/* SPDX-License-Identifier: GPL-2.0-or-later */
/*
 * SM4 Cipher Algorithm for ARMv8 with Crypto Extensions
 * as specified in
 * https://tools.ietf.org/id/draft-ribose-cfrg-sm4-10.html
 *
 * Copyright (C) 2022, Alibaba Group.
 * Copyright (C) 2022 Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
 */

#include <linux/linkage.h>
#include <asm/assembler.h>
#include "sm4-ce-asm.h"

.arch	armv8-a+crypto

.irp b, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, \
		20, 24, 25, 26, 27, 28, 29, 30, 31
	.set .Lv\b\().4s, \b
.endr

.macro sm4e, vd, vn
	.inst 0xcec08400 | (.L\vn << 5) | .L\vd
.endm

.macro sm4ekey, vd, vn, vm
	.inst 0xce60c800 | (.L\vm << 16) | (.L\vn << 5) | .L\vd
.endm

/* Register macros */

#define RTMP0	v16
#define RTMP1	v17
#define RTMP2	v18
#define RTMP3	v19

#define RIV	v20
#define RMAC	v20
#define RMASK	v21


.align 3
SYM_FUNC_START(sm4_ce_expand_key)
	/* input:
	 *   x0: 128-bit key
	 *   x1: rkey_enc
	 *   x2: rkey_dec
	 *   x3: fk array
	 *   x4: ck array
	 */
	ld1		{v0.16b}, [x0];
	rev32		v0.16b, v0.16b;
	ld1		{v1.16b}, [x3];
	/* load ck */
	ld1		{v24.16b-v27.16b}, [x4], #64;
	ld1		{v28.16b-v31.16b}, [x4];

	/* input ^ fk */
	eor		v0.16b, v0.16b, v1.16b;

	sm4ekey		v0.4s, v0.4s, v24.4s;
	sm4ekey		v1.4s, v0.4s, v25.4s;
	sm4ekey		v2.4s, v1.4s, v26.4s;
	sm4ekey		v3.4s, v2.4s, v27.4s;
	sm4ekey		v4.4s, v3.4s, v28.4s;
	sm4ekey		v5.4s, v4.4s, v29.4s;
	sm4ekey		v6.4s, v5.4s, v30.4s;
	sm4ekey		v7.4s, v6.4s, v31.4s;

	adr_l		x5, .Lbswap128_mask
	ld1		{v24.16b}, [x5]

	st1		{v0.16b-v3.16b}, [x1], #64;
	st1		{v4.16b-v7.16b}, [x1];

	tbl		v16.16b, {v7.16b}, v24.16b
	tbl		v17.16b, {v6.16b}, v24.16b
	tbl		v18.16b, {v5.16b}, v24.16b
	tbl		v19.16b, {v4.16b}, v24.16b
	tbl		v20.16b, {v3.16b}, v24.16b
	tbl		v21.16b, {v2.16b}, v24.16b
	tbl		v22.16b, {v1.16b}, v24.16b
	tbl		v23.16b, {v0.16b}, v24.16b

	st1		{v16.16b-v19.16b}, [x2], #64
	st1		{v20.16b-v23.16b}, [x2]

	ret;
SYM_FUNC_END(sm4_ce_expand_key)

.align 3
SYM_FUNC_START(sm4_ce_crypt_block)
	/* input:
	 *   x0: round key array, CTX
	 *   x1: dst
	 *   x2: src
	 */
	SM4_PREPARE(x0)

	ld1		{v0.16b}, [x2];
	SM4_CRYPT_BLK(v0);
	st1		{v0.16b}, [x1];

	ret;
SYM_FUNC_END(sm4_ce_crypt_block)

.align 3
SYM_FUNC_START(sm4_ce_crypt)
	/* input:
	 *   x0: round key array, CTX
	 *   x1: dst
	 *   x2: src
	 *   w3: nblocks
	 */
	SM4_PREPARE(x0)

.Lcrypt_loop_blk:
	sub		w3, w3, #8;
	tbnz		w3, #31, .Lcrypt_tail8;

	ld1		{v0.16b-v3.16b}, [x2], #64;
	ld1		{v4.16b-v7.16b}, [x2], #64;

	SM4_CRYPT_BLK8(v0, v1, v2, v3, v4, v5, v6, v7);

	st1		{v0.16b-v3.16b}, [x1], #64;
	st1		{v4.16b-v7.16b}, [x1], #64;

	cbz		w3, .Lcrypt_end;
	b		.Lcrypt_loop_blk;

