Add macro definitions for various function/variable attributes; aligned, inlining..

Cleaning of test application

2 files changed, 209 insertions, 117 deletions

diff --git a/src/ffts_attributes.h b/src/ffts_attributes.h
new file mode 100644
index 0000000..6ac2ac3
--- /dev/null
+++ b/src/ffts_attributes.h
@@ -0,0 +1,77 @@
+/*
+
+ This file is part of FFTS -- The Fastest Fourier Transform in the South
+
+ Copyright (c) 2012, Anthony M. Blake <amb@anthonix.com>
+ Copyright (c) 2012, The University of Waikato
+
+ All rights reserved.
+
+ Redistribution and use in source and binary forms, with or without
+ modification, are permitted provided that the following conditions are met:
+ 	* Redistributions of source code must retain the above copyright
+ 		notice, this list of conditions and the following disclaimer.
+ 	* Redistributions in binary form must reproduce the above copyright
+ 		notice, this list of conditions and the following disclaimer in the
+ 		documentation and/or other materials provided with the distribution.
+ 	* Neither the name of the organization nor the
+	  names of its contributors may be used to endorse or promote products
+ 		derived from this software without specific prior written permission.
+
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
+ ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
+ WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ DISCLAIMED. IN NO EVENT SHALL ANTHONY M. BLAKE BE LIABLE FOR ANY
+ DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
+ ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+*/
+
+#ifndef FFTS_ATTRIBUTES_H
+#define FFTS_ATTRIBUTES_H
+
+#if defined (_MSC_VER) && (_MSC_VER >= 1020)
+#pragma once
+#endif
+
+/* Macro definitions for various function/variable attributes */
+#ifdef __GNUC__
+#define GCC_VERSION_AT_LEAST(x,y) \
+	(__GNUC__ > x || __GNUC__ == x && __GNUC_MINOR__ >= y)
+#else
+#define GCC_VERSION_AT_LEAST(x,y) 0
+#endif
+
+#ifdef __GNUC__
+#define FFTS_ALIGN(x) __attribute__((aligned(x)))
+#elif defined(_MSC_VER)
+#define FFTS_ALIGN(x) __declspec(align(x))
+#else
+#define FFTS_ALIGN(x)
+#endif
+
+#if GCC_VERSION_AT_LEAST(3,1)
+#define FFTS_ALWAYS_INLINE __attribute__((always_inline)) inline
+#elif defined(_MSC_VER)
+#define FFTS_ALWAYS_INLINE __forceinline
+#else
+#define FFTS_ALWAYS_INLINE inline
+#endif
+
+#if defined(_MSC_VER)
+#define FFTS_INLINE __inline
+#else
+#define FFTS_INLINE inline
+#endif
+
+#if defined(_MSC_VER)
+#define FFTS_RESTRICT
+#else
+#define FFTS_RESTRICT __restrict
+#endif
+
+#endif /* FFTS_ATTRIBUTES_H */
diff --git a/tests/test.c b/tests/test.c
index 7ab79c6..b1e5509 100644
--- a/tests/test.c
+++ b/tests/test.c
@@ -1,7 +1,7 @@
 /*
- 
- This file is part of SFFT.
-  
+
+ This file is part of FFTS.
+
  Copyright (c) 2012, Anthony M. Blake
  All rights reserved.
 
@@ -29,149 +29,164 @@
 
 */
 
-#include <stdio.h>
-#include <math.h>
+#include "../include/ffts.h"
+#include "../src/ffts_attributes.h"
 
 #ifdef __ARM_NEON__
 #endif
-#ifdef HAVE_SSE
-	#include <xmmintrin.h>
-#endif 
 
-#include "../include/ffts.h"
+#ifdef HAVE_SSE
+#include <xmmintrin.h>
+#endif
 
+#include <stdio.h>
+#include <stdlib.h>
+#include <math.h>
 
-#define PI 3.1415926535897932384626433832795028841971693993751058209
+#ifndef M_PI
+#define M_PI 3.1415926535897932384626433832795028841971693993751058209
+#endif
 
