crypto: x86/aes-gcm - rewrite the AES-NI optimized AES-GCM

Rewrite the AES-NI implementations of AES-GCM, taking advantage of
things I learned while writing the VAES-AVX10 implementations.  This is
a complete rewrite that reduces the AES-NI GCM source code size by about
70% and the binary code size by about 95%, while not regressing
performance and in fact improving it significantly in many cases.

The following summarizes the state before this patch:

- The aesni-intel module registered algorithms "generic-gcm-aesni" and
  "rfc4106-gcm-aesni" with the crypto API that actually delegated to one
  of three underlying implementations according to the CPU capabilities
  detected at runtime: AES-NI, AES-NI + AVX, or AES-NI + AVX2.

- The AES-NI + AVX and AES-NI + AVX2 assembly code was in
  aesni-intel_avx-x86_64.S and consisted of 2804 lines of source and
  257 KB of binary.  This massive binary size was not really
  appropriate, and depending on the kconfig it could take up over 1% the
  size of the entire vmlinux.  The main loops did 8 blocks per
  iteration.  The AVX code minimized the use of carryless multiplication
  whereas the AVX2 code did not.  The "AVX2" code did not actually use
  AVX2; the check for AVX2 was really a check for Intel Haswell or later
  to detect support for fast carryless multiplication.  The long source
  length was caused by factors such as significant code duplication.

- The AES-NI only assembly code was in aesni-intel_asm.S and consisted
  of 1501 lines of source and 15 KB of binary.  The main loops did 4
  blocks per iteration and minimized the use of carryless multiplication
  by using Karatsuba multiplication and a multiplication-less reduction.

- The assembly code was contributed in 2010-2013.  Maintenance has been
  sporadic and most design choices haven't been revisited.

- The assembly function prototypes and the corresponding glue code were
  separate from and were not consistent with the new VAES-AVX10 code I
  recently added.  The older code had several issues such as not
  precomputing the GHASH key powers, which hurt performance.

This rewrite achieves the following goals:

- Much shorter source and binary sizes.  The assembly source shrinks
  from 4300 lines to 1130 lines, and it produces about 9 KB of binary
  instead of 272 KB.  This is achieved via a better designed AES-GCM
  implementation that doesn't excessively unroll the code and instead
  prioritizes the parts that really matter.  Sharing the C glue code
  with the VAES-AVX10 implementations also saves 250 lines of C source.

- Improve performance on most (possibly all) CPUs on which this code
  runs, for most (possibly all) message lengths.  Benchmark results are
  given in Tables 1 and 2 below.

- Use the same function prototypes and glue code as the new VAES-AVX10
  algorithms.  This fixes some issues with the integration of the
  assembly and results in some significant performance improvements,
  primarily on short messages.  Also, the AVX and non-AVX
  implementations are now registered as separate algorithms with the
  crypto API, which makes them both testable by the self-tests.

- Keep support for AES-NI without AVX (for Westmere, Silvermont,
  Goldmont, and Tremont), but unify the source code with AES-NI + AVX.
  Since 256-bit vectors cannot be used without VAES anyway, this is made
  feasible by just using the non-VEX coded form of most instructions.

- Use a unified approach where the main loop does 8 blocks per iteration
  and uses Karatsuba multiplication to save one pclmulqdq per block but
  does not use the multiplication-less reduction.  This strikes a good
  balance across the range of CPUs on which this code runs.

- Don't spam the kernel log with an informational message on every boot.

The following tables summarize the improvement in AES-GCM throughput on
various CPU microarchitectures as a result of this patch:

Table 1: AES-256-GCM encryption throughput improvement,
         CPU microarchitecture vs. message length in bytes:

                   | 16384 |  4096 |  4095 |  1420 |   512 |   500 |
-------------------+-------+-------+-------+-------+-------+-------+
Intel Broadwell    |    2% |    8% |   11% |   18% |   31% |   26% |
Intel Skylake      |    1% |    4% |    7% |   12% |   26% |   19% |
Intel Cascade Lake |    3% |    8% |   10% |   18% |   33% |   24% |
AMD Zen 1          |    6% |   12% |    6% |   15% |   27% |   24% |
AMD Zen 2          |    8% |   13% |   13% |   19% |   26% |   28% |
AMD Zen 3          |    8% |   14% |   13% |   19% |   26% |   25% |

                   |   300 |   200 |    64 |    63 |    16 |
-------------------+-------+-------+-------+-------+-------+
Intel Broadwell    |   35% |   29% |   45% |   55% |   54% |
Intel Skylake      |   25% |   19% |   28% |   33% |   27% |
Intel Cascade Lake |   36% |   28% |   39% |   49% |   54% |
AMD Zen 1          |   27% |   22% |   23% |   29% |   26% |
AMD Zen 2          |   32% |   24% |   22% |   25% |   31% |
AMD Zen 3          |   30% |   24% |   22% |   23% |   26% |

Table 2: AES-256-GCM decryption throughput improvement,
         CPU microarchitecture vs. message length in bytes:

                   | 16384 |  4096 |  4095 |  1420 |   512 |   500 |
-------------------+-------+-------+-------+-------+-------+-------+
Intel Broadwell    |    3% |    8% |   11% |   19% |   32% |   28% |
Intel Skylake      |    3% |    4% |    7% |   13% |   28% |   27% |
Intel Cascade Lake |    3% |    9% |   11% |   19% |   33% |   28% |
AMD Zen 1          |   15% |   18% |   14% |   20% |   36% |   33% |
AMD Zen 2          |    9% |   16% |   13% |   21% |   26% |   27% |
AMD Zen 3          |    8% |   15% |   12% |   18% |   23% |   23% |

                   |   300 |   200 |    64 |    63 |    16 |
-------------------+-------+-------+-------+-------+-------+
Intel Broadwell    |   36% |   31% |   40% |   51% |   53% |
Intel Skylake      |   28% |   21% |   23% |   30% |   30% |
Intel Cascade Lake |   36% |   29% |   36% |   47% |   53% |
AMD Zen 1          |   35% |   31% |   32% |   35% |   36% |
AMD Zen 2          |   31% |   30% |   27% |   38% |   30% |
AMD Zen 3          |   27% |   23% |   24% |   32% |   26% |

The above numbers are percentage improvements in single-thread
throughput, so e.g. an increase from 3000 MB/s to 3300 MB/s would be
listed as 10%.  They were collected by directly measuring the Linux
crypto API performance using a custom kernel module.  Note that indirect
benchmarks (e.g. 'cryptsetup benchmark' or benchmarking dm-crypt I/O)
include more overhead and won't see quite as much of a difference.  All
these benchmarks used an associated data length of 16 bytes.  Note that
AES-GCM is almost always used with short associated data lengths.

I didn't test Intel CPUs before Broadwell, AMD CPUs before Zen 1, or
Intel low-power CPUs, as these weren't readily available to me.
However, based on the design of the new code and the available
information about these other CPU microarchitectures, I wouldn't expect
any significant regressions, and there's a good chance performance is
improved just as it is above.

Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

1 files changed, 1 insertions, 1502 deletions

diff --git a/arch/x86/crypto/aesni-intel_asm.S b/arch/x86/crypto/aesni-intel_asm.S
index 39066b57a70e..eb153eff9331 100644
--- a/arch/x86/crypto/aesni-intel_asm.S
+++ b/arch/x86/crypto/aesni-intel_asm.S
@@ -10,16 +10,7 @@
  *            Vinodh Gopal <vinodh.gopal@intel.com>
  *            Kahraman Akdemir
  *
- * Added RFC4106 AES-GCM support for 128-bit keys under the AEAD
- * interface for 64-bit kernels.
- *    Authors: Erdinc Ozturk (erdinc.ozturk@intel.com)
- *             Aidan O'Mahony (aidan.o.mahony@intel.com)
- *             Adrian Hoban <adrian.hoban@intel.com>
- *             James Guilford (james.guilford@intel.com)
- *             Gabriele Paoloni <gabriele.paoloni@intel.com>
- *             Tadeusz Struk (tadeusz.struk@intel.com)
- *             Wajdi Feghali (wajdi.k.feghali@intel.com)
- *    Copyright (c) 2010, Intel Corporation.
+ * Copyright (c) 2010, Intel Corporation.
  *
  * Ported x86_64 version to x86:
  *    Author: Mathias Krause <minipli@googlemail.com>
@@ -27,95 +18,6 @@
 
 #include <linux/linkage.h>
 #include <asm/frame.h>
-#include <asm/nospec-branch.h>
-
-/*
- * The following macros are used to move an (un)aligned 16 byte value to/from
- * an XMM register.  This can done for either FP or integer values, for FP use
- * movaps (move aligned packed single) or integer use movdqa (move double quad
- * aligned).  It doesn't make a performance difference which instruction is used
- * since Nehalem (original Core i7) was released.  However, the movaps is a byte
- * shorter, so that is the one we'll use for now. (same for unaligned).
- */
-#define MOVADQ	movaps
-#define MOVUDQ	movups
-
-#ifdef __x86_64__
-
-# constants in mergeable sections, linker can reorder and merge
-.section	.rodata.cst16.POLY, "aM", @progbits, 16
-.align 16
-POLY:   .octa 0xC2000000000000000000000000000001
-.section	.rodata.cst16.TWOONE, "aM", @progbits, 16
-.align 16
-TWOONE: .octa 0x00000001000000000000000000000001
-
-.section	.rodata.cst16.SHUF_MASK, "aM", @progbits, 16
-.align 16
-SHUF_MASK:  .octa 0x000102030405060708090A0B0C0D0E0F
-.section	.rodata.cst16.MASK1, "aM", @progbits, 16
-.align 16
-MASK1:      .octa 0x0000000000000000ffffffffffffffff
-.section	.rodata.cst16.MASK2, "aM", @progbits, 16
-.align 16
-MASK2:      .octa 0xffffffffffffffff0000000000000000
-.section	.rodata.cst16.ONE, "aM", @progbits, 16
-.align 16
-ONE:        .octa 0x00000000000000000000000000000001
-.section	.rodata.cst16.F_MIN_MASK, "aM", @progbits, 16
-.align 16
-F_MIN_MASK: .octa 0xf1f2f3f4f5f6f7f8f9fafbfcfdfeff0
-.section	.rodata.cst16.dec, "aM", @progbits, 16
-.align 16
-dec:        .octa 0x1
-.section	.rodata.cst16.enc, "aM", @progbits, 16
-.align 16
-enc:        .octa 0x2
-
-# order of these constants should not change.
-# more specifically, ALL_F should follow SHIFT_MASK,
-# and zero should follow ALL_F
-.section	.rodata, "a", @progbits
-.align 16
-SHIFT_MASK: .octa 0x0f0e0d0c0b0a09080706050403020100
-ALL_F:      .octa 0xffffffffffffffffffffffffffffffff
-            .octa 0x00000000000000000000000000000000
-
-.text
-
-#define AadHash 16*0
-#define AadLen 16*1
-#define InLen (16*1)+8
-#define PBlockEncKey 16*2
-#define OrigIV 16*3
-#define CurCount 16*4
-#define PBlockLen 16*5
-#define	HashKey		16*6	// store HashKey <<1 mod poly here
-#define	HashKey_2	16*7	// store HashKey^2 <<1 mod poly here
-#define	HashKey_3	16*8	// store HashKey^3 <<1 mod poly here
-#define	HashKey_4	16*9	// store HashKey^4 <<1 mod poly here
-#define	HashKey_k	16*10	// store XOR of High 64 bits and Low 64
-				// bits of  HashKey <<1 mod poly here
-				//(for Karatsuba purposes)
-#define	HashKey_2_k	16*11	// store XOR of High 64 bits and Low 64
-				// bits of  HashKey^2 <<1 mod poly here
-				// (for Karatsuba purposes)
-#define	HashKey_3_k	16*12	// store XOR of High 64 bits and Low 64
-				// bits of  HashKey^3 <<1 mod poly here
-				// (for Karatsuba purposes)
-#define	HashKey_4_k	16*13	// store XOR of High 64 bits and Low 64
-				// bits of  HashKey^4 <<1 mod poly here
-				// (for Karatsuba purposes)
-
-#define arg1 rdi
-#define arg2 rsi
-#define arg3 rdx
-#define arg4 rcx
-#define arg5 r8
-#define arg6 r9
-#define keysize 2*15*16(%arg1)
-#endif
-
 
