diff options
| -rw-r--r-- | src/lib/libcrypto/sha/sha1_aarch64_ce.S | 12 | ||||
| -rw-r--r-- | src/lib/libcrypto/sha/sha1_amd64_generic.S | 150 | ||||
| -rw-r--r-- | src/lib/libcrypto/sha/sha1_amd64_shani.S | 20 | ||||
| -rw-r--r-- | src/lib/libcrypto/sha/sha256_aarch64_ce.S | 8 | ||||
| -rw-r--r-- | src/lib/libcrypto/sha/sha256_amd64_generic.S | 12 | ||||
| -rw-r--r-- | src/lib/libcrypto/sha/sha256_amd64_shani.S | 12 | ||||
| -rw-r--r-- | src/lib/libcrypto/sha/sha512_amd64_generic.S | 12 |
7 files changed, 113 insertions, 113 deletions
diff --git a/src/lib/libcrypto/sha/sha1_aarch64_ce.S b/src/lib/libcrypto/sha/sha1_aarch64_ce.S index ce7eb81115..853d467641 100644 --- a/src/lib/libcrypto/sha/sha1_aarch64_ce.S +++ b/src/lib/libcrypto/sha/sha1_aarch64_ce.S | |||
| @@ -1,4 +1,4 @@ | |||
| 1 | /* $OpenBSD: sha1_aarch64_ce.S,v 1.3 2026/01/17 06:31:45 jsing Exp $ */ | 1 | /* $OpenBSD: sha1_aarch64_ce.S,v 1.4 2026/01/24 14:20:52 jsing Exp $ */ |
| 2 | /* | 2 | /* |
| 3 | * Copyright (c) 2023,2025 Joel Sing <jsing@openbsd.org> | 3 | * Copyright (c) 2023,2025 Joel Sing <jsing@openbsd.org> |
| 4 | * | 4 | * |
| @@ -65,7 +65,7 @@ | |||
| 65 | */ | 65 | */ |
| 66 | #define sha1_message_schedule_update(m0, m1, m2, m3) \ | 66 | #define sha1_message_schedule_update(m0, m1, m2, m3) \ |
| 67 | sha1su0 m0.4s, m1.4s, m2.4s; \ | 67 | sha1su0 m0.4s, m1.4s, m2.4s; \ |
| 68 | sha1su1 m0.4s, m3.4s; | 68 | sha1su1 m0.4s, m3.4s |
| 69 | 69 | ||
| 70 | /* | 70 | /* |
| 71 | * Compute four SHA-1 rounds by adding W0:W1:W2:W3 + K0:K1:K2:K3, then | 71 | * Compute four SHA-1 rounds by adding W0:W1:W2:W3 + K0:K1:K2:K3, then |
| @@ -77,25 +77,25 @@ | |||
| 77 | add tmp0.4s, w.4s, k.4s; /* Tt = Wt + Kt */ \ | 77 | add tmp0.4s, w.4s, k.4s; /* Tt = Wt + Kt */ \ |
| 78 | mov tmp1, h0.s[0]; \ | 78 | mov tmp1, h0.s[0]; \ |
| 79 | sha1c h0, h1, tmp0.4s; \ | 79 | sha1c h0, h1, tmp0.4s; \ |
| 80 | sha1h h1, tmp1; | 80 | sha1h h1, tmp1 |
| 81 | 81 | ||
| 82 | #define sha1_round2(h0, h1, w, k) \ | 82 | #define sha1_round2(h0, h1, w, k) \ |
| 83 | add tmp0.4s, w.4s, k.4s; /* Tt = Wt + Kt */ \ | 83 | add tmp0.4s, w.4s, k.4s; /* Tt = Wt + Kt */ \ |
| 84 | mov tmp1, h0.s[0]; \ | 84 | mov tmp1, h0.s[0]; \ |
| 85 | sha1p h0, h1, tmp0.4s; \ | 85 | sha1p h0, h1, tmp0.4s; \ |
| 86 | sha1h h1, tmp1; | 86 | sha1h h1, tmp1 |
| 87 | 87 | ||
| 88 | #define sha1_round3(h0, h1, w, k) \ | 88 | #define sha1_round3(h0, h1, w, k) \ |
| 89 | add tmp0.4s, w.4s, k.4s; /* Tt = Wt + Kt */ \ | 89 | add tmp0.4s, w.4s, k.4s; /* Tt = Wt + Kt */ \ |
| 90 | mov tmp1, h0.s[0]; \ | 90 | mov tmp1, h0.s[0]; \ |
| 91 | sha1m h0, h1, tmp0.4s; \ | 91 | sha1m h0, h1, tmp0.4s; \ |
| 92 | sha1h h1, tmp1; | 92 | sha1h h1, tmp1 |
| 93 | 93 | ||
| 94 | #define sha1_round4(h0, h1, w, k) \ | 94 | #define sha1_round4(h0, h1, w, k) \ |
| 95 | add tmp0.4s, w.4s, k.4s; /* Tt = Wt + Kt */ \ | 95 | add tmp0.4s, w.4s, k.4s; /* Tt = Wt + Kt */ \ |
| 96 | mov tmp1, h0.s[0]; \ | 96 | mov tmp1, h0.s[0]; \ |
| 97 | sha1p h0, h1, tmp0.4s; \ | 97 | sha1p h0, h1, tmp0.4s; \ |
| 98 | sha1h h1, tmp1; | 98 | sha1h h1, tmp1 |
| 99 | 99 | ||
| 100 | .arch armv8-a+sha2 | 100 | .arch armv8-a+sha2 |
| 101 | 101 | ||
diff --git a/src/lib/libcrypto/sha/sha1_amd64_generic.S b/src/lib/libcrypto/sha/sha1_amd64_generic.S index 685d71edf8..a2088229ec 100644 --- a/src/lib/libcrypto/sha/sha1_amd64_generic.S +++ b/src/lib/libcrypto/sha/sha1_amd64_generic.S | |||
| @@ -1,4 +1,4 @@ | |||
| 1 | /* $OpenBSD: sha1_amd64_generic.S,v 1.3 2026/01/17 06:31:45 jsing Exp $ */ | 1 | /* $OpenBSD: sha1_amd64_generic.S,v 1.4 2026/01/24 14:20:52 jsing Exp $ */ |
| 2 | /* | 2 | /* |