.Lcrypt_tail8:
	add		w3, w3, #8;
	cmp		w3, #4;
	blt		.Lcrypt_tail4;

	sub		w3, w3, #4;

	ld1		{v0.16b-v3.16b}, [x2], #64;
	SM4_CRYPT_BLK4(v0, v1, v2, v3);
	st1		{v0.16b-v3.16b}, [x1], #64;

	cbz		w3, .Lcrypt_end;

.Lcrypt_tail4:
	sub		w3, w3, #1;

	ld1		{v0.16b}, [x2], #16;
	SM4_CRYPT_BLK(v0);
	st1		{v0.16b}, [x1], #16;

	cbnz		w3, .Lcrypt_tail4;

.Lcrypt_end:
	ret;
SYM_FUNC_END(sm4_ce_crypt)

.align 3
SYM_FUNC_START(sm4_ce_cbc_enc)
	/* input:
	 *   x0: round key array, CTX
	 *   x1: dst
	 *   x2: src
	 *   x3: iv (big endian, 128 bit)
	 *   w4: nblocks
	 */
	SM4_PREPARE(x0)

	ld1		{RIV.16b}, [x3]

.Lcbc_enc_loop_4x:
	cmp		w4, #4
	blt		.Lcbc_enc_loop_1x

	sub		w4, w4, #4

	ld1		{v0.16b-v3.16b}, [x2], #64

	eor		v0.16b, v0.16b, RIV.16b
	SM4_CRYPT_BLK(v0)
	eor		v1.16b, v1.16b, v0.16b
	SM4_CRYPT_BLK(v1)
	eor		v2.16b, v2.16b, v1.16b
	SM4_CRYPT_BLK(v2)
	eor		v3.16b, v3.16b, v2.16b
	SM4_CRYPT_BLK(v3)

	st1		{v0.16b-v3.16b}, [x1], #64
	mov		RIV.16b, v3.16b

	cbz		w4, .Lcbc_enc_end
	b		.Lcbc_enc_loop_4x

.Lcbc_enc_loop_1x:
	sub		w4, w4, #1

	ld1		{v0.16b}, [x2], #16

	eor		RIV.16b, RIV.16b, v0.16b
	SM4_CRYPT_BLK(RIV)

	st1		{RIV.16b}, [x1], #16

	cbnz		w4, .Lcbc_enc_loop_1x

.Lcbc_enc_end:
	/* store new IV */
	st1		{RIV.16b}, [x3]

	ret
SYM_FUNC_END(sm4_ce_cbc_enc)

.align 3
SYM_FUNC_START(sm4_ce_cbc_dec)
	/* input:
	 *   x0: round key array, CTX
	 *   x1: dst
	 *   x2: src
	 *   x3: iv (big endian, 128 bit)
	 *   w4: nblocks
	 */
	SM4_PREPARE(x0)

	ld1		{RIV.16b}, [x3]

.Lcbc_dec_loop_8x:
	sub		w4, w4, #8
	tbnz		w4, #31, .Lcbc_dec_4x

	ld1		{v0.16b-v3.16b}, [x2], #64
	ld1		{v4.16b-v7.16b}, [x2], #64

	rev32		v8.16b, v0.16b
	rev32		v9.16b, v1.16b
	rev32		v10.16b, v2.16b
	rev32		v11.16b, v3.16b
	rev32		v12.16b, v4.16b
	rev32		v13.16b, v5.16b
	rev32		v14.16b, v6.16b
	rev32		v15.16b, v7.16b

	SM4_CRYPT_BLK8_BE(v8, v9, v10, v11, v12, v13, v14, v15)

	eor		v8.16b, v8.16b, RIV.16b
	eor		v9.16b, v9.16b, v0.16b
	eor		v10.16b, v10.16b, v1.16b
	eor		v11.16b, v11.16b, v2.16b
	eor		v12.16b, v12.16b, v3.16b
	eor		v13.16b, v13.16b, v4.16b
	eor		v14.16b, v14.16b, v5.16b
	eor		v15.16b, v15.16b, v6.16b

	st1		{v8.16b-v11.16b}, [x1], #64
	st1		{v12.16b-v15.16b}, [x1], #64

	mov		RIV.16b, v7.16b

	cbz		w4, .Lcbc_dec_end
	b		.Lcbc_dec_loop_8x

.Lcbc_dec_4x:
	add		w4, w4, #8
	cmp		w4, #4
	blt		.Lcbc_dec_loop_1x

	sub		w4, w4, #4

	ld1		{v0.16b-v3.16b}, [x2], #64

	rev32		v8.16b, v0.16b
	rev32		v9.16b, v1.16b
	rev32		v10.16b, v2.16b
	rev32		v11.16b, v3.16b

