Filt_6k_7k_opt.s revision e17bb5cafdbc6089716d8e8c5afbb00f207a59bb
1@/* 2@ ** Copyright 2003-2010, VisualOn, Inc. 3@ ** 4@ ** Licensed under the Apache License, Version 2.0 (the "License"); 5@ ** you may not use this file except in compliance with the License. 6@ ** You may obtain a copy of the License at 7@ ** 8@ ** http://www.apache.org/licenses/LICENSE-2.0 9@ ** 10@ ** Unless required by applicable law or agreed to in writing, software 11@ ** distributed under the License is distributed on an "AS IS" BASIS, 12@ ** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 13@ ** See the License for the specific language governing permissions and 14@ ** limitations under the License. 15@ */ 16 17@**********************************************************************/ 18@void Filt_6k_7k( 19@ Word16 signal[], /* input: signal */ 20@ Word16 lg, /* input: length of input */ 21@ Word16 mem[] /* in/out: memory (size=30) */ 22@) 23@****************************************************************** 24@ r0 --- signal[] 25@ r1 --- lg 26@ r2 --- mem[] 27 28 .section .text 29 .global Filt_6k_7k_asm 30 .extern voAWB_Copy 31 .extern fir_6k_7k 32 .hidden fir_6k_7k 33 34Filt_6k_7k_asm: 35 36 STMFD r13!, {r4 - r12, r14} 37 SUB r13, r13, #240 @ x[L_SUBFR16k + (L_FIR - 1)] 38 MOV r8, r0 @ copy signal[] address 39 MOV r4, r1 @ copy lg address 40 MOV r5, r2 @ copy mem[] address 41 42 MOV r1, r13 43 MOV r0, r2 44 MOV r2, #30 @ L_FIR - 1 45 BL voAWB_Copy @ memcpy(x, mem, (L_FIR - 1)<<1) 46 47 ADR r3, Lable1 @ get fir_7k address 48 LDR r10, [r3] 49 ADD r10, r3 50 51 MOV r14, #0 52 MOV r3, r8 @ change myMemCopy to Copy, due to Copy will change r3 content 53 ADD r6, r13, #60 @ get x[L_FIR - 1] address 54 MOV r7, r3 @ get signal[i] 55LOOP1: 56 LDRSH r8, [r7], #2 57 LDRSH r9, [r7], #2 58 MOV r8, r8, ASR #2 59 MOV r9, r9, ASR #2 60 LDRSH r11, [r7], #2 61 LDRSH r12, [r7], #2 62 MOV r11, r11, ASR #2 63 MOV r12, r12, ASR #2 64 STRH r8, [r6], #2 65 STRH r9, [r6], #2 66 STRH r11, [r6], #2 67 STRH r12, [r6], #2 68 LDRSH r8, [r7], #2 69 LDRSH r9, [r7], #2 70 MOV r8, r8, ASR #2 71 MOV r9, r9, ASR #2 72 LDRSH r11, [r7], #2 73 LDRSH r12, [r7], #2 74 MOV r11, r11, ASR #2 75 MOV r12, r12, ASR #2 76 STRH r8, [r6], #2 77 STRH r9, [r6], #2 78 STRH r11, [r6], #2 79 STRH r12, [r6], #2 80 ADD r14, r14, #8 81 CMP r14, #80 82 BLT LOOP1 83 84 85 STR r5, [sp, #-4] @ PUSH r5 to stack 86 87 @ not use registers: r4, r10, r12, r14, r5 88 MOV r4, r13 89 MOV r5, #0 @ i = 0 90LOOP2: 91 LDR r0, [r10] 92 93 LDRSH r1, [r4] @ load x[i] 94 LDRSH r2, [r4, #60] @ load x[i + 30] 95 LDRSH r6, [r4, #2] @ load x[i + 1] 96 LDRSH r7, [r4, #58] @ load x[i + 29] 97 ADD r1, r1, r2 @ x[i] + x[i + 30] 98 ADD r6, r6, r7 @ x[i + 1] + x[i + 29] 99 LDRSH r8, [r4, #4] @ load x[i + 2] 100 LDRSH r9, [r4, #56] @ load x[i + 28] 101 102 SMULBB r14, r1, r0 @ (x[i] + x[i + 30]) * fir_7k[0] 103 ADD r8, r8, r9 @ x[i + 2] + x[i + 28] 104 SMLABT r14, r6, r0, r14 @ (x[i + 1] + x[i + 29]) * fir_7k[1] 105 106 LDR r0, [r10, #4] 107 LDRSH r1, [r4, #6] @ load x[i+3] 108 LDRSH r2, [r4, #54] @ load x[i+27] 109 LDRSH r6, [r4, #8] @ load x[i+4] 110 LDRSH r7, [r4, #52] @ load x[i+26] 111 ADD r1, r1, r2 @ x[i+3] + x[i+27] 112 ADD r6, r6, r7 @ x[i+4] + x[i+26] 113 SMLABB r14, r8, r0, r14 @ (x[i + 2] + x[i + 28]) * fir_7k[2] 114 LDRSH r8, [r4, #10] @ load x[i+5] 115 LDRSH r9, [r4, #50] @ load x[i+25] 116 SMLABT r14, r1, r0, r14 @ (x[i+3] + x[i+27]) * fir_7k[3] 117 ADD r8, r8, r9 @ x[i+5] + x[i+25] 118 119 LDR r0, [r10, #8] 120 LDRSH r1, [r4, #12] @ x[i+6] 121 LDRSH r2, [r4, #48] @ x[i+24] 122 SMLABB r14, r6, r0, r14 @ (x[i+4] + x[i+26]) * fir_7k[4] 123 LDRSH r6, [r4, #14] @ x[i+7] 124 LDRSH r7, [r4, #46] @ x[i+23] 125 SMLABT r14, r8, r0, r14 @ (x[i+5] + x[i+25]) * fir_7k[5] 126 LDR r0, [r10, #12] 127 ADD r1, r1, r2 @ (x[i+6] + x[i+24]) 128 ADD r6, r6, r7 @ (x[i+7] + x[i+23]) 129 SMLABB r14, r1, r0, r14 @ (x[i+6] + x[i+24]) * fir_7k[6] 130 LDRSH r8, [r4, #16] @ x[i+8] 131 LDRSH r9, [r4, #44] @ x[i+22] 132 SMLABT r14, r6, r0, r14 @ (x[i+7] + x[i+23]) * fir_7k[7] 133 LDR r0, [r10, #16] 134 LDRSH r1, [r4, #18] @ x[i+9] 135 LDRSH r2, [r4, #42] @ x[i+21] 136 LDRSH r6, [r4, #20] @ x[i+10] 137 LDRSH r7, [r4, #40] @ x[i+20] 138 ADD r8, r8, r9 @ (x[i+8] + x[i+22]) 139 ADD r1, r1, r2 @ (x[i+9] + x[i+21]) 140 ADD r6, r6, r7 @ (x[i+10] + x[i+20]) 141 SMLABB r14, r8, r0, r14 @ (x[i+8] + x[i+22]) * fir_7k[8] 142 LDRSH r8, [r4, #22] @ x[i+11] 143 LDRSH r9, [r4, #38] @ x[i+19] 144 SMLABT r14, r1, r0, r14 @ (x[i+9] + x[i+21]) * fir_7k[9] 145 LDR r0, [r10, #20] 146 LDRSH r1, [r4, #24] @ x[i+12] 147 LDRSH r2, [r4, #36] @ x[i+18] 148 SMLABB r14, r6, r0, r14 @ (x[i+10] + x[i+20]) * fir_7k[10] 149 LDRSH r6, [r4, #26] @ x[i+13] 150 ADD r8, r8, r9 @ (x[i+11] + x[i+19]) 151 LDRSH r7, [r4, #34] @ x[i+17] 152 SMLABT r14, r8, r0, r14 @ (x[i+11] + x[i+19]) * fir_7k[11] 153 LDR r0, [r10, #24] 154 ADD r1, r1, r2 @ x[i+12] + x[i+18] 155 LDRSH r8, [r4, #28] @ x[i+14] 156 SMLABB r14, r1, r0, r14 @ (x[i+12] + x[i+18]) * fir_7k[12] 157 ADD r6, r6, r7 @ (x[i+13] + x[i+17]) 158 LDRSH r9, [r4, #32] @ x[i+16] 159 SMLABT r14, r6, r0, r14 @ (x[i+13] + x[i+17]) * fir_7k[13] 160 LDR r0, [r10, #28] 161 ADD r8, r8, r9 @ (x[i+14] + x[i+16]) 162 LDRSH r1, [r4, #30] @ x[i+15] 163 SMLABB r14, r8, r0, r14 @ (x[i+14] + x[i+16]) * fir_7k[14] 164 SMLABT r14, r1, r0, r14 @ x[i+15] * fir_7k[15] 165 166 ADD r5, r5, #1 167 ADD r14, r14, #0x4000 168 ADD r4, r4, #2 169 MOV r1, r14, ASR #15 170 CMP r5, #80 171 STRH r1, [r3], #2 @signal[i] = (L_tmp + 0x4000) >> 15 172 BLT LOOP2 173 174 LDR r1, [sp, #-4] @mem address 175 ADD r0, r13, #160 @x + lg 176 MOV r2, #30 177 BL voAWB_Copy 178 179Filt_6k_7k_end: 180 ADD r13, r13, #240 181 LDMFD r13!, {r4 - r12, r15} 182 183Lable1: 184 .word fir_6k_7k-Lable1 185 @ENDFUNC 186 .END 187 188 189