| 3 | * Copyright (c) 2024 Joel Sing <jsing@openbsd.org> | 3 | * Copyright (c) 2024 Joel Sing <jsing@openbsd.org> |
| 4 | * | 4 | * |
| @@ -46,7 +46,7 @@ | |||
| 46 | #define sha1_message_schedule_load(idx, m, w, wt) \ | 46 | #define sha1_message_schedule_load(idx, m, w, wt) \ |
| 47 | movl ((idx&0xf)*4)(m), wt; \ | 47 | movl ((idx&0xf)*4)(m), wt; \ |
| 48 | bswapl wt; \ | 48 | bswapl wt; \ |
| 49 | movl wt, ((idx&0xf)*4)(w); | 49 | movl wt, ((idx&0xf)*4)(w) |
| 50 | 50 | ||
| 51 | /* | 51 | /* |
| 52 | * Update message schedule and return current value in wt: | 52 | * Update message schedule and return current value in wt: |
| @@ -60,7 +60,7 @@ | |||
| 60 | xorl (((idx)&0xf)*4)(w), wt; /* W0 */ \ | 60 | xorl (((idx)&0xf)*4)(w), wt; /* W0 */ \ |
| 61 | roll $1, wt; \ | 61 | roll $1, wt; \ |
| 62 | \ | 62 | \ |
| 63 | movl wt, ((idx&0xf)*4)(w); | 63 | movl wt, ((idx&0xf)*4)(w) |
| 64 | 64 | ||
| 65 | /* | 65 | /* |
| 66 | * Compute a SHA-1 round without logic function: | 66 | * Compute a SHA-1 round without logic function: |
| @@ -97,7 +97,7 @@ | |||
| 97 | xorl d, tmp2; /* Ch */ \ | 97 | xorl d, tmp2; /* Ch */ \ |
| 98 | addl tmp2, e; /* Ch */ \ | 98 | addl tmp2, e; /* Ch */ \ |
| 99 | \ | 99 | \ |
| 100 | sha1_round(a, b, c, d, e, kt, wt); | 100 | sha1_round(a, b, c, d, e, kt, wt) |
| 101 | 101 | ||
| 102 | /* | 102 | /* |
| 103 | * Compute a SHA-1 round with Parity: | 103 | * Compute a SHA-1 round with Parity: |
| @@ -114,7 +114,7 @@ | |||
| 114 | xorl d, tmp2; /* Parity */ \ | 114 | xorl d, tmp2; /* Parity */ \ |
| 115 | addl tmp2, e; /* Parity */ \ | 115 | addl tmp2, e; /* Parity */ \ |
| 116 | \ | 116 | \ |
| 117 | sha1_round(a, b, c, d, e, kt, wt); | 117 | sha1_round(a, b, c, d, e, kt, wt) |
| 118 | 118 | ||
| 119 | /* | 119 | /* |
| 120 | * Compute a SHA-1 round with Maj: | 120 | * Compute a SHA-1 round with Maj: |
| @@ -134,26 +134,26 @@ | |||
| 134 | xorl tmp2, tmp3; /* Maj */ \ | 134 | xorl tmp2, tmp3; /* Maj */ \ |
| 135 | addl tmp3, e; /* Maj */ \ | 135 | addl tmp3, e; /* Maj */ \ |
| 136 | \ | 136 | \ |
| 137 | sha1_round(a, b, c, d, e, kt, wt); | 137 | sha1_round(a, b, c, d, e, kt, wt) |
| 138 | 138 | ||
| 139 | #define sha1_round1_load(idx, a, b, c, d, e) \ | 139 | #define sha1_round1_load(idx, a, b, c, d, e) \ |
| 140 | sha1_message_schedule_load(idx, in, %rsp, tmp0) \ | 140 | sha1_message_schedule_load(idx, in, %rsp, tmp0); \ |
| 141 | sha1_round_ch(a, b, c, d, e, 0x5a827999, tmp0) | 141 | sha1_round_ch(a, b, c, d, e, 0x5a827999, tmp0) |
| 142 | 142 | ||
| 143 | #define sha1_round1_update(idx, a, b, c, d, e) \ | 143 | #define sha1_round1_update(idx, a, b, c, d, e) \ |
| 144 | sha1_message_schedule_update(idx, %rsp, tmp0) \ | 144 | sha1_message_schedule_update(idx, %rsp, tmp0); \ |
| 145 | sha1_round_ch(a, b, c, d, e, 0x5a827999, tmp0) | 145 | sha1_round_ch(a, b, c, d, e, 0x5a827999, tmp0) |
| 146 | 146 | ||
| 147 | #define sha1_round2_update(idx, a, b, c, d, e) \ | 147 | #define sha1_round2_update(idx, a, b, c, d, e) \ |
| 148 | sha1_message_schedule_update(idx, %rsp, tmp0) \ | 148 | sha1_message_schedule_update(idx, %rsp, tmp0); \ |
| 149 | sha1_round_parity(a, b, c, d, e, 0x6ed9eba1, tmp0) | 149 | sha1_round_parity(a, b, c, d, e, 0x6ed9eba1, tmp0) |
| 150 | 150 | ||
| 151 | #define sha1_round3_update(idx, a, b, c, d, e) \ | 151 | #define sha1_round3_update(idx, a, b, c, d, e) \ |
| 152 | sha1_message_schedule_update(idx, %rsp, tmp0) \ | 152 | sha1_message_schedule_update(idx, %rsp, tmp0); \ |
| 153 | sha1_round_maj(a, b, c, d, e, 0x8f1bbcdc, tmp0) | 153 | sha1_round_maj(a, b, c, d, e, 0x8f1bbcdc, tmp0) |
| 154 | 154 | ||