	SM4_CRYPT_BLK4_BE(v8, v9, v10, v11)

	eor		v8.16b, v8.16b, RIV.16b
	eor		v9.16b, v9.16b, v0.16b
	eor		v10.16b, v10.16b, v1.16b
	eor		v11.16b, v11.16b, v2.16b

	st1		{v8.16b-v11.16b}, [x1], #64

	mov		RIV.16b, v3.16b

	cbz		w4, .Lcbc_dec_end

.Lcbc_dec_loop_1x:
	sub		w4, w4, #1

	ld1		{v0.16b}, [x2], #16

	rev32		v8.16b, v0.16b

	SM4_CRYPT_BLK_BE(v8)

	eor		v8.16b, v8.16b, RIV.16b
	st1		{v8.16b}, [x1], #16

	mov		RIV.16b, v0.16b

	cbnz		w4, .Lcbc_dec_loop_1x

.Lcbc_dec_end:
	/* store new IV */
	st1		{RIV.16b}, [x3]

	ret
SYM_FUNC_END(sm4_ce_cbc_dec)

.align 3
SYM_FUNC_START(sm4_ce_cbc_cts_enc)
	/* input:
	 *   x0: round key array, CTX
	 *   x1: dst
	 *   x2: src
	 *   x3: iv (big endian, 128 bit)
	 *   w4: nbytes
	 */
	SM4_PREPARE(x0)

	sub		w5, w4, #16
	uxtw		x5, w5

	ld1		{RIV.16b}, [x3]

	ld1		{v0.16b}, [x2]
	eor		RIV.16b, RIV.16b, v0.16b
	SM4_CRYPT_BLK(RIV)

	/* load permute table */
	adr_l		x6, .Lcts_permute_table
	add		x7, x6, #32
	add		x6, x6, x5
	sub		x7, x7, x5
	ld1		{v3.16b}, [x6]
	ld1		{v4.16b}, [x7]

	/* overlapping loads */
	add		x2, x2, x5
	ld1		{v1.16b}, [x2]

	/* create Cn from En-1 */
	tbl		v0.16b, {RIV.16b}, v3.16b
	/* padding Pn with zeros */
	tbl		v1.16b, {v1.16b}, v4.16b

	eor		v1.16b, v1.16b, RIV.16b
	SM4_CRYPT_BLK(v1)

	/* overlapping stores */
	add		x5, x1, x5
	st1		{v0.16b}, [x5]
	st1		{v1.16b}, [x1]

	ret
SYM_FUNC_END(sm4_ce_cbc_cts_enc)

.align 3
SYM_FUNC_START(sm4_ce_cbc_cts_dec)
	/* input:
	 *   x0: round key array, CTX
	 *   x1: dst
	 *   x2: src
	 *   x3: iv (big endian, 128 bit)
	 *   w4: nbytes
	 */
	SM4_PREPARE(x0)

	sub		w5, w4, #16
	uxtw		x5, w5

	ld1		{RIV.16b}, [x3]

	/* load permute table */
	adr_l		x6, .Lcts_permute_table
	add		x7, x6, #32
	add		x6, x6, x5
	sub		x7, x7, x5
	ld1		{v3.16b}, [x6]
	ld1		{v4.16b}, [x7]

	/* overlapping loads */
	ld1		{v0.16b}, [x2], x5
	ld1		{v1.16b}, [x2]

	SM4_CRYPT_BLK(v0)
	/* select the first Ln bytes of Xn to create Pn */
	tbl		v2.16b, {v0.16b}, v3.16b
	eor		v2.16b, v2.16b, v1.16b

	/* overwrite the first Ln bytes with Cn to create En-1 */
	tbx		v0.16b, {v1.16b}, v4.16b
	SM4_CRYPT_BLK(v0)
	eor		v0.16b, v0.16b, RIV.16b

	/* overlapping stores */
	add		x5, x1, x5
	st1		{v2.16b}, [x5]
	st1		{v0.16b}, [x1]

	ret
SYM_FUNC_END(sm4_ce_cbc_cts_dec)

.align 3
SYM_FUNC_START(sm4_ce_cfb_enc)
	/* input:
	 *   x0: round key array, CTX
	 *   x1: dst
	 *   x2: src
	 *   x3: iv (big endian, 128 bit)
	 *   w4: nblocks
	 */
	SM4_PREPARE(x0)

	ld1		{RIV.16b}, [x3]