 #define STATE1	%xmm0
 #define STATE2	%xmm4
@@ -162,1409 +64,6 @@ ALL_F:      .octa 0xffffffffffffffffffffffffffffffff
 #define TKEYP	T1
 #endif
 
-.macro FUNC_SAVE
-	push	%r12
-	push	%r13
-	push	%r14
-#
-# states of %xmm registers %xmm6:%xmm15 not saved
-# all %xmm registers are clobbered
-#
-.endm
-
-
-.macro FUNC_RESTORE
-	pop	%r14
-	pop	%r13
-	pop	%r12
-.endm
-
-# Precompute hashkeys.
-# Input: Hash subkey.
-# Output: HashKeys stored in gcm_context_data.  Only needs to be called
-# once per key.
-# clobbers r12, and tmp xmm registers.
-.macro PRECOMPUTE SUBKEY TMP1 TMP2 TMP3 TMP4 TMP5 TMP6 TMP7
-	mov	\SUBKEY, %r12
-	movdqu	(%r12), \TMP3
-	movdqa	SHUF_MASK(%rip), \TMP2
-	pshufb	\TMP2, \TMP3
-
-	# precompute HashKey<<1 mod poly from the HashKey (required for GHASH)
-
-	movdqa	\TMP3, \TMP2
-	psllq	$1, \TMP3
-	psrlq	$63, \TMP2
-	movdqa	\TMP2, \TMP1
-	pslldq	$8, \TMP2
-	psrldq	$8, \TMP1
-	por	\TMP2, \TMP3
-
-	# reduce HashKey<<1
-
-	pshufd	$0x24, \TMP1, \TMP2
-	pcmpeqd TWOONE(%rip), \TMP2
-	pand	POLY(%rip), \TMP2
-	pxor	\TMP2, \TMP3
-	movdqu	\TMP3, HashKey(%arg2)
-
-	movdqa	   \TMP3, \TMP5
-	pshufd	   $78, \TMP3, \TMP1
-	pxor	   \TMP3, \TMP1
-	movdqu	   \TMP1, HashKey_k(%arg2)
-
-	GHASH_MUL  \TMP5, \TMP3, \TMP1, \TMP2, \TMP4, \TMP6, \TMP7
-# TMP5 = HashKey^2<<1 (mod poly)
-	movdqu	   \TMP5, HashKey_2(%arg2)
-# HashKey_2 = HashKey^2<<1 (mod poly)
-	pshufd	   $78, \TMP5, \TMP1
-	pxor	   \TMP5, \TMP1
-	movdqu	   \TMP1, HashKey_2_k(%arg2)
-
-	GHASH_MUL  \TMP5, \TMP3, \TMP1, \TMP2, \TMP4, \TMP6, \TMP7
-# TMP5 = HashKey^3<<1 (mod poly)
-	movdqu	   \TMP5, HashKey_3(%arg2)
-	pshufd	   $78, \TMP5, \TMP1
-	pxor	   \TMP5, \TMP1
-	movdqu	   \TMP1, HashKey_3_k(%arg2)
-
-	GHASH_MUL  \TMP5, \TMP3, \TMP1, \TMP2, \TMP4, \TMP6, \TMP7
-# TMP5 = HashKey^3<<1 (mod poly)
-	movdqu	   \TMP5, HashKey_4(%arg2)
-	pshufd	   $78, \TMP5, \TMP1
-	pxor	   \TMP5, \TMP1
-	movdqu	   \TMP1, HashKey_4_k(%arg2)
-.endm
-
-# GCM_INIT initializes a gcm_context struct to prepare for encoding/decoding.
-# Clobbers rax, r10-r13 and xmm0-xmm6, %xmm13
-.macro GCM_INIT Iv SUBKEY AAD AADLEN
-	mov \AADLEN, %r11
-	mov %r11, AadLen(%arg2) # ctx_data.aad_length = aad_length
-	xor %r11d, %r11d
-	mov %r11, InLen(%arg2) # ctx_data.in_length = 0
-	mov %r11, PBlockLen(%arg2) # ctx_data.partial_block_length = 0
-	mov %r11, PBlockEncKey(%arg2) # ctx_data.partial_block_enc_key = 0
-	mov \Iv, %rax
-	movdqu (%rax), %xmm0
-	movdqu %xmm0, OrigIV(%arg2) # ctx_data.orig_IV = iv
-
-	movdqa  SHUF_MASK(%rip), %xmm2
-	pshufb %xmm2, %xmm0
-	movdqu %xmm0, CurCount(%arg2) # ctx_data.current_counter = iv
-
-	PRECOMPUTE \SUBKEY, %xmm1, %xmm2, %xmm3, %xmm4, %xmm5, %xmm6, %xmm7
-	movdqu HashKey(%arg2), %xmm13
-
-	CALC_AAD_HASH %xmm13, \AAD, \AADLEN, %xmm0, %xmm1, %xmm2, %xmm3, \
-	%xmm4, %xmm5, %xmm6
-.endm
-
-# GCM_ENC_DEC Encodes/Decodes given data. Assumes that the passed gcm_context
-# struct has been initialized by GCM_INIT.
-# Requires the input data be at least 1 byte long because of READ_PARTIAL_BLOCK
-# Clobbers rax, r10-r13, and xmm0-xmm15
-.macro GCM_ENC_DEC operation
-	movdqu AadHash(%arg2), %xmm8
-	movdqu HashKey(%arg2), %xmm13
-	add %arg5, InLen(%arg2)
-
-	xor %r11d, %r11d # initialise the data pointer offset as zero
-	PARTIAL_BLOCK %arg3 %arg4 %arg5 %r11 %xmm8 \operation
-
-	sub %r11, %arg5		# sub partial block data used
-	mov %arg5, %r13		# save the number of bytes
-
-	and $-16, %r13		# %r13 = %r13 - (%r13 mod 16)
-	mov %r13, %r12
-	# Encrypt/Decrypt first few blocks
-
-	and	$(3<<4), %r12
-	jz	.L_initial_num_blocks_is_0_\@
-	cmp	$(2<<4), %r12
-	jb	.L_initial_num_blocks_is_1_\@
-	je	.L_initial_num_blocks_is_2_\@
-.L_initial_num_blocks_is_3_\@:
-	INITIAL_BLOCKS_ENC_DEC	%xmm9, %xmm10, %xmm13, %xmm11, %xmm12, %xmm0, \
-%xmm1, %xmm2, %xmm3, %xmm4, %xmm8, %xmm5, %xmm6, 5, 678, \operation
-	sub	$48, %r13
-	jmp	.L_initial_blocks_\@
-.L_initial_num_blocks_is_2_\@:
-	INITIAL_BLOCKS_ENC_DEC	%xmm9, %xmm10, %xmm13, %xmm11, %xmm12, %xmm0, \
-%xmm1, %xmm2, %xmm3, %xmm4, %xmm8, %xmm5, %xmm6, 6, 78, \operation
-	sub	$32, %r13
-	jmp	.L_initial_blocks_\@
-.L_initial_num_blocks_is_1_\@:
-	INITIAL_BLOCKS_ENC_DEC	%xmm9, %xmm10, %xmm13, %xmm11, %xmm12, %xmm0, \
-%xmm1, %xmm2, %xmm3, %xmm4, %xmm8, %xmm5, %xmm6, 7, 8, \operation
-	sub	$16, %r13
-	jmp	.L_initial_blocks_\@
-.L_initial_num_blocks_is_0_\@:
-	INITIAL_BLOCKS_ENC_DEC	%xmm9, %xmm10, %xmm13, %xmm11, %xmm12, %xmm0, \
-%xmm1, %xmm2, %xmm3, %xmm4, %xmm8, %xmm5, %xmm6, 8, 0, \operation
-.L_initial_blocks_\@:
-
-	# Main loop - Encrypt/Decrypt remaining blocks
-
-	test	%r13, %r13
-	je	.L_zero_cipher_left_\@
-	sub	$64, %r13
-	je	.L_four_cipher_left_\@
-.L_crypt_by_4_\@:
-	GHASH_4_ENCRYPT_4_PARALLEL_\operation	%xmm9, %xmm10, %xmm11, %xmm12, \
-	%xmm13, %xmm14, %xmm0, %xmm1, %xmm2, %xmm3, %xmm4, %xmm5, %xmm6, \
-	%xmm7, %xmm8, enc
-	add	$64, %r11
-	sub	$64, %r13
-	jne	.L_crypt_by_4_\@
-.L_four_cipher_left_\@:
-	GHASH_LAST_4	%xmm9, %xmm10, %xmm11, %xmm12, %xmm13, %xmm14, \
-%xmm15, %xmm1, %xmm2, %xmm3, %xmm4, %xmm8
-.L_zero_cipher_left_\@:
-	movdqu %xmm8, AadHash(%arg2)
-	movdqu %xmm0, CurCount(%arg2)
-
-	mov	%arg5, %r13
-	and	$15, %r13			# %r13 = arg5 (mod 16)
-	je	.L_multiple_of_16_bytes_\@
-
-	mov %r13, PBlockLen(%arg2)
-
-	# Handle the last <16 Byte block separately
-	paddd ONE(%rip), %xmm0                # INCR CNT to get Yn
-	movdqu %xmm0, CurCount(%arg2)
-	movdqa SHUF_MASK(%rip), %xmm10
-	pshufb %xmm10, %xmm0
-
-	ENCRYPT_SINGLE_BLOCK	%xmm0, %xmm1        # Encrypt(K, Yn)
-	movdqu %xmm0, PBlockEncKey(%arg2)
-
-	cmp	$16, %arg5
-	jge	.L_large_enough_update_\@
-
-	lea (%arg4,%r11,1), %r10
-	mov %r13, %r12
-	READ_PARTIAL_BLOCK %r10 %r12 %xmm2 %xmm1
-	jmp	.L_data_read_\@
-
-.L_large_enough_update_\@:
-	sub	$16, %r11
-	add	%r13, %r11
-
-	# receive the last <16 Byte block
-	movdqu	(%arg4, %r11, 1), %xmm1
-
-	sub	%r13, %r11
-	add	$16, %r11
-
-	lea	SHIFT_MASK+16(%rip), %r12
-	# adjust the shuffle mask pointer to be able to shift 16-r13 bytes
-	# (r13 is the number of bytes in plaintext mod 16)
-	sub	%r13, %r12
-	# get the appropriate shuffle mask
-	movdqu	(%r12), %xmm2
-	# shift right 16-r13 bytes
-	pshufb  %xmm2, %xmm1
-
-.L_data_read_\@:
-	lea ALL_F+16(%rip), %r12
-	sub %r13, %r12
-
-.ifc \operation, dec
-	movdqa  %xmm1, %xmm2
-.endif
-	pxor	%xmm1, %xmm0            # XOR Encrypt(K, Yn)
-	movdqu	(%r12), %xmm1
