1/* 2 * Copyright (C) 2013 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 */ 28 29 30/* 31 * This code assumes it is running on a processor that supports all arm v7 32 * instructions, that supports neon instructions, and that has a 32 byte 33 * cache line. 34 */ 35 36// Assumes neon instructions and a cache line size of 32 bytes. 37 38ENTRY_PRIVATE(MEMCPY_BASE) 39 .cfi_def_cfa_offset 8 40 .cfi_rel_offset r0, 0 41 .cfi_rel_offset lr, 4 42 43 /* do we have at least 16-bytes to copy (needed for alignment below) */ 44 cmp r2, #16 45 blo 5f 46 47 /* align destination to cache-line for the write-buffer */ 48 rsb r3, r0, #0 49 ands r3, r3, #0xF 50 beq 2f 51 52 /* copy up to 15-bytes (count in r3) */ 53 sub r2, r2, r3 54 movs ip, r3, lsl #31 55 itt mi 56 ldrbmi lr, [r1], #1 57 strbmi lr, [r0], #1 58 itttt cs 59 ldrbcs ip, [r1], #1 60 ldrbcs lr, [r1], #1 61 strbcs ip, [r0], #1 62 strbcs lr, [r0], #1 63 movs ip, r3, lsl #29 64 bge 1f 65 // copies 4 bytes, destination 32-bits aligned 66 vld4.8 {d0[0], d1[0], d2[0], d3[0]}, [r1]! 67 vst4.8 {d0[0], d1[0], d2[0], d3[0]}, [r0, :32]! 681: bcc 2f 69 // copies 8 bytes, destination 64-bits aligned 70 vld1.8 {d0}, [r1]! 71 vst1.8 {d0}, [r0, :64]! 72 732: /* make sure we have at least 64 bytes to copy */ 74 subs r2, r2, #64 75 blo 2f 76 771: /* The main loop copies 64 bytes at a time */ 78 vld1.8 {d0 - d3}, [r1]! 79 vld1.8 {d4 - d7}, [r1]! 80 pld [r1, #(32*8)] 81 subs r2, r2, #64 82 vst1.8 {d0 - d3}, [r0, :128]! 83 vst1.8 {d4 - d7}, [r0, :128]! 84 bhs 1b 85 862: /* fix-up the remaining count and make sure we have >= 32 bytes left */ 87 adds r2, r2, #32 88 blo 4f 89 90 /* Copy 32 bytes. These cache lines were already preloaded */ 91 vld1.8 {d0 - d3}, [r1]! 92 sub r2, r2, #32 93 vst1.8 {d0 - d3}, [r0, :128]! 94 954: /* less than 32 left */ 96 add r2, r2, #32 97 tst r2, #0x10 98 beq 5f 99 // copies 16 bytes, 128-bits aligned 100 vld1.8 {d0, d1}, [r1]! 101 vst1.8 {d0, d1}, [r0, :128]! 102 1035: /* copy up to 15-bytes (count in r2) */ 104 movs ip, r2, lsl #29 105 bcc 1f 106 vld1.8 {d0}, [r1]! 107 vst1.8 {d0}, [r0]! 1081: bge 2f 109 vld4.8 {d0[0], d1[0], d2[0], d3[0]}, [r1]! 110 vst4.8 {d0[0], d1[0], d2[0], d3[0]}, [r0]! 1112: movs ip, r2, lsl #31 112 itt mi 113 ldrbmi r3, [r1], #1 114 strbmi r3, [r0], #1 115 itttt cs 116 ldrbcs ip, [r1], #1 117 ldrbcs lr, [r1], #1 118 strbcs ip, [r0], #1 119 strbcs lr, [r0], #1 120 121 ldmfd sp!, {r0, lr} 122 bx lr 123END(MEMCPY_BASE) 124