bodr...@mail.dm.unipi.it writes:
You mean the BPSW primality test?
I wasn't aware of this BPSW suggestion.
What I am suggesting is similar, but I am not familiar with the Lucas
test. I also think one should do more trial dividing, and let its
limits be operand dependent.
Some months ago,
Ciao,
Il Gio, 14 Giugno 2012 2:41 pm, Torbjorn Granlund ha scritto:
bodr...@mail.dm.unipi.it writes:
I wrote a kind of implementation for the Lucas-Selfridge test, it
Cool. I suppose I need to read more about the maths to fully admire
your work. :-)
I plainly implemented the formulas
Torbjorn Granlund t...@gmplib.org writes:
I agree this method could make a faster SB mpn_tdiv_r, but I am
unconvinced that you can get the quotient with fewer operatons than the
current code.
The method specialized for div_qr_1 did beat the udiv_qrnnd_preinv loop
(at least when comparing the
Ciao!
1) mpz/generic/sizeinbase.c is unused.
Should we remove it? Or, at least, write it as:
size_t
mpn_sizeinbase (mp_srcptr xp, mp_size_t xsize, int base)
{
size_t result;
MPN_SIZEINBASE (result, xp, xsize, base);
return result;
}
2) gmp-impl.h defines the unused macro
bodr...@mail.dm.unipi.it writes:
1) mpz/generic/sizeinbase.c is unused.
s/mpz/mpn/, right?
Should we remove it? Or, at least, write it as:
size_t
mpn_sizeinbase (mp_srcptr xp, mp_size_t xsize, int base)
{
size_t result;
MPN_SIZEINBASE (result, xp, xsize, base);
return result;
}
bodr...@mail.dm.unipi.it writes:
About documentation... shall we explicitly document a supported way to
obtain an mpn form an mpz, and viceversa? The following should always
work, right?
/* {np, nn} - |z| */
mpz_export (np, nn, -1, sizeof(mp_limb_t), 0, sizeof(mp_limb_t) * 8 -