-	# get the appropriate mask to mask out top 16-r13 bytes of xmm0
-	pand	%xmm1, %xmm0            # mask out top 16-r13 bytes of xmm0
-.ifc \operation, dec
-	pand    %xmm1, %xmm2
-	movdqa SHUF_MASK(%rip), %xmm10
-	pshufb %xmm10 ,%xmm2
-
-	pxor %xmm2, %xmm8
-.else
-	movdqa SHUF_MASK(%rip), %xmm10
-	pshufb %xmm10,%xmm0
-
-	pxor	%xmm0, %xmm8
-.endif
-
-	movdqu %xmm8, AadHash(%arg2)
-.ifc \operation, enc
-	# GHASH computation for the last <16 byte block
-	movdqa SHUF_MASK(%rip), %xmm10
-	# shuffle xmm0 back to output as ciphertext
-	pshufb %xmm10, %xmm0
-.endif
-
-	# Output %r13 bytes
-	movq %xmm0, %rax
-	cmp $8, %r13
-	jle .L_less_than_8_bytes_left_\@
-	mov %rax, (%arg3 , %r11, 1)
-	add $8, %r11
-	psrldq $8, %xmm0
-	movq %xmm0, %rax
-	sub $8, %r13
-.L_less_than_8_bytes_left_\@:
-	mov %al,  (%arg3, %r11, 1)
-	add $1, %r11
-	shr $8, %rax
-	sub $1, %r13
-	jne .L_less_than_8_bytes_left_\@
-.L_multiple_of_16_bytes_\@:
-.endm
-
-# GCM_COMPLETE Finishes update of tag of last partial block
-# Output: Authorization Tag (AUTH_TAG)
-# Clobbers rax, r10-r12, and xmm0, xmm1, xmm5-xmm15
-.macro GCM_COMPLETE AUTHTAG AUTHTAGLEN
-	movdqu AadHash(%arg2), %xmm8
-	movdqu HashKey(%arg2), %xmm13
-
-	mov PBlockLen(%arg2), %r12
-
-	test %r12, %r12
-	je .L_partial_done\@
-
-	GHASH_MUL %xmm8, %xmm13, %xmm9, %xmm10, %xmm11, %xmm5, %xmm6
-
-.L_partial_done\@:
-	mov AadLen(%arg2), %r12  # %r13 = aadLen (number of bytes)
-	shl	$3, %r12		  # convert into number of bits
-	movd	%r12d, %xmm15		  # len(A) in %xmm15
-	mov InLen(%arg2), %r12
-	shl     $3, %r12                  # len(C) in bits (*128)
-	movq    %r12, %xmm1
-
-	pslldq	$8, %xmm15		  # %xmm15 = len(A)||0x0000000000000000
-	pxor	%xmm1, %xmm15		  # %xmm15 = len(A)||len(C)
-	pxor	%xmm15, %xmm8
-	GHASH_MUL	%xmm8, %xmm13, %xmm9, %xmm10, %xmm11, %xmm5, %xmm6
-	# final GHASH computation
-	movdqa SHUF_MASK(%rip), %xmm10
-	pshufb %xmm10, %xmm8
-
-	movdqu OrigIV(%arg2), %xmm0       # %xmm0 = Y0
-	ENCRYPT_SINGLE_BLOCK	%xmm0,  %xmm1	  # E(K, Y0)
-	pxor	%xmm8, %xmm0
-.L_return_T_\@:
-	mov	\AUTHTAG, %r10                     # %r10 = authTag
-	mov	\AUTHTAGLEN, %r11                    # %r11 = auth_tag_len
-	cmp	$16, %r11
-	je	.L_T_16_\@
-	cmp	$8, %r11
-	jl	.L_T_4_\@
-.L_T_8_\@:
-	movq	%xmm0, %rax
-	mov	%rax, (%r10)
-	add	$8, %r10
-	sub	$8, %r11
-	psrldq	$8, %xmm0
-	test	%r11, %r11
-	je	.L_return_T_done_\@
-.L_T_4_\@:
-	movd	%xmm0, %eax
-	mov	%eax, (%r10)
-	add	$4, %r10
-	sub	$4, %r11
-	psrldq	$4, %xmm0
-	test	%r11, %r11
-	je	.L_return_T_done_\@
-.L_T_123_\@:
-	movd	%xmm0, %eax
-	cmp	$2, %r11
-	jl	.L_T_1_\@
-	mov	%ax, (%r10)
-	cmp	$2, %r11
-	je	.L_return_T_done_\@
-	add	$2, %r10
-	sar	$16, %eax
-.L_T_1_\@:
-	mov	%al, (%r10)
-	jmp	.L_return_T_done_\@
-.L_T_16_\@:
-	movdqu	%xmm0, (%r10)
-.L_return_T_done_\@:
-.endm
-
-#ifdef __x86_64__
-/* GHASH_MUL MACRO to implement: Data*HashKey mod (128,127,126,121,0)
-*
-*
-* Input: A and B (128-bits each, bit-reflected)
-* Output: C = A*B*x mod poly, (i.e. >>1 )
-* To compute GH = GH*HashKey mod poly, give HK = HashKey<<1 mod poly as input
-* GH = GH * HK * x mod poly which is equivalent to GH*HashKey mod poly.
-*
-*/
-.macro GHASH_MUL GH HK TMP1 TMP2 TMP3 TMP4 TMP5
-	movdqa	  \GH, \TMP1
-	pshufd	  $78, \GH, \TMP2
-	pshufd	  $78, \HK, \TMP3
-	pxor	  \GH, \TMP2            # TMP2 = a1+a0
-	pxor	  \HK, \TMP3            # TMP3 = b1+b0
-	pclmulqdq $0x11, \HK, \TMP1     # TMP1 = a1*b1
-	pclmulqdq $0x00, \HK, \GH       # GH = a0*b0
-	pclmulqdq $0x00, \TMP3, \TMP2   # TMP2 = (a0+a1)*(b1+b0)
-	pxor	  \GH, \TMP2
-	pxor	  \TMP1, \TMP2          # TMP2 = (a0*b0)+(a1*b0)
-	movdqa	  \TMP2, \TMP3
-	pslldq	  $8, \TMP3             # left shift TMP3 2 DWs
-	psrldq	  $8, \TMP2             # right shift TMP2 2 DWs
-	pxor	  \TMP3, \GH
-	pxor	  \TMP2, \TMP1          # TMP2:GH holds the result of GH*HK
-
-        # first phase of the reduction
-
-	movdqa    \GH, \TMP2
-	movdqa    \GH, \TMP3
-	movdqa    \GH, \TMP4            # copy GH into TMP2,TMP3 and TMP4
-					# in in order to perform
-					# independent shifts
-	pslld     $31, \TMP2            # packed right shift <<31
-	pslld     $30, \TMP3            # packed right shift <<30
-	pslld     $25, \TMP4            # packed right shift <<25
-	pxor      \TMP3, \TMP2          # xor the shifted versions
-	pxor      \TMP4, \TMP2
-	movdqa    \TMP2, \TMP5
-	psrldq    $4, \TMP5             # right shift TMP5 1 DW
-	pslldq    $12, \TMP2            # left shift TMP2 3 DWs
-	pxor      \TMP2, \GH
-
-        # second phase of the reduction
-
-	movdqa    \GH,\TMP2             # copy GH into TMP2,TMP3 and TMP4
-					# in in order to perform
-					# independent shifts
-	movdqa    \GH,\TMP3
-	movdqa    \GH,\TMP4
-	psrld     $1,\TMP2              # packed left shift >>1
-	psrld     $2,\TMP3              # packed left shift >>2
-	psrld     $7,\TMP4              # packed left shift >>7
-	pxor      \TMP3,\TMP2		# xor the shifted versions
-	pxor      \TMP4,\TMP2
-	pxor      \TMP5, \TMP2
-	pxor      \TMP2, \GH
-	pxor      \TMP1, \GH            # result is in TMP1
-.endm
-
-# Reads DLEN bytes starting at DPTR and stores in XMMDst
-# where 0 < DLEN < 16
-# Clobbers %rax, DLEN and XMM1
-.macro READ_PARTIAL_BLOCK DPTR DLEN XMM1 XMMDst
-        cmp $8, \DLEN
-        jl .L_read_lt8_\@
-        mov (\DPTR), %rax
-        movq %rax, \XMMDst
-        sub $8, \DLEN
-        jz .L_done_read_partial_block_\@
-	xor %eax, %eax
-.L_read_next_byte_\@:
-        shl $8, %rax
-        mov 7(\DPTR, \DLEN, 1), %al
-        dec \DLEN
-        jnz .L_read_next_byte_\@
-        movq %rax, \XMM1
-	pslldq $8, \XMM1
-        por \XMM1, \XMMDst
-	jmp .L_done_read_partial_block_\@
-.L_read_lt8_\@:
-	xor %eax, %eax
-.L_read_next_byte_lt8_\@:
-        shl $8, %rax
-        mov -1(\DPTR, \DLEN, 1), %al
-        dec \DLEN
-        jnz .L_read_next_byte_lt8_\@
-        movq %rax, \XMMDst
-.L_done_read_partial_block_\@:
-.endm
-
-# CALC_AAD_HASH: Calculates the hash of the data which will not be encrypted.
-# clobbers r10-11, xmm14
-.macro CALC_AAD_HASH HASHKEY AAD AADLEN TMP1 TMP2 TMP3 TMP4 TMP5 \
-	TMP6 TMP7
-	MOVADQ	   SHUF_MASK(%rip), %xmm14
-	mov	   \AAD, %r10		# %r10 = AAD
-	mov	   \AADLEN, %r11		# %r11 = aadLen
-	pxor	   \TMP7, \TMP7
-	pxor	   \TMP6, \TMP6
-
-	cmp	   $16, %r11
-	jl	   .L_get_AAD_rest\@
-.L_get_AAD_blocks\@:
-	movdqu	   (%r10), \TMP7
-	pshufb	   %xmm14, \TMP7 # byte-reflect the AAD data
