
+	xor	r11d, r9d
+	movzx	ecx, sil
+	mov	r10d, DWORD PTR [r12+r10*4]
+	shr	esi, 8
+	add	esi, 256
+	xor	r10d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, bpl
+	xor	r10d, ebx
+	shr	ebp, 8
+	movd	xmm1, r11d
+	add	ebp, 256
+	movq	r11, xmm12
+	mov	r9d, DWORD PTR [r12+rcx*4]
+	xor	r9d, DWORD PTR [r12+rsi*4]
+	mov	eax, DWORD PTR [r12+rbp*4]
+	xor	r9d, edi
+	movzx	ecx, r14b
+	movd	xmm0, r10d
+	movd	xmm2, r9d
+	xor	eax, DWORD PTR [r12+rcx*4]
+	mov	rcx, rdx
+	xor	eax, r15d
+	punpckldq xmm2, xmm1
+	xor	rcx, 16
+	movd	xmm6, eax
+	mov	rax, rdx
+	punpckldq xmm6, xmm0
+	xor	rax, 32
+	punpckldq xmm6, xmm2
+	xor	rdx, 48
+	movdqu	xmm2, XMMWORD PTR [rcx+r11]
+	pxor	xmm6, xmm7
+	paddq	xmm2, xmm4
+	movdqu	xmm1, XMMWORD PTR [rax+r11]
+	movdqu	xmm0, XMMWORD PTR [rdx+r11]
+	paddq	xmm0, xmm5
+	movdqu	XMMWORD PTR [rcx+r11], xmm0
+	movdqu	XMMWORD PTR [rax+r11], xmm2
+	movq rcx, xmm13
+	paddq	xmm1, xmm7
+	movdqu	XMMWORD PTR [rdx+r11], xmm1
+	movq	rdi, xmm6
+	mov	r10, rdi
+	and	r10d, 2097136
+	xor	edx, edx
+	mov	rax, rcx
+	shl	rax, 32
+	movq	rbx, xmm10
+	xor	rbx, rax
+	lea	r9, QWORD PTR [rcx+rcx]
+	add	r9d, edi
+	movdqa	xmm0, xmm6
+	pxor	xmm0, xmm4
+	mov	ecx, -2147483647
+	movdqu	XMMWORD PTR [r13], xmm0
+	or	r9, rcx
+	movdqa	xmm0, xmm6
+	movaps	xmm1, xmm9
+	psrldq	xmm0, 8
+	movq	rax, xmm0
+	xor	rbx, QWORD PTR [r10+r11]
+	lea	r14, QWORD PTR [r10+r11]
+	mov	rbp, QWORD PTR [r14+8]
+	div	r9
+	shl	rdx, 32
+	mov	eax, eax
+	add	rdx, rax
+	lea	r9, QWORD PTR [rdx+rdi]
+	movq	xmm10, rdx
+	mov	rax, r9
+	shr	rax, 12
+	movq	xmm0, rax
+	paddq	xmm0, xmm8
+	sqrtsd	xmm1, xmm0
+	movq	rdx, xmm1
+	test	rdx, 524287
+	je	sqrt_fixup_originalv2_soft_aes_sandybridge
+	psrlq xmm1, 19
+sqrt_fixup_originalv2_soft_aes_sandybridge_ret:
+
+	mov	r9, r10
+	movdqa	xmm13, xmm1
+	xor	r9, 16
+	mov	rcx, r10
+	xor	rcx, 32
+	xor	r10, 48
+	mov	rax, rbx
+	mul	rdi
+	movdqu	xmm2, XMMWORD PTR [r9+r11]
+	movdqu	xmm1, XMMWORD PTR [rcx+r11]
+	paddq	xmm1, xmm7
+	movq	xmm0, rax
+	movq	xmm3, rdx
+	xor	rax, QWORD PTR [r11+rcx+8]
+	xor	rdx, QWORD PTR [rcx+r11]
+	punpcklqdq xmm3, xmm0
+	add	r8, rdx
+	movdqu	xmm0, XMMWORD PTR [r10+r11]
+	pxor	xmm2, xmm3
+	paddq	xmm0, xmm5
+	paddq	xmm2, xmm4
+	movdqu	XMMWORD PTR [r9+r11], xmm0
+	movdqa	xmm5, xmm4
+	mov	r9, QWORD PTR [rsp+240]
+	movdqa	xmm4, xmm6
+	add	r9, rax
+	movdqu	XMMWORD PTR [rcx+r11], xmm2
+	movdqu	XMMWORD PTR [r10+r11], xmm1
+	mov	r10, QWORD PTR [rsp+224]
+	movd r12d, xmm11
+	mov	QWORD PTR [r14], r8
+	xor	r8, rbx
+	mov	rax, r8
+	mov	QWORD PTR [r14+8], r9
+	and	eax, 2097136
+	xor	r9, rbp
+	mov	QWORD PTR [rsp+240], r9
+	mov	QWORD PTR [rsp+248], rax
+	sub	r12d, 1
+	jne	cnv2_main_loop_originalv2_soft_aes_sandybridge
+
+	ldmxcsr DWORD PTR [rsp+4]
+	movaps	xmm6, XMMWORD PTR [rsp+16]
