1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
|
#include <stdint.h> // for intptr_t
#include "pbc.h"
#include "pbc_test.h"
/* I've heard that sometimes automatic garbage collection can outperform
* manual collection, so I briefly tried using the Boehm-Demers-Weiser GC
* library. Both GMP and PBC support custom memory allocation routines so
* incorporating the GC library is trivial.
*
* Automatic garbage collection appears to slow this program down a little,
* even if only PBC collects automatically. (The case where PBC collects
* manually but GMP collects automatically cannot be achieved with the GC
* library because PBC objects point at GMP objects.)
*
* Perhaps specially-tailored memory allocation routines could shave off
* some time, but one would have to thoroughly analyze PBC and GMP memory usage
* patterns.
*
* Below is the commented-out code that collects garbage for PBC. Of course,
* if you want to use it you must also tell the build system where to find
* gc.h and to link with the GC library.
*
* Also, you may wish to write similar code for GMP (which I unfortunately
* deleted before thinking that it might be useful for others).
* Note GC_MALLOC_ATOMIC may be used for GMP since the mpz_t type does not
* store pointers in the memory it allocates.
*
* The malloc and realloc functions should exit on failure but I didn't
* bother since I was only seeing if GC could speed up this program.
#include <gc.h>
#include <pbc_utils.h>
void *gc_alloc(size_t size) {
return GC_MALLOC(size);
}
void *gc_realloc(void *ptr, size_t size) {
return GC_REALLOC(ptr, size);
}
void gc_free(void *ptr) {
UNUSED_VAR(ptr);
}
* The following should be the first two statements in main()
GC_INIT();
pbc_set_memory_functions(gc_alloc, gc_realloc, gc_free);
*/
int main(int argc, char **argv) {
pairing_t pairing;
element_t x, y, r, r2;
int i, n;
double t0, t1, ttotal, ttotalpp;
pairing_pp_t pp;
// Cheat for slightly faster times:
// pbc_set_memory_functions(malloc, realloc, free);
pbc_demo_pairing_init(pairing, argc, argv);
element_init_G1(x, pairing);
element_init_G2(y, pairing);
element_init_GT(r, pairing);
element_init_GT(r2, pairing);
n = 10;
ttotal = 0.0;
ttotalpp = 0.0;
for (i=0; i<n; i++) {
element_random(x);
element_random(y);
pairing_pp_init(pp, x, pairing);
t0 = pbc_get_time();
pairing_pp_apply(r, y, pp);
t1 = pbc_get_time();
ttotalpp += t1 - t0;
pairing_pp_clear(pp);
t0 = pbc_get_time();
element_pairing(r2, x, y);
t1 = pbc_get_time();
ttotal += t1 - t0;
element_printf("x = %B\n", x);
element_printf("y = %B\n", y);
element_printf("e(x,y) = %B\n", r);
if (element_cmp(r, r2)) {
printf("BUG!\n");
exit(1);
}
}
printf("average pairing time = %f\n", ttotal / n);
printf("average pairing time (preprocessed) = %f\n", ttotalpp / n);
element_clear(x);
element_clear(y);
element_clear(r);
element_clear(r2);
pairing_clear(pairing);
return 0;
}
|
