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
|
/* This Source Code Form is subject to the terms of the Mozilla Public
* License, v. 2.0. If a copy of the MPL was not distributed with this
* file, You can obtain one at http://mozilla.org/MPL/2.0/. */
/* The functions that are to be called from outside of the .s file have the
* following interfaces and array size requirements:
*/
void conv_i32_to_d32(double *d32, unsigned int *i32, int len);
/* Converts an array of int's to an array of doubles, so that each double
* corresponds to an int. len is the number of items converted.
* Does not allocate the output array.
* The pointers d32 and i32 should point to arrays of size at least len
* (doubles and unsigned ints, respectively)
*/
void conv_i32_to_d16(double *d16, unsigned int *i32, int len);
/* Converts an array of int's to an array of doubles so that each element
* of the int array is converted to a pair of doubles, the first one
* corresponding to the lower (least significant) 16 bits of the int and
* the second one corresponding to the upper (most significant) 16 bits of
* the 32-bit int. len is the number of ints converted.
* Does not allocate the output array.
* The pointer d16 should point to an array of doubles of size at least
* 2*len and i32 should point an array of ints of size at least len
*/
void conv_i32_to_d32_and_d16(double *d32, double *d16,
unsigned int *i32, int len);
/* Does the above two conversions together, it is much faster than doing
* both of those in succession
*/
void mont_mulf_noconv(unsigned int *result,
double *dm1, double *dm2, double *dt,
double *dn, unsigned int *nint,
int nlen, double dn0);
/* Does the Montgomery multiplication of the numbers stored in the arrays
* pointed to by dm1 and dm2, writing the result to the array pointed to by
* result. It uses the array pointed to by dt as a temporary work area.
* nint should point to the modulus in the array-of-integers representation,
* dn should point to its array-of-doubles as obtained as a result of the
* function call conv_i32_to_d32(dn, nint, nlen);
* nlen is the length of the array containing the modulus.
* The representation used for dm1 is the one that is a result of the function
* call conv_i32_to_d32(dm1, m1, nlen), the representation for dm2 is the
* result of the function call conv_i32_to_d16(dm2, m2, nlen).
* Note that m1 and m2 should both be of length nlen, so they should be
* padded with 0's if necessary before the conversion. The result comes in
* this form (int representation, padded with 0's).
* dn0 is the value of the 16 least significant bits of n0'.
* The function does not allocate memory for any of the arrays, so the
* pointers should point to arrays with the following minimal sizes:
* result - nlen+1
* dm1 - nlen
* dm2 - 2*nlen+1 ( the +1 is necessary for technical reasons )
* dt - 4*nlen+2
* dn - nlen
* nint - nlen
* No two arrays should point to overlapping areas of memory.
*/
|
