xref: /openbmc/linux/arch/x86/crypto/sm3-avx-asm_64.S (revision 8ba490d9)
1/* SPDX-License-Identifier: GPL-2.0-or-later */
2/*
3 * SM3 AVX accelerated transform.
4 * specified in: https://datatracker.ietf.org/doc/html/draft-sca-cfrg-sm3-02
5 *
6 * Copyright (C) 2021 Jussi Kivilinna <jussi.kivilinna@iki.fi>
7 * Copyright (C) 2021 Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
8 */
9
10/* Based on SM3 AES/BMI2 accelerated work by libgcrypt at:
11 *  https://gnupg.org/software/libgcrypt/index.html
12 */
13
14#include <linux/linkage.h>
15#include <linux/cfi_types.h>
16#include <asm/frame.h>
17
18/* Context structure */
19
20#define state_h0 0
21#define state_h1 4
22#define state_h2 8
23#define state_h3 12
24#define state_h4 16
25#define state_h5 20
26#define state_h6 24
27#define state_h7 28
28
29/* Constants */
30
31/* Round constant macros */
32
33#define K0   2043430169  /* 0x79cc4519 */
34#define K1   -208106958  /* 0xf3988a32 */
35#define K2   -416213915  /* 0xe7311465 */
36#define K3   -832427829  /* 0xce6228cb */
37#define K4  -1664855657  /* 0x9cc45197 */
38#define K5    965255983  /* 0x3988a32f */
39#define K6   1930511966  /* 0x7311465e */
40#define K7   -433943364  /* 0xe6228cbc */
41#define K8   -867886727  /* 0xcc451979 */
42#define K9  -1735773453  /* 0x988a32f3 */
43#define K10   823420391  /* 0x311465e7 */
44#define K11  1646840782  /* 0x6228cbce */
45#define K12 -1001285732  /* 0xc451979c */
46#define K13 -2002571463  /* 0x88a32f39 */
47#define K14   289824371  /* 0x11465e73 */
48#define K15   579648742  /* 0x228cbce6 */
49#define K16 -1651869049  /* 0x9d8a7a87 */
50#define K17   991229199  /* 0x3b14f50f */
51#define K18  1982458398  /* 0x7629ea1e */
52#define K19  -330050500  /* 0xec53d43c */
53#define K20  -660100999  /* 0xd8a7a879 */
54#define K21 -1320201997  /* 0xb14f50f3 */
55#define K22  1654563303  /* 0x629ea1e7 */
56#define K23  -985840690  /* 0xc53d43ce */
57#define K24 -1971681379  /* 0x8a7a879d */
58#define K25   351604539  /* 0x14f50f3b */
59#define K26   703209078  /* 0x29ea1e76 */
60#define K27  1406418156  /* 0x53d43cec */
61#define K28 -1482130984  /* 0xa7a879d8 */
62#define K29  1330705329  /* 0x4f50f3b1 */
63#define K30 -1633556638  /* 0x9ea1e762 */
64#define K31  1027854021  /* 0x3d43cec5 */
65#define K32  2055708042  /* 0x7a879d8a */
66#define K33  -183551212  /* 0xf50f3b14 */
67#define K34  -367102423  /* 0xea1e7629 */
68#define K35  -734204845  /* 0xd43cec53 */
69#define K36 -1468409689  /* 0xa879d8a7 */
70#define K37  1358147919  /* 0x50f3b14f */
71#define K38 -1578671458  /* 0xa1e7629e */
72#define K39  1137624381  /* 0x43cec53d */
73#define K40 -2019718534  /* 0x879d8a7a */