.Lcfb_enc_loop_4x:
	cmp		w4, #4
	blt		.Lcfb_enc_loop_1x

	sub		w4, w4, #4

	ld1		{v0.16b-v3.16b}, [x2], #64

	rev32		v8.16b, RIV.16b
	SM4_CRYPT_BLK_BE(v8)
	eor		v0.16b, v0.16b, v8.16b

	rev32		v8.16b, v0.16b
	SM4_CRYPT_BLK_BE(v8)
	eor		v1.16b, v1.16b, v8.16b

	rev32		v8.16b, v1.16b
	SM4_CRYPT_BLK_BE(v8)
	eor		v2.16b, v2.16b, v8.16b

	rev32		v8.16b, v2.16b
	SM4_CRYPT_BLK_BE(v8)
	eor		v3.16b, v3.16b, v8.16b

	st1		{v0.16b-v3.16b}, [x1], #64
	mov		RIV.16b, v3.16b

	cbz		w4, .Lcfb_enc_end
	b		.Lcfb_enc_loop_4x

.Lcfb_enc_loop_1x:
	sub		w4, w4, #1

	ld1		{v0.16b}, [x2], #16

	SM4_CRYPT_BLK(RIV)
	eor		RIV.16b, RIV.16b, v0.16b

	st1		{RIV.16b}, [x1], #16

	cbnz		w4, .Lcfb_enc_loop_1x

.Lcfb_enc_end:
	/* store new IV */
	st1		{RIV.16b}, [x3]

	ret
SYM_FUNC_END(sm4_ce_cfb_enc)

.align 3
SYM_FUNC_START(sm4_ce_cfb_dec)
	/* input:
	 *   x0: round key array, CTX
	 *   x1: dst
	 *   x2: src
	 *   x3: iv (big endian, 128 bit)
	 *   w4: nblocks
	 */
	SM4_PREPARE(x0)

	ld1		{RIV.16b}, [x3]

.Lcfb_dec_loop_8x:
	sub		w4, w4, #8
	tbnz		w4, #31, .Lcfb_dec_4x

	ld1		{v0.16b-v3.16b}, [x2], #64
	ld1		{v4.16b-v7.16b}, [x2], #64

	rev32		v8.16b, RIV.16b
	rev32		v9.16b, v0.16b
	rev32		v10.16b, v1.16b
	rev32		v11.16b, v2.16b
	rev32		v12.16b, v3.16b
	rev32		v13.16b, v4.16b
	rev32		v14.16b, v5.16b
	rev32		v15.16b, v6.16b

	SM4_CRYPT_BLK8_BE(v8, v9, v10, v11, v12, v13, v14, v15)

	mov		RIV.16b, v7.16b

	eor		v0.16b, v0.16b, v8.16b
	eor		v1.16b, v1.16b, v9.16b
	eor		v2.16b, v2.16b, v10.16b
	eor		v3.16b, v3.16b, v11.16b
	eor		v4.16b, v4.16b, v12.16b
	eor		v5.16b, v5.16b, v13.16b
	eor		v6.16b, v6.16b, v14.16b
	eor		v7.16b, v7.16b, v15.16b

	st1		{v0.16b-v3.16b}, [x1], #64
	st1		{v4.16b-v7.16b}, [x1], #64

	cbz		w4, .Lcfb_dec_end
	b		.Lcfb_dec_loop_8x

.Lcfb_dec_4x:
	add		w4, w4, #8
	cmp		w4, #4
	blt		.Lcfb_dec_loop_1x

	sub		w4, w4, #4

	ld1		{v0.16b-v3.16b}, [x2], #64

	rev32		v8.16b, RIV.16b
	rev32		v9.16b, v0.16b
	rev32		v10.16b, v1.16b
	rev32		v11.16b, v2.16b

	SM4_CRYPT_BLK4_BE(v8, v9, v10, v11)

	mov		RIV.16b, v3.16b

	eor		v0.16b, v0.16b, v8.16b
	eor		v1.16b, v1.16b, v9.16b
	eor		v2.16b, v2.16b, v10.16b
	eor		v3.16b, v3.16b, v11.16b

	st1		{v0.16b-v3.16b}, [x1], #64

	cbz		w4, .Lcfb_dec_end

.Lcfb_dec_loop_1x:
	sub		w4, w4, #1

	ld1		{v0.16b}, [x2], #16

	SM4_CRYPT_BLK(RIV)

	eor		RIV.16b, RIV.16b, v0.16b
	st1		{RIV.16b}, [x1], #16

	mov		RIV.16b, v0.16b

	cbnz		w4, .Lcfb_dec_loop_1x

.Lcfb_dec_end:
	/* store new IV */
	st1		{RIV.16b}, [x3]

	ret
SYM_FUNC_END(sm4_ce_cfb_dec)

