projects / thirdparty / glibc.git / blame
| Commit | Line | Data |
|---|---|---|
| a72cc2b2 | 1 | /* Optimized memcpy implementation for PowerPC476. |
| 568035b7 | 2 | Copyright (C) 2010-2013 Free Software Foundation, Inc. |
| a72cc2b2 LM |
3 | This file is part of the GNU C Library. |
| 4 | ||
| 5 | The GNU C Library is free software; you can redistribute it and/or | |
| 6 | modify it under the terms of the GNU Lesser General Public | |
| 7 | License as published by the Free Software Foundation; either | |
| 8 | version 2.1 of the License, or (at your option) any later version. | |
| 9 | ||
| 10 | The GNU C Library is distributed in the hope that it will be useful, | |
| 11 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
| 12 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | |
| 13 | Lesser General Public License for more details. | |
| 14 | ||
| 15 | You should have received a copy of the GNU Lesser General Public | |
| ab84e3ff PE |
16 | License along with the GNU C Library. If not, see |
| 17 | <http://www.gnu.org/licenses/>. */ | |
| a72cc2b2 LM |
18 | |
| 19 | #include <sysdep.h> | |
| 20 | #include <bp-sym.h> | |
| 21 | #include <bp-asm.h> | |
| 22 | ||
| 23 | /* memcpy | |
| 24 | ||
| 25 | r0:return address | |
| 26 | r3:destination address | |
| 27 | r4:source address | |
| 28 | r5:byte count | |
| 29 | ||
| 30 | Save return address in r0. | |
| 31 | If destinationn and source are unaligned and copy count is greater than 256 | |
| 32 | then copy 0-3 bytes to make destination aligned. | |
| 33 | If 32 or more bytes to copy we use 32 byte copy loop. | |
| 34 | Finaly we copy 0-31 extra bytes. */ | |
| 35 | ||
| 36 | EALIGN (BP_SYM (memcpy), 5, 0) | |
| 37 | /* Check if bytes to copy are greater than 256 and if | |
| 38 | source and destination are unaligned */ | |
| 39 | cmpwi r5,0x0100 | |
| 40 | addi r0,r3,0 | |
| 41 | ble L(string_count_loop) | |
| 42 | neg r6,r3 | |
| 43 | clrlwi. r6,r6,30 | |
| 44 | beq L(string_count_loop) | |
| 45 | neg r6,r4 | |
| 46 | clrlwi. r6,r6,30 | |
| 47 | beq L(string_count_loop) | |
| 48 | mtctr r6 | |
| 49 | subf r5,r6,r5 | |
| 50 | ||
| 51 | L(unaligned_bytecopy_loop): /* Align destination by coping 0-3 bytes */ | |
| 52 | lbz r8,0x0(r4) | |
| 53 | addi r4,r4,1 | |
| 54 | stb r8,0x0(r3) | |
| 55 | addi r3,r3,1 | |
| 56 | bdnz L(unaligned_bytecopy_loop) | |
| 57 | srwi. r7,r5,5 | |
| 58 | beq L(preword2_count_loop) | |
| 59 | mtctr r7 | |
| 60 | ||
| 61 | L(word8_count_loop_no_dcbt): /* Copy 32 bytes at a time */ | |
| 62 | lwz r6,0(r4) | |
| 63 | lwz r7,4(r4) | |
| 64 | lwz r8,8(r4) | |
| 65 | lwz r9,12(r4) | |
| 66 | subi r5,r5,0x20 | |
| 67 | stw r6,0(r3) | |
| 68 | stw r7,4(r3) | |
| 69 | stw r8,8(r3) | |
| 70 | stw r9,12(r3) | |
| 71 | lwz r6,16(r4) | |
| 72 | lwz r7,20(r4) | |
| 73 | lwz r8,24(r4) | |
| 74 | lwz r9,28(r4) | |
| 75 | addi r4,r4,0x20 | |
| 76 | stw r6,16(r3) | |
| 77 | stw r7,20(r3) | |
| 78 | stw r8,24(r3) | |
| 79 | stw r9,28(r3) | |
| 80 | addi r3,r3,0x20 | |
| 81 | bdnz L(word8_count_loop_no_dcbt) | |
| 82 | ||
| 83 | L(preword2_count_loop): /* Copy remaining 0-31 bytes */ | |
| 84 | clrlwi. r12,r5,27 | |
| 85 | beq L(end_memcpy) | |
| 86 | mtxer r12 | |
| 87 | lswx r5,0,r4 | |
| 88 | stswx r5,0,r3 | |
| 89 | mr r3,r0 | |
| 90 | blr | |
| 91 | ||
| 92 | L(string_count_loop): /* Copy odd 0-31 bytes */ | |
| 93 | clrlwi. r12,r5,28 | |
| 94 | add r3,r3,r5 | |
| 95 | add r4,r4,r5 | |
| 96 | beq L(pre_string_copy) | |
| 97 | mtxer r12 | |
| 98 | subf r4,r12,r4 | |
| 99 | subf r3,r12,r3 | |
| 100 | lswx r6,0,r4 | |
| 101 | stswx r6,0,r3 | |
| 102 | ||
| 103 | L(pre_string_copy): /* Check how many 32 byte chunck to copy */ | |
| 104 | srwi. r7,r5,4 | |
| 105 | beq L(end_memcpy) | |
| 106 | mtctr r7 | |
| 107 | ||
| 108 | L(word4_count_loop_no_dcbt): /* Copy 32 bytes at a time */ | |
| 109 | lwz r6,-4(r4) | |
| 110 | lwz r7,-8(r4) | |
| 111 | lwz r8,-12(r4) | |
| 112 | lwzu r9,-16(r4) | |
| 113 | stw r6,-4(r3) | |
| 114 | stw r7,-8(r3) | |
| 115 | stw r8,-12(r3) | |
| 116 | stwu r9,-16(r3) | |
| 117 | bdz L(end_memcpy) | |
| 118 | lwz r6,-4(r4) | |
| 119 | lwz r7,-8(r4) | |
| 120 | lwz r8,-12(r4) | |
| 121 | lwzu r9,-16(r4) | |
| 122 | stw r6,-4(r3) | |
| 123 | stw r7,-8(r3) | |
| 124 | stw r8,-12(r3) | |
| 125 | stwu r9,-16(r3) | |
| 126 | bdnz L(word4_count_loop_no_dcbt) | |
| 127 | ||
| 128 | L(end_memcpy): | |
| 129 | mr r3,r0 | |
| 130 | blr | |
| 131 | END (BP_SYM (memcpy)) | |
| 132 | libc_hidden_builtin_def (memcpy) |