+	movaps	xmm7, XMMWORD PTR [rsp+32]
+	movaps	xmm8, XMMWORD PTR [rsp+48]
+	movaps	xmm9, XMMWORD PTR [rsp+64]
+	movaps	xmm10, XMMWORD PTR [rsp+80]
+	movaps	xmm11, XMMWORD PTR [rsp+96]
+	movaps	xmm12, XMMWORD PTR [rsp+112]
+	movaps	xmm13, XMMWORD PTR [rsp+128]
+
+	add	rsp, 152
+	pop	r15
+	pop	r14
+	pop	r13
+	pop	r12
+	pop	rdi
+	pop	rsi
+	pop	rbp
+	pop	rbx
+	jmp cnv2_main_loop_originalv2_soft_aes_sandybridge_endp
+
+sqrt_fixup_originalv2_soft_aes_sandybridge:
+	dec	rdx
+	mov r15d, -1022
+	shl r15, 32
+	mov	rax, rdx
+	shr	rdx, 19
+	shr	rax, 20
+	mov	rcx, rdx
+	sub	rcx, rax
+	lea	rcx, [rcx+r15+1]
+	add	rax, r15
+	imul	rcx, rax
+	sub	rcx, r9
+	adc	rdx, 0
+	movq xmm1, rdx
+	jmp	sqrt_fixup_originalv2_soft_aes_sandybridge_ret
+
+cnv2_main_loop_originalv2_soft_aes_sandybridge_endp:
diff --git a/src/crypto/asm/win/cnv2_main_loop_ultralite_bulldozer.inc b/src/crypto/asm/win/cnv2_main_loop_ultralite_bulldozer.inc
new file mode 100644
index 00000000..ca48cae1
--- /dev/null
+++ b/src/crypto/asm/win/cnv2_main_loop_ultralite_bulldozer.inc
@@ -0,0 +1,180 @@
+	mov	QWORD PTR [rsp+16], rbx
+	mov	QWORD PTR [rsp+24], rbp
+	mov	QWORD PTR [rsp+32], rsi
+	push	rdi
+	push	r12
+	push	r13
+	push	r14
+	push	r15
+	sub	rsp, 64
+
+	stmxcsr DWORD PTR [rsp]
+	mov DWORD PTR [rsp+4], 24448
+	ldmxcsr DWORD PTR [rsp+4]
+
+	mov	rax, QWORD PTR [rcx+48]
+	mov	r9, rcx
+	xor	rax, QWORD PTR [rcx+16]
+	mov	ebp, 65536
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	mov	r11, QWORD PTR [rcx+40]
+	mov	r10, r8
+	mov	rdx, QWORD PTR [rcx+56]
+	movd	xmm3, rax
+	xor	rdx, QWORD PTR [rcx+24]
+	xor	r11, QWORD PTR [rcx+8]
+	mov	rbx, QWORD PTR [rcx+224]
+	mov	rax, QWORD PTR [r9+80]
+	xor	rax, QWORD PTR [r9+64]
+	movd	xmm0, rdx
+	mov	rcx, QWORD PTR [rcx+88]
+	xor	rcx, QWORD PTR [r9+72]
+	mov	rdi, QWORD PTR [r9+104]
+	and	r10d, 131056
+	movaps	XMMWORD PTR [rsp+48], xmm6
+	movd	xmm4, rax
+	movaps	XMMWORD PTR [rsp+32], xmm7
+	movaps	XMMWORD PTR [rsp+16], xmm8
+	xorps	xmm8, xmm8
+	mov ax, 1023
+	shl rax, 52
+	movd xmm7, rax
+	mov	r15, QWORD PTR [r9+96]
+	punpcklqdq xmm3, xmm0
+	movd	xmm0, rcx
+	punpcklqdq xmm4, xmm0
+
+	ALIGN 16
+cnv2_main_loop_ultralite_bulldozer:
+	movdqa	xmm5, XMMWORD PTR [r10+rbx]
+	movd xmm6, r8
+	pinsrq xmm6, r11, 1
+	lea	rdx, QWORD PTR [r10+rbx]
+	lea	r9, QWORD PTR [rdi+rdi]
+	shl	rdi, 32
+
+	mov	ecx, r10d
+	mov	eax, r10d
+	xor	ecx, 16
+	xor	eax, 32
+	xor	r10d, 48
+	aesenc	xmm5, xmm6
+	movdqa	xmm2, XMMWORD PTR [rcx+rbx]
+	movdqa	xmm1, XMMWORD PTR [rax+rbx]
+	movdqa	xmm0, XMMWORD PTR [r10+rbx]
+	paddq	xmm2, xmm3
+	paddq	xmm1, xmm6
+	paddq	xmm0, xmm4
+	movdqa	XMMWORD PTR [rcx+rbx], xmm0
+	movdqa	XMMWORD PTR [rax+rbx], xmm2
+	movdqa	XMMWORD PTR [r10+rbx], xmm1
+
+	movaps	xmm1, xmm8
+	mov	rsi, r15
+	xor	rsi, rdi
+
+	mov edi, 1023
