Commit | Line | Data |
---|---|---|
864cbeed AB |
1 | /* |
2 | * sha1-ce-core.S - SHA-1 secure hash using ARMv8 Crypto Extensions | |
3 | * | |
4 | * Copyright (C) 2015 Linaro Ltd. | |
5 | * Author: Ard Biesheuvel <ard.biesheuvel@linaro.org> | |
6 | * | |
7 | * This program is free software; you can redistribute it and/or modify | |
8 | * it under the terms of the GNU General Public License version 2 as | |
9 | * published by the Free Software Foundation. | |
10 | */ | |
11 | ||
12 | #include <linux/linkage.h> | |
13 | #include <asm/assembler.h> | |
14 | ||
15 | .text | |
16 | .fpu crypto-neon-fp-armv8 | |
17 | ||
18 | k0 .req q0 | |
19 | k1 .req q1 | |
20 | k2 .req q2 | |
21 | k3 .req q3 | |
22 | ||
23 | ta0 .req q4 | |
24 | ta1 .req q5 | |
25 | tb0 .req q5 | |
26 | tb1 .req q4 | |
27 | ||
28 | dga .req q6 | |
29 | dgb .req q7 | |
30 | dgbs .req s28 | |
31 | ||
32 | dg0 .req q12 | |
33 | dg1a0 .req q13 | |
34 | dg1a1 .req q14 | |
35 | dg1b0 .req q14 | |
36 | dg1b1 .req q13 | |
37 | ||
38 | .macro add_only, op, ev, rc, s0, dg1 | |
39 | .ifnb \s0 | |
40 | vadd.u32 tb\ev, q\s0, \rc | |
41 | .endif | |
42 | sha1h.32 dg1b\ev, dg0 | |
43 | .ifb \dg1 | |
44 | sha1\op\().32 dg0, dg1a\ev, ta\ev | |
45 | .else | |
46 | sha1\op\().32 dg0, \dg1, ta\ev | |
47 | .endif | |
48 | .endm | |
49 | ||
50 | .macro add_update, op, ev, rc, s0, s1, s2, s3, dg1 | |
51 | sha1su0.32 q\s0, q\s1, q\s2 | |
52 | add_only \op, \ev, \rc, \s1, \dg1 | |
53 | sha1su1.32 q\s0, q\s3 | |
54 | .endm | |
55 | ||
56 | .align 6 | |
57 | .Lsha1_rcon: | |
58 | .word 0x5a827999, 0x5a827999, 0x5a827999, 0x5a827999 | |
59 | .word 0x6ed9eba1, 0x6ed9eba1, 0x6ed9eba1, 0x6ed9eba1 | |
60 | .word 0x8f1bbcdc, 0x8f1bbcdc, 0x8f1bbcdc, 0x8f1bbcdc | |
61 | .word 0xca62c1d6, 0xca62c1d6, 0xca62c1d6, 0xca62c1d6 | |
62 | ||
63 | /* | |
dde00981 AB |
64 | * void sha1_ce_transform(struct sha1_state *sst, u8 const *src, |
65 | * int blocks); | |
864cbeed AB |
66 | */ |
67 | ENTRY(sha1_ce_transform) | |
68 | /* load round constants */ | |
69 | adr ip, .Lsha1_rcon | |
70 | vld1.32 {k0-k1}, [ip, :128]! | |
71 | vld1.32 {k2-k3}, [ip, :128] | |
72 | ||
73 | /* load state */ | |
dde00981 AB |
74 | vld1.32 {dga}, [r0] |
75 | vldr dgbs, [r0, #16] | |
864cbeed AB |
76 | |
77 | /* load input */ | |
78 | 0: vld1.32 {q8-q9}, [r1]! | |
79 | vld1.32 {q10-q11}, [r1]! | |
dde00981 | 80 | subs r2, r2, #1 |
864cbeed | 81 | |
864cbeed AB |
82 | #ifndef CONFIG_CPU_BIG_ENDIAN |
83 | vrev32.8 q8, q8 | |
84 | vrev32.8 q9, q9 | |
85 | vrev32.8 q10, q10 | |
86 | vrev32.8 q11, q11 | |
87 | #endif | |
88 | ||
89 | vadd.u32 ta0, q8, k0 | |
90 | vmov dg0, dga | |
91 | ||
92 | add_update c, 0, k0, 8, 9, 10, 11, dgb | |
93 | add_update c, 1, k0, 9, 10, 11, 8 | |
94 | add_update c, 0, k0, 10, 11, 8, 9 | |
95 | add_update c, 1, k0, 11, 8, 9, 10 | |
96 | add_update c, 0, k1, 8, 9, 10, 11 | |
97 | ||
98 | add_update p, 1, k1, 9, 10, 11, 8 | |
99 | add_update p, 0, k1, 10, 11, 8, 9 | |
100 | add_update p, 1, k1, 11, 8, 9, 10 | |
101 | add_update p, 0, k1, 8, 9, 10, 11 | |
102 | add_update p, 1, k2, 9, 10, 11, 8 | |
103 | ||
104 | add_update m, 0, k2, 10, 11, 8, 9 | |
105 | add_update m, 1, k2, 11, 8, 9, 10 | |
106 | add_update m, 0, k2, 8, 9, 10, 11 | |
107 | add_update m, 1, k2, 9, 10, 11, 8 | |
108 | add_update m, 0, k3, 10, 11, 8, 9 | |
109 | ||
110 | add_update p, 1, k3, 11, 8, 9, 10 | |
111 | add_only p, 0, k3, 9 | |
112 | add_only p, 1, k3, 10 | |
113 | add_only p, 0, k3, 11 | |
114 | add_only p, 1 | |
115 | ||
116 | /* update state */ | |
117 | vadd.u32 dga, dga, dg0 | |
118 | vadd.u32 dgb, dgb, dg1a0 | |
119 | bne 0b | |
120 | ||
121 | /* store new state */ | |
dde00981 AB |
122 | vst1.32 {dga}, [r0] |
123 | vstr dgbs, [r0, #16] | |
864cbeed AB |
124 | bx lr |
125 | ENDPROC(sha1_ce_transform) |