xref: /linux/arch/powerpc/crypto/sha1-powerpc-asm.S (revision e5c86679d5e864947a52fb31e45a425dea3e7fa9)
1/*
2 * SHA-1 implementation for PowerPC.
3 *
4 * Copyright (C) 2005 Paul Mackerras <paulus@samba.org>
5 */
6
7#include <asm/ppc_asm.h>
8#include <asm/asm-offsets.h>
9
10#ifdef __BIG_ENDIAN__
11#define LWZ(rt, d, ra)	\
12	lwz	rt,d(ra)
13#else
14#define LWZ(rt, d, ra)	\
15	li	rt,d;	\
16	lwbrx	rt,rt,ra
17#endif
18
19/*
20 * We roll the registers for T, A, B, C, D, E around on each
21 * iteration; T on iteration t is A on iteration t+1, and so on.
22 * We use registers 7 - 12 for this.
23 */
24#define RT(t)	((((t)+5)%6)+7)
25#define RA(t)	((((t)+4)%6)+7)
26#define RB(t)	((((t)+3)%6)+7)
27#define RC(t)	((((t)+2)%6)+7)
28#define RD(t)	((((t)+1)%6)+7)
29#define RE(t)	((((t)+0)%6)+7)
30
31/* We use registers 16 - 31 for the W values */
32#define W(t)	(((t)%16)+16)
33
34#define LOADW(t)				\
35	LWZ(W(t),(t)*4,r4)
36
37#define STEPD0_LOAD(t)				\
38	andc	r0,RD(t),RB(t);		\
39	and	r6,RB(t),RC(t);		\
40	rotlwi	RT(t),RA(t),5;			\
41	or	r6,r6,r0;			\
42	add	r0,RE(t),r15;			\
43	add	RT(t),RT(t),r6;		\
44	add	r14,r0,W(t);			\
45	LWZ(W((t)+4),((t)+4)*4,r4);	\
46	rotlwi	RB(t),RB(t),30;			\
47	add	RT(t),RT(t),r14
48
49#define STEPD0_UPDATE(t)			\
50	and	r6,RB(t),RC(t);		\
51	andc	r0,RD(t),RB(t);		\
52	rotlwi	RT(t),RA(t),5;			\
53	rotlwi	RB(t),RB(t),30;			\
54	or	r6,r6,r0;			\
55	add	r0,RE(t),r15;			\
56	xor	r5,W((t)+4-3),W((t)+4-8);		\
57	add	RT(t),RT(t),r6;		\
58	xor	W((t)+4),W((t)+4-16),W((t)+4-14);	\
59	add	r0,r0,W(t);			\
60	xor	W((t)+4),W((t)+4),r5;			\
61	add	RT(t),RT(t),r0;		\
62	rotlwi	W((t)+4),W((t)+4),1
63
64#define STEPD1(t)				\
65	xor	r6,RB(t),RC(t);		\
66	rotlwi	RT(t),RA(t),5;			\
67	rotlwi	RB(t),RB(t),30;			\
68	xor	r6,r6,RD(t);			\
69	add	r0,RE(t),r15;			\
70	add	RT(t),RT(t),r6;		\
71	add	r0,r0,W(t);			\
72	add	RT(t),RT(t),r0
73
74#define STEPD1_UPDATE(t)				\
75	xor	r6,RB(t),RC(t);		\
76	rotlwi	RT(t),RA(t),5;			\
77	rotlwi	RB(t),RB(t),30;			\
78	xor	r6,r6,RD(t);			\
79	add	r0,RE(t),r15;			\
80	xor	r5,W((t)+4-3),W((t)+4-8);		\
81	add	RT(t),RT(t),r6;		\
82	xor	W((t)+4),W((t)+4-16),W((t)+4-14);	\
83	add	r0,r0,W(t);			\
84	xor	W((t)+4),W((t)+4),r5;			\
85	add	RT(t),RT(t),r0;		\