.align 3
SYM_FUNC_START(sm4_ce_ctr_enc)
	/* input:
	 *   x0: round key array, CTX
	 *   x1: dst
	 *   x2: src
	 *   x3: ctr (big endian, 128 bit)
	 *   w4: nblocks
	 */
	SM4_PREPARE(x0)

	ldp		x7, x8, [x3]
	rev		x7, x7
	rev		x8, x8

.Lctr_loop_8x:
	sub		w4, w4, #8
	tbnz		w4, #31, .Lctr_4x

#define inc_le128(vctr)					\
		mov		vctr.d[1], x8;		\
		mov		vctr.d[0], x7;		\
		adds		x8, x8, #1;		\
		rev64		vctr.16b, vctr.16b;	\
		adc		x7, x7, xzr;

	/* construct CTRs */
	inc_le128(v0)			/* +0 */
	inc_le128(v1)			/* +1 */
	inc_le128(v2)			/* +2 */
	inc_le128(v3)			/* +3 */
	inc_le128(v4)			/* +4 */
	inc_le128(v5)			/* +5 */
	inc_le128(v6)			/* +6 */
	inc_le128(v7)			/* +7 */

	ld1		{v8.16b-v11.16b}, [x2], #64
	ld1		{v12.16b-v15.16b}, [x2], #64

	SM4_CRYPT_BLK8(v0, v1, v2, v3, v4, v5, v6, v7)

	eor		v0.16b, v0.16b, v8.16b
	eor		v1.16b, v1.16b, v9.16b
	eor		v2.16b, v2.16b, v10.16b
	eor		v3.16b, v3.16b, v11.16b
	eor		v4.16b, v4.16b, v12.16b
	eor		v5.16b, v5.16b, v13.16b
	eor		v6.16b, v6.16b, v14.16b
	eor		v7.16b, v7.16b, v15.16b

	st1		{v0.16b-v3.16b}, [x1], #64
	st1		{v4.16b-v7.16b}, [x1], #64

	cbz		w4, .Lctr_end
	b		.Lctr_loop_8x

.Lctr_4x:
	add		w4, w4, #8
	cmp		w4, #4
	blt		.Lctr_loop_1x

	sub		w4, w4, #4

	/* construct CTRs */
	inc_le128(v0)			/* +0 */
	inc_le128(v1)			/* +1 */
	inc_le128(v2)			/* +2 */
	inc_le128(v3)			/* +3 */

	ld1		{v8.16b-v11.16b}, [x2], #64

	SM4_CRYPT_BLK4(v0, v1, v2, v3)

	eor		v0.16b, v0.16b, v8.16b
	eor		v1.16b, v1.16b, v9.16b
	eor		v2.16b, v2.16b, v10.16b
	eor		v3.16b, v3.16b, v11.16b

	st1		{v0.16b-v3.16b}, [x1], #64

	cbz		w4, .Lctr_end

.Lctr_loop_1x:
	sub		w4, w4, #1

	/* construct CTRs */
	inc_le128(v0)

	ld1		{v8.16b}, [x2], #16

	SM4_CRYPT_BLK(v0)

	eor		v0.16b, v0.16b, v8.16b
	st1		{v0.16b}, [x1], #16

	cbnz		w4, .Lctr_loop_1x

.Lctr_end:
	/* store new CTR */
	rev		x7, x7
	rev		x8, x8
	stp		x7, x8, [x3]

	ret
SYM_FUNC_END(sm4_ce_ctr_enc)


#define tweak_next(vt, vin, RTMP)					\
		sshr		RTMP.2d, vin.2d, #63;			\
		and		RTMP.16b, RTMP.16b, RMASK.16b;		\
		add		vt.2d, vin.2d, vin.2d;			\
		ext		RTMP.16b, RTMP.16b, RTMP.16b, #8;	\
		eor		vt.16b, vt.16b, RTMP.16b;

.align 3
SYM_FUNC_START(sm4_ce_xts_enc)
	/* input:
	 *   x0: round key array, CTX
	 *   x1: dst
	 *   x2: src
	 *   x3: tweak (big endian, 128 bit)
	 *   w4: nbytes
	 *   x5: round key array for IV
	 */
	ld1		{v8.16b}, [x3]

	cbz		x5, .Lxts_enc_nofirst

	SM4_PREPARE(x5)

	/* Generate first tweak */
	SM4_CRYPT_BLK(v8)

