1/* SPDX-License-Identifier: GPL-2.0 */
2/*
3 * SHA-1 implementation for PowerPC.
4 *
5 * Copyright (C) 2005 Paul Mackerras <paulus@samba.org>
6 */
7
8#include <asm/ppc_asm.h>
9#include <asm/asm-offsets.h>
10#include <asm/asm-compat.h>
11
12#ifdef __BIG_ENDIAN__
13#define LWZ(rt, d, ra) \
14 lwz rt,d(ra)
15#else
16#define LWZ(rt, d, ra) \
17 li rt,d; \
18 lwbrx rt,rt,ra
19#endif
20
21/*
22 * We roll the registers for T, A, B, C, D, E around on each
23 * iteration; T on iteration t is A on iteration t+1, and so on.
24 * We use registers 7 - 12 for this.
25 */
26#define RT(t) ((((t)+5)%6)+7)
27#define RA(t) ((((t)+4)%6)+7)
28#define RB(t) ((((t)+3)%6)+7)
29#define RC(t) ((((t)+2)%6)+7)
30#define RD(t) ((((t)+1)%6)+7)
31#define RE(t) ((((t)+0)%6)+7)
32
33/* We use registers 16 - 31 for the W values */
34#define W(t) (((t)%16)+16)
35
36#define LOADW(t) \
37 LWZ(W(t),(t)*4,r4)
38
39#define STEPD0_LOAD(t) \
40 andc r0,RD(t),RB(t); \
41 and r6,RB(t),RC(t); \
42 rotlwi RT(t),RA(t),5; \
43 or r6,r6,r0; \
44 add r0,RE(t),r15; \
45 add RT(t),RT(t),r6; \
46 add r14,r0,W(t); \
47 LWZ(W((t)+4),((t)+4)*4,r4); \
48 rotlwi RB(t),RB(t),30; \
49 add RT(t),RT(t),r14
50
51#define STEPD0_UPDATE(t) \
52 and r6,RB(t),RC(t); \
53 andc r0,RD(t),RB(t); \
54 rotlwi RT(t),RA(t),5; \
55 rotlwi RB(t),RB(t),30; \
56 or r6,r6,r0; \
57 add r0,RE(t),r15; \
58 xor r5,W((t)+4-3),W((t)+4-8); \
59 add RT(t),RT(t),r6; \
60 xor W((t)+4),W((t)+4-16),W((t)+4-14); \
61 add r0,r0,W(t); \
62 xor W((t)+4),W((t)+4),r5; \
63 add RT(t),RT(t),r0; \
64 rotlwi W((t)+4),W((t)+4),1
65
66#define STEPD1(t) \
67 xor r6,RB(t),RC(t); \
68 rotlwi RT(t),RA(t),5; \
69 rotlwi RB(t),RB(t),30; \
70 xor r6,r6,RD(t); \
71 add r0,RE(t),r15; \
72 add RT(t),RT(t),r6; \
73 add r0,r0,W(t); \
74 add RT(t),RT(t),r0
75
76#define STEPD1_UPDATE(t) \
77 xor r6,RB(t),RC(t); \
78 rotlwi RT(t),RA(t),5; \
79 rotlwi RB(t),RB(t),30; \
80 xor r6,r6,RD(t); \
81 add r0,RE(t),r15; \
82 xor r5,W((t)+4-3),W((t)+4-8); \
83 add RT(t),RT(t),r6; \
84 xor W((t)+4),W((t)+4-16),W((t)+4-14); \
85 add r0,r0,W(t); \
86 xor W((t)+4),W((t)+4),r5; \
87 add RT(t),RT(t),r0; \
88 rotlwi W((t)+4),W((t)+4),1
89
90#define STEPD2_UPDATE(t) \
91 and r6,RB(t),RC(t); \
92 and r0,RB(t),RD(t); \
93 rotlwi RT(t),RA(t),5; \
94 or r6,r6,r0; \
95 rotlwi RB(t),RB(t),30; \
96 and r0,RC(t),RD(t); \
97 xor r5,W((t)+4-3),W((t)+4-8); \
98 or r6,r6,r0; \
99 xor W((t)+4),W((t)+4-16),W((t)+4-14); \
100 add r0,RE(t),r15; \
101 add RT(t),RT(t),r6; \
102 add r0,r0,W(t); \
103 xor W((t)+4),W((t)+4),r5; \
104 add RT(t),RT(t),r0; \
105 rotlwi W((t)+4),W((t)+4),1
106
107#define STEP0LD4(t) \
108 STEPD0_LOAD(t); \
109 STEPD0_LOAD((t)+1); \
110 STEPD0_LOAD((t)+2); \
111 STEPD0_LOAD((t)+3)
112
113#define STEPUP4(t, fn) \
114 STEP##fn##_UPDATE(t); \
115 STEP##fn##_UPDATE((t)+1); \
116 STEP##fn##_UPDATE((t)+2); \
117 STEP##fn##_UPDATE((t)+3)
118
119#define STEPUP20(t, fn) \
120 STEPUP4(t, fn); \
121 STEPUP4((t)+4, fn); \
122 STEPUP4((t)+8, fn); \
123 STEPUP4((t)+12, fn); \
124 STEPUP4((t)+16, fn)
125
126_GLOBAL(powerpc_sha_transform)
127 PPC_STLU r1,-INT_FRAME_SIZE(r1)
128 SAVE_GPRS(14, 31, r1)
129
130 /* Load up A - E */
131 lwz RA(0),0(r3) /* A */
132 lwz RB(0),4(r3) /* B */
133 lwz RC(0),8(r3) /* C */
134 lwz RD(0),12(r3) /* D */
135 lwz RE(0),16(r3) /* E */
136
137 LOADW(0)
138 LOADW(1)
139 LOADW(2)
140 LOADW(3)
141
142 lis r15,0x5a82 /* K0-19 */
143 ori r15,r15,0x7999
144 STEP0LD4(0)
145 STEP0LD4(4)
146 STEP0LD4(8)
147 STEPUP4(12, D0)
148 STEPUP4(16, D0)
149
150 lis r15,0x6ed9 /* K20-39 */
151 ori r15,r15,0xeba1
152 STEPUP20(20, D1)
153
154 lis r15,0x8f1b /* K40-59 */
155 ori r15,r15,0xbcdc
156 STEPUP20(40, D2)
157
158 lis r15,0xca62 /* K60-79 */
159 ori r15,r15,0xc1d6
160 STEPUP4(60, D1)
161 STEPUP4(64, D1)
162 STEPUP4(68, D1)
163 STEPUP4(72, D1)
164 lwz r20,16(r3)
165 STEPD1(76)
166 lwz r19,12(r3)
167 STEPD1(77)
168 lwz r18,8(r3)
169 STEPD1(78)
170 lwz r17,4(r3)
171 STEPD1(79)
172
173 lwz r16,0(r3)
174 add r20,RE(80),r20
175 add RD(0),RD(80),r19
176 add RC(0),RC(80),r18
177 add RB(0),RB(80),r17
178 add RA(0),RA(80),r16
179 mr RE(0),r20
180 stw RA(0),0(r3)
181 stw RB(0),4(r3)
182 stw RC(0),8(r3)
183 stw RD(0),12(r3)
184 stw RE(0),16(r3)
185
186 REST_GPRS(14, 31, r1)
187 addi r1,r1,INT_FRAME_SIZE
188 blr
189

source code of linux/arch/powerpc/crypto/sha1-powerpc-asm.S