-float impulse_error(int N, int sign, float *data) {
+static float impulse_error(int N, int sign, float *data)
+{
 #ifdef __ANDROID__
-	double delta_sum = 0.0f;
-	double sum = 0.0f;
+    double delta_sum = 0.0f;
+    double sum = 0.0f;
 #else
-	long double delta_sum = 0.0f;
-	long double sum = 0.0f;
-#endif 
+    long double delta_sum = 0.0f;
+    long double sum = 0.0f;
+#endif
+    int i;
 
-	int i;
-	for(i=0;i<N;i++) {
+    for (i = 0; i < N; i++) {
 #ifdef __ANDROID__
-		double re, im;
-		if(sign < 0) {
-			re = cos(2 * PI * (double)i / (double)N); 
-			im = -sin(2 * PI * (double)i / (double)N); 
-		}else{
-			re = cos(2 * PI * (double)i / (double)N); 
-			im = sin(2 * PI * (double)i / (double)N); 
-		}
+        double re, im;
+
+        if (sign < 0) {
+            re = cos(2 * M_PI * (double) i / (double) N);
+            im = -sin(2 * M_PI * (double) i / (double) N);
+        } else {
+            re = cos(2 * M_PI * (double) i / (double) N);
+            im = sin(2 * M_PI * (double) i / (double) N);
+        }
 #else
-		long double re, im;
-		if(sign < 0) {
-			re = cosl(2 * PI * (long double)i / (long double)N); 
-			im = -sinl(2 * PI * (long double)i / (long double)N); 
-		}else{
-			re = cosl(2 * PI * (long double)i / (long double)N); 
-			im = sinl(2 * PI * (long double)i / (long double)N); 
-		}
+        long double re, im;
+
+        if (sign < 0) {
+            re = cosl(2 * M_PI * (long double) i / (long double) N);
+            im = -sinl(2 * M_PI * (long double) i / (long double) N);
+        } else {
+            re = cosl(2 * M_PI * (long double) i / (long double) N);
+            im = sinl(2 * M_PI * (long double) i / (long double) N);
+        }
 #endif
-		sum += re * re + im * im;
 
-		re = re - data[2*i];
-		im = im - data[2*i+1];
-		
-		delta_sum += re * re + im * im;
+        sum += re * re + im * im;
+
+        re = re - data[2*i];
+        im = im - data[2*i+1];
+
+        delta_sum += re * re + im * im;
+    }
 
-	}
 #ifdef __ANDROID__
-	return sqrt(delta_sum) / sqrt(sum);
+    return (float) (sqrt(delta_sum) / sqrt(sum));
 #else
-	return sqrtl(delta_sum) / sqrtl(sum);
+    return (float) (sqrtl(delta_sum) / sqrtl(sum));
 #endif
 }
 
-int 
-test_transform(int n, int sign) {
+int test_transform(int n, int sign)
+{
+    ffts_plan_t *p;
 
-#ifdef HAVE_SSE 
-	float __attribute__ ((aligned(32))) *input = _mm_malloc(2 * n * sizeof(float), 32);
-  float __attribute__ ((aligned(32))) *output = _mm_malloc(2 * n * sizeof(float), 32);
+#ifdef HAVE_SSE
+    float FFTS_ALIGN(32) *input = _mm_malloc(2 * n * sizeof(float), 32);
+    float FFTS_ALIGN(32) *output = _mm_malloc(2 * n * sizeof(float), 32);
 #else
-	float __attribute__ ((aligned(32))) *input = valloc(2 * n * sizeof(float));
-  float __attribute__ ((aligned(32))) *output = valloc(2 * n * sizeof(float));
+    float FFTS_ALIGN(32) *input = valloc(2 * n * sizeof(float));
+    float FFTS_ALIGN(32) *output = valloc(2 * n * sizeof(float));
 #endif
-	int i;	
-	for(i=0;i<n;i++) {
-		input[2*i]   = 0.0f;
-		input[2*i+1] = 0.0f;
-	}
-
-	input[2] = 1.0f;
-
-	ffts_plan_t *p = ffts_init_1d(i, sign);
-	if(p) {
-		ffts_execute(p, input, output);
-		printf(" %3d  | %9d | %10E\n", sign, n, impulse_error(n, sign, output));
-  	ffts_free(p);
-	}else{
-		printf("Plan unsupported\n");
-		return 0;
-	}
-
-	return 1;
+    int i;
+
+    for (i = 0; i < n; i++) {
+        input[2*i + 0] = 0.0f;
+        input[2*i + 1] = 0.0f;
+    }
+
+    input[2] = 1.0f;
+
+    p = ffts_init_1d(i, sign);
+    if (!p) {
+        printf("Plan unsupported\n");
+        return 0;
+    }
+
+    ffts_execute(p, input, output);
+    printf(" %3d  | %9d | %10E\n", sign, n, impulse_error(n, sign, output));
+    ffts_free(p);
+    return 1;
 }
 