-	pxor	   \TMP7, \TMP6
-	GHASH_MUL  \TMP6, \HASHKEY, \TMP1, \TMP2, \TMP3, \TMP4, \TMP5
-	add	   $16, %r10
-	sub	   $16, %r11
-	cmp	   $16, %r11
-	jge	   .L_get_AAD_blocks\@
-
-	movdqu	   \TMP6, \TMP7
-
-	/* read the last <16B of AAD */
-.L_get_AAD_rest\@:
-	test	   %r11, %r11
-	je	   .L_get_AAD_done\@
-
-	READ_PARTIAL_BLOCK %r10, %r11, \TMP1, \TMP7
-	pshufb	   %xmm14, \TMP7 # byte-reflect the AAD data
-	pxor	   \TMP6, \TMP7
-	GHASH_MUL  \TMP7, \HASHKEY, \TMP1, \TMP2, \TMP3, \TMP4, \TMP5
-	movdqu \TMP7, \TMP6
-
-.L_get_AAD_done\@:
-	movdqu \TMP6, AadHash(%arg2)
-.endm
-
-# PARTIAL_BLOCK: Handles encryption/decryption and the tag partial blocks
-# between update calls.
-# Requires the input data be at least 1 byte long due to READ_PARTIAL_BLOCK
-# Outputs encrypted bytes, and updates hash and partial info in gcm_data_context
-# Clobbers rax, r10, r12, r13, xmm0-6, xmm9-13
-.macro PARTIAL_BLOCK CYPH_PLAIN_OUT PLAIN_CYPH_IN PLAIN_CYPH_LEN DATA_OFFSET \
-	AAD_HASH operation
-	mov 	PBlockLen(%arg2), %r13
-	test	%r13, %r13
-	je	.L_partial_block_done_\@	# Leave Macro if no partial blocks
-	# Read in input data without over reading
-	cmp	$16, \PLAIN_CYPH_LEN
-	jl	.L_fewer_than_16_bytes_\@
-	movups	(\PLAIN_CYPH_IN), %xmm1	# If more than 16 bytes, just fill xmm
-	jmp	.L_data_read_\@
-
-.L_fewer_than_16_bytes_\@:
-	lea	(\PLAIN_CYPH_IN, \DATA_OFFSET, 1), %r10
-	mov	\PLAIN_CYPH_LEN, %r12
-	READ_PARTIAL_BLOCK %r10 %r12 %xmm0 %xmm1
-
-	mov PBlockLen(%arg2), %r13
-
-.L_data_read_\@:				# Finished reading in data
-
-	movdqu	PBlockEncKey(%arg2), %xmm9
-	movdqu	HashKey(%arg2), %xmm13
-
-	lea	SHIFT_MASK(%rip), %r12
-
-	# adjust the shuffle mask pointer to be able to shift r13 bytes
-	# r16-r13 is the number of bytes in plaintext mod 16)
-	add	%r13, %r12
-	movdqu	(%r12), %xmm2		# get the appropriate shuffle mask
-	pshufb	%xmm2, %xmm9		# shift right r13 bytes
-
-.ifc \operation, dec
-	movdqa	%xmm1, %xmm3
-	pxor	%xmm1, %xmm9		# Ciphertext XOR E(K, Yn)
-
-	mov	\PLAIN_CYPH_LEN, %r10
-	add	%r13, %r10
-	# Set r10 to be the amount of data left in CYPH_PLAIN_IN after filling
-	sub	$16, %r10
-	# Determine if partial block is not being filled and
-	# shift mask accordingly
-	jge	.L_no_extra_mask_1_\@
-	sub	%r10, %r12
-.L_no_extra_mask_1_\@:
-
-	movdqu	ALL_F-SHIFT_MASK(%r12), %xmm1
-	# get the appropriate mask to mask out bottom r13 bytes of xmm9
-	pand	%xmm1, %xmm9		# mask out bottom r13 bytes of xmm9
-
-	pand	%xmm1, %xmm3
-	movdqa	SHUF_MASK(%rip), %xmm10
-	pshufb	%xmm10, %xmm3
-	pshufb	%xmm2, %xmm3
-	pxor	%xmm3, \AAD_HASH
-
-	test	%r10, %r10
-	jl	.L_partial_incomplete_1_\@
-
-	# GHASH computation for the last <16 Byte block
-	GHASH_MUL \AAD_HASH, %xmm13, %xmm0, %xmm10, %xmm11, %xmm5, %xmm6
-	xor	%eax, %eax
-
-	mov	%rax, PBlockLen(%arg2)
-	jmp	.L_dec_done_\@
-.L_partial_incomplete_1_\@:
-	add	\PLAIN_CYPH_LEN, PBlockLen(%arg2)
-.L_dec_done_\@:
-	movdqu	\AAD_HASH, AadHash(%arg2)
-.else
-	pxor	%xmm1, %xmm9			# Plaintext XOR E(K, Yn)
-
-	mov	\PLAIN_CYPH_LEN, %r10
-	add	%r13, %r10
-	# Set r10 to be the amount of data left in CYPH_PLAIN_IN after filling
-	sub	$16, %r10
-	# Determine if partial block is not being filled and
-	# shift mask accordingly
-	jge	.L_no_extra_mask_2_\@
-	sub	%r10, %r12
-.L_no_extra_mask_2_\@:
-
-	movdqu	ALL_F-SHIFT_MASK(%r12), %xmm1
-	# get the appropriate mask to mask out bottom r13 bytes of xmm9
-	pand	%xmm1, %xmm9
-
-	movdqa	SHUF_MASK(%rip), %xmm1
-	pshufb	%xmm1, %xmm9
-	pshufb	%xmm2, %xmm9
-	pxor	%xmm9, \AAD_HASH
-
-	test	%r10, %r10
-	jl	.L_partial_incomplete_2_\@
-
-	# GHASH computation for the last <16 Byte block
-	GHASH_MUL \AAD_HASH, %xmm13, %xmm0, %xmm10, %xmm11, %xmm5, %xmm6
-	xor	%eax, %eax
-
-	mov	%rax, PBlockLen(%arg2)
-	jmp	.L_encode_done_\@
-.L_partial_incomplete_2_\@:
-	add	\PLAIN_CYPH_LEN, PBlockLen(%arg2)
-.L_encode_done_\@:
-	movdqu	\AAD_HASH, AadHash(%arg2)
-
-	movdqa	SHUF_MASK(%rip), %xmm10
-	# shuffle xmm9 back to output as ciphertext
-	pshufb	%xmm10, %xmm9
-	pshufb	%xmm2, %xmm9
-.endif
-	# output encrypted Bytes
-	test	%r10, %r10
-	jl	.L_partial_fill_\@
-	mov	%r13, %r12
-	mov	$16, %r13
-	# Set r13 to be the number of bytes to write out
-	sub	%r12, %r13
-	jmp	.L_count_set_\@
-.L_partial_fill_\@:
-	mov	\PLAIN_CYPH_LEN, %r13
-.L_count_set_\@:
-	movdqa	%xmm9, %xmm0
-	movq	%xmm0, %rax
-	cmp	$8, %r13
-	jle	.L_less_than_8_bytes_left_\@
-
-	mov	%rax, (\CYPH_PLAIN_OUT, \DATA_OFFSET, 1)
-	add	$8, \DATA_OFFSET
-	psrldq	$8, %xmm0
-	movq	%xmm0, %rax
-	sub	$8, %r13
-.L_less_than_8_bytes_left_\@:
-	movb	%al, (\CYPH_PLAIN_OUT, \DATA_OFFSET, 1)
-	add	$1, \DATA_OFFSET
-	shr	$8, %rax
-	sub	$1, %r13
-	jne	.L_less_than_8_bytes_left_\@
-.L_partial_block_done_\@:
-.endm # PARTIAL_BLOCK
-
-/*
-* if a = number of total plaintext bytes
-* b = floor(a/16)
-* num_initial_blocks = b mod 4
-* encrypt the initial num_initial_blocks blocks and apply ghash on
-* the ciphertext
-* %r10, %r11, %r12, %rax, %xmm5, %xmm6, %xmm7, %xmm8, %xmm9 registers
-* are clobbered
-* arg1, %arg2, %arg3 are used as a pointer only, not modified
-*/
-
-
-.macro INITIAL_BLOCKS_ENC_DEC TMP1 TMP2 TMP3 TMP4 TMP5 XMM0 XMM1 \
-	XMM2 XMM3 XMM4 XMMDst TMP6 TMP7 i i_seq operation
-	MOVADQ		SHUF_MASK(%rip), %xmm14
-
-	movdqu AadHash(%arg2), %xmm\i		    # XMM0 = Y0
-
-	# start AES for num_initial_blocks blocks
-
-	movdqu CurCount(%arg2), \XMM0                # XMM0 = Y0
-
-.if (\i == 5) || (\i == 6) || (\i == 7)
-
-	MOVADQ		ONE(%RIP),\TMP1
-	MOVADQ		0(%arg1),\TMP2
-.irpc index, \i_seq
-	paddd		\TMP1, \XMM0                 # INCR Y0
-.ifc \operation, dec
-        movdqa     \XMM0, %xmm\index
-.else
-	MOVADQ		\XMM0, %xmm\index
-.endif
-	pshufb	%xmm14, %xmm\index      # perform a 16 byte swap
-	pxor		\TMP2, %xmm\index
-.endr
-	lea	0x10(%arg1),%r10
-	mov	keysize,%eax
-	shr	$2,%eax				# 128->4, 192->6, 256->8
-	add	$5,%eax			      # 128->9, 192->11, 256->13
-
-.Laes_loop_initial_\@:
-	MOVADQ	(%r10),\TMP1
-.irpc	index, \i_seq
-	aesenc	\TMP1, %xmm\index
-.endr
-	add	$16,%r10
-	sub	$1,%eax
-	jnz	.Laes_loop_initial_\@
-
-	MOVADQ	(%r10), \TMP1
-.irpc index, \i_seq
-	aesenclast \TMP1, %xmm\index         # Last Round
-.endr
-.irpc index, \i_seq
-	movdqu	   (%arg4 , %r11, 1), \TMP1
-	pxor	   \TMP1, %xmm\index
-	movdqu	   %xmm\index, (%arg3 , %r11, 1)
-	# write back plaintext/ciphertext for num_initial_blocks
-	add	   $16, %r11
-
-.ifc \operation, dec