+	shl rdi, 52
+
+	movd	r14, xmm5
+	pextrq rax, xmm5, 1
+
+	movdqa	xmm0, xmm5
+	pxor	xmm0, xmm3
+	mov	r10, r14
+	and	r10d, 131056
+	movdqa	XMMWORD PTR [rdx], xmm0
+	xor	rsi, QWORD PTR [r10+rbx]
+	lea	r12, QWORD PTR [r10+rbx]
+	mov	r13, QWORD PTR [r10+rbx+8]
+
+	add	r9d, r14d
+	or	r9d, -2147483647
+	xor	edx, edx
+	div	r9
+	mov	eax, eax
+	shl	rdx, 32
+	lea	r15, [rax+rdx]
+	lea	rax, [r14+r15]
+	shr	rax, 12
+	add	rax, rdi
+	movd	xmm0, rax
+	sqrtsd	xmm1, xmm0
+	movd	rdi, xmm1
+	test	rdi, 524287
+	je	sqrt_fixup_ultralite_bulldozer
+	shr	rdi, 19
+
+sqrt_fixup_ultralite_bulldozer_ret:
+	mov	rax, rsi
+	mul	r14
+	movd xmm1, rax
+	movd xmm0, rdx
+	punpcklqdq xmm0, xmm1
+
+	mov	r9d, r10d
+	mov	ecx, r10d
+	xor	r9d, 16
+	xor	ecx, 32
+	xor	r10d, 48
+	movdqa	xmm1, XMMWORD PTR [rcx+rbx]
+	xor rdx, [rcx+rbx]
+	xor rax, [rcx+rbx+8]
+	movdqa	xmm2, XMMWORD PTR [r9+rbx]
+	pxor xmm2, xmm0
+	paddq xmm4, XMMWORD PTR [r10+rbx]
+	paddq	xmm2, xmm3
+	paddq	xmm1, xmm6
+	movdqa	XMMWORD PTR [r9+rbx], xmm4
+	movdqa	XMMWORD PTR [rcx+rbx], xmm2
+	movdqa	XMMWORD PTR [r10+rbx], xmm1
+
+	movdqa	xmm4, xmm3
+	add	r8, rdx
+	add	r11, rax
+	mov	QWORD PTR [r12], r8
+	xor	r8, rsi
+	mov	QWORD PTR [r12+8], r11
+	mov	r10, r8
+	xor	r11, r13
+	and	r10d, 131056
+	movdqa	xmm3, xmm5
+	dec	ebp
+	jne	cnv2_main_loop_ultralite_bulldozer
+
+	ldmxcsr DWORD PTR [rsp]
+	movaps	xmm6, XMMWORD PTR [rsp+48]
+	lea	r11, QWORD PTR [rsp+64]
+	mov	rbx, QWORD PTR [r11+56]
+	mov	rbp, QWORD PTR [r11+64]
+	mov	rsi, QWORD PTR [r11+72]
+	movaps	xmm8, XMMWORD PTR [r11-48]
+	movaps	xmm7, XMMWORD PTR [rsp+32]
+	mov	rsp, r11
+	pop	r15
+	pop	r14
+	pop	r13
+	pop	r12
+	pop	rdi
+	jmp cnv2_main_loop_ultralite_bulldozer_endp
+
+sqrt_fixup_ultralite_bulldozer:
+	movd r9, xmm5
+	add r9, r15
+	dec	rdi
+	mov edx, -1022
+	shl rdx, 32
+	mov	rax, rdi
+	shr	rdi, 19
+	shr	rax, 20
+	mov	rcx, rdi
+	sub	rcx, rax
+	lea	rcx, [rcx+rdx+1]
+	add	rax, rdx
+	imul	rcx, rax
+	sub	rcx, r9
+	adc	rdi, 0
+	jmp	sqrt_fixup_ultralite_bulldozer_ret
+
+cnv2_main_loop_ultralite_bulldozer_endp:
diff --git a/src/crypto/asm/win/cnv2_main_loop_ultralite_ivybridge.inc b/src/crypto/asm/win/cnv2_main_loop_ultralite_ivybridge.inc
new file mode 100644
index 00000000..61aa261c
--- /dev/null
+++ b/src/crypto/asm/win/cnv2_main_loop_ultralite_ivybridge.inc
@@ -0,0 +1,182 @@
+	mov	 QWORD PTR [rsp+24], rbx
+	push	 rbp
+	push	 rsi
+	push	 rdi
+	push	 r12
+	push	 r13
+	push	 r14
+	push	 r15
+	sub	 rsp, 80
+
+	stmxcsr DWORD PTR [rsp]
+	mov DWORD PTR [rsp+4], 24448
+	ldmxcsr DWORD PTR [rsp+4]
+
+	mov	 rax, QWORD PTR [rcx+48]
+	mov	 r9, rcx
+	xor	 rax, QWORD PTR [rcx+16]
+	mov	 esi, 65536
+	mov	 r8, QWORD PTR [rcx+32]
+	mov	 r13d, -2147483647
+	xor	 r8, QWORD PTR [rcx]
+	mov	 r11, QWORD PTR [rcx+40]
