45 #define TEST_LENGTH_SAMPLES 320 46 #define MAX_BLOCKSIZE 320 47 #define MAX_NUMTAPS 100 49 #define TEST_COUNT 5000 67 #if defined (SMOKE_TEST)||(REGRESSION_TEST) 70 #ifdef PERFORMANCE_TEST 92 -0.110273, -0.042966, -0.043804, 0.087350, -0.085191, 0.009420, 0.086440
106 0.068186, 0.064344, -0.162450, 0.057015, 0.029743, 0.010066, 0.047792, 0.021273,
107 -0.096447, -0.211652, -0.086613, 0.057501, -0.187605, -0.167199, -0.026983, -0.025464,
108 -0.061495, 0.110914, -0.081973, -0.055231, -0.074430, -0.196536, 0.016845, -0.096493,
109 0.039625, -0.110273, -0.042966, -0.043804, 0.087350, -0.085191, 0.009420, 0.086440
118 -0.432565, -1.665584, 0.125332, 0.287676, -1.146471, 1.190915, 1.189164, -0.037633,
119 0.327292, 0.174639, -0.186709, 0.725791, -0.588317, 2.183186, -0.136396, 0.113931,
120 1.066768, 0.059281, -0.095648, -0.832349, 0.294411, -1.336182, 0.714325, 1.623562,
121 -0.691776, 0.857997, 1.254001, -1.593730, -1.440964, 0.571148, -0.399886, 0.689997,
122 0.815622, 0.711908, 1.290250, 0.668601, 1.190838, -1.202457, -0.019790, -0.156717,
123 -1.604086, 0.257304, -1.056473, 1.415141, -0.805090, 0.528743, 0.219321, -0.921902,
124 -2.170674, -0.059188, -1.010634, 0.614463, 0.507741, 1.692430, 0.591283, -0.643595,
125 0.380337, -1.009116, -0.019511, -0.048221, 0.000043, -0.317859, 1.095004, -1.873990,
126 0.428183, 0.895638, 0.730957, 0.577857, 0.040314, 0.677089, 0.568900, -0.255645,
127 -0.377469, -0.295887, -1.475135, -0.234004, 0.118445, 0.314809, 1.443508, -0.350975,
128 0.623234, 0.799049, 0.940890, -0.992092, 0.212035, 0.237882, -1.007763, -0.742045,
129 1.082295, -0.131500, 0.389880, 0.087987, -0.635465, -0.559573, 0.443653, -0.949904,
130 0.781182, 0.568961, -0.821714, -0.265607, -1.187777, -2.202321, 0.986337, -0.518635,
131 0.327368, 0.234057, 0.021466, -1.003944, -0.947146, -0.374429, -1.185886, -1.055903,
132 1.472480, 0.055744, -1.217317, -0.041227, -1.128344, -1.349278, -0.261102, 0.953465,
133 0.128644, 0.656468, -1.167819, -0.460605, -0.262440, -1.213152, -1.319437, 0.931218,
134 0.011245, -0.645146, 0.805729, 0.231626, -0.989760, 1.339586, 0.289502, 1.478917,
135 1.138028, -0.684139, -1.291936, -0.072926, -0.330599, -0.843628, 0.497770, 1.488490,
136 -0.546476, -0.846758, -0.246337, 0.663024, -0.854197, -1.201315, -0.119869, -0.065294,
137 0.485296, -0.595491, -0.149668, -0.434752, -0.079330, 1.535152, -0.606483, -1.347363,
138 0.469383, -0.903567, 0.035880, -0.627531, 0.535398, 0.552884, -0.203690, -2.054325,
139 0.132561, 1.592941, 1.018412, -1.580402, -0.078662, -0.681657, -1.024553, -1.234353,
140 0.288807, -0.429303, 0.055801, -0.367874, -0.464973, 0.370961, 0.728283, 2.112160,
141 -1.357298, -1.022610, 1.037834, -0.389800, -1.381266, 0.315543, 1.553243, 0.707894,
142 1.957385, 0.504542, 1.864529, -0.339812, -1.139779, -0.211123, 1.190245, -1.116209,
143 0.635274, -0.601412, 0.551185, -1.099840, 0.085991, -2.004563, -0.493088, 0.462048,
144 -0.321005, 1.236556, -0.631280, -2.325211, -1.231637, 1.055648, -0.113224, 0.379224,
145 0.944200, -2.120427, -0.644679, -0.704302, -1.018137, -0.182082, 1.521013, -0.038439,