.Lxts_enc_nofirst:
	SM4_PREPARE(x0)

	ands		w5, w4, #15
	lsr		w4, w4, #4
	sub		w6, w4, #1
	csel		w4, w4, w6, eq
	uxtw		x5, w5

	movi		RMASK.2s, #0x1
	movi		RTMP0.2s, #0x87
	uzp1		RMASK.4s, RMASK.4s, RTMP0.4s

	cbz		w4, .Lxts_enc_cts

.Lxts_enc_loop_8x:
	sub		w4, w4, #8
	tbnz		w4, #31, .Lxts_enc_4x

	tweak_next( v9,  v8, RTMP0)
	tweak_next(v10,  v9, RTMP1)
	tweak_next(v11, v10, RTMP2)
	tweak_next(v12, v11, RTMP3)
	tweak_next(v13, v12, RTMP0)
	tweak_next(v14, v13, RTMP1)
	tweak_next(v15, v14, RTMP2)

	ld1		{v0.16b-v3.16b}, [x2], #64
	ld1		{v4.16b-v7.16b}, [x2], #64
	eor		v0.16b, v0.16b,  v8.16b
	eor		v1.16b, v1.16b,  v9.16b
	eor		v2.16b, v2.16b, v10.16b
	eor		v3.16b, v3.16b, v11.16b
	eor		v4.16b, v4.16b, v12.16b
	eor		v5.16b, v5.16b, v13.16b
	eor		v6.16b, v6.16b, v14.16b
	eor		v7.16b, v7.16b, v15.16b

	SM4_CRYPT_BLK8(v0, v1, v2, v3, v4, v5, v6, v7)

	eor		v0.16b, v0.16b,  v8.16b
	eor		v1.16b, v1.16b,  v9.16b
	eor		v2.16b, v2.16b, v10.16b
	eor		v3.16b, v3.16b, v11.16b
	eor		v4.16b, v4.16b, v12.16b
	eor		v5.16b, v5.16b, v13.16b
	eor		v6.16b, v6.16b, v14.16b
	eor		v7.16b, v7.16b, v15.16b
	st1		{v0.16b-v3.16b}, [x1], #64
	st1		{v4.16b-v7.16b}, [x1], #64

	tweak_next(v8, v15, RTMP3)

	cbz		w4, .Lxts_enc_cts
	b		.Lxts_enc_loop_8x

.Lxts_enc_4x:
	add		w4, w4, #8
	cmp		w4, #4
	blt		.Lxts_enc_loop_1x

	sub		w4, w4, #4

	tweak_next( v9,  v8, RTMP0)
	tweak_next(v10,  v9, RTMP1)
	tweak_next(v11, v10, RTMP2)

	ld1		{v0.16b-v3.16b}, [x2], #64
	eor		v0.16b, v0.16b,  v8.16b
	eor		v1.16b, v1.16b,  v9.16b
	eor		v2.16b, v2.16b, v10.16b
	eor		v3.16b, v3.16b, v11.16b

	SM4_CRYPT_BLK4(v0, v1, v2, v3)

	eor		v0.16b, v0.16b,  v8.16b
	eor		v1.16b, v1.16b,  v9.16b
	eor		v2.16b, v2.16b, v10.16b
	eor		v3.16b, v3.16b, v11.16b
	st1		{v0.16b-v3.16b}, [x1], #64

	tweak_next(v8, v11, RTMP3)

	cbz		w4, .Lxts_enc_cts

.Lxts_enc_loop_1x:
	sub		w4, w4, #1

	ld1		{v0.16b}, [x2], #16
	eor		v0.16b, v0.16b, v8.16b

	SM4_CRYPT_BLK(v0)

	eor		v0.16b, v0.16b, v8.16b
	st1		{v0.16b}, [x1], #16

	tweak_next(v8, v8, RTMP0)

	cbnz		w4, .Lxts_enc_loop_1x

.Lxts_enc_cts:
	cbz		x5, .Lxts_enc_end

	/* cipher text stealing */

	tweak_next(v9, v8, RTMP0)
	ld1		{v0.16b}, [x2]
	eor		v0.16b, v0.16b, v8.16b
	SM4_CRYPT_BLK(v0)
	eor		v0.16b, v0.16b, v8.16b

	/* load permute table */
	adr_l		x6, .Lcts_permute_table
	add		x7, x6, #32
	add		x6, x6, x5
	sub		x7, x7, x5
	ld1		{v3.16b}, [x6]
	ld1		{v4.16b}, [x7]

	/* overlapping loads */
	add		x2, x2, x5
	ld1		{v1.16b}, [x2]