+	mov	 r10, r8
+	mov	 rdx, QWORD PTR [rcx+56]
+	movq	 xmm4, rax
+	xor	 rdx, QWORD PTR [rcx+24]
+	xor	 r11, QWORD PTR [rcx+8]
+	mov	 rbx, QWORD PTR [rcx+224]
+	mov	 rax, QWORD PTR [r9+80]
+	xor	 rax, QWORD PTR [r9+64]
+	movq	 xmm0, rdx
+	mov	 rcx, QWORD PTR [rcx+88]
+	xor	 rcx, QWORD PTR [r9+72]
+	movq	 xmm3, QWORD PTR [r9+104]
+	movaps	 XMMWORD PTR [rsp+64], xmm6
+	movaps	 XMMWORD PTR [rsp+48], xmm7
+	movaps	 XMMWORD PTR [rsp+32], xmm8
+	and	 r10d, 131056
+	movq	 xmm5, rax
+
+	mov ax, 1023
+	shl rax, 52
+	movq xmm8, rax
+	mov r15, QWORD PTR [r9+96]
+	punpcklqdq xmm4, xmm0
+	movq	 xmm0, rcx
+	punpcklqdq xmm5, xmm0
+	movdqu	 xmm6, XMMWORD PTR [r10+rbx]
+
+    ALIGN 64
+cnv2_main_loop_ultralite_ivybridge:
+	lea	 rdx, QWORD PTR [r10+rbx]
+	mov	 ecx, r10d
+	mov	 eax, r10d
+	mov rdi, r15
+	xor	 ecx, 16
+	xor	 eax, 32
+	xor	 r10d, 48
+	movq	 xmm0, r11
+	movq	 xmm7, r8
+	punpcklqdq xmm7, xmm0
+	aesenc	 xmm6, xmm7
+	movq	 rbp, xmm6
+	mov	 r9, rbp
+	and	 r9d, 131056
+	movdqu	 xmm2, XMMWORD PTR [rcx+rbx]
+	movdqu	 xmm1, XMMWORD PTR [rax+rbx]
+	movdqu	 xmm0, XMMWORD PTR [r10+rbx]
+	paddq	 xmm1, xmm7
+	paddq	 xmm0, xmm5
+	paddq	 xmm2, xmm4
+	movdqu	 XMMWORD PTR [rcx+rbx], xmm0
+	movdqu	 XMMWORD PTR [rax+rbx], xmm2
+	movdqu	 XMMWORD PTR [r10+rbx], xmm1
+	mov r10, r9
+	xor r10d, 32
+	movq	 rcx, xmm3
+	mov	 rax, rcx
+	shl	 rax, 32
+	xor	 rdi, rax
+	movdqa	 xmm0, xmm6
+	pxor	 xmm0, xmm4
+	movdqu	 XMMWORD PTR [rdx], xmm0
+	xor	 rdi, QWORD PTR [r9+rbx]
+	lea	 r14, QWORD PTR [r9+rbx]
+	mov	 r12, QWORD PTR [r14+8]
+	xor	 edx, edx
+	lea	 r9d, DWORD PTR [ecx+ecx]
+	add	 r9d, ebp
+	movdqa	 xmm0, xmm6
+	psrldq	 xmm0, 8
+	or	 r9d, r13d
+	movq	 rax, xmm0
+	div	 r9
+	xorps xmm3, xmm3
+	mov	 eax, eax
+	shl	 rdx, 32
+	add	 rdx, rax
+	lea	 r9, QWORD PTR [rdx+rbp]
+	mov r15, rdx
+	mov	 rax, r9
+	shr	 rax, 12
+	movq	 xmm0, rax
+	paddq	 xmm0, xmm8
+	sqrtsd	 xmm3, xmm0
+	movq	 rdx, xmm3
+	test	 edx, 524287
+	je	 sqrt_fixup_ultralite_ivybridge
+	psrlq	 xmm3, 19
+sqrt_fixup_ultralite_ivybridge_ret:
+
+	mov	 ecx, r10d
+	mov	 rax, rdi
+	mul	 rbp
+	movq xmm2, rdx
+	xor rdx, [rcx+rbx]
+	add	 r8, rdx
+	mov	 QWORD PTR [r14], r8
+	xor	 r8, rdi
+	mov edi, r8d
+	and edi, 131056
+	movq xmm0, rax
+	xor rax, [rcx+rbx+8]
+	add	 r11, rax
+	mov	 QWORD PTR [r14+8], r11
+	punpcklqdq xmm2, xmm0
+
+	mov	 r9d, r10d
+	xor	 r9d, 48
+	xor	 r10d, 16
+	pxor	 xmm2, XMMWORD PTR [r9+rbx]
+	movdqu	 xmm0, XMMWORD PTR [r10+rbx]
+	paddq	 xmm0, xmm5
+	movdqu	 xmm1, XMMWORD PTR [rcx+rbx]
+	paddq	 xmm2, xmm4
+	paddq	 xmm1, xmm7
+	movdqa	 xmm5, xmm4
+	movdqu	 XMMWORD PTR [r9+rbx], xmm0
+	movdqa	 xmm4, xmm6
+	movdqu	 XMMWORD PTR [rcx+rbx], xmm2
+	movdqu	 XMMWORD PTR [r10+rbx], xmm1
+	movdqu xmm6, [rdi+rbx]
+	mov	 r10d, edi
+	xor	 r11, r12