86	rotlwi	W((t)+4),W((t)+4),1
87
88#define STEPD2_UPDATE(t)			\
89	and	r6,RB(t),RC(t);		\
90	and	r0,RB(t),RD(t);		\
91	rotlwi	RT(t),RA(t),5;			\
92	or	r6,r6,r0;			\
93	rotlwi	RB(t),RB(t),30;			\
94	and	r0,RC(t),RD(t);		\
95	xor	r5,W((t)+4-3),W((t)+4-8);	\
96	or	r6,r6,r0;			\
97	xor	W((t)+4),W((t)+4-16),W((t)+4-14);	\
98	add	r0,RE(t),r15;			\
99	add	RT(t),RT(t),r6;		\
100	add	r0,r0,W(t);			\
101	xor	W((t)+4),W((t)+4),r5;		\
102	add	RT(t),RT(t),r0;		\
103	rotlwi	W((t)+4),W((t)+4),1
104
105#define STEP0LD4(t)				\
106	STEPD0_LOAD(t);				\
107	STEPD0_LOAD((t)+1);			\
108	STEPD0_LOAD((t)+2);			\
109	STEPD0_LOAD((t)+3)
110
111#define STEPUP4(t, fn)				\
112	STEP##fn##_UPDATE(t);			\
113	STEP##fn##_UPDATE((t)+1);		\
114	STEP##fn##_UPDATE((t)+2);		\
115	STEP##fn##_UPDATE((t)+3)
116
117#define STEPUP20(t, fn)				\
118	STEPUP4(t, fn);				\
119	STEPUP4((t)+4, fn);			\
120	STEPUP4((t)+8, fn);			\
121	STEPUP4((t)+12, fn);			\
122	STEPUP4((t)+16, fn)
123
124_GLOBAL(powerpc_sha_transform)
125	PPC_STLU r1,-INT_FRAME_SIZE(r1)
126	SAVE_8GPRS(14, r1)
127	SAVE_10GPRS(22, r1)
128
129	/* Load up A - E */
130	lwz	RA(0),0(r3)	/* A */
131	lwz	RB(0),4(r3)	/* B */
132	lwz	RC(0),8(r3)	/* C */
133	lwz	RD(0),12(r3)	/* D */
134	lwz	RE(0),16(r3)	/* E */
135
136	LOADW(0)
137	LOADW(1)
138	LOADW(2)
139	LOADW(3)
140
141	lis	r15,0x5a82	/* K0-19 */
142	ori	r15,r15,0x7999
143	STEP0LD4(0)
144	STEP0LD4(4)
145	STEP0LD4(8)
146	STEPUP4(12, D0)
147	STEPUP4(16, D0)
148
149	lis	r15,0x6ed9	/* K20-39 */
150	ori	r15,r15,0xeba1
151	STEPUP20(20, D1)
152
153	lis	r15,0x8f1b	/* K40-59 */
154	ori	r15,r15,0xbcdc
155	STEPUP20(40, D2)
156
157	lis	r15,0xca62	/* K60-79 */
158	ori	r15,r15,0xc1d6
159	STEPUP4(60, D1)
160	STEPUP4(64, D1)
161	STEPUP4(68, D1)
162	STEPUP4(72, D1)
163	lwz	r20,16(r3)
164	STEPD1(76)
165	lwz	r19,12(r3)
166	STEPD1(77)
167	lwz	r18,8(r3)
168	STEPD1(78)
169	lwz	r17,4(r3)
170	STEPD1(79)
171
172	lwz	r16,0(r3)
173	add	r20,RE(80),r20
174	add	RD(0),RD(80),r19
175	add	RC(0),RC(80),r18
176	add	RB(0),RB(80),r17
177	add	RA(0),RA(80),r16
178	mr	RE(0),r20
179	stw	RA(0),0(r3)
180	stw	RB(0),4(r3)
181	stw	RC(0),8(r3)
182	stw	RD(0),12(r3)
183	stw	RE(0),16(r3)
184
185	REST_8GPRS(14, r1)
186	REST_10GPRS(22, r1)
187	addi	r1,r1,INT_FRAME_SIZE
188	blr
189