time-to-botec

Benchmark sampling in different programming languages
Log | Files | Refs | README

drev.c (1853B)


      1 /**
      2 * @license Apache-2.0
      3 *
      4 * Copyright (c) 2020 The Stdlib Authors.
      5 *
      6 * Licensed under the Apache License, Version 2.0 (the "License");
      7 * you may not use this file except in compliance with the License.
      8 * You may obtain a copy of the License at
      9 *
     10 *    http://www.apache.org/licenses/LICENSE-2.0
     11 *
     12 * Unless required by applicable law or agreed to in writing, software
     13 * distributed under the License is distributed on an "AS IS" BASIS,
     14 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
     15 * See the License for the specific language governing permissions and
     16 * limitations under the License.
     17 */
     18 
     19 #include "stdlib/blas/ext/base/drev.h"
     20 #include <stdint.h>
     21 
     22 /**
     23 * Reverses a double-precision floating-point strided array in-place.
     24 *
     25 * @param N       number of indexed elements
     26 * @param X       input array
     27 * @param stride  index increment
     28 */
     29 void c_drev( const int64_t N, double *X, const int64_t stride ) {
     30 	double tmp;
     31 	int64_t ix;
     32 	int64_t iy;
     33 	int64_t m;
     34 	int64_t n;
     35 	int64_t i;
     36 
     37 	if ( N <= 0 ) {
     38 		return;
     39 	}
     40 	n = N / 2;
     41 
     42 	// Use loop unrolling if the stride is equal to `1`...
     43 	if ( stride == 1 ) {
     44 		m = n % 3;
     45 		iy = N - 1;
     46 
     47 		// If we have a remainder, run a clean-up loop...
     48 		if ( m > 0 ) {
     49 			for ( ix = 0; ix < m; ix++ ) {
     50 				tmp = X[ ix ];
     51 				X[ ix ] = X[ iy ];
     52 				X[ iy ] = tmp;
     53 				iy -= 1;
     54 			}
     55 		}
     56 		if ( n < 3 ) {
     57 			return;
     58 		}
     59 		for ( ix = m; ix < n; ix += 3 ) {
     60 			tmp = X[ ix ];
     61 			X[ ix ] = X[ iy ];
     62 			X[ iy ] = tmp;
     63 
     64 			tmp = X[ ix+1 ];
     65 			X[ ix+1 ] = X[ iy-1 ];
     66 			X[ iy-1 ] = tmp;
     67 
     68 			tmp = X[ ix+2 ];
     69 			X[ ix+2 ] = X[ iy-2 ];
     70 			X[ iy-2 ] = tmp;
     71 
     72 			iy -= 3;
     73 		}
     74 		return;
     75 	}
     76 	if ( stride < 0 ) {
     77 		ix = (1-N) * stride;
     78 	} else {
     79 		ix = 0;
     80 	}
     81 	iy = ix + ((N-1)*stride);
     82 	for ( i = 0; i < n; i++ ) {
     83 		tmp = X[ ix ];
     84 		X[ ix ] = X[ iy ];
     85 		X[ iy ] = tmp;
     86 		ix += stride;
     87 		iy -= stride;
     88 	}
     89 	return;
     90 }