Christopher Ferris | 753eb7f | 2014-06-27 15:26:15 -0700 | [diff] [blame] | 1 | /* |
| 2 | * Copyright (C) 2014 The Android Open Source Project |
| 3 | * All rights reserved. |
| 4 | * |
| 5 | * Redistribution and use in source and binary forms, with or without |
| 6 | * modification, are permitted provided that the following conditions |
| 7 | * are met: |
| 8 | * * Redistributions of source code must retain the above copyright |
| 9 | * notice, this list of conditions and the following disclaimer. |
| 10 | * * Redistributions in binary form must reproduce the above copyright |
| 11 | * notice, this list of conditions and the following disclaimer in |
| 12 | * the documentation and/or other materials provided with the |
| 13 | * distribution. |
| 14 | * |
| 15 | * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS |
| 16 | * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT |
| 17 | * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS |
| 18 | * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE |
| 19 | * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, |
| 20 | * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, |
| 21 | * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS |
| 22 | * OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED |
| 23 | * AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, |
| 24 | * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT |
| 25 | * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF |
| 26 | * SUCH DAMAGE. |
| 27 | */ |
Elliott Hughes | 1558138 | 2014-07-07 15:42:06 -0700 | [diff] [blame] | 28 | |
| 29 | /* |
| 30 | Copyright (c) 2014, Linaro Limited |
Christopher Ferris | 753eb7f | 2014-06-27 15:26:15 -0700 | [diff] [blame] | 31 | All rights reserved. |
| 32 | |
| 33 | Redistribution and use in source and binary forms, with or without |
| 34 | modification, are permitted provided that the following conditions are met: |
| 35 | * Redistributions of source code must retain the above copyright |
| 36 | notice, this list of conditions and the following disclaimer. |
| 37 | * Redistributions in binary form must reproduce the above copyright |
| 38 | notice, this list of conditions and the following disclaimer in the |
| 39 | documentation and/or other materials provided with the distribution. |
| 40 | * Neither the name of the Linaro nor the |
| 41 | names of its contributors may be used to endorse or promote products |
| 42 | derived from this software without specific prior written permission. |
| 43 | |
| 44 | THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS |
| 45 | "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT |
| 46 | LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR |
| 47 | A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT |
| 48 | HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, |
| 49 | SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT |
| 50 | LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, |
| 51 | DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY |
| 52 | THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT |
| 53 | (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE |
| 54 | OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
| 55 | */ |
| 56 | |
| 57 | /* Assumptions: |
| 58 | * |
| 59 | * ARMv8-a, AArch64 |
| 60 | */ |
| 61 | |
| 62 | #if !defined(STPCPY) && !defined(STRCPY) |
| 63 | #error "Either STPCPY or STRCPY must be defined." |
| 64 | #endif |
| 65 | |
| 66 | #include <private/bionic_asm.h> |
| 67 | |
| 68 | /* Arguments and results. */ |
| 69 | #if defined(STPCPY) |
| 70 | #define dst x0 |
| 71 | #elif defined(STRCPY) |
| 72 | #define dstin x0 |
| 73 | #endif |
| 74 | #define src x1 |
| 75 | |
| 76 | /* Locals and temporaries. */ |
| 77 | #if defined(STRCPY) |
| 78 | #define dst x2 |
| 79 | #endif |
| 80 | #define data1 x3 |
| 81 | #define data1_w w3 |
| 82 | #define data2 x4 |
| 83 | #define data2_w w4 |
| 84 | #define has_nul1 x5 |
| 85 | #define has_nul1_w w5 |
| 86 | #define has_nul2 x6 |
| 87 | #define tmp1 x7 |
| 88 | #define tmp2 x8 |
| 89 | #define tmp3 x9 |
| 90 | #define tmp4 x10 |
| 91 | #define zeroones x11 |
| 92 | #define zeroones_w w11 |
| 93 | #define pos x12 |
| 94 | |
| 95 | #define REP8_01 0x0101010101010101 |
| 96 | #define REP8_7f 0x7f7f7f7f7f7f7f7f |
| 97 | #define REP8_80 0x8080808080808080 |
| 98 | |
| 99 | #if defined(STPCPY) |
| 100 | ENTRY(stpcpy) |
| 101 | #elif defined(STRCPY) |
| 102 | ENTRY(strcpy) |
| 103 | #endif |
| 104 | mov zeroones, #REP8_01 |
| 105 | #if defined(STRCPY) |
| 106 | mov dst, dstin |
| 107 | #endif |
| 108 | ands tmp1, src, #15 |
| 109 | b.ne .Lmisaligned |
| 110 | // NUL detection works on the principle that (X - 1) & (~X) & 0x80 |
| 111 | // (=> (X - 1) & ~(X | 0x7f)) is non-zero iff a byte is zero, and |
| 112 | // can be done in parallel across the entire word. |
| 113 | // The inner loop deals with two Dwords at a time. This has a |
| 114 | // slightly higher start-up cost, but we should win quite quickly, |
| 115 | // especially on cores with a high number of issue slots per |
| 116 | // cycle, as we get much better parallelism out of the operations. |
| 117 | .Lloop: |
| 118 | ldp data1, data2, [src], #16 |
| 119 | sub tmp1, data1, zeroones |
| 120 | orr tmp2, data1, #REP8_7f |
| 121 | bic has_nul1, tmp1, tmp2 |
| 122 | cbnz has_nul1, .Lnul_in_data1 |
| 123 | sub tmp3, data2, zeroones |
| 124 | orr tmp4, data2, #REP8_7f |
| 125 | bic has_nul2, tmp3, tmp4 |
| 126 | cbnz has_nul2, .Lnul_in_data2 |
| 127 | // No NUL in either register, copy it in a single instruction. |
| 128 | stp data1, data2, [dst], #16 |
| 129 | b .Lloop |
| 130 | |
| 131 | .Lnul_in_data1: |
| 132 | rev has_nul1, has_nul1 |
| 133 | clz pos, has_nul1 |
| 134 | add tmp1, pos, #0x8 |
| 135 | |
| 136 | tbz tmp1, #6, 1f |
| 137 | #if defined(STPCPY) |
| 138 | str data1, [dst], #7 |
| 139 | #elif defined(STRCPY) |
| 140 | str data1, [dst] |
| 141 | #endif |
| 142 | ret |
| 143 | 1: |
| 144 | tbz tmp1, #5, 1f |
| 145 | str data1_w, [dst], #4 |
| 146 | lsr data1, data1, #32 |
| 147 | 1: |
| 148 | tbz tmp1, #4, 1f |
| 149 | strh data1_w, [dst], #2 |
| 150 | lsr data1, data1, #16 |
| 151 | 1: |
| 152 | tbz tmp1, #3, 1f |
| 153 | strb data1_w, [dst] |
| 154 | #if defined(STPCPY) |
| 155 | ret |
| 156 | #endif |
| 157 | 1: |
| 158 | #if defined(STPCPY) |
| 159 | // Back up one so that dst points to the '\0' string terminator. |
| 160 | sub dst, dst, #1 |
| 161 | #endif |
| 162 | ret |
| 163 | |
| 164 | .Lnul_in_data2: |
| 165 | str data1, [dst], #8 |
| 166 | rev has_nul2, has_nul2 |
| 167 | clz pos, has_nul2 |
| 168 | add tmp1, pos, #0x8 |
| 169 | |
| 170 | tbz tmp1, #6, 1f |
| 171 | #if defined(STPCPY) |
| 172 | str data2, [dst], #7 |
| 173 | #elif defined(STRCPY) |
| 174 | str data2, [dst] |
| 175 | #endif |
| 176 | ret |
| 177 | 1: |
| 178 | tbz tmp1, #5, 1f |
| 179 | str data2_w, [dst], #4 |
| 180 | lsr data2, data2, #32 |
| 181 | 1: |
| 182 | tbz tmp1, #4, 1f |
| 183 | strh data2_w, [dst], #2 |
| 184 | lsr data2, data2, #16 |
| 185 | 1: |
| 186 | tbz tmp1, #3, 1f |
| 187 | strb data2_w, [dst] |
| 188 | #if defined(STPCPY) |
| 189 | ret |
| 190 | #endif |
| 191 | 1: |
| 192 | #if defined(STPCPY) |
| 193 | // Back up one so that dst points to the '\0' string terminator. |
| 194 | sub dst, dst, #1 |
| 195 | #endif |
| 196 | ret |
| 197 | |
| 198 | .Lmisaligned: |
| 199 | tbz src, #0, 1f |
| 200 | ldrb data1_w, [src], #1 |
| 201 | strb data1_w, [dst], #1 |
| 202 | cbnz data1_w, 1f |
| 203 | #if defined(STPCPY) |
| 204 | // Back up one so that dst points to the '\0' string terminator. |
| 205 | sub dst, dst, #1 |
| 206 | #endif |
| 207 | ret |
| 208 | 1: |
| 209 | tbz src, #1, 1f |
| 210 | ldrb data1_w, [src], #1 |
| 211 | strb data1_w, [dst], #1 |
| 212 | cbz data1_w, .Ldone |
| 213 | ldrb data2_w, [src], #1 |
| 214 | strb data2_w, [dst], #1 |
| 215 | cbnz data2_w, 1f |
| 216 | .Ldone: |
| 217 | #if defined(STPCPY) |
| 218 | // Back up one so that dst points to the '\0' string terminator. |
| 219 | sub dst, dst, #1 |
| 220 | #endif |
| 221 | ret |
| 222 | 1: |
| 223 | tbz src, #2, 1f |
| 224 | ldr data1_w, [src], #4 |
| 225 | // Check for a zero. |
| 226 | sub has_nul1_w, data1_w, zeroones_w |
| 227 | bic has_nul1_w, has_nul1_w, data1_w |
| 228 | ands has_nul1_w, has_nul1_w, #0x80808080 |
| 229 | b.ne .Lnul_in_data1 |
| 230 | str data1_w, [dst], #4 |
| 231 | 1: |
| 232 | tbz src, #3, .Lloop |
| 233 | ldr data1, [src], #8 |
| 234 | // Check for a zero. |
| 235 | sub tmp1, data1, zeroones |
| 236 | orr tmp2, data1, #REP8_7f |
| 237 | bics has_nul1, tmp1, tmp2 |
| 238 | b.ne .Lnul_in_data1 |
| 239 | str data1, [dst], #8 |
| 240 | b .Lloop |
| 241 | #if defined(STPCPY) |
| 242 | END(stpcpy) |
| 243 | #elif defined(STRCPY) |
| 244 | END(strcpy) |
| 245 | #endif |