-int
-main(int argc, char *argv[]) {
-	
-	if(argc == 3) {
-		// test specific transform with test pattern and display output
-		int n = atoi(argv[1]);
-		int sign = atoi(argv[2]);
+int main(int argc, char *argv[])
+{
+    if (argc == 3) {
+        ffts_plan_t *p;
+        int i;
+
+        /* test specific transform with test pattern and display output */
+        int n = atoi(argv[1]);
+        int sign = atoi(argv[2]);
 
 #ifdef HAVE_SSE
-		float __attribute__ ((aligned(32))) *input = _mm_malloc(2 * n * sizeof(float), 32);
-		float __attribute__ ((aligned(32))) *output = _mm_malloc(2 * n * sizeof(float), 32);
+        float FFTS_ALIGN(32) *input = _mm_malloc(2 * n * sizeof(float), 32);
+        float FFTS_ALIGN(32) *output = _mm_malloc(2 * n * sizeof(float), 32);
 #else
-		float __attribute__ ((aligned(32))) *input = valloc(2 * n * sizeof(float));
-		float __attribute__ ((aligned(32))) *output = valloc(2 * n * sizeof(float));
+        float FFTS_ALIGN(32) *input = valloc(2 * n * sizeof(float));
+        float FFTS_ALIGN(32) *output = valloc(2 * n * sizeof(float));
 #endif
-		int i;	
-		for(i=0;i<n;i++) {
-			input[2*i]   = i;
-			input[2*i+1] = 0.0f;
-		}
-
-	//	input[2] = 1.0f;
-
-		ffts_plan_t *p = ffts_init_1d(i, sign);
-		if(p) {
-			ffts_execute(p, input, output);
-			for(i=0;i<n;i++) printf("%d %d %f %f\n", i, sign, output[2*i], output[2*i+1]);
-			ffts_free(p);
-		}else{
-			printf("Plan unsupported\n");
-			return 0;
-		}
-
-#ifdef HAVE_NEON 
-		_mm_free(input);
-		_mm_free(output);
+
+        for (i = 0; i < n; i++) {
+            input[2*i + 0] = (float) i;
+            input[2*i + 1] = 0.0f;
+        }
+
+        // input[2] = 1.0f;
+
+        p = ffts_init_1d(i, sign);
+        if (!p) {
+            printf("Plan unsupported\n");
+            return 0;
+        }
+
+        ffts_execute(p, input, output);
+
+        for (i = 0; i < n; i++)
+            printf("%d %d %f %f\n", i, sign, output[2*i], output[2*i+1]);
+        ffts_free(p);
+
+#ifdef HAVE_NEON
+        _mm_free(input);
+        _mm_free(output);
 #else
-		free(input);
-		free(output);
+        free(input);
+        free(output);
 #endif
+    } else {
+        int n, power2;
 
-	}else{
-		// test various sizes and display error
-		printf(" Sign |      Size |     L2 Error\n");
-		printf("------+-----------+-------------\n");
-		int n;
-		for(n=1;n<=18;n++) {
-			test_transform(pow(2,n), -1);
-		}
-		for(n=1;n<=18;n++) {
-			test_transform(pow(2,n), 1);
-		}
-	}
-  return 0;
-}
-// vim: set autoindent noexpandtab tabstop=3 shiftwidth=3:
+        /* test various sizes and display error */
+        printf(" Sign |      Size |     L2 Error\n");
+        printf("------+-----------+-------------\n");
+
+        for (n = 1, power2 = 2; n <= 18; n++, power2 <<= 1) {
+            test_transform(power2, -1);
+        }
+
+        for (n = 1, power2 = 2; n <= 18; n++, power2 <<= 1) {
+            test_transform(power2, 1);
+        }
+    }
+
+    return 0;
+}
\ No newline at end of file