-	movdqa     \TMP1, %xmm\index
-.endif
-	pshufb	   %xmm14, %xmm\index
-
-		# prepare plaintext/ciphertext for GHASH computation
-.endr
-.endif
-
-        # apply GHASH on num_initial_blocks blocks
-
-.if \i == 5
-        pxor       %xmm5, %xmm6
-	GHASH_MUL  %xmm6, \TMP3, \TMP1, \TMP2, \TMP4, \TMP5, \XMM1
-        pxor       %xmm6, %xmm7
-	GHASH_MUL  %xmm7, \TMP3, \TMP1, \TMP2, \TMP4, \TMP5, \XMM1
-        pxor       %xmm7, %xmm8
-	GHASH_MUL  %xmm8, \TMP3, \TMP1, \TMP2, \TMP4, \TMP5, \XMM1
-.elseif \i == 6
-        pxor       %xmm6, %xmm7
-	GHASH_MUL  %xmm7, \TMP3, \TMP1, \TMP2, \TMP4, \TMP5, \XMM1
-        pxor       %xmm7, %xmm8
-	GHASH_MUL  %xmm8, \TMP3, \TMP1, \TMP2, \TMP4, \TMP5, \XMM1
-.elseif \i == 7
-        pxor       %xmm7, %xmm8
-	GHASH_MUL  %xmm8, \TMP3, \TMP1, \TMP2, \TMP4, \TMP5, \XMM1
-.endif
-	cmp	   $64, %r13
-	jl	.L_initial_blocks_done\@
-	# no need for precomputed values
-/*
-*
-* Precomputations for HashKey parallel with encryption of first 4 blocks.
-* Haskey_i_k holds XORed values of the low and high parts of the Haskey_i
-*/
-	MOVADQ	   ONE(%RIP),\TMP1
-	paddd	   \TMP1, \XMM0              # INCR Y0
-	MOVADQ	   \XMM0, \XMM1
-	pshufb  %xmm14, \XMM1        # perform a 16 byte swap
-
-	paddd	   \TMP1, \XMM0              # INCR Y0
-	MOVADQ	   \XMM0, \XMM2
-	pshufb  %xmm14, \XMM2        # perform a 16 byte swap
-
-	paddd	   \TMP1, \XMM0              # INCR Y0
-	MOVADQ	   \XMM0, \XMM3
-	pshufb %xmm14, \XMM3        # perform a 16 byte swap
-
-	paddd	   \TMP1, \XMM0              # INCR Y0
-	MOVADQ	   \XMM0, \XMM4
-	pshufb %xmm14, \XMM4        # perform a 16 byte swap
-
-	MOVADQ	   0(%arg1),\TMP1
-	pxor	   \TMP1, \XMM1
-	pxor	   \TMP1, \XMM2
-	pxor	   \TMP1, \XMM3
-	pxor	   \TMP1, \XMM4
-.irpc index, 1234 # do 4 rounds
-	movaps 0x10*\index(%arg1), \TMP1
-	aesenc	   \TMP1, \XMM1
-	aesenc	   \TMP1, \XMM2
-	aesenc	   \TMP1, \XMM3
-	aesenc	   \TMP1, \XMM4
-.endr
-.irpc index, 56789 # do next 5 rounds
-	movaps 0x10*\index(%arg1), \TMP1
-	aesenc	   \TMP1, \XMM1
-	aesenc	   \TMP1, \XMM2
-	aesenc	   \TMP1, \XMM3
-	aesenc	   \TMP1, \XMM4
-.endr
-	lea	   0xa0(%arg1),%r10
-	mov	   keysize,%eax
-	shr	   $2,%eax			# 128->4, 192->6, 256->8
-	sub	   $4,%eax			# 128->0, 192->2, 256->4
-	jz	   .Laes_loop_pre_done\@
-
-.Laes_loop_pre_\@:
-	MOVADQ	   (%r10),\TMP2
-.irpc	index, 1234
-	aesenc	   \TMP2, %xmm\index
-.endr
-	add	   $16,%r10
-	sub	   $1,%eax
-	jnz	   .Laes_loop_pre_\@
-
-.Laes_loop_pre_done\@:
-	MOVADQ	   (%r10), \TMP2
-	aesenclast \TMP2, \XMM1
-	aesenclast \TMP2, \XMM2
-	aesenclast \TMP2, \XMM3
-	aesenclast \TMP2, \XMM4
-	movdqu	   16*0(%arg4 , %r11 , 1), \TMP1
-	pxor	   \TMP1, \XMM1
-.ifc \operation, dec
-	movdqu     \XMM1, 16*0(%arg3 , %r11 , 1)
-	movdqa     \TMP1, \XMM1
-.endif
-	movdqu	   16*1(%arg4 , %r11 , 1), \TMP1
-	pxor	   \TMP1, \XMM2
-.ifc \operation, dec
-	movdqu     \XMM2, 16*1(%arg3 , %r11 , 1)
-	movdqa     \TMP1, \XMM2
-.endif
-	movdqu	   16*2(%arg4 , %r11 , 1), \TMP1
-	pxor	   \TMP1, \XMM3
-.ifc \operation, dec
-	movdqu     \XMM3, 16*2(%arg3 , %r11 , 1)
-	movdqa     \TMP1, \XMM3
-.endif
-	movdqu	   16*3(%arg4 , %r11 , 1), \TMP1
-	pxor	   \TMP1, \XMM4
-.ifc \operation, dec
-	movdqu     \XMM4, 16*3(%arg3 , %r11 , 1)
-	movdqa     \TMP1, \XMM4
-.else
-	movdqu     \XMM1, 16*0(%arg3 , %r11 , 1)
-	movdqu     \XMM2, 16*1(%arg3 , %r11 , 1)
-	movdqu     \XMM3, 16*2(%arg3 , %r11 , 1)
-	movdqu     \XMM4, 16*3(%arg3 , %r11 , 1)
-.endif
-
-	add	   $64, %r11
-	pshufb %xmm14, \XMM1 # perform a 16 byte swap
-	pxor	   \XMMDst, \XMM1
-# combine GHASHed value with the corresponding ciphertext
-	pshufb %xmm14, \XMM2 # perform a 16 byte swap
-	pshufb %xmm14, \XMM3 # perform a 16 byte swap
-	pshufb %xmm14, \XMM4 # perform a 16 byte swap
-
-.L_initial_blocks_done\@:
-
-.endm
-
-/*
-* encrypt 4 blocks at a time
-* ghash the 4 previously encrypted ciphertext blocks
-* arg1, %arg3, %arg4 are used as pointers only, not modified
-* %r11 is the data offset value
-*/
-.macro GHASH_4_ENCRYPT_4_PARALLEL_enc TMP1 TMP2 TMP3 TMP4 TMP5 \
-TMP6 XMM0 XMM1 XMM2 XMM3 XMM4 XMM5 XMM6 XMM7 XMM8 operation
-
-	movdqa	  \XMM1, \XMM5
-	movdqa	  \XMM2, \XMM6
-	movdqa	  \XMM3, \XMM7
-	movdqa	  \XMM4, \XMM8
-
-        movdqa    SHUF_MASK(%rip), %xmm15
-        # multiply TMP5 * HashKey using karatsuba
-
-	movdqa	  \XMM5, \TMP4
-	pshufd	  $78, \XMM5, \TMP6
-	pxor	  \XMM5, \TMP6
-	paddd     ONE(%rip), \XMM0		# INCR CNT
-	movdqu	  HashKey_4(%arg2), \TMP5
-	pclmulqdq $0x11, \TMP5, \TMP4           # TMP4 = a1*b1
-	movdqa    \XMM0, \XMM1
-	paddd     ONE(%rip), \XMM0		# INCR CNT
-	movdqa    \XMM0, \XMM2
-	paddd     ONE(%rip), \XMM0		# INCR CNT
-	movdqa    \XMM0, \XMM3
-	paddd     ONE(%rip), \XMM0		# INCR CNT
-	movdqa    \XMM0, \XMM4
-	pshufb %xmm15, \XMM1	# perform a 16 byte swap
-	pclmulqdq $0x00, \TMP5, \XMM5           # XMM5 = a0*b0
-	pshufb %xmm15, \XMM2	# perform a 16 byte swap
-	pshufb %xmm15, \XMM3	# perform a 16 byte swap
-	pshufb %xmm15, \XMM4	# perform a 16 byte swap
-
-	pxor	  (%arg1), \XMM1
-	pxor	  (%arg1), \XMM2
-	pxor	  (%arg1), \XMM3
-	pxor	  (%arg1), \XMM4
-	movdqu	  HashKey_4_k(%arg2), \TMP5
-	pclmulqdq $0x00, \TMP5, \TMP6       # TMP6 = (a1+a0)*(b1+b0)
-	movaps 0x10(%arg1), \TMP1
-	aesenc	  \TMP1, \XMM1              # Round 1
-	aesenc	  \TMP1, \XMM2
-	aesenc	  \TMP1, \XMM3
-	aesenc	  \TMP1, \XMM4
-	movaps 0x20(%arg1), \TMP1
-	aesenc	  \TMP1, \XMM1              # Round 2
-	aesenc	  \TMP1, \XMM2
-	aesenc	  \TMP1, \XMM3
-	aesenc	  \TMP1, \XMM4
-	movdqa	  \XMM6, \TMP1
-	pshufd	  $78, \XMM6, \TMP2
-	pxor	  \XMM6, \TMP2
-	movdqu	  HashKey_3(%arg2), \TMP5
-	pclmulqdq $0x11, \TMP5, \TMP1       # TMP1 = a1 * b1
-	movaps 0x30(%arg1), \TMP3
-	aesenc    \TMP3, \XMM1              # Round 3
-	aesenc    \TMP3, \XMM2
-	aesenc    \TMP3, \XMM3
-	aesenc    \TMP3, \XMM4
-	pclmulqdq $0x00, \TMP5, \XMM6       # XMM6 = a0*b0
-	movaps 0x40(%arg1), \TMP3
-	aesenc	  \TMP3, \XMM1              # Round 4
-	aesenc	  \TMP3, \XMM2
-	aesenc	  \TMP3, \XMM3
-	aesenc	  \TMP3, \XMM4
-	movdqu	  HashKey_3_k(%arg2), \TMP5
-	pclmulqdq $0x00, \TMP5, \TMP2       # TMP2 = (a1+a0)*(b1+b0)
-	movaps 0x50(%arg1), \TMP3
-	aesenc	  \TMP3, \XMM1              # Round 5
-	aesenc	  \TMP3, \XMM2
-	aesenc	  \TMP3, \XMM3
-	aesenc	  \TMP3, \XMM4
-	pxor	  \TMP1, \TMP4
-# accumulate the results in TMP4:XMM5, TMP6 holds the middle part
-	pxor	  \XMM6, \XMM5
-	pxor	  \TMP2, \TMP6
