Primo - Page 3

jasonp · 2012-12-09, 15:16

I think we would all be interested in seeing Marcel's reply!

RichD · 2013-04-22, 19:31

Primo v4.0.3 (18 APR 2013) for Linux is posted here.

yoyo · 2013-04-24, 15:29

Would be nice if a command line version is available ;)
yoyo

bearnol · 2013-04-24, 16:30

I don't know if GMP-ECPP would suit - but I'd be happy for you to try
J

yoyo · 2013-04-27, 07:13

Quote:

Originally Posted by bearnol

I don't know if GMP-ECPP would suit - but I'd be happy for you to try
J

I compiled gmp-ecpp on a Linux 64 bit system, but didn't got it static compiled, libgmpxx.a is missing. Is there a way to get a static version of it for Suse Linux?

How long does it needs to compute a prp certificate for a 500 digit prp?

yoyo

bearnol · 2013-04-27, 10:48

You are correct that it needs libgmpxx.a as well as libgmp.a
If you compile with the flags -lgmpxx -lgmp then this will dynamically link, as you say.
I _thought_ that if you added the flag -static you would get a static build, but this probably requires that both the gmp libs are available in static form somewhere, so maybe one would need to specifically recompile the gmp library to create these (with --enable-cxx in the ./configure step).
However the way I've always got a static build is to pass the full paths to the gmp libs at link time.
On a modern core 500-digit will take (order of magnitude) 1 hour, but it can be variable.
J

yoyo · 2013-04-27, 16:50

For the time being I compiled it with shared libs and -O3. I run it with
./atkin243 -Q -s 1367903930 < in1 > out1
for a 461 digit PRP. But it is now running since 5 hours on a i7?

yoyo

bearnol · 2013-04-27, 16:59

Hmmmmm that does seem a little slow. Are you getting any output at all? For a number that small you might find the (additional) switch -B 2000 will speed things up.
J

Stargate38 · 2013-04-27, 17:10

Marcel STILL hasn't replied to my E-mail!

If he doesn't reply, I'm going to keep sending him mail every 4-6 months UNTIL either I get a reply OR he makes another Windows version!

yoyo · 2013-04-27, 19:47

Quote:

Originally Posted by bearnol

Hmmmmm that does seem a little slow. Are you getting any output at all? For a number that small you might find the (additional) switch -B 2000 will speed things up.
J

Now it finished after ~ 2 hours. Input and output are here https://dl.dropboxusercontent.com/u/6717960/result1.zip.

For which numbers should I use -B 2000?

yoyo

bearnol · 2013-04-27, 20:26

Thanks for that output, which looks correct. (though you can also run with the -q rather than -Q if you want a little more detail)
I would (vaguely) suggest -B 2000 for any number under 1000-digits that otherwise seems to be taking too long
J

2012-12-09, 15:16	#23
jasonp Tribal Bullet Oct 2004 110111001110₂ Posts	I think we would all be interested in seeing Marcel's reply! Last fiddled with by jasonp on 2012-12-09 at 15:17

2013-04-22, 19:31	#24
RichD Sep 2008 Kansas 2²×3²×7×13 Posts	Primo Primo v4.0.3 (18 APR 2013) for Linux is posted here.

2013-04-27, 16:50	#29
yoyo Oct 2006 Berlin, Germany 2·3·101 Posts	gmp-ecpp For the time being I compiled it with shared libs and -O3. I run it with ./atkin243 -Q -s 1367903930 < in1 > out1 for a 461 digit PRP. But it is now running since 5 hours on a i7? yoyo

2013-04-27, 17:10	#31
Stargate38 "Daniel Jackson" May 2011 14285714285714285714 643 Posts	Still no reply, EVEN AFTER 4-5 MONTHS!! Marcel STILL hasn't replied to my E-mail! If he doesn't reply, I'm going to keep sending him mail every 4-6 months UNTIL either I get a reply OR he makes another Windows version!

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Primo Browser?	PawnProver44	Information & Answers	14	2016-04-09 05:49
Primo Verifier...	WraithX	Software	15	2013-09-10 07:24
Fixup Old Primo Certificate?	wblipp	FactorDB	1	2012-05-28 03:16
PRIMO 3.0.7	Cybertronic	Five or Bust - The Dual Sierpinski Problem	17	2009-08-13 20:42
primo question	fivemack	Math	35	2009-04-28 15:03

2013-04-24, 15:29	#25
yoyo Oct 2006 Berlin, Germany 1001011110₂ Posts	Would be nice if a command line version is available ;) yoyo

2013-04-24, 16:30	#26
bearnol Sep 2005 127 Posts	I don't know if GMP-ECPP would suit - but I'd be happy for you to try J

2013-04-27, 10:48	#28
bearnol Sep 2005 127 Posts	You are correct that it needs libgmpxx.a as well as libgmp.a If you compile with the flags -lgmpxx -lgmp then this will dynamically link, as you say. I _thought_ that if you added the flag -static you would get a static build, but this probably requires that both the gmp libs are available in static form somewhere, so maybe one would need to specifically recompile the gmp library to create these (with --enable-cxx in the ./configure step). However the way I've always got a static build is to pass the full paths to the gmp libs at link time. On a modern core 500-digit will take (order of magnitude) 1 hour, but it can be variable. J

2013-04-27, 16:59	#30
bearnol Sep 2005 127 Posts	Hmmmmm that does seem a little slow. Are you getting any output at all? For a number that small you might find the (additional) switch -B 2000 will speed things up. J

2013-04-27, 20:26	#33
bearnol Sep 2005 127 Posts	Thanks for that output, which looks correct. (though you can also run with the -q rather than -Q if you want a little more detail) I would (vaguely) suggest -B 2000 for any number under 1000-digits that otherwise seems to be taking too long J