74#define K41   255530229  /* 0x0f3b14f5 */
75#define K42   511060458  /* 0x1e7629ea */
76#define K43  1022120916  /* 0x3cec53d4 */
77#define K44  2044241832  /* 0x79d8a7a8 */
78#define K45  -206483632  /* 0xf3b14f50 */
79#define K46  -412967263  /* 0xe7629ea1 */
80#define K47  -825934525  /* 0xcec53d43 */
81#define K48 -1651869049  /* 0x9d8a7a87 */
82#define K49   991229199  /* 0x3b14f50f */
83#define K50  1982458398  /* 0x7629ea1e */
84#define K51  -330050500  /* 0xec53d43c */
85#define K52  -660100999  /* 0xd8a7a879 */
86#define K53 -1320201997  /* 0xb14f50f3 */
87#define K54  1654563303  /* 0x629ea1e7 */
88#define K55  -985840690  /* 0xc53d43ce */
89#define K56 -1971681379  /* 0x8a7a879d */
90#define K57   351604539  /* 0x14f50f3b */
91#define K58   703209078  /* 0x29ea1e76 */
92#define K59  1406418156  /* 0x53d43cec */
93#define K60 -1482130984  /* 0xa7a879d8 */
94#define K61  1330705329  /* 0x4f50f3b1 */
95#define K62 -1633556638  /* 0x9ea1e762 */
96#define K63  1027854021  /* 0x3d43cec5 */
97
98/* Register macros */
99
100#define RSTATE %rdi
101#define RDATA  %rsi
102#define RNBLKS %rdx
103
104#define t0 %eax
105#define t1 %ebx
106#define t2 %ecx
107
108#define a %r8d
109#define b %r9d
110#define c %r10d
111#define d %r11d
112#define e %r12d
113#define f %r13d
114#define g %r14d
115#define h %r15d
116
117#define W0 %xmm0
118#define W1 %xmm1
119#define W2 %xmm2
120#define W3 %xmm3
121#define W4 %xmm4
122#define W5 %xmm5
123
124#define XTMP0 %xmm6
125#define XTMP1 %xmm7
126#define XTMP2 %xmm8
127#define XTMP3 %xmm9
128#define XTMP4 %xmm10
129#define XTMP5 %xmm11
130#define XTMP6 %xmm12
131
132#define BSWAP_REG %xmm15
133
134/* Stack structure */
135
136#define STACK_W_SIZE        (32 * 2 * 3)
137#define STACK_REG_SAVE_SIZE (64)
138
139#define STACK_W             (0)
140#define STACK_REG_SAVE      (STACK_W + STACK_W_SIZE)
141#define STACK_SIZE          (STACK_REG_SAVE + STACK_REG_SAVE_SIZE)
142
143/* Instruction helpers. */
144
145#define roll2(v, reg)		\
146	roll $(v), reg;
147
148#define roll3mov(v, src, dst)	\
149	movl src, dst;		\
150	roll $(v), dst;
151
152#define roll3(v, src, dst)	\
153	rorxl $(32-(v)), src, dst;
154
155#define addl2(a, out)		\
156	leal (a, out), out;
157
158/* Round function macros. */
159
160#define GG1(x, y, z, o, t)	\
161	movl x, o;		\
162	xorl y, o;		\
163	xorl z, o;
164
165#define FF1(x, y, z, o, t) GG1(x, y, z, o, t)
166
167#define GG2(x, y, z, o, t)	\
168	andnl z, x, o;		\
169	movl y, t;		\
170	andl x, t;		\
171	addl2(t, o);
172
173#define FF2(x, y, z, o, t)	\
174	movl y, o;		\
175	xorl x, o;		\
176	movl y, t;		\
177	andl x, t;		\
178	andl z, o;		\
179	xorl t, o;
