/* Optimized strcmp implementation for PowerPC476. Copyright (C) 2010 Free Software Foundation, Inc. This file is part of the GNU C Library. The GNU C Library is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; either version 2.1 of the License, or (at your option) any later version. The GNU C Library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details. You should have received a copy of the GNU Lesser General Public License along with the GNU C Library. If not, see . */ #include #include #include /* strcmp Register Use r0:temp return equality r3:source1 address, return equality r4:source2 address Implementation description Check 2 words from src1 and src2. If unequal jump to end and return src1 > src2 or src1 < src2. If null check bytes before null and then jump to end and return src1 > src2, src1 < src2 or src1 = src2. If src1 = src2 and no null, repeat. */ EALIGN (BP_SYM(strcmp),5,0) neg r7,r3 clrlwi r7,r7,20 neg r8,r4 clrlwi r8,r8,20 srwi. r7,r7,5 beq L(byte_loop) srwi. r8,r8,5 beq L(byte_loop) cmplw r7,r8 mtctr r7 ble L(big_loop) mtctr r8 L(big_loop): lwz r5,0(r3) lwz r6,4(r3) lwz r8,0(r4) lwz r9,4(r4) dlmzb. r12,r5,r6 bne L(end_check) cmplw r5,r8 bne L(st1) cmplw r6,r9 bne L(st1) lwz r5,8(r3) lwz r6,12(r3) lwz r8,8(r4) lwz r9,12(r4) dlmzb. r12,r5,r6 bne L(end_check) cmplw r5,r8 bne L(st1) cmplw r6,r9 bne L(st1) lwz r5,16(r3) lwz r6,20(r3) lwz r8,16(r4) lwz r9,20(r4) dlmzb. r12,r5,r6 bne L(end_check) cmplw r5,r8 bne L(st1) cmplw r6,r9 bne L(st1) lwz r5,24(r3) lwz r6,28(r3) addi r3,r3,0x20 lwz r8,24(r4) lwz r9,28(r4) addi r4,r4,0x20 dlmzb. r12,r5,r6 bne L(end_check) cmplw r5,r8 bne L(st1) cmplw r6,r9 bne L(st1) bdnz L(big_loop) b L(byte_loop) L(end_check): subfic r12,r12,4 blt L(end_check2) rlwinm r12,r12,3,0,31 srw r5,r5,r12 srw r8,r8,r12 cmplw r5,r8 bne L(st1) b L(end_strcmp) L(end_check2): addi r12,r12,4 cmplw r5,r8 rlwinm r12,r12,3,0,31 bne L(st1) srw r6,r6,r12 srw r9,r9,r12 cmplw r6,r9 bne L(st1) L(end_strcmp): addi r3,r0,0 blr L(st1): mfcr r3 blr L(byte_loop): lbz r5,0(r3) addi r3,r3,1 lbz r6,0(r4) addi r4,r4,1 cmplw r5,r6 bne L(st1) cmpwi r5,0 beq L(end_strcmp) b L(byte_loop) END (BP_SYM (strcmp)) libc_hidden_builtin_def (strcmp)