+	dec rsi
+	jne	 cnv2_main_loop_ultralite_ivybridge
+
+	ldmxcsr DWORD PTR [rsp]
+	mov	 rbx, QWORD PTR [rsp+160]
+	movaps	 xmm6, XMMWORD PTR [rsp+64]
+	movaps	 xmm7, XMMWORD PTR [rsp+48]
+	movaps	 xmm8, XMMWORD PTR [rsp+32]
+	add	 rsp, 80
+	pop	 r15
+	pop	 r14
+	pop	 r13
+	pop	 r12
+	pop	 rdi
+	pop	 rsi
+	pop	 rbp
+	jmp cnv2_main_loop_ultralite_ivybridge_endp
+
+sqrt_fixup_ultralite_ivybridge:
+	dec	 rdx
+	mov r13d, -1022
+	shl r13, 32
+	mov	 rax, rdx
+	shr	 rdx, 19
+	shr	 rax, 20
+	mov	 rcx, rdx
+	sub	 rcx, rax
+	add	 rax, r13
+	not r13
+	sub	 rcx, r13
+	mov	 r13d, -2147483647
+	imul	 rcx, rax
+	sub	 rcx, r9
+	adc	 rdx, 0
+	movq	 xmm3, rdx
+	jmp	 sqrt_fixup_ultralite_ivybridge_ret
+
+cnv2_main_loop_ultralite_ivybridge_endp:
diff --git a/src/crypto/asm/win/cnv2_main_loop_ultralite_ryzen.inc b/src/crypto/asm/win/cnv2_main_loop_ultralite_ryzen.inc
new file mode 100644
index 00000000..0933ec35
--- /dev/null
+++ b/src/crypto/asm/win/cnv2_main_loop_ultralite_ryzen.inc
@@ -0,0 +1,179 @@
+	mov	QWORD PTR [rsp+16], rbx
+	mov	QWORD PTR [rsp+24], rbp
+	mov	QWORD PTR [rsp+32], rsi
+	push	rdi
+	push	r12
+	push	r13
+	push	r14
+	push	r15
+	sub	rsp, 64
+
+	stmxcsr DWORD PTR [rsp]
+	mov DWORD PTR [rsp+4], 24448
+	ldmxcsr DWORD PTR [rsp+4]
+
+	mov	rax, QWORD PTR [rcx+48]
+	mov	r9, rcx
+	xor	rax, QWORD PTR [rcx+16]
+	mov	ebp, 65536
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	mov	r11, QWORD PTR [rcx+40]
+	mov	r10, r8
+	mov	rdx, QWORD PTR [rcx+56]
+	movq	xmm3, rax
+	xor	rdx, QWORD PTR [rcx+24]
+	xor	r11, QWORD PTR [rcx+8]
+	mov	rbx, QWORD PTR [rcx+224]
+	mov	rax, QWORD PTR [r9+80]
+	xor	rax, QWORD PTR [r9+64]
+	movq	xmm0, rdx
+	mov	rcx, QWORD PTR [rcx+88]
+	xor	rcx, QWORD PTR [r9+72]
+	mov	rdi, QWORD PTR [r9+104]
+	and	r10d, 131056
+	movaps	XMMWORD PTR [rsp+48], xmm6
+	movq	xmm4, rax
+	movaps	XMMWORD PTR [rsp+32], xmm7
+	movaps	XMMWORD PTR [rsp+16], xmm8
+	xorps	xmm8, xmm8
+	mov ax, 1023
+	shl rax, 52
+	movq xmm7, rax
+	mov	r15, QWORD PTR [r9+96]
+	punpcklqdq xmm3, xmm0
+	movq	xmm0, rcx
+	punpcklqdq xmm4, xmm0
+
+    ALIGN 64
+cnv2_main_loop_ultralite_ryzen:
+	movdqa	xmm5, XMMWORD PTR [r10+rbx]
+	movq	xmm0, r11
+	movq	xmm6, r8
+	punpcklqdq xmm6, xmm0
+	lea	rdx, QWORD PTR [r10+rbx]
+	lea	r9, QWORD PTR [rdi+rdi]
+	shl	rdi, 32
+
+	mov	ecx, r10d
+	mov	eax, r10d
+	xor	ecx, 16
+	xor	eax, 32
+	xor	r10d, 48
+	aesenc	xmm5, xmm6
+	movdqa	xmm2, XMMWORD PTR [rcx+rbx]
+	movdqa	xmm1, XMMWORD PTR [rax+rbx]
+	movdqa	xmm0, XMMWORD PTR [r10+rbx]
+	paddq	xmm2, xmm3
+	paddq	xmm1, xmm6
+	paddq	xmm0, xmm4
+	movdqa	XMMWORD PTR [rcx+rbx], xmm0
+	movdqa	XMMWORD PTR [rax+rbx], xmm2
+	movdqa	XMMWORD PTR [r10+rbx], xmm1
+
+	movaps	xmm1, xmm8
+	mov	rsi, r15
+	xor	rsi, rdi
+	movq	r14, xmm5
+	movdqa	xmm0, xmm5