180
181#define R(i, a, b, c, d, e, f, g, h, round, widx, wtype)		\
182	/* rol(a, 12) => t0 */						\
183	roll3mov(12, a, t0); /* rorxl here would reduce perf by 6% on zen3 */ \
184	/* rol (t0 + e + t), 7) => t1 */				\
185	leal K##round(t0, e, 1), t1;					\
186	roll2(7, t1);							\
187	/* h + w1 => h */						\
188	addl wtype##_W1_ADDR(round, widx), h;				\
189	/* h + t1 => h */						\
190	addl2(t1, h);							\
191	/* t1 ^ t0 => t0 */						\
192	xorl t1, t0;							\
193	/* w1w2 + d => d */						\
194	addl wtype##_W1W2_ADDR(round, widx), d;				\
195	/* FF##i(a,b,c) => t1 */					\
196	FF##i(a, b, c, t1, t2);						\
197	/* d + t1 => d */						\
198	addl2(t1, d);							\
199	/* GG#i(e,f,g) => t2 */						\
200	GG##i(e, f, g, t2, t1);						\
201	/* h + t2 => h */						\
202	addl2(t2, h);							\
203	/* rol (f, 19) => f */						\
204	roll2(19, f);							\
205	/* d + t0 => d */						\
206	addl2(t0, d);							\
207	/* rol (b, 9) => b */						\
208	roll2(9, b);							\
209	/* P0(h) => h */						\
210	roll3(9, h, t2);						\
211	roll3(17, h, t1);						\
212	xorl t2, h;							\
213	xorl t1, h;
214
215#define R1(a, b, c, d, e, f, g, h, round, widx, wtype) \
216	R(1, a, b, c, d, e, f, g, h, round, widx, wtype)
217
218#define R2(a, b, c, d, e, f, g, h, round, widx, wtype) \
219	R(2, a, b, c, d, e, f, g, h, round, widx, wtype)
220
221/* Input expansion macros. */
222
223/* Byte-swapped input address. */
224#define IW_W_ADDR(round, widx, offs) \
225	(STACK_W + ((round) / 4) * 64 + (offs) + ((widx) * 4))(%rsp)
226
227/* Expanded input address. */
228#define XW_W_ADDR(round, widx, offs) \
229	(STACK_W + ((((round) / 3) - 4) % 2) * 64 + (offs) + ((widx) * 4))(%rsp)
230
231/* Rounds 1-12, byte-swapped input block addresses. */
232#define IW_W1_ADDR(round, widx)   IW_W_ADDR(round, widx, 0)
233#define IW_W1W2_ADDR(round, widx) IW_W_ADDR(round, widx, 32)
234
235/* Rounds 1-12, expanded input block addresses. */
236#define XW_W1_ADDR(round, widx)   XW_W_ADDR(round, widx, 0)
237#define XW_W1W2_ADDR(round, widx) XW_W_ADDR(round, widx, 32)
238
239/* Input block loading. */
240#define LOAD_W_XMM_1()							\
241	vmovdqu 0*16(RDATA), XTMP0; /* XTMP0: w3, w2, w1, w0 */		\
242	vmovdqu 1*16(RDATA), XTMP1; /* XTMP1: w7, w6, w5, w4 */		\
243	vmovdqu 2*16(RDATA), XTMP2; /* XTMP2: w11, w10, w9, w8 */	\
244	vmovdqu 3*16(RDATA), XTMP3; /* XTMP3: w15, w14, w13, w12 */	\
245	vpshufb BSWAP_REG, XTMP0, XTMP0;				\
246	vpshufb BSWAP_REG, XTMP1, XTMP1;				\
247	vpshufb BSWAP_REG, XTMP2, XTMP2;				\
248	vpshufb BSWAP_REG, XTMP3, XTMP3;				\
249	vpxor XTMP0, XTMP1, XTMP4;					\
250	vpxor XTMP1, XTMP2, XTMP5;					\
251	vpxor XTMP2, XTMP3, XTMP6;					\