	/* create Cn from En-1 */
	tbl		v2.16b, {v0.16b}, v3.16b
	/* padding Pn with En-1 at the end */
	tbx		v0.16b, {v1.16b}, v4.16b

	eor		v0.16b, v0.16b, v9.16b
	SM4_CRYPT_BLK(v0)
	eor		v0.16b, v0.16b, v9.16b


	/* overlapping stores */
	add		x5, x1, x5
	st1		{v2.16b}, [x5]
	st1		{v0.16b}, [x1]

	b		.Lxts_enc_ret

.Lxts_enc_end:
	/* store new tweak */
	st1		{v8.16b}, [x3]

.Lxts_enc_ret:
	ret
SYM_FUNC_END(sm4_ce_xts_enc)

.align 3
SYM_FUNC_START(sm4_ce_xts_dec)
	/* input:
	 *   x0: round key array, CTX
	 *   x1: dst
	 *   x2: src
	 *   x3: tweak (big endian, 128 bit)
	 *   w4: nbytes
	 *   x5: round key array for IV
	 */
	ld1		{v8.16b}, [x3]

	cbz		x5, .Lxts_dec_nofirst

	SM4_PREPARE(x5)

	/* Generate first tweak */
	SM4_CRYPT_BLK(v8)

.Lxts_dec_nofirst:
	SM4_PREPARE(x0)

	ands		w5, w4, #15
	lsr		w4, w4, #4
	sub		w6, w4, #1
	csel		w4, w4, w6, eq
	uxtw		x5, w5

	movi		RMASK.2s, #0x1
	movi		RTMP0.2s, #0x87
	uzp1		RMASK.4s, RMASK.4s, RTMP0.4s

	cbz		w4, .Lxts_dec_cts

.Lxts_dec_loop_8x:
	sub		w4, w4, #8
	tbnz		w4, #31, .Lxts_dec_4x

	tweak_next( v9,  v8, RTMP0)
	tweak_next(v10,  v9, RTMP1)
	tweak_next(v11, v10, RTMP2)
	tweak_next(v12, v11, RTMP3)
	tweak_next(v13, v12, RTMP0)
	tweak_next(v14, v13, RTMP1)
	tweak_next(v15, v14, RTMP2)

	ld1		{v0.16b-v3.16b}, [x2], #64
	ld1		{v4.16b-v7.16b}, [x2], #64
	eor		v0.16b, v0.16b,  v8.16b
	eor		v1.16b, v1.16b,  v9.16b
	eor		v2.16b, v2.16b, v10.16b
	eor		v3.16b, v3.16b, v11.16b
	eor		v4.16b, v4.16b, v12.16b
	eor		v5.16b, v5.16b, v13.16b
	eor		v6.16b, v6.16b, v14.16b
	eor		v7.16b, v7.16b, v15.16b

	SM4_CRYPT_BLK8(v0, v1, v2, v3, v4, v5, v6, v7)

	eor		v0.16b, v0.16b,  v8.16b
	eor		v1.16b, v1.16b,  v9.16b
	eor		v2.16b, v2.16b, v10.16b
	eor		v3.16b, v3.16b, v11.16b
	eor		v4.16b, v4.16b, v12.16b
	eor		v5.16b, v5.16b, v13.16b
	eor		v6.16b, v6.16b, v14.16b
	eor		v7.16b, v7.16b, v15.16b
	st1		{v0.16b-v3.16b}, [x1], #64
	st1		{v4.16b-v7.16b}, [x1], #64

	tweak_next(v8, v15, RTMP3)

	cbz		w4, .Lxts_dec_cts
	b		.Lxts_dec_loop_8x

.Lxts_dec_4x:
	add		w4, w4, #8
	cmp		w4, #4
	blt		.Lxts_dec_loop_1x

	sub		w4, w4, #4

	tweak_next( v9,  v8, RTMP0)
	tweak_next(v10,  v9, RTMP1)
	tweak_next(v11, v10, RTMP2)

	ld1		{v0.16b-v3.16b}, [x2], #64
	eor		v0.16b, v0.16b,  v8.16b
	eor		v1.16b, v1.16b,  v9.16b
	eor		v2.16b, v2.16b, v10.16b
	eor		v3.16b, v3.16b, v11.16b