-	movdqa	  \XMM7, \TMP1
-	pshufd	  $78, \XMM7, \TMP2
-	pxor	  \XMM7, \TMP2
-	movdqu	  HashKey_2(%arg2), \TMP5
-
-        # Multiply TMP5 * HashKey using karatsuba
-
-	pclmulqdq $0x11, \TMP5, \TMP1       # TMP1 = a1*b1
-	movaps 0x60(%arg1), \TMP3
-	aesenc	  \TMP3, \XMM1              # Round 6
-	aesenc	  \TMP3, \XMM2
-	aesenc	  \TMP3, \XMM3
-	aesenc	  \TMP3, \XMM4
-	pclmulqdq $0x00, \TMP5, \XMM7       # XMM7 = a0*b0
-	movaps 0x70(%arg1), \TMP3
-	aesenc	  \TMP3, \XMM1              # Round 7
-	aesenc	  \TMP3, \XMM2
-	aesenc	  \TMP3, \XMM3
-	aesenc	  \TMP3, \XMM4
-	movdqu	  HashKey_2_k(%arg2), \TMP5
-	pclmulqdq $0x00, \TMP5, \TMP2       # TMP2 = (a1+a0)*(b1+b0)
-	movaps 0x80(%arg1), \TMP3
-	aesenc	  \TMP3, \XMM1              # Round 8
-	aesenc	  \TMP3, \XMM2
-	aesenc	  \TMP3, \XMM3
-	aesenc	  \TMP3, \XMM4
-	pxor	  \TMP1, \TMP4
-# accumulate the results in TMP4:XMM5, TMP6 holds the middle part
-	pxor	  \XMM7, \XMM5
-	pxor	  \TMP2, \TMP6
-
-        # Multiply XMM8 * HashKey
-        # XMM8 and TMP5 hold the values for the two operands
-
-	movdqa	  \XMM8, \TMP1
-	pshufd	  $78, \XMM8, \TMP2
-	pxor	  \XMM8, \TMP2
-	movdqu	  HashKey(%arg2), \TMP5
-	pclmulqdq $0x11, \TMP5, \TMP1      # TMP1 = a1*b1
-	movaps 0x90(%arg1), \TMP3
-	aesenc	  \TMP3, \XMM1             # Round 9
-	aesenc	  \TMP3, \XMM2
-	aesenc	  \TMP3, \XMM3
-	aesenc	  \TMP3, \XMM4
-	pclmulqdq $0x00, \TMP5, \XMM8      # XMM8 = a0*b0
-	lea	  0xa0(%arg1),%r10
-	mov	  keysize,%eax
-	shr	  $2,%eax			# 128->4, 192->6, 256->8
-	sub	  $4,%eax			# 128->0, 192->2, 256->4
-	jz	  .Laes_loop_par_enc_done\@
-
-.Laes_loop_par_enc\@:
-	MOVADQ	  (%r10),\TMP3
-.irpc	index, 1234
-	aesenc	  \TMP3, %xmm\index
-.endr
-	add	  $16,%r10
-	sub	  $1,%eax
-	jnz	  .Laes_loop_par_enc\@
-
-.Laes_loop_par_enc_done\@:
-	MOVADQ	  (%r10), \TMP3
-	aesenclast \TMP3, \XMM1           # Round 10
-	aesenclast \TMP3, \XMM2
-	aesenclast \TMP3, \XMM3
-	aesenclast \TMP3, \XMM4
-	movdqu    HashKey_k(%arg2), \TMP5
-	pclmulqdq $0x00, \TMP5, \TMP2          # TMP2 = (a1+a0)*(b1+b0)
-	movdqu	  (%arg4,%r11,1), \TMP3
-	pxor	  \TMP3, \XMM1                 # Ciphertext/Plaintext XOR EK
-	movdqu	  16(%arg4,%r11,1), \TMP3
-	pxor	  \TMP3, \XMM2                 # Ciphertext/Plaintext XOR EK
-	movdqu	  32(%arg4,%r11,1), \TMP3
-	pxor	  \TMP3, \XMM3                 # Ciphertext/Plaintext XOR EK
-	movdqu	  48(%arg4,%r11,1), \TMP3
-	pxor	  \TMP3, \XMM4                 # Ciphertext/Plaintext XOR EK
-        movdqu    \XMM1, (%arg3,%r11,1)        # Write to the ciphertext buffer
-        movdqu    \XMM2, 16(%arg3,%r11,1)      # Write to the ciphertext buffer
-        movdqu    \XMM3, 32(%arg3,%r11,1)      # Write to the ciphertext buffer
-        movdqu    \XMM4, 48(%arg3,%r11,1)      # Write to the ciphertext buffer
-	pshufb %xmm15, \XMM1        # perform a 16 byte swap
-	pshufb %xmm15, \XMM2	# perform a 16 byte swap
-	pshufb %xmm15, \XMM3	# perform a 16 byte swap
-	pshufb %xmm15, \XMM4	# perform a 16 byte swap
-
-	pxor	  \TMP4, \TMP1
-	pxor	  \XMM8, \XMM5
-	pxor	  \TMP6, \TMP2
-	pxor	  \TMP1, \TMP2
-	pxor	  \XMM5, \TMP2
-	movdqa	  \TMP2, \TMP3
-	pslldq	  $8, \TMP3                    # left shift TMP3 2 DWs
-	psrldq	  $8, \TMP2                    # right shift TMP2 2 DWs
-	pxor	  \TMP3, \XMM5
-	pxor	  \TMP2, \TMP1	  # accumulate the results in TMP1:XMM5
-
-        # first phase of reduction
-
-	movdqa    \XMM5, \TMP2
-	movdqa    \XMM5, \TMP3
-	movdqa    \XMM5, \TMP4
-# move XMM5 into TMP2, TMP3, TMP4 in order to perform shifts independently
-	pslld     $31, \TMP2                   # packed right shift << 31
-	pslld     $30, \TMP3                   # packed right shift << 30
-	pslld     $25, \TMP4                   # packed right shift << 25
-	pxor      \TMP3, \TMP2	               # xor the shifted versions
-	pxor      \TMP4, \TMP2
-	movdqa    \TMP2, \TMP5
-	psrldq    $4, \TMP5                    # right shift T5 1 DW
-	pslldq    $12, \TMP2                   # left shift T2 3 DWs
-	pxor      \TMP2, \XMM5
-
-        # second phase of reduction
-
-	movdqa    \XMM5,\TMP2 # make 3 copies of XMM5 into TMP2, TMP3, TMP4
-	movdqa    \XMM5,\TMP3
-	movdqa    \XMM5,\TMP4
-	psrld     $1, \TMP2                    # packed left shift >>1
-	psrld     $2, \TMP3                    # packed left shift >>2
-	psrld     $7, \TMP4                    # packed left shift >>7
-	pxor      \TMP3,\TMP2		       # xor the shifted versions
-	pxor      \TMP4,\TMP2
-	pxor      \TMP5, \TMP2
-	pxor      \TMP2, \XMM5
-	pxor      \TMP1, \XMM5                 # result is in TMP1
-
-	pxor	  \XMM5, \XMM1
-.endm
-
-/*
-* decrypt 4 blocks at a time
-* ghash the 4 previously decrypted ciphertext blocks
-* arg1, %arg3, %arg4 are used as pointers only, not modified
-* %r11 is the data offset value
-*/
-.macro GHASH_4_ENCRYPT_4_PARALLEL_dec TMP1 TMP2 TMP3 TMP4 TMP5 \
-TMP6 XMM0 XMM1 XMM2 XMM3 XMM4 XMM5 XMM6 XMM7 XMM8 operation
-
-	movdqa	  \XMM1, \XMM5
-	movdqa	  \XMM2, \XMM6
-	movdqa	  \XMM3, \XMM7
-	movdqa	  \XMM4, \XMM8
-
-        movdqa    SHUF_MASK(%rip), %xmm15
-        # multiply TMP5 * HashKey using karatsuba
-
-	movdqa	  \XMM5, \TMP4
-	pshufd	  $78, \XMM5, \TMP6
-	pxor	  \XMM5, \TMP6
-	paddd     ONE(%rip), \XMM0		# INCR CNT
-	movdqu	  HashKey_4(%arg2), \TMP5
-	pclmulqdq $0x11, \TMP5, \TMP4           # TMP4 = a1*b1
-	movdqa    \XMM0, \XMM1
-	paddd     ONE(%rip), \XMM0		# INCR CNT
-	movdqa    \XMM0, \XMM2
-	paddd     ONE(%rip), \XMM0		# INCR CNT
-	movdqa    \XMM0, \XMM3
-	paddd     ONE(%rip), \XMM0		# INCR CNT
-	movdqa    \XMM0, \XMM4
-	pshufb %xmm15, \XMM1	# perform a 16 byte swap
-	pclmulqdq $0x00, \TMP5, \XMM5           # XMM5 = a0*b0
-	pshufb %xmm15, \XMM2	# perform a 16 byte swap
-	pshufb %xmm15, \XMM3	# perform a 16 byte swap
-	pshufb %xmm15, \XMM4	# perform a 16 byte swap
-
-	pxor	  (%arg1), \XMM1
-	pxor	  (%arg1), \XMM2
-	pxor	  (%arg1), \XMM3
-	pxor	  (%arg1), \XMM4
-	movdqu	  HashKey_4_k(%arg2), \TMP5
-	pclmulqdq $0x00, \TMP5, \TMP6       # TMP6 = (a1+a0)*(b1+b0)
-	movaps 0x10(%arg1), \TMP1
-	aesenc	  \TMP1, \XMM1              # Round 1
-	aesenc	  \TMP1, \XMM2
-	aesenc	  \TMP1, \XMM3
-	aesenc	  \TMP1, \XMM4
-	movaps 0x20(%arg1), \TMP1
-	aesenc	  \TMP1, \XMM1              # Round 2
-	aesenc	  \TMP1, \XMM2
-	aesenc	  \TMP1, \XMM3
-	aesenc	  \TMP1, \XMM4
-	movdqa	  \XMM6, \TMP1
-	pshufd	  $78, \XMM6, \TMP2
-	pxor	  \XMM6, \TMP2
-	movdqu	  HashKey_3(%arg2), \TMP5