252	leaq 64(RDATA), RDATA;						\
253	vmovdqa XTMP0, IW_W1_ADDR(0, 0);				\
254	vmovdqa XTMP4, IW_W1W2_ADDR(0, 0);				\
255	vmovdqa XTMP1, IW_W1_ADDR(4, 0);				\
256	vmovdqa XTMP5, IW_W1W2_ADDR(4, 0);
257
258#define LOAD_W_XMM_2()				\
259	vmovdqa XTMP2, IW_W1_ADDR(8, 0);	\
260	vmovdqa XTMP6, IW_W1W2_ADDR(8, 0);
261
262#define LOAD_W_XMM_3()							\
263	vpshufd $0b00000000, XTMP0, W0; /* W0: xx, w0, xx, xx */	\
264	vpshufd $0b11111001, XTMP0, W1; /* W1: xx, w3, w2, w1 */	\
265	vmovdqa XTMP1, W2;              /* W2: xx, w6, w5, w4 */	\
266	vpalignr $12, XTMP1, XTMP2, W3; /* W3: xx, w9, w8, w7 */	\
267	vpalignr $8, XTMP2, XTMP3, W4;  /* W4: xx, w12, w11, w10 */	\
268	vpshufd $0b11111001, XTMP3, W5; /* W5: xx, w15, w14, w13 */
269
270/* Message scheduling. Note: 3 words per XMM register. */
271#define SCHED_W_0(round, w0, w1, w2, w3, w4, w5)			\
272	/* Load (w[i - 16]) => XTMP0 */					\
273	vpshufd $0b10111111, w0, XTMP0;					\
274	vpalignr $12, XTMP0, w1, XTMP0; /* XTMP0: xx, w2, w1, w0 */	\
275	/* Load (w[i - 13]) => XTMP1 */					\
276	vpshufd $0b10111111, w1, XTMP1;					\
277	vpalignr $12, XTMP1, w2, XTMP1;					\
278	/* w[i - 9] == w3 */						\
279	/* XMM3 ^ XTMP0 => XTMP0 */					\
280	vpxor w3, XTMP0, XTMP0;
281
282#define SCHED_W_1(round, w0, w1, w2, w3, w4, w5)	\
283	/* w[i - 3] == w5 */				\
284	/* rol(XMM5, 15) ^ XTMP0 => XTMP0 */		\
285	vpslld $15, w5, XTMP2;				\
286	vpsrld $(32-15), w5, XTMP3;			\
287	vpxor XTMP2, XTMP3, XTMP3;			\
288	vpxor XTMP3, XTMP0, XTMP0;			\
289	/* rol(XTMP1, 7) => XTMP1 */			\
290	vpslld $7, XTMP1, XTMP5;			\
291	vpsrld $(32-7), XTMP1, XTMP1;			\
292	vpxor XTMP5, XTMP1, XTMP1;			\
293	/* XMM4 ^ XTMP1 => XTMP1 */			\
294	vpxor w4, XTMP1, XTMP1;				\
295	/* w[i - 6] == XMM4 */				\
296	/* P1(XTMP0) ^ XTMP1 => XMM0 */			\
297	vpslld $15, XTMP0, XTMP5;			\
298	vpsrld $(32-15), XTMP0, XTMP6;			\
299	vpslld $23, XTMP0, XTMP2;			\
300	vpsrld $(32-23), XTMP0, XTMP3;			\
301	vpxor XTMP0, XTMP1, XTMP1;			\
302	vpxor XTMP6, XTMP5, XTMP5;			\
303	vpxor XTMP3, XTMP2, XTMP2;			\
304	vpxor XTMP2, XTMP5, XTMP5;			\
305	vpxor XTMP5, XTMP1, w0;
306
307#define SCHED_W_2(round, w0, w1, w2, w3, w4, w5)	\
308	/* W1 in XMM12 */				\
309	vpshufd $0b10111111, w4, XTMP4;			\
310	vpalignr $12, XTMP4, w5, XTMP4;			\
311	vmovdqa XTMP4, XW_W1_ADDR((round), 0);		\
312	/* W1 ^ W2 => XTMP1 */				\
313	vpxor w0, XTMP4, XTMP1;				\
314	vmovdqa XTMP1, XW_W1W2_ADDR((round), 0);
315
316
317.section	.rodata.cst16, "aM", @progbits, 16
318.align 16
319
320.Lbe32mask:
321	.long 0x00010203, 0x04050607, 0x08090a0b, 0x0c0d0e0f
322
323.text
324
325/*
326 * Transform nblocks*64 bytes (nblocks*16 32-bit words) at DATA.