146 1.227448, -0.696205, 0.007524, -0.782893, 0.586939, -0.251207, 0.480136, 0.668155,
147 -0.078321, 0.889173, 2.309287, 0.524639, -0.011787, 0.913141, 0.055941, -1.107070,
148 0.485498, -0.005005, -0.276218, 1.276452, 1.863401, -0.522559, 0.103424, -0.807649,
149 0.680439, -2.364590, 0.990115, 0.218899, 0.261662, 1.213444, -0.274667, -0.133134,
150 -1.270500, -1.663606, -0.703554, 0.280880, -0.541209, -1.333531, 1.072686, -0.712085,
151 -0.011286, -0.000817, -0.249436, 0.396575, -0.264013, -1.664011, -1.028975, 0.243095,
152 -1.256590, -0.347183, -0.941372, -1.174560, -1.021142, -0.401667, 0.173666, -0.116118,
153 1.064119, -0.245386, -1.517539, 0.009734, 0.071373, 0.316536, 0.499826, 1.278084,
154 -0.547816, 0.260808, -0.013177, -0.580264, 2.136308, -0.257617, -1.409528, 1.770101,
155 0.325546, -1.119040, 0.620350, 1.269782, -0.896043, 0.135175, -0.139040, -1.163395,
156 1.183720, -0.015430, 0.536219, -0.716429, -0.655559, 0.314363, 0.106814, 1.848216,
157 -0.275106, 2.212554, 1.508526, -1.945079, -1.680543, -0.573534, -0.185817, 0.008934
174 #if defined (SMOKE_TEST)||(REGRESSION_TEST) 175 static test_config CONFIG[] = {{0, 1, 2, 80, &testCoeffs1_f32, &testInput_f32[0]},
177 {4, 1, 2, 80, &testCoeffs1_f32, &testInput_f32[0]},
178 {4, 1, 3, 80, &testCoeffs1_f32, &testInput_f32[0]},
181 {2, 7, 2, 160, &testCoeffs7_f32[0], &testInput_f32[0]},
182 {64, 7, 2, 5, &testCoeffs7_f32[0], &testInput_f32[0]},
183 {64, 32, 4, 5, &testCoeffs32_f32[0], &testInput_f32[0]},
184 {32, 32, 4, 10, &testCoeffs32_f32[0], &testInput_f32[0]}
186 #define NUM_TESTS (sizeof(CONFIG) / sizeof(CONFIG[0]) ) 188 #ifdef PERFORMANCE_TEST 189 static test_config CONFIG_PERF[] =
191 {64, 7, 2, 5, &testCoeffs7_f32[0], &testInput_f32[0]},
192 {64, 32, 4, 5, &testCoeffs32_f32[0], &testInput_f32[0]},
193 {32, 32, 4, 10, &testCoeffs32_f32[0], &testInput_f32[0]}
195 #define NUM_PERF_TESTS (sizeof(CONFIG_PERF) / sizeof(CONFIG_PERF[0]) ) 198 void test_fir_decimate_case0()
211 fprintf (stdout,
"----------%30s start\n", __FUNCTION__);
225 #ifdef ENABLE_NE10_FIR_DECIMATE_FLOAT_NEON 226 #if defined (SMOKE_TEST)||(REGRESSION_TEST) 228 for (loop = 0; loop < NUM_TESTS; loop++)
230 config = &CONFIG[loop];
231 length = config->numFrames * config->blockSize / config->D;
241 fprintf (stdout,
"length of input data is wrong!\n");
246 fprintf (stdout,
"initialization error\n");
252 in_c[i] = testInput_f32[i];
253 in_neon[i] = testInput_f32[i];
256 GUARD_ARRAY (out_c, length);
257 GUARD_ARRAY (out_neon, length);
259 for (block = 0; block < config->numFrames; block++)
261 ne10_fir_decimate_float_c (&SC, in_c + (block * config->blockSize), out_c + (block * config->blockSize / config->D), config->blockSize);
264 for (block = 0; block < config->numFrames; block++)
266 ne10_fir_decimate_float_neon (&SN, in_neon + (block * config->blockSize), out_neon + (block * config->blockSize / config->D), config->blockSize);
269 assert_true (CHECK_ARRAY_GUARD (out_c, length));
270 assert_true (CHECK_ARRAY_GUARD (out_neon, length));
273 snr = CAL_SNR_FLOAT32 (out_c, out_neon, length);
274 assert_false ( (snr < SNR_THRESHOLD));
277 #if defined (DEBUG_TRACE) 278 printf (
"--------------------config %d\n", loop);
279 printf (
"snr %f\n", snr);
281 for (pos = 0; pos < length; pos++)
283 #if defined (DEBUG_TRACE) 284 printf (
"pos %d \n", pos);
285 printf (
"c %f (0x%04X) neon %f (0x%04X)\n", out_c[pos], * (
ne10_uint32_t*) &out_c[pos], out_neon[pos], * (
ne10_uint32_t*) &out_neon[pos]);
287 assert_float_vec_equal (&out_c[pos], &out_neon[pos], ERROR_MARGIN_LARGE, 1);
291 #endif // ENABLE_NE10_FIR_DECIMATE_FLOAT_NEON 293 #ifdef PERFORMANCE_TEST 295 fprintf (stdout,
"%25s%20s%20s%20s%20s\n",
"FIR Length&Taps",
"C Time (micro-s)",
"NEON Time (micro-s)",
"Time Savings",
"Performance Ratio");
296 for (loop = 0; loop < NUM_PERF_TESTS; loop++)
298 config = &CONFIG_PERF[loop];
299 length = config->numFrames * config->blockSize / config->D;
309 fprintf (stdout,
"length of input data is wrong!\n");
314 fprintf (stdout,
"initialization error\n");
321 in_c[i] = testInput_f32[i];
322 in_neon[i] = testInput_f32[i];
331 for (block = 0; block < config->numFrames; block++)
333 ne10_fir_decimate_float_c (&SC, in_c + (block * config->blockSize), out_c + (block * config->blockSize / config->D), config->blockSize);
338 #ifdef ENABLE_NE10_FIR_DECIMATE_FLOAT_NEON 345 for (block = 0; block < config->numFrames; block++)
347 ne10_fir_decimate_float_neon (&SN, in_neon + (block * config->blockSize), out_neon + (block * config->blockSize / config->D), config->blockSize);
352 #endif // ENABLE_NE10_FIR_DECIMATE_FLOAT_NEON 355 time_savings = ( ( (
ne10_float32_t) (time_c - time_neon)) / time_c) * 100;
356 ne10_log (__FUNCTION__,
"%20d,%4d%20lld%20lld%19.2f%%%18.2f:1\n", config->numTaps, time_c, time_neon, time_savings, time_speedup);
361 free (guarded_in_neon);
362 free (guarded_out_c);
363 free (guarded_out_neon);
364 free (guarded_fir_state_c);
365 free (guarded_fir_state_neon);
366 fprintf (stdout,
"----------%30s end\n", __FUNCTION__);
369 void test_fir_decimate()
371 test_fir_decimate_case0();
376 ne10_log_buffer_ptr = ne10_log_buffer;
379 void test_fixture_fir_decimate (
void)
381 test_fixture_start();
385 run_test (test_fir_decimate);
ne10_uint16_t numTaps
Length of the filter.
void ne10_fir_decimate_float_neon(const ne10_fir_decimate_instance_f32_t *S, ne10_float32_t *pSrc, ne10_float32_t *pDst, ne10_uint32_t blockSize) asm("ne10_fir_decimate_float_neon")
Specific implementation of ne10_fir_decimate_float using NEON SIMD capabilities.
void ne10_fir_decimate_float_c(const ne10_fir_decimate_instance_f32_t *S, ne10_float32_t *pSrc, ne10_float32_t *pDst, ne10_uint32_t blockSize)
Specific implementation of ne10_fir_decimate_float using plain C.
ne10_result_t ne10_fir_decimate_init_float(ne10_fir_decimate_instance_f32_t *S, ne10_uint16_t numTaps, ne10_uint8_t M, ne10_float32_t *pCoeffs, ne10_float32_t *pState, ne10_uint32_t blockSize)
Initialization function for the floating-point FIR decimator.
#define TEST_LENGTH_SAMPLES
Instance structure for the floating-point FIR Decimation.