2 * sha1-ce-core.S - SHA-1 secure hash using ARMv8 Crypto Extensions
4 * Copyright (C) 2014 Linaro Ltd <ard.biesheuvel@linaro.org>
6 * This program is free software; you can redistribute it and/or modify
7 * it under the terms of the GNU General Public License version 2 as
8 * published by the Free Software Foundation.
11 #include <linux/linkage.h>
12 #include <asm/assembler.h>
37 .macro add_only, op, ev, rc, s0, dg1
39 add t1.4s, v\s0\().4s, \rc\().4s
42 sha1\op dg0q, \dg1, t0.4s
44 sha1\op dg0q, dg1s, t0.4s
48 add t0.4s, v\s0\().4s, \rc\().4s
51 sha1\op dg0q, dg2s, t1.4s
55 .macro add_update, op, ev, rc, s0, s1, s2, s3, dg1
56 sha1su0 v\s0\().4s, v\s1\().4s, v\s2\().4s
57 add_only \op, \ev, \rc, \s1, \dg1
58 sha1su1 v\s0\().4s, v\s3\().4s
61 .macro loadrc, k, val, tmp
62 movz \tmp, :abs_g0_nc:\val
63 movk \tmp, :abs_g1:\val
68 * void sha1_ce_transform(struct sha1_ce_state *sst, u8 const *src,
71 ENTRY(sha1_ce_transform)
78 /* load round constants */
79 0: loadrc k0.4s, 0x5a827999, w6
80 loadrc k1.4s, 0x6ed9eba1, w6
81 loadrc k2.4s, 0x8f1bbcdc, w6
82 loadrc k3.4s, 0xca62c1d6, w6
88 /* load sha1_ce_state::finalize */
89 ldr_l w4, sha1_ce_offsetof_finalize, x4
93 1: ld1 {v8.4s-v11.4s}, [x20], #64
96 CPU_LE( rev32 v8.16b, v8.16b )
97 CPU_LE( rev32 v9.16b, v9.16b )
98 CPU_LE( rev32 v10.16b, v10.16b )
99 CPU_LE( rev32 v11.16b, v11.16b )
101 2: add t0.4s, v8.4s, k0.4s
102 mov dg0v.16b, dgav.16b
104 add_update c, ev, k0, 8, 9, 10, 11, dgb
105 add_update c, od, k0, 9, 10, 11, 8
106 add_update c, ev, k0, 10, 11, 8, 9
107 add_update c, od, k0, 11, 8, 9, 10
108 add_update c, ev, k1, 8, 9, 10, 11
110 add_update p, od, k1, 9, 10, 11, 8
111 add_update p, ev, k1, 10, 11, 8, 9
112 add_update p, od, k1, 11, 8, 9, 10
113 add_update p, ev, k1, 8, 9, 10, 11
114 add_update p, od, k2, 9, 10, 11, 8
116 add_update m, ev, k2, 10, 11, 8, 9
117 add_update m, od, k2, 11, 8, 9, 10
118 add_update m, ev, k2, 8, 9, 10, 11
119 add_update m, od, k2, 9, 10, 11, 8
120 add_update m, ev, k3, 10, 11, 8, 9
122 add_update p, od, k3, 11, 8, 9, 10
123 add_only p, ev, k3, 9
124 add_only p, od, k3, 10
125 add_only p, ev, k3, 11
129 add dgbv.2s, dgbv.2s, dg1v.2s
130 add dgav.4s, dgav.4s, dg0v.4s
134 if_will_cond_yield_neon
144 * Final block: add padding and total bit count.
145 * Skip if the input size was not a round multiple of the block size,
146 * the padding is handled by the C code in that case.
149 ldr_l w4, sha1_ce_offsetof_count, x4
154 ror x7, x4, #29 // ror(lsl(x4, 3), 32)
161 /* store new state */
162 4: st1 {dgav.4s}, [x19]
166 ENDPROC(sha1_ce_transform)