	SM4_CRYPT_BLK4(v0, v1, v2, v3)

	eor		v0.16b, v0.16b,  v8.16b
	eor		v1.16b, v1.16b,  v9.16b
	eor		v2.16b, v2.16b, v10.16b
	eor		v3.16b, v3.16b, v11.16b
	st1		{v0.16b-v3.16b}, [x1], #64

	tweak_next(v8, v11, RTMP3)

	cbz		w4, .Lxts_dec_cts

.Lxts_dec_loop_1x:
	sub		w4, w4, #1

	ld1		{v0.16b}, [x2], #16
	eor		v0.16b, v0.16b, v8.16b

	SM4_CRYPT_BLK(v0)

	eor		v0.16b, v0.16b, v8.16b
	st1		{v0.16b}, [x1], #16

	tweak_next(v8, v8, RTMP0)

	cbnz		w4, .Lxts_dec_loop_1x

.Lxts_dec_cts:
	cbz		x5, .Lxts_dec_end

	/* cipher text stealing */

	tweak_next(v9, v8, RTMP0)
	ld1		{v0.16b}, [x2]
	eor		v0.16b, v0.16b, v9.16b
	SM4_CRYPT_BLK(v0)
	eor		v0.16b, v0.16b, v9.16b

	/* load permute table */
	adr_l		x6, .Lcts_permute_table
	add		x7, x6, #32
	add		x6, x6, x5
	sub		x7, x7, x5
	ld1		{v3.16b}, [x6]
	ld1		{v4.16b}, [x7]

	/* overlapping loads */
	add		x2, x2, x5
	ld1		{v1.16b}, [x2]

	/* create Cn from En-1 */
	tbl		v2.16b, {v0.16b}, v3.16b
	/* padding Pn with En-1 at the end */
	tbx		v0.16b, {v1.16b}, v4.16b

	eor		v0.16b, v0.16b, v8.16b
	SM4_CRYPT_BLK(v0)
	eor		v0.16b, v0.16b, v8.16b


	/* overlapping stores */
	add		x5, x1, x5
	st1		{v2.16b}, [x5]
	st1		{v0.16b}, [x1]

	b		.Lxts_dec_ret

.Lxts_dec_end:
	/* store new tweak */
	st1		{v8.16b}, [x3]

.Lxts_dec_ret:
	ret
SYM_FUNC_END(sm4_ce_xts_dec)

.align 3
SYM_FUNC_START(sm4_ce_mac_update)
	/* input:
	 *   x0: round key array, CTX
	 *   x1: digest
	 *   x2: src
	 *   w3: nblocks
	 *   w4: enc_before
	 *   w5: enc_after
	 */
	SM4_PREPARE(x0)

	ld1		{RMAC.16b}, [x1]

	cbz		w4, .Lmac_update

	SM4_CRYPT_BLK(RMAC)

.Lmac_update:
	cbz		w3, .Lmac_ret

	sub		w6, w3, #1
	cmp		w5, wzr
	csel		w3, w3, w6, ne

	cbz		w3, .Lmac_end

.Lmac_loop_4x:
	cmp		w3, #4
	blt		.Lmac_loop_1x

	sub		w3, w3, #4

	ld1		{v0.16b-v3.16b}, [x2], #64

	eor		RMAC.16b, RMAC.16b, v0.16b
	SM4_CRYPT_BLK(RMAC)
	eor		RMAC.16b, RMAC.16b, v1.16b
	SM4_CRYPT_BLK(RMAC)
	eor		RMAC.16b, RMAC.16b, v2.16b
	SM4_CRYPT_BLK(RMAC)
	eor		RMAC.16b, RMAC.16b, v3.16b
	SM4_CRYPT_BLK(RMAC)

	cbz		w3, .Lmac_end
	b		.Lmac_loop_4x

.Lmac_loop_1x:
	sub		w3, w3, #1

	ld1		{v0.16b}, [x2], #16

	eor		RMAC.16b, RMAC.16b, v0.16b
	SM4_CRYPT_BLK(RMAC)

	cbnz		w3, .Lmac_loop_1x


.Lmac_end:
	cbnz		w5, .Lmac_ret

	ld1		{v0.16b}, [x2], #16
	eor		RMAC.16b, RMAC.16b, v0.16b

.Lmac_ret:
	st1		{RMAC.16b}, [x1]
	ret
SYM_FUNC_END(sm4_ce_mac_update)


	.section	".rodata", "a"
	.align 4
.Lbswap128_mask:
	.byte		0x0c, 0x0d, 0x0e, 0x0f, 0x08, 0x09, 0x0a, 0x0b
	.byte		0x04, 0x05, 0x06, 0x07, 0x00, 0x01, 0x02, 0x03

.Lcts_permute_table:
	.byte		0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff
	.byte		0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff
	.byte		 0x0,  0x1,  0x2,  0x3,  0x4,  0x5,  0x6,  0x7
	.byte		 0x8,  0x9,  0xa,  0xb,  0xc,  0xd,  0xe,  0xf
	.byte		0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff
	.byte		0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff