1// 2// Copyright (c) 2013 - 2016, Linaro Limited 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 are met: 7// * Redistributions of source code must retain the above copyright 8// notice, this list of conditions and the following disclaimer. 9// * Redistributions in binary form must reproduce the above copyright 10// notice, this list of conditions and the following disclaimer in the 11// documentation and/or other materials provided with the distribution. 12// * Neither the name of the Linaro nor the 13// names of its contributors may be used to endorse or promote products 14// derived from this software without specific prior written permission. 15// 16// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 17// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 18// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR 19// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT 20// HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, 21// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT 22// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, 23// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY 24// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 25// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE 26// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 27// 28 29// Parameters and result. 30#define src1 r0 31#define src2 r1 32#define limit r2 33#define result r0 34 35// Internal variables. 36#define data1 r3 37#define data2 r4 38#define limit_wd r5 39#define diff r6 40#define tmp1 r7 41#define tmp2 r12 42#define pos r8 43#define mask r14 44 45 .text 46 .thumb 47 .syntax unified 48 .align 5 49ASM_GLOBAL ASM_PFX(InternalMemCompareMem) 50ASM_PFX(InternalMemCompareMem): 51 push {r4-r8, lr} 52 eor tmp1, src1, src2 53 tst tmp1, #3 54 bne .Lmisaligned4 55 ands tmp1, src1, #3 56 bne .Lmutual_align 57 add limit_wd, limit, #3 58 nop.w 59 lsr limit_wd, limit_wd, #2 60 61 // Start of performance-critical section -- one 32B cache line. 62.Lloop_aligned: 63 ldr data1, [src1], #4 64 ldr data2, [src2], #4 65.Lstart_realigned: 66 subs limit_wd, limit_wd, #1 67 eor diff, data1, data2 // Non-zero if differences found. 68 cbnz diff, 0f 69 bne .Lloop_aligned 70 // End of performance-critical section -- one 32B cache line. 71 72 // Not reached the limit, must have found a diff. 730: cbnz limit_wd, .Lnot_limit 74 75 // Limit % 4 == 0 => all bytes significant. 76 ands limit, limit, #3 77 beq .Lnot_limit 78 79 lsl limit, limit, #3 // Bits -> bytes. 80 mov mask, #~0 81 lsl mask, mask, limit 82 bic data1, data1, mask 83 bic data2, data2, mask 84 85 orr diff, diff, mask 86 87.Lnot_limit: 88 rev diff, diff 89 rev data1, data1 90 rev data2, data2 91 92 // The MS-non-zero bit of DIFF marks either the first bit 93 // that is different, or the end of the significant data. 94 // Shifting left now will bring the critical information into the 95 // top bits. 96 clz pos, diff 97 lsl data1, data1, pos 98 lsl data2, data2, pos 99 100 // But we need to zero-extend (char is unsigned) the value and then 101 // perform a signed 32-bit subtraction. 102 lsr data1, data1, #28 103 sub result, data1, data2, lsr #28 104 pop {r4-r8, pc} 105 106.Lmutual_align: 107 // Sources are mutually aligned, but are not currently at an 108 // alignment boundary. Round down the addresses and then mask off 109 // the bytes that precede the start point. 110 bic src1, src1, #3 111 bic src2, src2, #3 112 add limit, limit, tmp1 // Adjust the limit for the extra. 113 lsl tmp1, tmp1, #3 // Bytes beyond alignment -> bits. 114 ldr data1, [src1], #4 115 rsb tmp1, tmp1, #32 // Bits to alignment -32. 116 ldr data2, [src2], #4 117 mov tmp2, #~0 118 119 // Little-endian. Early bytes are at LSB. 120 lsr tmp2, tmp2, tmp1 // Shift (tmp1 & 31). 121 add limit_wd, limit, #3 122 orr data1, data1, tmp2 123 orr data2, data2, tmp2 124 lsr limit_wd, limit_wd, #2 125 b .Lstart_realigned 126 127.Lmisaligned4: 128 sub limit, limit, #1 1291: 130 // Perhaps we can do better than this. 131 ldrb data1, [src1], #1 132 ldrb data2, [src2], #1 133 subs limit, limit, #1 134 it cs 135 cmpcs.n data1, data2 136 beq 1b 137 sub result, data1, data2 138 pop {r4-r8, pc} 139