| 155 | #define sha1_round4_update(idx, a, b, c, d, e) \ | 155 | #define sha1_round4_update(idx, a, b, c, d, e) \ |
| 156 | sha1_message_schedule_update(idx, %rsp, tmp0) \ | 156 | sha1_message_schedule_update(idx, %rsp, tmp0); \ |
| 157 | sha1_round_parity(a, b, c, d, e, 0xca62c1d6, tmp0) | 157 | sha1_round_parity(a, b, c, d, e, 0xca62c1d6, tmp0) |
| 158 | 158 | ||
| 159 | .section .text | 159 | .section .text |
| @@ -215,76 +215,76 @@ sha1_block_generic: | |||
| 215 | sha1_round1_load(15, hs0, hs1, hs2, hs3, hs4) | 215 | sha1_round1_load(15, hs0, hs1, hs2, hs3, hs4) |
| 216 | 216 | ||
| 217 | /* Round 16 through 31. */ | 217 | /* Round 16 through 31. */ |
| 218 | sha1_round1_update(16, hs4, hs0, hs1, hs2, hs3); | 218 | sha1_round1_update(16, hs4, hs0, hs1, hs2, hs3) |
| 219 | sha1_round1_update(17, hs3, hs4, hs0, hs1, hs2); | 219 | sha1_round1_update(17, hs3, hs4, hs0, hs1, hs2) |
| 220 | sha1_round1_update(18, hs2, hs3, hs4, hs0, hs1); | 220 | sha1_round1_update(18, hs2, hs3, hs4, hs0, hs1) |
| 221 | sha1_round1_update(19, hs1, hs2, hs3, hs4, hs0); | 221 | sha1_round1_update(19, hs1, hs2, hs3, hs4, hs0) |
| 222 | sha1_round2_update(20, hs0, hs1, hs2, hs3, hs4); | 222 | sha1_round2_update(20, hs0, hs1, hs2, hs3, hs4) |
| 223 | sha1_round2_update(21, hs4, hs0, hs1, hs2, hs3); | 223 | sha1_round2_update(21, hs4, hs0, hs1, hs2, hs3) |
| 224 | sha1_round2_update(22, hs3, hs4, hs0, hs1, hs2); | 224 | sha1_round2_update(22, hs3, hs4, hs0, hs1, hs2) |
| 225 | sha1_round2_update(23, hs2, hs3, hs4, hs0, hs1); | 225 | sha1_round2_update(23, hs2, hs3, hs4, hs0, hs1) |
| 226 | sha1_round2_update(24, hs1, hs2, hs3, hs4, hs0); | 226 | sha1_round2_update(24, hs1, hs2, hs3, hs4, hs0) |
| 227 | sha1_round2_update(25, hs0, hs1, hs2, hs3, hs4); | 227 | sha1_round2_update(25, hs0, hs1, hs2, hs3, hs4) |
| 228 | sha1_round2_update(26, hs4, hs0, hs1, hs2, hs3); | 228 | sha1_round2_update(26, hs4, hs0, hs1, hs2, hs3) |
| 229 | sha1_round2_update(27, hs3, hs4, hs0, hs1, hs2); | 229 | sha1_round2_update(27, hs3, hs4, hs0, hs1, hs2) |
| 230 | sha1_round2_update(28, hs2, hs3, hs4, hs0, hs1); | 230 | sha1_round2_update(28, hs2, hs3, hs4, hs0, hs1) |
| 231 | sha1_round2_update(29, hs1, hs2, hs3, hs4, hs0); | 231 | sha1_round2_update(29, hs1, hs2, hs3, hs4, hs0) |
| 232 | sha1_round2_update(30, hs0, hs1, hs2, hs3, hs4); | 232 | sha1_round2_update(30, hs0, hs1, hs2, hs3, hs4) |
| 233 | sha1_round2_update(31, hs4, hs0, hs1, hs2, hs3); | 233 | sha1_round2_update(31, hs4, hs0, hs1, hs2, hs3) |
| 234 | 234 | ||
| 235 | /* Round 32 through 47. */ | 235 | /* Round 32 through 47. */ |
| 236 | sha1_round2_update(32, hs3, hs4, hs0, hs1, hs2); | 236 | sha1_round2_update(32, hs3, hs4, hs0, hs1, hs2) |
| 237 | sha1_round2_update(33, hs2, hs3, hs4, hs0, hs1); | 237 | sha1_round2_update(33, hs2, hs3, hs4, hs0, hs1) |
| 238 | sha1_round2_update(34, hs1, hs2, hs3, hs4, hs0); | 238 | sha1_round2_update(34, hs1, hs2, hs3, hs4, hs0) |
| 239 | sha1_round2_update(35, hs0, hs1, hs2, hs3, hs4); | 239 | sha1_round2_update(35, hs0, hs1, hs2, hs3, hs4) |
| 240 | sha1_round2_update(36, hs4, hs0, hs1, hs2, hs3); | 240 | sha1_round2_update(36, hs4, hs0, hs1, hs2, hs3) |
| 241 | sha1_round2_update(37, hs3, hs4, hs0, hs1, hs2); | 241 | sha1_round2_update(37, hs3, hs4, hs0, hs1, hs2) |
| 242 | sha1_round2_update(38, hs2, hs3, hs4, hs0, hs1); | 242 | sha1_round2_update(38, hs2, hs3, hs4, hs0, hs1) |
| 243 | sha1_round2_update(39, hs1, hs2, hs3, hs4, hs0); | 243 | sha1_round2_update(39, hs1, hs2, hs3, hs4, hs0) |
| 244 | sha1_round3_update(40, hs0, hs1, hs2, hs3, hs4); | 244 | sha1_round3_update(40, hs0, hs1, hs2, hs3, hs4) |