+	pxor	xmm0, xmm3
+	mov	r10, r14
+	and	r10d, 131056
+	movdqa	XMMWORD PTR [rdx], xmm0
+	xor	rsi, QWORD PTR [r10+rbx]
+	lea	r12, QWORD PTR [r10+rbx]
+	mov	r13, QWORD PTR [r10+rbx+8]
+
+	add	r9d, r14d
+	or	r9d, -2147483647
+	xor	edx, edx
+	movdqa	xmm0, xmm5
+	psrldq	xmm0, 8
+	movq	rax, xmm0
+
+	div	r9
+	movq xmm0, rax
+	movq xmm1, rdx
+	punpckldq xmm0, xmm1
+	movq r15, xmm0
+	paddq xmm0, xmm5
+	movdqa xmm2, xmm0
+	psrlq xmm0, 12
+	paddq	xmm0, xmm7
+	sqrtsd	xmm1, xmm0
+	movq	rdi, xmm1
+	test	rdi, 524287
+	je	sqrt_fixup_ultralite_ryzen
+	shr	rdi, 19
+
+sqrt_fixup_ultralite_ryzen_ret:
+	mov	rax, rsi
+	mul	r14
+	movq xmm1, rax
+	movq xmm0, rdx
+	punpcklqdq xmm0, xmm1
+
+	mov	r9d, r10d
+	mov	ecx, r10d
+	xor	r9d, 16
+	xor	ecx, 32
+	xor	r10d, 48
+	movdqa	xmm1, XMMWORD PTR [rcx+rbx]
+	xor rdx, [rcx+rbx]
+	xor rax, [rcx+rbx+8]
+	movdqa	xmm2, XMMWORD PTR [r9+rbx]
+	pxor xmm2, xmm0
+	paddq xmm4, XMMWORD PTR [r10+rbx]
+	paddq	xmm2, xmm3
+	paddq	xmm1, xmm6
+	movdqa	XMMWORD PTR [r9+rbx], xmm4
+	movdqa	XMMWORD PTR [rcx+rbx], xmm2
+	movdqa	XMMWORD PTR [r10+rbx], xmm1
+
+	movdqa	xmm4, xmm3
+	add	r8, rdx
+	add	r11, rax
+	mov	QWORD PTR [r12], r8
+	xor	r8, rsi
+	mov	QWORD PTR [r12+8], r11
+	mov	r10, r8
+	xor	r11, r13
+	and	r10d, 131056
+	movdqa	xmm3, xmm5
+	dec	ebp
+	jne	cnv2_main_loop_ultralite_ryzen
+
+	ldmxcsr DWORD PTR [rsp]
+	movaps	xmm6, XMMWORD PTR [rsp+48]
+	lea	r11, QWORD PTR [rsp+64]
+	mov	rbx, QWORD PTR [r11+56]
+	mov	rbp, QWORD PTR [r11+64]
+	mov	rsi, QWORD PTR [r11+72]
+	movaps	xmm8, XMMWORD PTR [r11-48]
+	movaps	xmm7, XMMWORD PTR [rsp+32]
+	mov	rsp, r11
+	pop	r15
+	pop	r14
+	pop	r13
+	pop	r12
+	pop	rdi
+	jmp cnv2_main_loop_ultralite_ryzen_endp
+
+sqrt_fixup_ultralite_ryzen:
+	movq r9, xmm2
+	dec	rdi
+	mov edx, -1022
+	shl rdx, 32
+	mov	rax, rdi
+	shr	rdi, 19
+	shr	rax, 20
+	mov	rcx, rdi
+	sub	rcx, rax
+	lea	rcx, [rcx+rdx+1]
+	add	rax, rdx
+	imul	rcx, rax
+	sub	rcx, r9
+	adc	rdi, 0
+	jmp	sqrt_fixup_ultralite_ryzen_ret
+
+cnv2_main_loop_ultralite_ryzen_endp:
diff --git a/src/crypto/asm/win/cnv2_main_loop_ultralite_soft_aes_sandybridge.inc b/src/crypto/asm/win/cnv2_main_loop_ultralite_soft_aes_sandybridge.inc
new file mode 100644
index 00000000..408254ee
--- /dev/null
+++ b/src/crypto/asm/win/cnv2_main_loop_ultralite_soft_aes_sandybridge.inc
@@ -0,0 +1,267 @@
+	mov	QWORD PTR [rsp+8], rcx
+	push	rbx
+	push	rbp
+	push	rsi
+	push	rdi
+	push	r12
+	push	r13
+	push	r14
+	push	r15
+	sub	rsp, 152
+
+	stmxcsr DWORD PTR [rsp+4]
+	mov DWORD PTR [rsp], 24448
+	ldmxcsr DWORD PTR [rsp]
+
+	mov	rax, QWORD PTR [rcx+48]
+	mov	r10, rcx
+	xor	rax, QWORD PTR [rcx+16]
+	mov	r8, QWORD PTR [rcx+32]
+	xor	r8, QWORD PTR [rcx]
+	mov	r9, QWORD PTR [rcx+40]
+	xor	r9, QWORD PTR [rcx+8]
+	movq	xmm4, rax
+	mov	rdx, QWORD PTR [rcx+56]
+	xor	rdx, QWORD PTR [rcx+24]
+	mov	r11, QWORD PTR [rcx+224]
+	mov	rcx, QWORD PTR [rcx+88]
+	xor	rcx, QWORD PTR [r10+72]
+	mov	rax, QWORD PTR [r10+80]
+	movq	xmm0, rdx
+	xor	rax, QWORD PTR [r10+64]
+
+	movaps	XMMWORD PTR [rsp+16], xmm6
+	movaps	XMMWORD PTR [rsp+32], xmm7
+	movaps	XMMWORD PTR [rsp+48], xmm8
+	movaps	XMMWORD PTR [rsp+64], xmm9
+	movaps	XMMWORD PTR [rsp+80], xmm10
+	movaps	XMMWORD PTR [rsp+96], xmm11
+	movaps	XMMWORD PTR [rsp+112], xmm12
+	movaps	XMMWORD PTR [rsp+128], xmm13
+
+	movq	xmm5, rax
+
+	mov ax, 1023
+	shl rax, 52
+	movq xmm8, rax
+
+	mov	rax, r8
+	punpcklqdq xmm4, xmm0
+	and	eax, 131056
+	movq	xmm10, QWORD PTR [r10+96]
+	movq	xmm0, rcx
+	mov	rcx, QWORD PTR [r10+104]
+	xorps	xmm9, xmm9
+	mov	QWORD PTR [rsp+248], rax
+	movq	xmm12, r11
+	mov	QWORD PTR [rsp+240], r9
+	punpcklqdq xmm5, xmm0
+	movq xmm13, rcx
+	mov r12d, 65536
+
+	ALIGN 64
+cnv2_main_loop_ultralite_soft_aes_sandybridge:
+	movd xmm11, r12d
+	mov	r12, QWORD PTR [r10+272]
+	lea	r13, QWORD PTR [rax+r11]
+	mov	esi, DWORD PTR [r13]
+	movq	xmm0, r9
+	mov	r10d, DWORD PTR [r13+4]
+	movq	xmm7, r8
+	mov	ebp, DWORD PTR [r13+12]
+	mov	r14d, DWORD PTR [r13+8]
+	mov	rdx, QWORD PTR [rsp+248]
+	movzx	ecx, sil
+	shr	esi, 8
+	punpcklqdq xmm7, xmm0
+	mov	r15d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r10b
+	shr	r10d, 8
+	mov	edi, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r14b
+	shr	r14d, 8
+	mov	ebx, DWORD PTR [r12+rcx*4]
+	movzx	ecx, bpl
+	shr	ebp, 8
+	mov	r9d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, r10b
+	shr	r10d, 8
+	xor	r15d, DWORD PTR [r12+rcx*4+1024]
+	movzx	ecx, r14b
+	shr	r14d, 8
+	mov	eax, r14d
+	shr	eax, 8
+	xor	edi, DWORD PTR [r12+rcx*4+1024]
+	add	eax, 256
+	movzx	ecx, bpl
+	shr	ebp, 8
+	xor	ebx, DWORD PTR [r12+rcx*4+1024]
+	movzx	ecx, sil
+	shr	esi, 8
+	xor	r9d, DWORD PTR [r12+rcx*4+1024]
+	add	r12, 2048
+	movzx	ecx, r10b
+	shr	r10d, 8
+	add	r10d, 256
+	mov	r11d, DWORD PTR [r12+rax*4]
+	xor	r11d, DWORD PTR [r12+rcx*4]
+	xor	r11d, r9d
+	movzx	ecx, sil
+	mov	r10d, DWORD PTR [r12+r10*4]
+	shr	esi, 8
+	add	esi, 256
+	xor	r10d, DWORD PTR [r12+rcx*4]
+	movzx	ecx, bpl
+	xor	r10d, ebx
+	shr	ebp, 8
+	movd	xmm1, r11d
+	add	ebp, 256
+	movq	r11, xmm12
+	mov	r9d, DWORD PTR [r12+rcx*4]
+	xor	r9d, DWORD PTR [r12+rsi*4]
+	mov	eax, DWORD PTR [r12+rbp*4]
+	xor	r9d, edi
+	movzx	ecx, r14b
+	movd	xmm0, r10d
+	movd	xmm2, r9d
+	xor	eax, DWORD PTR [r12+rcx*4]
+	mov	rcx, rdx
+	xor	eax, r15d
+	punpckldq xmm2, xmm1
+	xor	rcx, 16
+	movd	xmm6, eax
+	mov	rax, rdx
+	punpckldq xmm6, xmm0
+	xor	rax, 32
+	punpckldq xmm6, xmm2
+	xor	rdx, 48
+	movdqu	xmm2, XMMWORD PTR [rcx+r11]
+	pxor	xmm6, xmm7
+	paddq	xmm2, xmm4
+	movdqu	xmm1, XMMWORD PTR [rax+r11]
+	movdqu	xmm0, XMMWORD PTR [rdx+r11]