-	pclmulqdq $0x11, \TMP5, \TMP1       # TMP1 = a1 * b1
-	movaps 0x30(%arg1), \TMP3
-	aesenc    \TMP3, \XMM1              # Round 3
-	aesenc    \TMP3, \XMM2
-	aesenc    \TMP3, \XMM3
-	aesenc    \TMP3, \XMM4
-	pclmulqdq $0x00, \TMP5, \XMM6       # XMM6 = a0*b0
-	movaps 0x40(%arg1), \TMP3
-	aesenc	  \TMP3, \XMM1              # Round 4
-	aesenc	  \TMP3, \XMM2
-	aesenc	  \TMP3, \XMM3
-	aesenc	  \TMP3, \XMM4
-	movdqu	  HashKey_3_k(%arg2), \TMP5
-	pclmulqdq $0x00, \TMP5, \TMP2       # TMP2 = (a1+a0)*(b1+b0)
-	movaps 0x50(%arg1), \TMP3
-	aesenc	  \TMP3, \XMM1              # Round 5
-	aesenc	  \TMP3, \XMM2
-	aesenc	  \TMP3, \XMM3
-	aesenc	  \TMP3, \XMM4
-	pxor	  \TMP1, \TMP4
-# accumulate the results in TMP4:XMM5, TMP6 holds the middle part
-	pxor	  \XMM6, \XMM5
-	pxor	  \TMP2, \TMP6
-	movdqa	  \XMM7, \TMP1
-	pshufd	  $78, \XMM7, \TMP2
-	pxor	  \XMM7, \TMP2
-	movdqu	  HashKey_2(%arg2), \TMP5
-
-        # Multiply TMP5 * HashKey using karatsuba
-
-	pclmulqdq $0x11, \TMP5, \TMP1       # TMP1 = a1*b1
-	movaps 0x60(%arg1), \TMP3
-	aesenc	  \TMP3, \XMM1              # Round 6
-	aesenc	  \TMP3, \XMM2
-	aesenc	  \TMP3, \XMM3
-	aesenc	  \TMP3, \XMM4
-	pclmulqdq $0x00, \TMP5, \XMM7       # XMM7 = a0*b0
-	movaps 0x70(%arg1), \TMP3
-	aesenc	  \TMP3, \XMM1              # Round 7
-	aesenc	  \TMP3, \XMM2
-	aesenc	  \TMP3, \XMM3
-	aesenc	  \TMP3, \XMM4
-	movdqu	  HashKey_2_k(%arg2), \TMP5
-	pclmulqdq $0x00, \TMP5, \TMP2       # TMP2 = (a1+a0)*(b1+b0)
-	movaps 0x80(%arg1), \TMP3
-	aesenc	  \TMP3, \XMM1              # Round 8
-	aesenc	  \TMP3, \XMM2
-	aesenc	  \TMP3, \XMM3
-	aesenc	  \TMP3, \XMM4
-	pxor	  \TMP1, \TMP4
-# accumulate the results in TMP4:XMM5, TMP6 holds the middle part
-	pxor	  \XMM7, \XMM5
-	pxor	  \TMP2, \TMP6
-
-        # Multiply XMM8 * HashKey
-        # XMM8 and TMP5 hold the values for the two operands
-
-	movdqa	  \XMM8, \TMP1
-	pshufd	  $78, \XMM8, \TMP2
-	pxor	  \XMM8, \TMP2
-	movdqu	  HashKey(%arg2), \TMP5
-	pclmulqdq $0x11, \TMP5, \TMP1      # TMP1 = a1*b1
-	movaps 0x90(%arg1), \TMP3
-	aesenc	  \TMP3, \XMM1             # Round 9
-	aesenc	  \TMP3, \XMM2
-	aesenc	  \TMP3, \XMM3
-	aesenc	  \TMP3, \XMM4
-	pclmulqdq $0x00, \TMP5, \XMM8      # XMM8 = a0*b0
-	lea	  0xa0(%arg1),%r10
-	mov	  keysize,%eax
-	shr	  $2,%eax		        # 128->4, 192->6, 256->8
-	sub	  $4,%eax			# 128->0, 192->2, 256->4
-	jz	  .Laes_loop_par_dec_done\@
-
-.Laes_loop_par_dec\@:
-	MOVADQ	  (%r10),\TMP3
-.irpc	index, 1234
-	aesenc	  \TMP3, %xmm\index
-.endr
-	add	  $16,%r10
-	sub	  $1,%eax
-	jnz	  .Laes_loop_par_dec\@
-
-.Laes_loop_par_dec_done\@:
-	MOVADQ	  (%r10), \TMP3
-	aesenclast \TMP3, \XMM1           # last round
-	aesenclast \TMP3, \XMM2
-	aesenclast \TMP3, \XMM3
-	aesenclast \TMP3, \XMM4
-	movdqu    HashKey_k(%arg2), \TMP5
-	pclmulqdq $0x00, \TMP5, \TMP2          # TMP2 = (a1+a0)*(b1+b0)
-	movdqu	  (%arg4,%r11,1), \TMP3
-	pxor	  \TMP3, \XMM1                 # Ciphertext/Plaintext XOR EK
-	movdqu	  \XMM1, (%arg3,%r11,1)        # Write to plaintext buffer
-	movdqa    \TMP3, \XMM1
-	movdqu	  16(%arg4,%r11,1), \TMP3
-	pxor	  \TMP3, \XMM2                 # Ciphertext/Plaintext XOR EK
-	movdqu	  \XMM2, 16(%arg3,%r11,1)      # Write to plaintext buffer
-	movdqa    \TMP3, \XMM2
-	movdqu	  32(%arg4,%r11,1), \TMP3
-	pxor	  \TMP3, \XMM3                 # Ciphertext/Plaintext XOR EK
-	movdqu	  \XMM3, 32(%arg3,%r11,1)      # Write to plaintext buffer
-	movdqa    \TMP3, \XMM3
-	movdqu	  48(%arg4,%r11,1), \TMP3
-	pxor	  \TMP3, \XMM4                 # Ciphertext/Plaintext XOR EK
-	movdqu	  \XMM4, 48(%arg3,%r11,1)      # Write to plaintext buffer
-	movdqa    \TMP3, \XMM4
-	pshufb %xmm15, \XMM1        # perform a 16 byte swap
-	pshufb %xmm15, \XMM2	# perform a 16 byte swap
-	pshufb %xmm15, \XMM3	# perform a 16 byte swap
-	pshufb %xmm15, \XMM4	# perform a 16 byte swap
-
-	pxor	  \TMP4, \TMP1
-	pxor	  \XMM8, \XMM5
-	pxor	  \TMP6, \TMP2
-	pxor	  \TMP1, \TMP2
-	pxor	  \XMM5, \TMP2
-	movdqa	  \TMP2, \TMP3
-	pslldq	  $8, \TMP3                    # left shift TMP3 2 DWs
-	psrldq	  $8, \TMP2                    # right shift TMP2 2 DWs
-	pxor	  \TMP3, \XMM5
-	pxor	  \TMP2, \TMP1	  # accumulate the results in TMP1:XMM5
-
-        # first phase of reduction
-
-	movdqa    \XMM5, \TMP2
-	movdqa    \XMM5, \TMP3
-	movdqa    \XMM5, \TMP4
-# move XMM5 into TMP2, TMP3, TMP4 in order to perform shifts independently
-	pslld     $31, \TMP2                   # packed right shift << 31
-	pslld     $30, \TMP3                   # packed right shift << 30
-	pslld     $25, \TMP4                   # packed right shift << 25
-	pxor      \TMP3, \TMP2	               # xor the shifted versions
-	pxor      \TMP4, \TMP2
-	movdqa    \TMP2, \TMP5
-	psrldq    $4, \TMP5                    # right shift T5 1 DW
-	pslldq    $12, \TMP2                   # left shift T2 3 DWs
-	pxor      \TMP2, \XMM5
-
-        # second phase of reduction
-
-	movdqa    \XMM5,\TMP2 # make 3 copies of XMM5 into TMP2, TMP3, TMP4
-	movdqa    \XMM5,\TMP3
-	movdqa    \XMM5,\TMP4
-	psrld     $1, \TMP2                    # packed left shift >>1
-	psrld     $2, \TMP3                    # packed left shift >>2
-	psrld     $7, \TMP4                    # packed left shift >>7
-	pxor      \TMP3,\TMP2		       # xor the shifted versions
-	pxor      \TMP4,\TMP2
-	pxor      \TMP5, \TMP2
-	pxor      \TMP2, \XMM5
-	pxor      \TMP1, \XMM5                 # result is in TMP1
-
-	pxor	  \XMM5, \XMM1
-.endm
-
-/* GHASH the last 4 ciphertext blocks. */
-.macro	GHASH_LAST_4 TMP1 TMP2 TMP3 TMP4 TMP5 TMP6 \
-TMP7 XMM1 XMM2 XMM3 XMM4 XMMDst
-
-        # Multiply TMP6 * HashKey (using Karatsuba)
-
-	movdqa	  \XMM1, \TMP6
-	pshufd	  $78, \XMM1, \TMP2
-	pxor	  \XMM1, \TMP2
-	movdqu	  HashKey_4(%arg2), \TMP5
-	pclmulqdq $0x11, \TMP5, \TMP6       # TMP6 = a1*b1
-	pclmulqdq $0x00, \TMP5, \XMM1       # XMM1 = a0*b0
-	movdqu	  HashKey_4_k(%arg2), \TMP4
-	pclmulqdq $0x00, \TMP4, \TMP2       # TMP2 = (a1+a0)*(b1+b0)
-	movdqa	  \XMM1, \XMMDst
-	movdqa	  \TMP2, \XMM1              # result in TMP6, XMMDst, XMM1
-
-        # Multiply TMP1 * HashKey (using Karatsuba)
-
-	movdqa	  \XMM2, \TMP1
-	pshufd	  $78, \XMM2, \TMP2
-	pxor	  \XMM2, \TMP2
-	movdqu	  HashKey_3(%arg2), \TMP5
-	pclmulqdq $0x11, \TMP5, \TMP1       # TMP1 = a1*b1
-	pclmulqdq $0x00, \TMP5, \XMM2       # XMM2 = a0*b0
-	movdqu	  HashKey_3_k(%arg2), \TMP4