| 245 | sha1_round3_update(41, hs4, hs0, hs1, hs2, hs3); | 245 | sha1_round3_update(41, hs4, hs0, hs1, hs2, hs3) |
| 246 | sha1_round3_update(42, hs3, hs4, hs0, hs1, hs2); | 246 | sha1_round3_update(42, hs3, hs4, hs0, hs1, hs2) |
| 247 | sha1_round3_update(43, hs2, hs3, hs4, hs0, hs1); | 247 | sha1_round3_update(43, hs2, hs3, hs4, hs0, hs1) |
| 248 | sha1_round3_update(44, hs1, hs2, hs3, hs4, hs0); | 248 | sha1_round3_update(44, hs1, hs2, hs3, hs4, hs0) |
| 249 | sha1_round3_update(45, hs0, hs1, hs2, hs3, hs4); | 249 | sha1_round3_update(45, hs0, hs1, hs2, hs3, hs4) |
| 250 | sha1_round3_update(46, hs4, hs0, hs1, hs2, hs3); | 250 | sha1_round3_update(46, hs4, hs0, hs1, hs2, hs3) |
| 251 | sha1_round3_update(47, hs3, hs4, hs0, hs1, hs2); | 251 | sha1_round3_update(47, hs3, hs4, hs0, hs1, hs2) |
| 252 | 252 | ||
| 253 | /* Round 48 through 63. */ | 253 | /* Round 48 through 63. */ |
| 254 | sha1_round3_update(48, hs2, hs3, hs4, hs0, hs1); | 254 | sha1_round3_update(48, hs2, hs3, hs4, hs0, hs1) |
| 255 | sha1_round3_update(49, hs1, hs2, hs3, hs4, hs0); | 255 | sha1_round3_update(49, hs1, hs2, hs3, hs4, hs0) |
| 256 | sha1_round3_update(50, hs0, hs1, hs2, hs3, hs4); | 256 | sha1_round3_update(50, hs0, hs1, hs2, hs3, hs4) |
| 257 | sha1_round3_update(51, hs4, hs0, hs1, hs2, hs3); | 257 | sha1_round3_update(51, hs4, hs0, hs1, hs2, hs3) |
| 258 | sha1_round3_update(52, hs3, hs4, hs0, hs1, hs2); | 258 | sha1_round3_update(52, hs3, hs4, hs0, hs1, hs2) |
| 259 | sha1_round3_update(53, hs2, hs3, hs4, hs0, hs1); | 259 | sha1_round3_update(53, hs2, hs3, hs4, hs0, hs1) |
| 260 | sha1_round3_update(54, hs1, hs2, hs3, hs4, hs0); | 260 | sha1_round3_update(54, hs1, hs2, hs3, hs4, hs0) |
| 261 | sha1_round3_update(55, hs0, hs1, hs2, hs3, hs4); | 261 | sha1_round3_update(55, hs0, hs1, hs2, hs3, hs4) |
| 262 | sha1_round3_update(56, hs4, hs0, hs1, hs2, hs3); | 262 | sha1_round3_update(56, hs4, hs0, hs1, hs2, hs3) |
| 263 | sha1_round3_update(57, hs3, hs4, hs0, hs1, hs2); | 263 | sha1_round3_update(57, hs3, hs4, hs0, hs1, hs2) |
| 264 | sha1_round3_update(58, hs2, hs3, hs4, hs0, hs1); | 264 | sha1_round3_update(58, hs2, hs3, hs4, hs0, hs1) |
| 265 | sha1_round3_update(59, hs1, hs2, hs3, hs4, hs0); | 265 | sha1_round3_update(59, hs1, hs2, hs3, hs4, hs0) |
| 266 | sha1_round4_update(60, hs0, hs1, hs2, hs3, hs4); | 266 | sha1_round4_update(60, hs0, hs1, hs2, hs3, hs4) |
| 267 | sha1_round4_update(61, hs4, hs0, hs1, hs2, hs3); | 267 | sha1_round4_update(61, hs4, hs0, hs1, hs2, hs3) |
| 268 | sha1_round4_update(62, hs3, hs4, hs0, hs1, hs2); | 268 | sha1_round4_update(62, hs3, hs4, hs0, hs1, hs2) |
| 269 | sha1_round4_update(63, hs2, hs3, hs4, hs0, hs1); | 269 | sha1_round4_update(63, hs2, hs3, hs4, hs0, hs1) |
| 270 | 270 | ||
| 271 | /* Round 64 through 79. */ | 271 | /* Round 64 through 79. */ |
| 272 | sha1_round4_update(64, hs1, hs2, hs3, hs4, hs0); | 272 | sha1_round4_update(64, hs1, hs2, hs3, hs4, hs0) |
| 273 | sha1_round4_update(65, hs0, hs1, hs2, hs3, hs4); | 273 | sha1_round4_update(65, hs0, hs1, hs2, hs3, hs4) |
| 274 | sha1_round4_update(66, hs4, hs0, hs1, hs2, hs3); | 274 | sha1_round4_update(66, hs4, hs0, hs1, hs2, hs3) |
| 275 | sha1_round4_update(67, hs3, hs4, hs0, hs1, hs2); | 275 | sha1_round4_update(67, hs3, hs4, hs0, hs1, hs2) |
| 276 | sha1_round4_update(68, hs2, hs3, hs4, hs0, hs1); | 276 | sha1_round4_update(68, hs2, hs3, hs4, hs0, hs1) |
| 277 | sha1_round4_update(69, hs1, hs2, hs3, hs4, hs0); | 277 | sha1_round4_update(69, hs1, hs2, hs3, hs4, hs0) |
| 278 | sha1_round4_update(70, hs0, hs1, hs2, hs3, hs4); | 278 | sha1_round4_update(70, hs0, hs1, hs2, hs3, hs4) |
| 279 | sha1_round4_update(71, hs4, hs0, hs1, hs2, hs3); | 279 | sha1_round4_update(71, hs4, hs0, hs1, hs2, hs3) |
| 280 | sha1_round4_update(72, hs3, hs4, hs0, hs1, hs2); | 280 | sha1_round4_update(72, hs3, hs4, hs0, hs1, hs2) |
| 281 | sha1_round4_update(73, hs2, hs3, hs4, hs0, hs1); | 281 | sha1_round4_update(73, hs2, hs3, hs4, hs0, hs1) |
| 282 | sha1_round4_update(74, hs1, hs2, hs3, hs4, hs0); | 282 | sha1_round4_update(74, hs1, hs2, hs3, hs4, hs0) |
| 283 | sha1_round4_update(75, hs0, hs1, hs2, hs3, hs4); | 283 | sha1_round4_update(75, hs0, hs1, hs2, hs3, hs4) |
| 284 | sha1_round4_update(76, hs4, hs0, hs1, hs2, hs3); | 284 | sha1_round4_update(76, hs4, hs0, hs1, hs2, hs3) |
| 285 | sha1_round4_update(77, hs3, hs4, hs0, hs1, hs2); | 285 | sha1_round4_update(77, hs3, hs4, hs0, hs1, hs2) |
| 286 | sha1_round4_update(78, hs2, hs3, hs4, hs0, hs1); | 286 | sha1_round4_update(78, hs2, hs3, hs4, hs0, hs1) |
| 287 | sha1_round4_update(79, hs1, hs2, hs3, hs4, hs0); | 287 | sha1_round4_update(79, hs1, hs2, hs3, hs4, hs0) |
| 288 | 288 | ||
| 289 | /* Add intermediate state to hash state. */ | 289 | /* Add intermediate state to hash state. */ |
| 290 | addl (0*4)(ctx), hs0 | 290 | addl (0*4)(ctx), hs0 |
diff --git a/src/lib/libcrypto/sha/sha1_amd64_shani.S b/src/lib/libcrypto/sha/sha1_amd64_shani.S index 751554f1d5..e31143887a 100644 --- a/src/lib/libcrypto/sha/sha1_amd64_shani.S +++ b/src/lib/libcrypto/sha/sha1_amd64_shani.S | |||
| @@ -1,4 +1,4 @@ | |||
| 1 | /* $OpenBSD: sha1_amd64_shani.S,v 1.2 2026/01/17 06:31:45 jsing Exp $ */ | 1 | /* $OpenBSD: sha1_amd64_shani.S,v 1.3 2026/01/24 14:20:52 jsing Exp $ */ |
| 2 | /* | 2 | /* |
| 3 | * Copyright (c) 2024 Joel Sing <jsing@openbsd.org> | 3 | * Copyright (c) 2024 Joel Sing <jsing@openbsd.org> |
| 4 | * | 4 | * |
| @@ -50,25 +50,25 @@ | |||
| 50 | 50 | ||
| 51 | #define sha1_message_schedule_load(idx, m, xmsg) \ | 51 | #define sha1_message_schedule_load(idx, m, xmsg) \ |
| 52 | movdqu (idx*16)(m), xmsg; \ | 52 | movdqu (idx*16)(m), xmsg; \ |
| 53 | pshufb xshufmask, xmsg; | 53 | pshufb xshufmask, xmsg |
| 54 | 54 | ||
| 55 | #define sha1_message_schedule_update(xm0, xm1, xm2, xm3) \ | 55 | #define sha1_message_schedule_update(xm0, xm1, xm2, xm3) \ |
| 56 | sha1msg1 xm1, xm0; \ | 56 | sha1msg1 xm1, xm0; \ |
| 57 | pxor xm2, xm0; \ | 57 | pxor xm2, xm0; \ |
| 58 | sha1msg2 xm3, xm0; | 58 | sha1msg2 xm3, xm0 |
| 59 | 59 | ||
| 60 | #define sha1_shani_round(fn, xmsg, xe, xe_next) \ | 60 | #define sha1_shani_round(fn, xmsg, xe, xe_next) \ |
| 61 | sha1nexte xmsg, xe; \ | 61 | sha1nexte xmsg, xe; \ |
| 62 | movdqa xabcd, xe_next; \ | 62 | movdqa xabcd, xe_next; \ |
| 63 | sha1rnds4 fn, xe, xabcd; | 63 | sha1rnds4 fn, xe, xabcd |
| 64 | 64 | ||
| 65 | #define sha1_shani_round_load(fn, idx, m, xmsg, xe, xe_next) \ | 65 | #define sha1_shani_round_load(fn, idx, m, xmsg, xe, xe_next) \ |
| 66 | sha1_message_schedule_load(idx, m, xmsg); \ | 66 | sha1_message_schedule_load(idx, m, xmsg); \ |
| 67 | sha1_shani_round(fn, xmsg, xe, xe_next); | 67 | sha1_shani_round(fn, xmsg, xe, xe_next) |
| 68 | 68 | ||
| 69 | #define sha1_shani_round_update(fn, xm0, xm1, xm2, xm3, xe, xe_next) \ | 69 | #define sha1_shani_round_update(fn, xm0, xm1, xm2, xm3, xe, xe_next) \ |
| 70 | sha1_message_schedule_update(xm0, xm1, xm2, xm3); \ | 70 | sha1_message_schedule_update(xm0, xm1, xm2, xm3); \ |
| 71 | sha1_shani_round(fn, xm0, xe, xe_next); | 71 | sha1_shani_round(fn, xm0, xe, xe_next) |
| 72 | 72 | ||
| 73 | 73 | ||
| 74 | .section .text | 74 | .section .text |
| @@ -109,14 +109,14 @@ sha1_block_shani: | |||
| 109 | movdqa xe0, xe_save | 109 | movdqa xe0, xe_save |
| 110 | 110 | ||
| 111 | /* Rounds 0 through 15 (four rounds at a time). */ | 111 | /* Rounds 0 through 15 (four rounds at a time). */ |
| 112 | sha1_message_schedule_load(0, in, xmsg0); | 112 | sha1_message_schedule_load(0, in, xmsg0) |
| 113 | paddd xmsg0, xe0 | 113 | paddd xmsg0, xe0 |
| 114 | movdqa xabcd, xe1 | 114 | movdqa xabcd, xe1 |
| 115 | sha1rnds4 $0, xe0, xabcd | 115 | sha1rnds4 $0, xe0, xabcd |
| 116 | 116 | ||
| 117 | sha1_shani_round_load($0, 1, in, xmsg1, xe1, xe0); | 117 | sha1_shani_round_load($0, 1, in, xmsg1, xe1, xe0) |
| 118 | sha1_shani_round_load($0, 2, in, xmsg2, xe0, xe1); | 118 | sha1_shani_round_load($0, 2, in, xmsg2, xe0, xe1) |
| 119 | sha1_shani_round_load($0, 3, in, xmsg3, xe1, xe0); | 119 | sha1_shani_round_load($0, 3, in, xmsg3, xe1, xe0) |
| 120 | 120 | ||
| 121 | /* Rounds 16 through 79 (four rounds at a time). */ | 121 | /* Rounds 16 through 79 (four rounds at a time). */ |
| 122 | sha1_shani_round_update($0, xmsg0, xmsg1, xmsg2, xmsg3, xe0, xe1) | 122 | sha1_shani_round_update($0, xmsg0, xmsg1, xmsg2, xmsg3, xe0, xe1) |
diff --git a/src/lib/libcrypto/sha/sha256_aarch64_ce.S b/src/lib/libcrypto/sha/sha256_aarch64_ce.S index b66969427b..343f338390 100644 --- a/src/lib/libcrypto/sha/sha256_aarch64_ce.S +++ b/src/lib/libcrypto/sha/sha256_aarch64_ce.S | |||
| @@ -1,4 +1,4 @@ | |||
| 1 | /* $OpenBSD: sha256_aarch64_ce.S,v 1.4 2026/01/17 06:31:45 jsing Exp $ */ | 1 | /* $OpenBSD: sha256_aarch64_ce.S,v 1.5 2026/01/24 14:20:52 jsing Exp $ */ |
| 2 | /* | 2 | /* |
| 3 | * Copyright (c) 2023,2025 Joel Sing <jsing@openbsd.org> | 3 | * Copyright (c) 2023,2025 Joel Sing <jsing@openbsd.org> |
| 4 | * | 4 | * |
| @@ -65,7 +65,7 @@ | |||
| 65 | */ | 65 | */ |
| 66 | #define sha256_message_schedule_update(m0, m1, m2, m3) \ | 66 | #define sha256_message_schedule_update(m0, m1, m2, m3) \ |
| 67 | sha256su0 m0.4s, m1.4s; \ | 67 | sha256su0 m0.4s, m1.4s; \ |
| 68 | sha256su1 m0.4s, m2.4s, m3.4s; | 68 | sha256su1 m0.4s, m2.4s, m3.4s |
| 69 | 69 | ||
| 70 | /* | 70 | /* |
| 71 | * Compute four SHA-256 rounds by adding W0:W1:W2:W3 + K0:K1:K2:K3, then | 71 | * Compute four SHA-256 rounds by adding W0:W1:W2:W3 + K0:K1:K2:K3, then |
| @@ -76,10 +76,10 @@ | |||
| 76 | add tmp0.4s, w.4s, k.4s; /* Tt = Wt + Kt */ \ | 76 | add tmp0.4s, w.4s, k.4s; /* Tt = Wt + Kt */ \ |
| 77 | mov tmp1.4s, h0.4s; \ | 77 | mov tmp1.4s, h0.4s; \ |
| 78 | sha256h h0, h1, tmp0.4s; \ | 78 | sha256h h0, h1, tmp0.4s; \ |
| 79 | sha256h2 h1, tmp1, tmp0.4s; | 79 | sha256h2 h1, tmp1, tmp0.4s |
| 80 | 80 | ||
| 81 | #define sha256_round_update(h0, h1, m0, m1, m2, m3, k) \ | 81 | #define sha256_round_update(h0, h1, m0, m1, m2, m3, k) \ |
| 82 | sha256_message_schedule_update(m0, m1, m2, m3) \ | 82 | sha256_message_schedule_update(m0, m1, m2, m3); \ |
| 83 | sha256_round(h0, h1, m0, k) | 83 | sha256_round(h0, h1, m0, k) |
| 84 | 84 | ||
| 85 | .arch armv8-a+sha2 | 85 | .arch armv8-a+sha2 |
diff --git a/src/lib/libcrypto/sha/sha256_amd64_generic.S b/src/lib/libcrypto/sha/sha256_amd64_generic.S index a5bb3eca42..a860b014f2 100644 --- a/src/lib/libcrypto/sha/sha256_amd64_generic.S +++ b/src/lib/libcrypto/sha/sha256_amd64_generic.S | |||
| @@ -1,4 +1,4 @@ | |||
| 1 | /* $OpenBSD: sha256_amd64_generic.S,v 1.4 2026/01/17 06:31:45 jsing Exp $ */ | 1 | /* $OpenBSD: sha256_amd64_generic.S,v 1.5 2026/01/24 14:20:52 jsing Exp $ */ |
| 2 | /* | 2 | /* |
| 3 | * Copyright (c) 2024 Joel Sing <jsing@openbsd.org> | 3 | * Copyright (c) 2024 Joel Sing <jsing@openbsd.org> |
| 4 | * | 4 | * |
| @@ -51,7 +51,7 @@ | |||
| 51 | #define sha256_message_schedule_load(idx, m, w, wt) \ | 51 | #define sha256_message_schedule_load(idx, m, w, wt) \ |
| 52 | movl (m, round, 4), wt; \ | 52 | movl (m, round, 4), wt; \ |
| 53 | bswapl wt; \ | 53 | bswapl wt; \ |
| 54 | movl wt, ((idx&0xf)*4)(w); | 54 | movl wt, ((idx&0xf)*4)(w) |
| 55 | 55 | ||
| 56 | /* | 56 | /* |
| 57 | * Update message schedule and return current value in wt: | 57 | * Update message schedule and return current value in wt: |
| @@ -82,7 +82,7 @@ | |||
| 82 | xorl tmp3, tmp2; /* sigma0 */ \ | 82 | xorl tmp3, tmp2; /* sigma0 */ \ |
| 83 | addl tmp2, wt; /* sigma0 */ \ | 83 | addl tmp2, wt; /* sigma0 */ \ |
| 84 | \ | 84 | \ |
| 85 | movl wt, ((idx&0xf)*4)(w); | 85 | movl wt, ((idx&0xf)*4)(w) |
| 86 | 86 | ||
| 87 | /* | 87 | /* |
| 88 | * Compute a SHA-256 round: | 88 | * Compute a SHA-256 round: |
| @@ -133,14 +133,14 @@ | |||
| 133 | xorl tmp2, tmp3; /* T2 Maj */ \ | 133 | xorl tmp2, tmp3; /* T2 Maj */ \ |
| 134 | addl tmp3, h; /* T2 Maj */ \ | 134 | addl tmp3, h; /* T2 Maj */ \ |
| 135 | \ | 135 | \ |
| 136 | addq $1, round; | 136 | addq $1, round |
| 137 | 137 | ||
| 138 | #define sha256_round_load(idx, a, b, c, d, e, f, g, h) \ | 138 | #define sha256_round_load(idx, a, b, c, d, e, f, g, h) \ |
| 139 | sha256_message_schedule_load(idx, in, %rsp, tmp0) \ | 139 | sha256_message_schedule_load(idx, in, %rsp, tmp0); \ |
| 140 | sha256_round(idx, a, b, c, d, e, f, g, h, k256, %rsp, tmp0) | 140 | sha256_round(idx, a, b, c, d, e, f, g, h, k256, %rsp, tmp0) |
| 141 | 141 | ||
| 142 | #define sha256_round_update(idx, a, b, c, d, e, f, g, h) \ | 142 | #define sha256_round_update(idx, a, b, c, d, e, f, g, h) \ |
| 143 | sha256_message_schedule_update(idx, %rsp, tmp0) \ | 143 | sha256_message_schedule_update(idx, %rsp, tmp0); \ |
| 144 | sha256_round(idx, a, b, c, d, e, f, g, h, k256, %rsp, tmp0) | 144 | sha256_round(idx, a, b, c, d, e, f, g, h, k256, %rsp, tmp0) |
| 145 | 145 | ||
| 146 | .section .text | 146 | .section .text |
diff --git a/src/lib/libcrypto/sha/sha256_amd64_shani.S b/src/lib/libcrypto/sha/sha256_amd64_shani.S index e43ecfa51f..b6f32b72cf 100644 --- a/src/lib/libcrypto/sha/sha256_amd64_shani.S +++ b/src/lib/libcrypto/sha/sha256_amd64_shani.S | |||
| @@ -1,4 +1,4 @@ | |||
| 1 | /* $OpenBSD: sha256_amd64_shani.S,v 1.2 2026/01/17 06:31:45 jsing Exp $ */ | 1 | /* $OpenBSD: sha256_amd64_shani.S,v 1.3 2026/01/24 14:20:52 jsing Exp $ */ |
| 2 | /* | 2 | /* |
| 3 | * Copyright (c) 2024 Joel Sing <jsing@openbsd.org> | 3 | * Copyright (c) 2024 Joel Sing <jsing@openbsd.org> |
| 4 | * | 4 | * |
| @@ -56,29 +56,29 @@ | |||
| 56 | #define sha256_message_schedule_load(idx, m, xmsgtmp) \ | 56 | #define sha256_message_schedule_load(idx, m, xmsgtmp) \ |
| 57 | movdqu (idx*16)(m), xmsg; \ | 57 | movdqu (idx*16)(m), xmsg; \ |
| 58 | pshufb xshufmask, xmsg; \ | 58 | pshufb xshufmask, xmsg; \ |
| 59 | movdqa xmsg, xmsgtmp; | 59 | movdqa xmsg, xmsgtmp |
| 60 | 60 | ||
| 61 | #define sha256_message_schedule_update(xmt0, xmt1, xmt2, xmt3) \ | 61 | #define sha256_message_schedule_update(xmt0, xmt1, xmt2, xmt3) \ |
| 62 | sha256msg1 xmt1, xmt0; \ | 62 | sha256msg1 xmt1, xmt0; \ |
| 63 | movdqa xmt3, xmsgtmp4; \ | 63 | movdqa xmt3, xmsgtmp4; \ |
| 64 | palignr $4, xmt2, xmsgtmp4; \ | 64 | palignr $4, xmt2, xmsgtmp4; \ |
| 65 | paddd xmsgtmp4, xmt0; \ | 65 | paddd xmsgtmp4, xmt0; \ |
| 66 | sha256msg2 xmt3, xmt0; | 66 | sha256msg2 xmt3, xmt0 |
| 67 | 67 | ||
| 68 | #define sha256_shani_round(idx) \ | 68 | #define sha256_shani_round(idx) \ |
| 69 | paddd (idx*16)(k256), xmsg; \ | 69 | paddd (idx*16)(k256), xmsg; \ |