327 *
328 * void sm3_transform_avx(struct sm3_state *state,
329 *                        const u8 *data, int nblocks);
330 */
331SYM_TYPED_FUNC_START(sm3_transform_avx)
332	/* input:
333	 *	%rdi: ctx, CTX
334	 *	%rsi: data (64*nblks bytes)
335	 *	%rdx: nblocks
336	 */
337	vzeroupper;
338
339	pushq %rbp;
340	movq %rsp, %rbp;
341
342	movq %rdx, RNBLKS;
343
344	subq $STACK_SIZE, %rsp;
345	andq $(~63), %rsp;
346
347	movq %rbx, (STACK_REG_SAVE + 0 * 8)(%rsp);
348	movq %r15, (STACK_REG_SAVE + 1 * 8)(%rsp);
349	movq %r14, (STACK_REG_SAVE + 2 * 8)(%rsp);
350	movq %r13, (STACK_REG_SAVE + 3 * 8)(%rsp);
351	movq %r12, (STACK_REG_SAVE + 4 * 8)(%rsp);
352
353	vmovdqa .Lbe32mask (%rip), BSWAP_REG;
354
355	/* Get the values of the chaining variables. */
356	movl state_h0(RSTATE), a;
357	movl state_h1(RSTATE), b;
358	movl state_h2(RSTATE), c;
359	movl state_h3(RSTATE), d;
360	movl state_h4(RSTATE), e;
361	movl state_h5(RSTATE), f;
362	movl state_h6(RSTATE), g;
363	movl state_h7(RSTATE), h;
364
365.align 16
366.Loop:
367	/* Load data part1. */
368	LOAD_W_XMM_1();
369
370	leaq -1(RNBLKS), RNBLKS;
371
372	/* Transform 0-3 + Load data part2. */
373	R1(a, b, c, d, e, f, g, h, 0, 0, IW); LOAD_W_XMM_2();
374	R1(d, a, b, c, h, e, f, g, 1, 1, IW);
375	R1(c, d, a, b, g, h, e, f, 2, 2, IW);
376	R1(b, c, d, a, f, g, h, e, 3, 3, IW); LOAD_W_XMM_3();
377
378	/* Transform 4-7 + Precalc 12-14. */
379	R1(a, b, c, d, e, f, g, h, 4, 0, IW);
380	R1(d, a, b, c, h, e, f, g, 5, 1, IW);
381	R1(c, d, a, b, g, h, e, f, 6, 2, IW); SCHED_W_0(12, W0, W1, W2, W3, W4, W5);
382	R1(b, c, d, a, f, g, h, e, 7, 3, IW); SCHED_W_1(12, W0, W1, W2, W3, W4, W5);
383
384	/* Transform 8-11 + Precalc 12-17. */
385	R1(a, b, c, d, e, f, g, h, 8, 0, IW); SCHED_W_2(12, W0, W1, W2, W3, W4, W5);
386	R1(d, a, b, c, h, e, f, g, 9, 1, IW); SCHED_W_0(15, W1, W2, W3, W4, W5, W0);
387	R1(c, d, a, b, g, h, e, f, 10, 2, IW); SCHED_W_1(15, W1, W2, W3, W4, W5, W0);
388	R1(b, c, d, a, f, g, h, e, 11, 3, IW); SCHED_W_2(15, W1, W2, W3, W4, W5, W0);
389
390	/* Transform 12-14 + Precalc 18-20 */
391	R1(a, b, c, d, e, f, g, h, 12, 0, XW); SCHED_W_0(18, W2, W3, W4, W5, W0, W1);
392	R1(d, a, b, c, h, e, f, g, 13, 1, XW); SCHED_W_1(18, W2, W3, W4, W5, W0, W1);
393	R1(c, d, a, b, g, h, e, f, 14, 2, XW); SCHED_W_2(18, W2, W3, W4, W5, W0, W1);
394
395	/* Transform 15-17 + Precalc 21-23 */
396	R1(b, c, d, a, f, g, h, e, 15, 0, XW); SCHED_W_0(21, W3, W4, W5, W0, W1, W2);
397	R2(a, b, c, d, e, f, g, h, 16, 1, XW); SCHED_W_1(21, W3, W4, W5, W0, W1, W2);
398	R2(d, a, b, c, h, e, f, g, 17, 2, XW); SCHED_W_2(21, W3, W4, W5, W0, W1, W2);
399
400	/* Transform 18-20 + Precalc 24-26 */
401	R2(c, d, a, b, g, h, e, f, 18, 0, XW); SCHED_W_0(24, W4, W5, W0, W1, W2, W3);
402	R2(b, c, d, a, f, g, h, e, 19, 1, XW); SCHED_W_1(24, W4, W5, W0, W1, W2, W3);
403	R2(a, b, c, d, e, f, g, h, 20, 2, XW); SCHED_W_2(24, W4, W5, W0, W1, W2, W3);
404
405	/* Transform 21-23 + Precalc 27-29 */
406	R2(d, a, b, c, h, e, f, g, 21, 0, XW); SCHED_W_0(27, W5, W0, W1, W2, W3, W4);
407	R2(c, d, a, b, g, h, e, f, 22, 1, XW); SCHED_W_1(27, W5, W0, W1, W2, W3, W4);
408	R2(b, c, d, a, f, g, h, e, 23, 2, XW); SCHED_W_2(27, W5, W0, W1, W2, W3, W4);
409
410	/* Transform 24-26 + Precalc 30-32 */
411	R2(a, b, c, d, e, f, g, h, 24, 0, XW); SCHED_W_0(30, W0, W1, W2, W3, W4, W5);
412	R2(d, a, b, c, h, e, f, g, 25, 1, XW); SCHED_W_1(30, W0, W1, W2, W3, W4, W5);
413	R2(c, d, a, b, g, h, e, f, 26, 2, XW); SCHED_W_2(30, W0, W1, W2, W3, W4, W5);
414
415	/* Transform 27-29 + Precalc 33-35 */
416	R2(b, c, d, a, f, g, h, e, 27, 0, XW); SCHED_W_0(33, W1, W2, W3, W4, W5, W0);
417	R2(a, b, c, d, e, f, g, h, 28, 1, XW); SCHED_W_1(33, W1, W2, W3, W4, W5, W0);
418	R2(d, a, b, c, h, e, f, g, 29, 2, XW); SCHED_W_2(33, W1, W2, W3, W4, W5, W0);
419
420	/* Transform 30-32 + Precalc 36-38 */
421	R2(c, d, a, b, g, h, e, f, 30, 0, XW); SCHED_W_0(36, W2, W3, W4, W5, W0, W1);
422	R2(b, c, d, a, f, g, h, e, 31, 1, XW); SCHED_W_1(36, W2, W3, W4, W5, W0, W1);
423	R2(a, b, c, d, e, f, g, h, 32, 2, XW); SCHED_W_2(36, W2, W3, W4, W5, W0, W1);
424
425	/* Transform 33-35 + Precalc 39-41 */
426	R2(d, a, b, c, h, e, f, g, 33, 0, XW); SCHED_W_0(39, W3, W4, W5, W0, W1, W2);
427	R2(c, d, a, b, g, h, e, f, 34, 1, XW); SCHED_W_1(39, W3, W4, W5, W0, W1, W2);
428	R2(b, c, d, a, f, g, h, e, 35, 2, XW); SCHED_W_2(39, W3, W4, W5, W0, W1, W2);
429
430	/* Transform 36-38 + Precalc 42-44 */
431	R2(a, b, c, d, e, f, g, h, 36, 0, XW); SCHED_W_0(42, W4, W5, W0, W1, W2, W3);
432	R2(d, a, b, c, h, e, f, g, 37, 1, XW); SCHED_W_1(42, W4, W5, W0, W1, W2, W3);
433	R2(c, d, a, b, g, h, e, f, 38, 2, XW); SCHED_W_2(42, W4, W5, W0, W1, W2, W3);
434
435	/* Transform 39-41 + Precalc 45-47 */
436	R2(b, c, d, a, f, g, h, e, 39, 0, XW); SCHED_W_0(45, W5, W0, W1, W2, W3, W4);
437	R2(a, b, c, d, e, f, g, h, 40, 1, XW); SCHED_W_1(45, W5, W0, W1, W2, W3, W4);
438	R2(d, a, b, c, h, e, f, g, 41, 2, XW); SCHED_W_2(45, W5, W0, W1, W2, W3, W4);
439
440	/* Transform 42-44 + Precalc 48-50 */
441	R2(c, d, a, b, g, h, e, f, 42, 0, XW); SCHED_W_0(48, W0, W1, W2, W3, W4, W5);
442	R2(b, c, d, a, f, g, h, e, 43, 1, XW); SCHED_W_1(48, W0, W1, W2, W3, W4, W5);
443	R2(a, b, c, d, e, f, g, h, 44, 2, XW); SCHED_W_2(48, W0, W1, W2, W3, W4, W5);
444
445	/* Transform 45-47 + Precalc 51-53 */
446	R2(d, a, b, c, h, e, f, g, 45, 0, XW); SCHED_W_0(51, W1, W2, W3, W4, W5, W0);
447	R2(c, d, a, b, g, h, e, f, 46, 1, XW); SCHED_W_1(51, W1, W2, W3, W4, W5, W0);
448	R2(b, c, d, a, f, g, h, e, 47, 2, XW); SCHED_W_2(51, W1, W2, W3, W4, W5, W0);
449
450	/* Transform 48-50 + Precalc 54-56 */
451	R2(a, b, c, d, e, f, g, h, 48, 0, XW); SCHED_W_0(54, W2, W3, W4, W5, W0, W1);
452	R2(d, a, b, c, h, e, f, g, 49, 1, XW); SCHED_W_1(54, W2, W3, W4, W5, W0, W1);
453	R2(c, d, a, b, g, h, e, f, 50, 2, XW); SCHED_W_2(54, W2, W3, W4, W5, W0, W1);
454
455	/* Transform 51-53 + Precalc 57-59 */
456	R2(b, c, d, a, f, g, h, e, 51, 0, XW); SCHED_W_0(57, W3, W4, W5, W0, W1, W2);
457	R2(a, b, c, d, e, f, g, h, 52, 1, XW); SCHED_W_1(57, W3, W4, W5, W0, W1, W2);
458	R2(d, a, b, c, h, e, f, g, 53, 2, XW); SCHED_W_2(57, W3, W4, W5, W0, W1, W2);
459
460	/* Transform 54-56 + Precalc 60-62 */
461	R2(c, d, a, b, g, h, e, f, 54, 0, XW); SCHED_W_0(60, W4, W5, W0, W1, W2, W3);
462	R2(b, c, d, a, f, g, h, e, 55, 1, XW); SCHED_W_1(60, W4, W5, W0, W1, W2, W3);
463	R2(a, b, c, d, e, f, g, h, 56, 2, XW); SCHED_W_2(60, W4, W5, W0, W1, W2, W3);
464
465	/* Transform 57-59 + Precalc 63 */
466	R2(d, a, b, c, h, e, f, g, 57, 0, XW); SCHED_W_0(63, W5, W0, W1, W2, W3, W4);
467	R2(c, d, a, b, g, h, e, f, 58, 1, XW);
468	R2(b, c, d, a, f, g, h, e, 59, 2, XW); SCHED_W_1(63, W5, W0, W1, W2, W3, W4);
469
470	/* Transform 60-62 + Precalc 63 */
471	R2(a, b, c, d, e, f, g, h, 60, 0, XW);
472	R2(d, a, b, c, h, e, f, g, 61, 1, XW); SCHED_W_2(63, W5, W0, W1, W2, W3, W4);
473	R2(c, d, a, b, g, h, e, f, 62, 2, XW);
474
475	/* Transform 63 */
476	R2(b, c, d, a, f, g, h, e, 63, 0, XW);
477
478	/* Update the chaining variables. */
479	xorl state_h0(RSTATE), a;
480	xorl state_h1(RSTATE), b;
481	xorl state_h2(RSTATE), c;
482	xorl state_h3(RSTATE), d;
483	movl a, state_h0(RSTATE);
484	movl b, state_h1(RSTATE);
485	movl c, state_h2(RSTATE);
486	movl d, state_h3(RSTATE);
487	xorl state_h4(RSTATE), e;
488	xorl state_h5(RSTATE), f;
489	xorl state_h6(RSTATE), g;
490	xorl state_h7(RSTATE), h;
491	movl e, state_h4(RSTATE);
492	movl f, state_h5(RSTATE);
493	movl g, state_h6(RSTATE);
494	movl h, state_h7(RSTATE);
495
496	cmpq $0, RNBLKS;
497	jne .Loop;
498
499	vzeroall;
500
501	movq (STACK_REG_SAVE + 0 * 8)(%rsp), %rbx;
502	movq (STACK_REG_SAVE + 1 * 8)(%rsp), %r15;
503	movq (STACK_REG_SAVE + 2 * 8)(%rsp), %r14;
504	movq (STACK_REG_SAVE + 3 * 8)(%rsp), %r13;
505	movq (STACK_REG_SAVE + 4 * 8)(%rsp), %r12;
506
507	vmovdqa %xmm0, IW_W1_ADDR(0, 0);
508	vmovdqa %xmm0, IW_W1W2_ADDR(0, 0);
509	vmovdqa %xmm0, IW_W1_ADDR(4, 0);
510	vmovdqa %xmm0, IW_W1W2_ADDR(4, 0);
511	vmovdqa %xmm0, IW_W1_ADDR(8, 0);
512	vmovdqa %xmm0, IW_W1W2_ADDR(8, 0);
513
514	movq %rbp, %rsp;
515	popq %rbp;
516	RET;
517SYM_FUNC_END(sm3_transform_avx)
518