Loading...
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 | /* Copyright (C) 1991,1993,1995,1997,1998,2003,2004 Free Software Foundation, Inc. This file is part of the GNU C Library. Contributed by Torbjorn Granlund (tege@sics.se). The GNU C Library is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; either version 2.1 of the License, or (at your option) any later version. The GNU C Library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details. You should have received a copy of the GNU Lesser General Public License along with the GNU C Library; if not, see <http://www.gnu.org/licenses/>. */ #include <string.h> #include "memcopy.h" #include <endian.h> #if __BYTE_ORDER == __BIG_ENDIAN # define CMP_LT_OR_GT(a, b) ((a) > (b) ? 1 : -1) #else # define CMP_LT_OR_GT(a, b) memcmp_bytes ((a), (b)) #endif /* BE VERY CAREFUL IF YOU CHANGE THIS CODE! */ /* The strategy of this memcmp is: 1. Compare bytes until one of the block pointers is aligned. 2. Compare using memcmp_common_alignment or memcmp_not_common_alignment, regarding the alignment of the other block after the initial byte operations. The maximum number of full words (of type op_t) are compared in this way. 3. Compare the few remaining bytes. */ #if __BYTE_ORDER != __BIG_ENDIAN /* memcmp_bytes -- Compare A and B bytewise in the byte order of the machine. A and B are known to be different. This is needed only on little-endian machines. */ static __inline__ int memcmp_bytes (op_t a, op_t b) { long int srcp1 = (long int) &a; long int srcp2 = (long int) &b; op_t a0, b0; do { a0 = ((byte *) srcp1)[0]; b0 = ((byte *) srcp2)[0]; srcp1 += 1; srcp2 += 1; } while (a0 == b0); return a0 - b0; } #endif /* memcmp_common_alignment -- Compare blocks at SRCP1 and SRCP2 with LEN `op_t' objects (not LEN bytes!). Both SRCP1 and SRCP2 should be aligned for memory operations on `op_t's. */ static int memcmp_common_alignment (long int srcp1, long int srcp2, size_t len) { op_t a0, a1; op_t b0, b1; switch (len % 4) { default: /* Avoid warning about uninitialized local variables. */ case 2: a0 = ((op_t *) srcp1)[0]; b0 = ((op_t *) srcp2)[0]; srcp1 -= 2 * OPSIZ; srcp2 -= 2 * OPSIZ; len += 2; goto do1; case 3: a1 = ((op_t *) srcp1)[0]; b1 = ((op_t *) srcp2)[0]; srcp1 -= OPSIZ; srcp2 -= OPSIZ; len += 1; goto do2; case 0: if (OP_T_THRES <= 3 * OPSIZ && len == 0) return 0; a0 = ((op_t *) srcp1)[0]; b0 = ((op_t *) srcp2)[0]; goto do3; case 1: a1 = ((op_t *) srcp1)[0]; b1 = ((op_t *) srcp2)[0]; srcp1 += OPSIZ; srcp2 += OPSIZ; len -= 1; if (OP_T_THRES <= 3 * OPSIZ && len == 0) goto do0; /* Fall through. */ } do { a0 = ((op_t *) srcp1)[0]; b0 = ((op_t *) srcp2)[0]; if (a1 != b1) return CMP_LT_OR_GT (a1, b1); do3: a1 = ((op_t *) srcp1)[1]; b1 = ((op_t *) srcp2)[1]; if (a0 != b0) return CMP_LT_OR_GT (a0, b0); do2: a0 = ((op_t *) srcp1)[2]; b0 = ((op_t *) srcp2)[2]; if (a1 != b1) return CMP_LT_OR_GT (a1, b1); do1: a1 = ((op_t *) srcp1)[3]; b1 = ((op_t *) srcp2)[3]; if (a0 != b0) return CMP_LT_OR_GT (a0, b0); srcp1 += 4 * OPSIZ; srcp2 += 4 * OPSIZ; len -= 4; } while (len != 0); /* This is the right position for do0. Please don't move it into the loop. */ do0: if (a1 != b1) return CMP_LT_OR_GT (a1, b1); return 0; } /* memcmp_not_common_alignment -- Compare blocks at SRCP1 and SRCP2 with LEN `op_t' objects (not LEN bytes!). SRCP2 should be aligned for memory operations on `op_t', but SRCP1 *should be unaligned*. */ static int memcmp_not_common_alignment (long int srcp1, long int srcp2, size_t len) { op_t a0, a1, a2, a3; op_t b0, b1, b2, b3; op_t x; int shl, shr; /* Calculate how to shift a word read at the memory operation aligned srcp1 to make it aligned for comparison. */ shl = 8 * (srcp1 % OPSIZ); shr = 8 * OPSIZ - shl; /* Make SRCP1 aligned by rounding it down to the beginning of the `op_t' it points in the middle of. */ srcp1 &= -OPSIZ; switch (len % 4) { default: /* Avoid warning about uninitialized local variables. */ case 2: a1 = ((op_t *) srcp1)[0]; a2 = ((op_t *) srcp1)[1]; b2 = ((op_t *) srcp2)[0]; srcp1 -= 1 * OPSIZ; srcp2 -= 2 * OPSIZ; len += 2; goto do1; case 3: a0 = ((op_t *) srcp1)[0]; a1 = ((op_t *) srcp1)[1]; b1 = ((op_t *) srcp2)[0]; srcp2 -= 1 * OPSIZ; len += 1; goto do2; case 0: if (OP_T_THRES <= 3 * OPSIZ && len == 0) return 0; a3 = ((op_t *) srcp1)[0]; a0 = ((op_t *) srcp1)[1]; b0 = ((op_t *) srcp2)[0]; srcp1 += 1 * OPSIZ; goto do3; case 1: a2 = ((op_t *) srcp1)[0]; a3 = ((op_t *) srcp1)[1]; b3 = ((op_t *) srcp2)[0]; srcp1 += 2 * OPSIZ; srcp2 += 1 * OPSIZ; len -= 1; if (OP_T_THRES <= 3 * OPSIZ && len == 0) goto do0; /* Fall through. */ } do { a0 = ((op_t *) srcp1)[0]; b0 = ((op_t *) srcp2)[0]; x = MERGE(a2, shl, a3, shr); if (x != b3) return CMP_LT_OR_GT (x, b3); do3: a1 = ((op_t *) srcp1)[1]; b1 = ((op_t *) srcp2)[1]; x = MERGE(a3, shl, a0, shr); if (x != b0) return CMP_LT_OR_GT (x, b0); do2: a2 = ((op_t *) srcp1)[2]; b2 = ((op_t *) srcp2)[2]; x = MERGE(a0, shl, a1, shr); if (x != b1) return CMP_LT_OR_GT (x, b1); do1: a3 = ((op_t *) srcp1)[3]; b3 = ((op_t *) srcp2)[3]; x = MERGE(a1, shl, a2, shr); if (x != b2) return CMP_LT_OR_GT (x, b2); srcp1 += 4 * OPSIZ; srcp2 += 4 * OPSIZ; len -= 4; } while (len != 0); /* This is the right position for do0. Please don't move it into the loop. */ do0: x = MERGE(a2, shl, a3, shr); if (x != b3) return CMP_LT_OR_GT (x, b3); return 0; } int memcmp (const __ptr_t s1, const __ptr_t s2, size_t len) { op_t a0; op_t b0; long int srcp1 = (long int) s1; long int srcp2 = (long int) s2; op_t res; if (len >= OP_T_THRES) { /* There are at least some bytes to compare. No need to test for LEN == 0 in this alignment loop. */ while (srcp2 % OPSIZ != 0) { a0 = ((byte *) srcp1)[0]; b0 = ((byte *) srcp2)[0]; srcp1 += 1; srcp2 += 1; res = a0 - b0; if (res != 0) return res; len -= 1; } /* SRCP2 is now aligned for memory operations on `op_t'. SRCP1 alignment determines if we can do a simple, aligned compare or need to shuffle bits. */ if (srcp1 % OPSIZ == 0) res = memcmp_common_alignment (srcp1, srcp2, len / OPSIZ); else res = memcmp_not_common_alignment (srcp1, srcp2, len / OPSIZ); if (res != 0) return res; /* Number of bytes remaining in the interval [0..OPSIZ-1]. */ srcp1 += len & -OPSIZ; srcp2 += len & -OPSIZ; len %= OPSIZ; } /* There are just a few bytes to compare. Use byte memory operations. */ while (len != 0) { a0 = ((byte *) srcp1)[0]; b0 = ((byte *) srcp2)[0]; srcp1 += 1; srcp2 += 1; res = a0 - b0; if (res != 0) return res; len -= 1; } return 0; } libc_hidden_weak(memcmp) #ifdef __UCLIBC_SUSV3_LEGACY__ strong_alias(memcmp,bcmp) #endif |