| 70 | sha256rnds2 xmsg, xhs0, xhs1; \ | 70 | sha256rnds2 xmsg, xhs0, xhs1; \ |
| 71 | pshufd $0x0e, xmsg, xmsg; \ | 71 | pshufd $0x0e, xmsg, xmsg; \ |
| 72 | sha256rnds2 xmsg, xhs1, xhs0; | 72 | sha256rnds2 xmsg, xhs1, xhs0 |
| 73 | 73 | ||
| 74 | #define sha256_shani_round_load(idx, m, xmsgtmp) \ | 74 | #define sha256_shani_round_load(idx, m, xmsgtmp) \ |
| 75 | sha256_message_schedule_load(idx, m, xmsgtmp); \ | 75 | sha256_message_schedule_load(idx, m, xmsgtmp); \ |
| 76 | sha256_shani_round(idx); | 76 | sha256_shani_round(idx) |
| 77 | 77 | ||
| 78 | #define sha256_shani_round_update(idx, xmt0, xmt1, xmt2, xmt3) \ | 78 | #define sha256_shani_round_update(idx, xmt0, xmt1, xmt2, xmt3) \ |
| 79 | sha256_message_schedule_update(xmt0, xmt1, xmt2, xmt3); \ | 79 | sha256_message_schedule_update(xmt0, xmt1, xmt2, xmt3); \ |
| 80 | movdqa xmt0, xmsg; \ | 80 | movdqa xmt0, xmsg; \ |
| 81 | sha256_shani_round(idx); | 81 | sha256_shani_round(idx) |
| 82 | 82 | ||
| 83 | .section .text | 83 | .section .text |
| 84 | 84 | ||
diff --git a/src/lib/libcrypto/sha/sha512_amd64_generic.S b/src/lib/libcrypto/sha/sha512_amd64_generic.S index 3b6a9719e1..052367df07 100644 --- a/src/lib/libcrypto/sha/sha512_amd64_generic.S +++ b/src/lib/libcrypto/sha/sha512_amd64_generic.S | |||
| @@ -1,4 +1,4 @@ | |||
| 1 | /* $OpenBSD: sha512_amd64_generic.S,v 1.2 2026/01/17 06:31:45 jsing Exp $ */ | 1 | /* $OpenBSD: sha512_amd64_generic.S,v 1.3 2026/01/24 14:20:52 jsing Exp $ */ |
| 2 | /* | 2 | /* |
| 3 | * Copyright (c) 2024 Joel Sing <jsing@openbsd.org> | 3 | * Copyright (c) 2024 Joel Sing <jsing@openbsd.org> |
| 4 | * | 4 | * |
| @@ -51,7 +51,7 @@ | |||
| 51 | #define sha512_message_schedule_load(idx, m, w, wt) \ | 51 | #define sha512_message_schedule_load(idx, m, w, wt) \ |
| 52 | movq (m, round, 8), wt; \ | 52 | movq (m, round, 8), wt; \ |
| 53 | bswapq wt; \ | 53 | bswapq wt; \ |
| 54 | movq wt, ((idx&0xf)*8)(w); | 54 | movq wt, ((idx&0xf)*8)(w) |
| 55 | 55 | ||
| 56 | /* | 56 | /* |
| 57 | * Update message schedule and return current value in wt: | 57 | * Update message schedule and return current value in wt: |
| @@ -83,7 +83,7 @@ | |||
| 83 | xorq tmp3, tmp2; /* sigma0 */ \ | 83 | xorq tmp3, tmp2; /* sigma0 */ \ |
| 84 | addq tmp2, wt; /* sigma0 */ \ | 84 | addq tmp2, wt; /* sigma0 */ \ |
| 85 | \ | 85 | \ |
| 86 | movq wt, ((idx&0xf)*8)(w); | 86 | movq wt, ((idx&0xf)*8)(w) |
| 87 | 87 | ||
| 88 | /* | 88 | /* |
| 89 | * Compute a SHA-512 round: | 89 | * Compute a SHA-512 round: |
| @@ -134,14 +134,14 @@ | |||
| 134 | xorq tmp2, tmp3; /* T2 Maj */ \ | 134 | xorq tmp2, tmp3; /* T2 Maj */ \ |
| 135 | addq tmp3, h; /* T2 Maj */ \ | 135 | addq tmp3, h; /* T2 Maj */ \ |
| 136 | \ | 136 | \ |
| 137 | addq $1, round; | 137 | addq $1, round |
| 138 | 138 | ||
| 139 | #define sha512_round_load(idx, a, b, c, d, e, f, g, h) \ | 139 | #define sha512_round_load(idx, a, b, c, d, e, f, g, h) \ |
| 140 | sha512_message_schedule_load(idx, in, %rsp, tmp0) \ | 140 | sha512_message_schedule_load(idx, in, %rsp, tmp0); \ |
| 141 | sha512_round(idx, a, b, c, d, e, f, g, h, k512, %rsp, tmp0) | 141 | sha512_round(idx, a, b, c, d, e, f, g, h, k512, %rsp, tmp0) |
| 142 | 142 | ||
| 143 | #define sha512_round_update(idx, a, b, c, d, e, f, g, h) \ | 143 | #define sha512_round_update(idx, a, b, c, d, e, f, g, h) \ |
| 144 | sha512_message_schedule_update(idx, %rsp, tmp0) \ | 144 | sha512_message_schedule_update(idx, %rsp, tmp0); \ |
| 145 | sha512_round(idx, a, b, c, d, e, f, g, h, k512, %rsp, tmp0) | 145 | sha512_round(idx, a, b, c, d, e, f, g, h, k512, %rsp, tmp0) |
| 146 | 146 | ||
| 147 | .section .text | 147 | .section .text |