+	paddq	xmm0, xmm5
+	movdqu	XMMWORD PTR [rcx+r11], xmm0
+	movdqu	XMMWORD PTR [rax+r11], xmm2
+	movq rcx, xmm13
+	paddq	xmm1, xmm7
+	movdqu	XMMWORD PTR [rdx+r11], xmm1
+	movq	rdi, xmm6
+	mov	r10, rdi
+	and	r10d, 131056
+	xor	edx, edx
+	mov	rax, rcx
+	shl	rax, 32
+	movq	rbx, xmm10
+	xor	rbx, rax
+	lea	r9, QWORD PTR [rcx+rcx]
+	add	r9d, edi
+	movdqa	xmm0, xmm6
+	pxor	xmm0, xmm4
+	mov	ecx, -2147483647
+	movdqu	XMMWORD PTR [r13], xmm0
+	or	r9, rcx
+	movdqa	xmm0, xmm6
+	movaps	xmm1, xmm9
+	psrldq	xmm0, 8
+	movq	rax, xmm0
+	xor	rbx, QWORD PTR [r10+r11]
+	lea	r14, QWORD PTR [r10+r11]
+	mov	rbp, QWORD PTR [r14+8]
+	div	r9
+	shl	rdx, 32
+	mov	eax, eax
+	add	rdx, rax
+	lea	r9, QWORD PTR [rdx+rdi]
+	movq	xmm10, rdx
+	mov	rax, r9
+	shr	rax, 12
+	movq	xmm0, rax
+	paddq	xmm0, xmm8
+	sqrtsd	xmm1, xmm0
+	movq	rdx, xmm1
+	test	rdx, 524287
+	je	sqrt_fixup_ultralite_soft_aes_sandybridge
+	psrlq xmm1, 19
+sqrt_fixup_ultralite_soft_aes_sandybridge_ret:
+
+	mov	r9, r10
+	movdqa	xmm13, xmm1
+	xor	r9, 16
+	mov	rcx, r10
+	xor	rcx, 32
+	xor	r10, 48
+	mov	rax, rbx
+	mul	rdi
+	movdqu	xmm2, XMMWORD PTR [r9+r11]
+	movdqu	xmm1, XMMWORD PTR [rcx+r11]
+	paddq	xmm1, xmm7
+	movq	xmm0, rax
+	movq	xmm3, rdx
+	xor	rax, QWORD PTR [r11+rcx+8]
+	xor	rdx, QWORD PTR [rcx+r11]
+	punpcklqdq xmm3, xmm0
+	add	r8, rdx
+	movdqu	xmm0, XMMWORD PTR [r10+r11]
+	pxor	xmm2, xmm3
+	paddq	xmm0, xmm5
+	paddq	xmm2, xmm4
+	movdqu	XMMWORD PTR [r9+r11], xmm0
+	movdqa	xmm5, xmm4
+	mov	r9, QWORD PTR [rsp+240]
+	movdqa	xmm4, xmm6
+	add	r9, rax
+	movdqu	XMMWORD PTR [rcx+r11], xmm2
+	movdqu	XMMWORD PTR [r10+r11], xmm1
+	mov	r10, QWORD PTR [rsp+224]
+	movd r12d, xmm11
+	mov	QWORD PTR [r14], r8
+	xor	r8, rbx
+	mov	rax, r8
+	mov	QWORD PTR [r14+8], r9
+	and	eax, 131056
+	xor	r9, rbp
+	mov	QWORD PTR [rsp+240], r9
+	mov	QWORD PTR [rsp+248], rax
+	sub	r12d, 1
+	jne	cnv2_main_loop_ultralite_soft_aes_sandybridge
+
+	ldmxcsr DWORD PTR [rsp+4]
+	movaps	xmm6, XMMWORD PTR [rsp+16]
+	movaps	xmm7, XMMWORD PTR [rsp+32]
+	movaps	xmm8, XMMWORD PTR [rsp+48]
+	movaps	xmm9, XMMWORD PTR [rsp+64]
+	movaps	xmm10, XMMWORD PTR [rsp+80]
+	movaps	xmm11, XMMWORD PTR [rsp+96]
+	movaps	xmm12, XMMWORD PTR [rsp+112]
+	movaps	xmm13, XMMWORD PTR [rsp+128]
+
+	add	rsp, 152
+	pop	r15
+	pop	r14
+	pop	r13
+	pop	r12
+	pop	rdi
+	pop	rsi
+	pop	rbp
+	pop	rbx
+	jmp cnv2_main_loop_ultralite_soft_aes_sandybridge_endp
+
+sqrt_fixup_ultralite_soft_aes_sandybridge:
+	dec	rdx
+	mov r15d, -1022
+	shl r15, 32
+	mov	rax, rdx
+	shr	rdx, 19
+	shr	rax, 20
+	mov	rcx, rdx
+	sub	rcx, rax
+	lea	rcx, [rcx+r15+1]
+	add	rax, r15
+	imul	rcx, rax
+	sub	rcx, r9
+	adc	rdx, 0
+	movq xmm1, rdx
+	jmp	sqrt_fixup_ultralite_soft_aes_sandybridge_ret
+
+cnv2_main_loop_ultralite_soft_aes_sandybridge_endp: