^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) /* SPDX-License-Identifier: GPL-2.0-only */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3) * Copyright (c) 2010-2011, The Linux Foundation. All rights reserved.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) * Description
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) * library function for memcpy where length bytes are copied from
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10) * ptr_in to ptr_out. ptr_out is returned unchanged.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11) * Allows any combination of alignment on input and output pointers
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12) * and length from 0 to 2^32-1
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) * Restrictions
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) * The arrays should not overlap, the program will produce undefined output
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16) * if they do.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17) * For blocks less than 16 bytes a byte by byte copy is performed. For
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) * 8byte alignments, and length multiples, a dword copy is performed up to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) * 96bytes
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20) * History
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) * DJH 5/15/09 Initial version 1.0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23) * DJH 6/ 1/09 Version 1.1 modified ABI to inlcude R16-R19
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24) * DJH 7/12/09 Version 1.2 optimized codesize down to 760 was 840
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25) * DJH 10/14/09 Version 1.3 added special loop for aligned case, was
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) * overreading bloated codesize back up to 892
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) * DJH 4/20/10 Version 1.4 fixed Ldword_loop_epilog loop to prevent loads
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) * occurring if only 1 left outstanding, fixes bug
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) * # 3888, corrected for all alignments. Peeled off
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30) * 1 32byte chunk from kernel loop and extended 8byte
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) * loop at end to solve all combinations and prevent
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32) * over read. Fixed Ldword_loop_prolog to prevent
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33) * overread for blocks less than 48bytes. Reduced
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) * codesize to 752 bytes
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) * DJH 4/21/10 version 1.5 1.4 fix broke code for input block ends not
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) * aligned to dword boundaries,underwriting by 1
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) * byte, added detection for this and fixed. A
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38) * little bloat.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) * DJH 4/23/10 version 1.6 corrected stack error, R20 was not being restored
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40) * always, fixed the error of R20 being modified
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41) * before it was being saved
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42) * Natural c model
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) * ===============
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44) * void * memcpy(char * ptr_out, char * ptr_in, int length) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) * int i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46) * if(length) for(i=0; i < length; i++) { ptr_out[i] = ptr_in[i]; }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47) * return(ptr_out);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) * }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50) * Optimized memcpy function
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51) * =========================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52) * void * memcpy(char * ptr_out, char * ptr_in, int len) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53) * int i, prolog, kernel, epilog, mask;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54) * u8 offset;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55) * s64 data0, dataF8, data70;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57) * s64 * ptr8_in;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) * s64 * ptr8_out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59) * s32 * ptr4;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60) * s16 * ptr2;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62) * offset = ((int) ptr_in) & 7;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63) * ptr8_in = (s64 *) &ptr_in[-offset]; //read in the aligned pointers
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) * data70 = *ptr8_in++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66) * dataF8 = *ptr8_in++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) * data0 = HEXAGON_P_valignb_PPp(dataF8, data70, offset);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70) * prolog = 32 - ((int) ptr_out);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71) * mask = 0x7fffffff >> HEXAGON_R_cl0_R(len);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72) * prolog = prolog & mask;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) * kernel = len - prolog;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74) * epilog = kernel & 0x1F;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75) * kernel = kernel>>5;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) * if (prolog & 1) { ptr_out[0] = (u8) data0; data0 >>= 8; ptr_out += 1;}
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78) * ptr2 = (s16 *) &ptr_out[0];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) * if (prolog & 2) { ptr2[0] = (u16) data0; data0 >>= 16; ptr_out += 2;}
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80) * ptr4 = (s32 *) &ptr_out[0];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81) * if (prolog & 4) { ptr4[0] = (u32) data0; data0 >>= 32; ptr_out += 4;}
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83) * offset = offset + (prolog & 7);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84) * if (offset >= 8) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85) * data70 = dataF8;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86) * dataF8 = *ptr8_in++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87) * }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88) * offset = offset & 0x7;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90) * prolog = prolog >> 3;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91) * if (prolog) for (i=0; i < prolog; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92) * data0 = HEXAGON_P_valignb_PPp(dataF8, data70, offset);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93) * ptr8_out = (s64 *) &ptr_out[0]; *ptr8_out = data0; ptr_out += 8;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94) * data70 = dataF8;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95) * dataF8 = *ptr8_in++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96) * }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97) * if(kernel) { kernel -= 1; epilog += 32; }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 98) * if(kernel) for(i=0; i < kernel; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 99) * data0 = HEXAGON_P_valignb_PPp(dataF8, data70, offset);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100) * ptr8_out = (s64 *) &ptr_out[0]; *ptr8_out = data0; ptr_out += 8;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101) * data70 = *ptr8_in++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103) * data0 = HEXAGON_P_valignb_PPp(data70, dataF8, offset);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104) * ptr8_out = (s64 *) &ptr_out[0]; *ptr8_out = data0; ptr_out += 8;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105) * dataF8 = *ptr8_in++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) * data0 = HEXAGON_P_valignb_PPp(dataF8, data70, offset);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) * ptr8_out = (s64 *) &ptr_out[0]; *ptr8_out = data0; ptr_out += 8;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109) * data70 = *ptr8_in++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) * data0 = HEXAGON_P_valignb_PPp(data70, dataF8, offset);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112) * ptr8_out = (s64 *) &ptr_out[0]; *ptr8_out = data0; ptr_out += 8;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113) * dataF8 = *ptr8_in++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114) * }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115) * epilogdws = epilog >> 3;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116) * if (epilogdws) for (i=0; i < epilogdws; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117) * data0 = HEXAGON_P_valignb_PPp(dataF8, data70, offset);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) * ptr8_out = (s64 *) &ptr_out[0]; *ptr8_out = data0; ptr_out += 8;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119) * data70 = dataF8;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120) * dataF8 = *ptr8_in++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121) * }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122) * data0 = HEXAGON_P_valignb_PPp(dataF8, data70, offset);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124) * ptr4 = (s32 *) &ptr_out[0];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125) * if (epilog & 4) { ptr4[0] = (u32) data0; data0 >>= 32; ptr_out += 4;}
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126) * ptr2 = (s16 *) &ptr_out[0];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127) * if (epilog & 2) { ptr2[0] = (u16) data0; data0 >>= 16; ptr_out += 2;}
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128) * if (epilog & 1) { *ptr_out++ = (u8) data0; }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130) * return(ptr_out - length);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131) * }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133) * Codesize : 784 bytes
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) #define ptr_out R0 /* destination pounter */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138) #define ptr_in R1 /* source pointer */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139) #define len R2 /* length of copy in bytes */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141) #define data70 R13:12 /* lo 8 bytes of non-aligned transfer */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142) #define dataF8 R11:10 /* hi 8 bytes of non-aligned transfer */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143) #define ldata0 R7:6 /* even 8 bytes chunks */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144) #define ldata1 R25:24 /* odd 8 bytes chunks */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145) #define data1 R7 /* lower 8 bytes of ldata1 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146) #define data0 R6 /* lower 8 bytes of ldata0 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148) #define ifbyte p0 /* if transfer has bytes in epilog/prolog */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149) #define ifhword p0 /* if transfer has shorts in epilog/prolog */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150) #define ifword p0 /* if transfer has words in epilog/prolog */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151) #define noprolog p0 /* no prolog, xfer starts at 32byte */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152) #define nokernel p1 /* no 32byte multiple block in the transfer */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153) #define noepilog p0 /* no epilog, xfer ends on 32byte boundary */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154) #define align p2 /* alignment of input rel to 8byte boundary */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) #define kernel1 p0 /* kernel count == 1 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) #define dalign R25 /* rel alignment of input to output data */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158) #define star3 R16 /* number bytes in prolog - dwords */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159) #define rest R8 /* length - prolog bytes */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160) #define back R7 /* nr bytes > dword boundary in src block */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) #define epilog R3 /* bytes in epilog */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162) #define inc R15:14 /* inc kernel by -1 and defetch ptr by 32 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163) #define kernel R4 /* number of 32byte chunks in kernel */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164) #define ptr_in_p_128 R5 /* pointer for prefetch of input data */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165) #define mask R8 /* mask used to determine prolog size */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) #define shift R8 /* used to work a shifter to extract bytes */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167) #define shift2 R5 /* in epilog to workshifter to extract bytes */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168) #define prolog R15 /* bytes in prolog */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169) #define epilogdws R15 /* number dwords in epilog */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170) #define shiftb R14 /* used to extract bytes */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171) #define offset R9 /* same as align in reg */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172) #define ptr_out_p_32 R17 /* pointer to output dczero */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173) #define align888 R14 /* if simple dword loop can be used */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174) #define len8 R9 /* number of dwords in length */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175) #define over R20 /* nr of bytes > last inp buf dword boundary */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177) #define ptr_in_p_128kernel R5:4 /* packed fetch pointer & kernel cnt */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179) .section .text
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180) .p2align 4
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181) .global memcpy
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 182) .type memcpy, @function
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 183) memcpy:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 184) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 185) p2 = cmp.eq(len, #0); /* =0 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 186) align888 = or(ptr_in, ptr_out); /* %8 < 97 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 187) p0 = cmp.gtu(len, #23); /* %1, <24 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 188) p1 = cmp.eq(ptr_in, ptr_out); /* attempt to overwrite self */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 189) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 190) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 191) p1 = or(p2, p1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 192) p3 = cmp.gtu(len, #95); /* %8 < 97 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 193) align888 = or(align888, len); /* %8 < 97 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 194) len8 = lsr(len, #3); /* %8 < 97 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 195) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 196) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 197) dcfetch(ptr_in); /* zero/ptrin=ptrout causes fetch */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 198) p2 = bitsclr(align888, #7); /* %8 < 97 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 199) if(p1) jumpr r31; /* =0 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 200) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 201) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 202) p2 = and(p2,!p3); /* %8 < 97 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 203) if (p2.new) len = add(len, #-8); /* %8 < 97 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 204) if (p2.new) jump:NT .Ldwordaligned; /* %8 < 97 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 205) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 206) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 207) if(!p0) jump .Lbytes23orless; /* %1, <24 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 208) mask.l = #LO(0x7fffffff);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 209) /* all bytes before line multiples of data */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 210) prolog = sub(#0, ptr_out);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 211) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 212) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 213) /* save r31 on stack, decrement sp by 16 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 214) allocframe(#24);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 215) mask.h = #HI(0x7fffffff);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 216) ptr_in_p_128 = add(ptr_in, #32);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 217) back = cl0(len);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 218) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 219) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 220) memd(sp+#0) = R17:16; /* save r16,r17 on stack6 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 221) r31.l = #LO(.Lmemcpy_return); /* set up final return pointer */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 222) prolog &= lsr(mask, back);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 223) offset = and(ptr_in, #7);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 224) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 225) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 226) memd(sp+#8) = R25:24; /* save r25,r24 on stack */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 227) dalign = sub(ptr_out, ptr_in);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 228) r31.h = #HI(.Lmemcpy_return); /* set up final return pointer */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 229) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 230) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 231) /* see if there if input buffer end if aligned */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 232) over = add(len, ptr_in);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 233) back = add(len, offset);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 234) memd(sp+#16) = R21:20; /* save r20,r21 on stack */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 235) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 236) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 237) noprolog = bitsclr(prolog, #7);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 238) prolog = and(prolog, #31);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 239) dcfetch(ptr_in_p_128);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 240) ptr_in_p_128 = add(ptr_in_p_128, #32);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 241) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 242) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 243) kernel = sub(len, prolog);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 244) shift = asl(prolog, #3);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 245) star3 = and(prolog, #7);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 246) ptr_in = and(ptr_in, #-8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 247) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 248) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 249) prolog = lsr(prolog, #3);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 250) epilog = and(kernel, #31);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 251) ptr_out_p_32 = add(ptr_out, prolog);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 252) over = and(over, #7);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 253) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 254) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 255) p3 = cmp.gtu(back, #8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 256) kernel = lsr(kernel, #5);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 257) dcfetch(ptr_in_p_128);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 258) ptr_in_p_128 = add(ptr_in_p_128, #32);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 259) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 260) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 261) p1 = cmp.eq(prolog, #0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 262) if(!p1.new) prolog = add(prolog, #1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 263) dcfetch(ptr_in_p_128); /* reserve the line 64bytes on */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 264) ptr_in_p_128 = add(ptr_in_p_128, #32);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 265) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 266) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 267) nokernel = cmp.eq(kernel,#0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 268) dcfetch(ptr_in_p_128); /* reserve the line 64bytes on */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 269) ptr_in_p_128 = add(ptr_in_p_128, #32);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 270) shiftb = and(shift, #8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 271) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 272) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 273) dcfetch(ptr_in_p_128); /* reserve the line 64bytes on */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 274) ptr_in_p_128 = add(ptr_in_p_128, #32);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 275) if(nokernel) jump .Lskip64;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 276) p2 = cmp.eq(kernel, #1); /* skip ovr if kernel == 0 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 277) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 278) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 279) dczeroa(ptr_out_p_32);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 280) /* don't advance pointer */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 281) if(!p2) ptr_out_p_32 = add(ptr_out_p_32, #32);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 282) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 283) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 284) dalign = and(dalign, #31);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 285) dczeroa(ptr_out_p_32);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 286) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 287) .Lskip64:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 288) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 289) data70 = memd(ptr_in++#16);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 290) if(p3) dataF8 = memd(ptr_in+#8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 291) if(noprolog) jump .Lnoprolog32;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 292) align = offset;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 293) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 294) /* upto initial 7 bytes */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 295) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 296) ldata0 = valignb(dataF8, data70, align);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 297) ifbyte = tstbit(shift,#3);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 298) offset = add(offset, star3);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 299) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 300) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 301) if(ifbyte) memb(ptr_out++#1) = data0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 302) ldata0 = lsr(ldata0, shiftb);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 303) shiftb = and(shift, #16);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 304) ifhword = tstbit(shift,#4);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 305) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 306) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 307) if(ifhword) memh(ptr_out++#2) = data0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 308) ldata0 = lsr(ldata0, shiftb);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 309) ifword = tstbit(shift,#5);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 310) p2 = cmp.gtu(offset, #7);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 311) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 312) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 313) if(ifword) memw(ptr_out++#4) = data0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 314) if(p2) data70 = dataF8;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 315) if(p2) dataF8 = memd(ptr_in++#8); /* another 8 bytes */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 316) align = offset;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 317) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 318) .Lnoprolog32:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 319) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 320) p3 = sp1loop0(.Ldword_loop_prolog, prolog)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 321) rest = sub(len, star3); /* whats left after the loop */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 322) p0 = cmp.gt(over, #0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 323) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 324) if(p0) rest = add(rest, #16);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 325) .Ldword_loop_prolog:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 326) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 327) if(p3) memd(ptr_out++#8) = ldata0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 328) ldata0 = valignb(dataF8, data70, align);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 329) p0 = cmp.gt(rest, #16);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 330) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 331) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 332) data70 = dataF8;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 333) if(p0) dataF8 = memd(ptr_in++#8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 334) rest = add(rest, #-8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 335) }:endloop0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 336) .Lkernel:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 337) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 338) /* kernel is at least 32bytes */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 339) p3 = cmp.gtu(kernel, #0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 340) /* last itn. remove edge effects */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 341) if(p3.new) kernel = add(kernel, #-1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 342) /* dealt with in last dword loop */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 343) if(p3.new) epilog = add(epilog, #32);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 344) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 345) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 346) nokernel = cmp.eq(kernel, #0); /* after adjustment, recheck */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 347) if(nokernel.new) jump:NT .Lepilog; /* likely not taken */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 348) inc = combine(#32, #-1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 349) p3 = cmp.gtu(dalign, #24);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 350) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 351) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 352) if(p3) jump .Lodd_alignment;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 353) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 354) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 355) loop0(.Loword_loop_25to31, kernel);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 356) kernel1 = cmp.gtu(kernel, #1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 357) rest = kernel;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 358) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 359) .falign
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 360) .Loword_loop_25to31:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 361) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 362) dcfetch(ptr_in_p_128); /* prefetch 4 lines ahead */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 363) if(kernel1) ptr_out_p_32 = add(ptr_out_p_32, #32);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 364) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 365) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 366) dczeroa(ptr_out_p_32); /* reserve the next 32bytes in cache */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 367) p3 = cmp.eq(kernel, rest);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 368) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 369) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 370) /* kernel -= 1 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 371) ptr_in_p_128kernel = vaddw(ptr_in_p_128kernel, inc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 372) /* kill write on first iteration */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 373) if(!p3) memd(ptr_out++#8) = ldata1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 374) ldata1 = valignb(dataF8, data70, align);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 375) data70 = memd(ptr_in++#8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 376) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 377) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 378) memd(ptr_out++#8) = ldata0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 379) ldata0 = valignb(data70, dataF8, align);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 380) dataF8 = memd(ptr_in++#8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 381) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 382) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 383) memd(ptr_out++#8) = ldata1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 384) ldata1 = valignb(dataF8, data70, align);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 385) data70 = memd(ptr_in++#8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 386) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 387) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 388) memd(ptr_out++#8) = ldata0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 389) ldata0 = valignb(data70, dataF8, align);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 390) dataF8 = memd(ptr_in++#8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 391) kernel1 = cmp.gtu(kernel, #1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 392) }:endloop0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 393) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 394) memd(ptr_out++#8) = ldata1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 395) jump .Lepilog;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 396) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 397) .Lodd_alignment:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 398) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 399) loop0(.Loword_loop_00to24, kernel);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 400) kernel1 = cmp.gtu(kernel, #1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 401) rest = add(kernel, #-1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 402) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 403) .falign
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 404) .Loword_loop_00to24:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 405) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 406) dcfetch(ptr_in_p_128); /* prefetch 4 lines ahead */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 407) ptr_in_p_128kernel = vaddw(ptr_in_p_128kernel, inc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 408) if(kernel1) ptr_out_p_32 = add(ptr_out_p_32, #32);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 409) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 410) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 411) dczeroa(ptr_out_p_32); /* reserve the next 32bytes in cache */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 412) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 413) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 414) memd(ptr_out++#8) = ldata0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 415) ldata0 = valignb(dataF8, data70, align);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 416) data70 = memd(ptr_in++#8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 417) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 418) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 419) memd(ptr_out++#8) = ldata0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 420) ldata0 = valignb(data70, dataF8, align);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 421) dataF8 = memd(ptr_in++#8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 422) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 423) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 424) memd(ptr_out++#8) = ldata0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 425) ldata0 = valignb(dataF8, data70, align);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 426) data70 = memd(ptr_in++#8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 427) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 428) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 429) memd(ptr_out++#8) = ldata0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 430) ldata0 = valignb(data70, dataF8, align);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 431) dataF8 = memd(ptr_in++#8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 432) kernel1 = cmp.gtu(kernel, #1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 433) }:endloop0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 434) .Lepilog:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 435) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 436) noepilog = cmp.eq(epilog,#0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 437) epilogdws = lsr(epilog, #3);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 438) kernel = and(epilog, #7);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 439) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 440) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 441) if(noepilog) jumpr r31;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 442) if(noepilog) ptr_out = sub(ptr_out, len);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 443) p3 = cmp.eq(epilogdws, #0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 444) shift2 = asl(epilog, #3);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 445) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 446) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 447) shiftb = and(shift2, #32);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 448) ifword = tstbit(epilog,#2);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 449) if(p3) jump .Lepilog60;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 450) if(!p3) epilog = add(epilog, #-16);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 451) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 452) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 453) loop0(.Ldword_loop_epilog, epilogdws);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 454) /* stop criteria is lsbs unless = 0 then its 8 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 455) p3 = cmp.eq(kernel, #0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 456) if(p3.new) kernel= #8;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 457) p1 = cmp.gt(over, #0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 458) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 459) /* if not aligned to end of buffer execute 1 more iteration */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 460) if(p1) kernel= #0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 461) .Ldword_loop_epilog:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 462) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 463) memd(ptr_out++#8) = ldata0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 464) ldata0 = valignb(dataF8, data70, align);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 465) p3 = cmp.gt(epilog, kernel);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 466) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 467) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 468) data70 = dataF8;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 469) if(p3) dataF8 = memd(ptr_in++#8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 470) epilog = add(epilog, #-8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 471) }:endloop0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 472) /* copy last 7 bytes */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 473) .Lepilog60:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 474) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 475) if(ifword) memw(ptr_out++#4) = data0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 476) ldata0 = lsr(ldata0, shiftb);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 477) ifhword = tstbit(epilog,#1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 478) shiftb = and(shift2, #16);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 479) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 480) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 481) if(ifhword) memh(ptr_out++#2) = data0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 482) ldata0 = lsr(ldata0, shiftb);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 483) ifbyte = tstbit(epilog,#0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 484) if(ifbyte.new) len = add(len, #-1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 485) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 486) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 487) if(ifbyte) memb(ptr_out) = data0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 488) ptr_out = sub(ptr_out, len); /* return dest pointer */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 489) jumpr r31;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 490) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 491) /* do byte copy for small n */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 492) .Lbytes23orless:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 493) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 494) p3 = sp1loop0(.Lbyte_copy, len);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 495) len = add(len, #-1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 496) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 497) .Lbyte_copy:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 498) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 499) data0 = memb(ptr_in++#1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 500) if(p3) memb(ptr_out++#1) = data0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 501) }:endloop0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 502) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 503) memb(ptr_out) = data0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 504) ptr_out = sub(ptr_out, len);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 505) jumpr r31;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 506) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 507) /* do dword copies for aligned in, out and length */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 508) .Ldwordaligned:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 509) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 510) p3 = sp1loop0(.Ldword_copy, len8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 511) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 512) .Ldword_copy:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 513) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 514) if(p3) memd(ptr_out++#8) = ldata0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 515) ldata0 = memd(ptr_in++#8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 516) }:endloop0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 517) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 518) memd(ptr_out) = ldata0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 519) ptr_out = sub(ptr_out, len);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 520) jumpr r31; /* return to function caller */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 521) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 522) .Lmemcpy_return:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 523) r21:20 = memd(sp+#16); /* restore r20+r21 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 524) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 525) r25:24 = memd(sp+#8); /* restore r24+r25 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 526) r17:16 = memd(sp+#0); /* restore r16+r17 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 527) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 528) deallocframe; /* restore r31 and incrment stack by 16 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 529) jumpr r31
Orange Pi5 kernel
Deprecated Linux kernel 5.10.110 for OrangePi 5/5B/5+ boards
3 Commits
0 Branches
0 Tags