-	pclmulqdq $0x00, \TMP4, \TMP2       # TMP2 = (a1+a0)*(b1+b0)
-	pxor	  \TMP1, \TMP6
-	pxor	  \XMM2, \XMMDst
-	pxor	  \TMP2, \XMM1
-# results accumulated in TMP6, XMMDst, XMM1
-
-        # Multiply TMP1 * HashKey (using Karatsuba)
-
-	movdqa	  \XMM3, \TMP1
-	pshufd	  $78, \XMM3, \TMP2
-	pxor	  \XMM3, \TMP2
-	movdqu	  HashKey_2(%arg2), \TMP5
-	pclmulqdq $0x11, \TMP5, \TMP1       # TMP1 = a1*b1
-	pclmulqdq $0x00, \TMP5, \XMM3       # XMM3 = a0*b0
-	movdqu	  HashKey_2_k(%arg2), \TMP4
-	pclmulqdq $0x00, \TMP4, \TMP2       # TMP2 = (a1+a0)*(b1+b0)
-	pxor	  \TMP1, \TMP6
-	pxor	  \XMM3, \XMMDst
-	pxor	  \TMP2, \XMM1   # results accumulated in TMP6, XMMDst, XMM1
-
-        # Multiply TMP1 * HashKey (using Karatsuba)
-	movdqa	  \XMM4, \TMP1
-	pshufd	  $78, \XMM4, \TMP2
-	pxor	  \XMM4, \TMP2
-	movdqu	  HashKey(%arg2), \TMP5
-	pclmulqdq $0x11, \TMP5, \TMP1	    # TMP1 = a1*b1
-	pclmulqdq $0x00, \TMP5, \XMM4       # XMM4 = a0*b0
-	movdqu	  HashKey_k(%arg2), \TMP4
-	pclmulqdq $0x00, \TMP4, \TMP2       # TMP2 = (a1+a0)*(b1+b0)
-	pxor	  \TMP1, \TMP6
-	pxor	  \XMM4, \XMMDst
-	pxor	  \XMM1, \TMP2
-	pxor	  \TMP6, \TMP2
-	pxor	  \XMMDst, \TMP2
-	# middle section of the temp results combined as in karatsuba algorithm
-	movdqa	  \TMP2, \TMP4
-	pslldq	  $8, \TMP4                 # left shift TMP4 2 DWs
-	psrldq	  $8, \TMP2                 # right shift TMP2 2 DWs
-	pxor	  \TMP4, \XMMDst
-	pxor	  \TMP2, \TMP6
-# TMP6:XMMDst holds the result of the accumulated carry-less multiplications
-	# first phase of the reduction
-	movdqa    \XMMDst, \TMP2
-	movdqa    \XMMDst, \TMP3
-	movdqa    \XMMDst, \TMP4
-# move XMMDst into TMP2, TMP3, TMP4 in order to perform 3 shifts independently
-	pslld     $31, \TMP2                # packed right shifting << 31
-	pslld     $30, \TMP3                # packed right shifting << 30
-	pslld     $25, \TMP4                # packed right shifting << 25
-	pxor      \TMP3, \TMP2              # xor the shifted versions
-	pxor      \TMP4, \TMP2
-	movdqa    \TMP2, \TMP7
-	psrldq    $4, \TMP7                 # right shift TMP7 1 DW
-	pslldq    $12, \TMP2                # left shift TMP2 3 DWs
-	pxor      \TMP2, \XMMDst
-
-        # second phase of the reduction
-	movdqa    \XMMDst, \TMP2
-	# make 3 copies of XMMDst for doing 3 shift operations
-	movdqa    \XMMDst, \TMP3
-	movdqa    \XMMDst, \TMP4
-	psrld     $1, \TMP2                 # packed left shift >> 1
-	psrld     $2, \TMP3                 # packed left shift >> 2
-	psrld     $7, \TMP4                 # packed left shift >> 7
-	pxor      \TMP3, \TMP2              # xor the shifted versions
-	pxor      \TMP4, \TMP2
-	pxor      \TMP7, \TMP2
-	pxor      \TMP2, \XMMDst
-	pxor      \TMP6, \XMMDst            # reduced result is in XMMDst
-.endm
-
-
-/* Encryption of a single block
-* uses eax & r10
-*/
-
-.macro ENCRYPT_SINGLE_BLOCK XMM0 TMP1
-
-	pxor		(%arg1), \XMM0
-	mov		keysize,%eax
-	shr		$2,%eax			# 128->4, 192->6, 256->8
-	add		$5,%eax			# 128->9, 192->11, 256->13
-	lea		16(%arg1), %r10	  # get first expanded key address
-
-_esb_loop_\@:
-	MOVADQ		(%r10),\TMP1
-	aesenc		\TMP1,\XMM0
-	add		$16,%r10
-	sub		$1,%eax
-	jnz		_esb_loop_\@
-
-	MOVADQ		(%r10),\TMP1
-	aesenclast	\TMP1,\XMM0
-.endm
-
-/*****************************************************************************
-* void aesni_gcm_init(void *aes_ctx,      // AES Key schedule. Starts on a 16 byte boundary.
-*                     struct gcm_context_data *data,
-*                                         // context data
-*                     u8 *iv,             // Pre-counter block j0: 4 byte salt (from Security Association)
-*                                         // concatenated with 8 byte Initialisation Vector (from IPSec ESP Payload)
-*                                         // concatenated with 0x00000001. 16-byte aligned pointer.
-*                     u8 *hash_subkey,    // H, the Hash sub key input. Data starts on a 16-byte boundary.
-*                     const u8 *aad,      // Additional Authentication Data (AAD)
-*                     u64 aad_len)        // Length of AAD in bytes.
-*/
-SYM_FUNC_START(aesni_gcm_init)
-	FUNC_SAVE
-	GCM_INIT %arg3, %arg4,%arg5, %arg6
-	FUNC_RESTORE
-	RET
-SYM_FUNC_END(aesni_gcm_init)
-
-/*****************************************************************************
-* void aesni_gcm_enc_update(void *aes_ctx,      // AES Key schedule. Starts on a 16 byte boundary.
-*                    struct gcm_context_data *data,
-*                                        // context data
-*                    u8 *out,            // Ciphertext output. Encrypt in-place is allowed.
-*                    const u8 *in,       // Plaintext input
-*                    u64 plaintext_len,  // Length of data in bytes for encryption.
-*/
-SYM_FUNC_START(aesni_gcm_enc_update)
-	FUNC_SAVE
-	GCM_ENC_DEC enc
-	FUNC_RESTORE
-	RET
-SYM_FUNC_END(aesni_gcm_enc_update)
-
-/*****************************************************************************
-* void aesni_gcm_dec_update(void *aes_ctx,      // AES Key schedule. Starts on a 16 byte boundary.
-*                    struct gcm_context_data *data,
-*                                        // context data
-*                    u8 *out,            // Ciphertext output. Encrypt in-place is allowed.
-*                    const u8 *in,       // Plaintext input
-*                    u64 plaintext_len,  // Length of data in bytes for encryption.
-*/
-SYM_FUNC_START(aesni_gcm_dec_update)
-	FUNC_SAVE
-	GCM_ENC_DEC dec
-	FUNC_RESTORE
-	RET
-SYM_FUNC_END(aesni_gcm_dec_update)
-
-/*****************************************************************************
-* void aesni_gcm_finalize(void *aes_ctx,      // AES Key schedule. Starts on a 16 byte boundary.
-*                    struct gcm_context_data *data,
-*                                        // context data
-*                    u8 *auth_tag,       // Authenticated Tag output.
-*                    u64 auth_tag_len);  // Authenticated Tag Length in bytes. Valid values are 16 (most likely),
-*                                        // 12 or 8.
-*/
-SYM_FUNC_START(aesni_gcm_finalize)
-	FUNC_SAVE
-	GCM_COMPLETE %arg3 %arg4
-	FUNC_RESTORE
-	RET
-SYM_FUNC_END(aesni_gcm_finalize)
-
-#endif
-
 SYM_FUNC_START_LOCAL(_key_expansion_256a)
 	pshufd $0b11111111, %xmm1, %xmm1
 	shufps $0b00010000, %xmm0, %xmm4