lh | 9ed821d | 2023-04-07 01:36:19 -0700 | [diff] [blame] | 1 | /* Optimized version of the standard strcpy() function. |
| 2 | This file is part of the GNU C Library. |
| 3 | Copyright (C) 2000-2015 Free Software Foundation, Inc. |
| 4 | Contributed by Dan Pop <Dan.Pop@cern.ch>. |
| 5 | |
| 6 | The GNU C Library is free software; you can redistribute it and/or |
| 7 | modify it under the terms of the GNU Lesser General Public |
| 8 | License as published by the Free Software Foundation; either |
| 9 | version 2.1 of the License, or (at your option) any later version. |
| 10 | |
| 11 | The GNU C Library is distributed in the hope that it will be useful, |
| 12 | but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 13 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
| 14 | Lesser General Public License for more details. |
| 15 | |
| 16 | You should have received a copy of the GNU Lesser General Public |
| 17 | License along with the GNU C Library; if not, see |
| 18 | <http://www.gnu.org/licenses/>. */ |
| 19 | |
| 20 | /* Return: dest |
| 21 | |
| 22 | Inputs: |
| 23 | in0: dest |
| 24 | in1: src |
| 25 | |
| 26 | In this form, it assumes little endian mode. For big endian mode, |
| 27 | the two shifts in .l2 must be inverted: |
| 28 | |
| 29 | shl value = r[1], sh1 // value = w0 << sh1 |
| 30 | shr.u tmp = r[0], sh2 // tmp = w1 >> sh2 |
| 31 | */ |
| 32 | |
| 33 | #include <sysdep.h> |
| 34 | #undef ret |
| 35 | |
| 36 | #define saved_lc r15 |
| 37 | #define saved_pr r16 |
| 38 | #define thresh r17 |
| 39 | #define dest r19 |
| 40 | #define src r20 |
| 41 | #define len r21 |
| 42 | #define asrc r22 |
| 43 | #define tmp r23 |
| 44 | #define pos r24 |
| 45 | #define w0 r25 |
| 46 | #define w1 r26 |
| 47 | #define c r27 |
| 48 | #define sh2 r28 |
| 49 | #define sh1 r29 |
| 50 | #define loopcnt r30 |
| 51 | #define value r31 |
| 52 | |
| 53 | ENTRY(strcpy) |
| 54 | .prologue |
| 55 | alloc r2 = ar.pfs, 2, 0, 30, 32 |
| 56 | |
| 57 | #define MEMLAT 2 |
| 58 | .rotr r[MEMLAT + 2] |
| 59 | .rotp p[MEMLAT + 1] |
| 60 | |
| 61 | mov ret0 = in0 // return value = dest |
| 62 | .save pr, saved_pr |
| 63 | mov saved_pr = pr // save the predicate registers |
| 64 | .save ar.lc, saved_lc |
| 65 | mov saved_lc = ar.lc // save the loop counter |
| 66 | .body |
| 67 | sub tmp = r0, in0 ;; // tmp = -dest |
| 68 | mov dest = in0 // dest |
| 69 | mov src = in1 // src |
| 70 | and loopcnt = 7, tmp ;; // loopcnt = -dest % 8 |
| 71 | cmp.eq p6, p0 = loopcnt, r0 |
| 72 | adds loopcnt = -1, loopcnt // --loopcnt |
| 73 | (p6) br.cond.sptk .dest_aligned ;; |
| 74 | mov ar.lc = loopcnt |
| 75 | .l1: // copy -dest % 8 bytes |
| 76 | ld1 c = [src], 1 // c = *src++ |
| 77 | ;; |
| 78 | st1 [dest] = c, 1 // *dest++ = c |
| 79 | cmp.eq p6, p0 = c, r0 |
| 80 | (p6) br.cond.dpnt .restore_and_exit |
| 81 | br.cloop.dptk .l1 ;; |
| 82 | .dest_aligned: |
| 83 | and sh1 = 7, src // sh1 = src % 8 |
| 84 | mov ar.lc = -1 // "infinite" loop |
| 85 | and asrc = -8, src ;; // asrc = src & -OPSIZ -- align src |
| 86 | sub thresh = 8, sh1 |
| 87 | mov pr.rot = 1 << 16 // set rotating predicates |
| 88 | cmp.ne p7, p0 = r0, r0 // clear p7 |
| 89 | shl sh1 = sh1, 3 ;; // sh1 = 8 * (src % 8) |
| 90 | sub sh2 = 64, sh1 // sh2 = 64 - sh1 |
| 91 | cmp.eq p6, p0 = sh1, r0 // is the src aligned? |
| 92 | (p6) br.cond.sptk .src_aligned ;; |
| 93 | ld8 r[1] = [asrc],8 ;; |
| 94 | |
| 95 | .align 32 |
| 96 | .l2: |
| 97 | ld8.s r[0] = [asrc], 8 |
| 98 | shr.u value = r[1], sh1 ;; // value = w0 >> sh1 |
| 99 | czx1.r pos = value ;; // do we have an "early" zero |
| 100 | cmp.lt p7, p0 = pos, thresh // in w0 >> sh1? |
| 101 | (p7) br.cond.dpnt .found0 |
| 102 | chk.s r[0], .recovery2 // it is safe to do that only |
| 103 | .back2: // after the previous test |
| 104 | shl tmp = r[0], sh2 // tmp = w1 << sh2 |
| 105 | ;; |
| 106 | or value = value, tmp ;; // value |= tmp |
| 107 | czx1.r pos = value ;; |
| 108 | cmp.ne p7, p0 = 8, pos |
| 109 | (p7) br.cond.dpnt .found0 |
| 110 | st8 [dest] = value, 8 // store val to dest |
| 111 | br.ctop.dptk .l2 ;; |
| 112 | .src_aligned: |
| 113 | .l3: |
| 114 | (p[0]) ld8.s r[0] = [src], 8 |
| 115 | (p[MEMLAT]) chk.s r[MEMLAT], .recovery3 |
| 116 | .back3: |
| 117 | (p[MEMLAT]) mov value = r[MEMLAT] |
| 118 | (p[MEMLAT]) czx1.r pos = r[MEMLAT] ;; |
| 119 | (p[MEMLAT]) cmp.ne p7, p0 = 8, pos |
| 120 | (p7) br.cond.dpnt .found0 |
| 121 | (p[MEMLAT]) st8 [dest] = r[MEMLAT], 8 |
| 122 | br.ctop.dptk .l3 ;; |
| 123 | .found0: |
| 124 | mov ar.lc = pos |
| 125 | .l4: |
| 126 | extr.u c = value, 0, 8 // c = value & 0xff |
| 127 | shr.u value = value, 8 |
| 128 | ;; |
| 129 | st1 [dest] = c, 1 |
| 130 | br.cloop.dptk .l4 ;; |
| 131 | .restore_and_exit: |
| 132 | mov ar.lc = saved_lc // restore the loop counter |
| 133 | mov pr = saved_pr, -1 // restore the predicate registers |
| 134 | br.ret.sptk.many b0 |
| 135 | .recovery2: |
| 136 | add tmp = -8, asrc ;; |
| 137 | ld8 r[0] = [tmp] |
| 138 | br.cond.sptk .back2 |
| 139 | .recovery3: |
| 140 | add tmp = -(MEMLAT + 1) * 8, src ;; |
| 141 | ld8 r[MEMLAT] = [tmp] |
| 142 | br.cond.sptk .back3 |
| 143 | END(strcpy) |
| 144 | libc_hidden_builtin_def (strcpy) |