Okay, previous error with matrix was, in my opinion, hardware failure - rerunning matsolve gets a result at 100.6% of estimated iterations count. The next...
1617
Bob Backstrom
bobb7772004
Nov 1, 2005 1:36 pm
Anton, This version is a binary-only distribution, so we cannot help debug it. [ We thought it was fixed :-) ] See below for the depnum 30 sqrt attempt: The...
1618
wavesky17
Nov 1, 2005 5:38 pm
... assembler ... 0.77.1 ... value. ... Thanks much! I'm a newbie.I'm not very clear about "ARCH=ev6 make commonm". I must to change the Makefile of the ggnfs?...
1619
Anton Korobeynikov
aslsoft
Nov 1, 2005 8:56 pm
Hello Bob, Tuesday, November 1, 2005, 4:35:07 PM, you wrote: BB> Anton, This version is a binary-only distribution, so we cannot help BB> debug it. [ We...
1620
trilliwig
Nov 1, 2005 10:38 pm
... On the command line in your top-level ggnfs directory, type ARCH=ev6 make common and press Enter. Or you can add a Makefile entry that looks like this: ...
1621
Mark Rodenkirch
mgrogue
Nov 4, 2005 10:13 pm
I finally used ggnfs to factor some numbers (via gnfs) on a PPC G5 at 1.5 GHz. These are from Sander Hoogendoorn's Home Prime search. [10/31 01:39:10]...
1622
Tyler Cadigan
t_cadigan
Nov 5, 2005 12:26 am
Mark, not sure if you bothered to check but congratulations on finding a new end prime for 992! ...
1623
Mark Rodenkirch
mgrogue
Nov 5, 2005 12:58 am
which is now: 128833103449336493596603820110507 * 108810160806387798072497441309848022012254957060031013847521680549237664 915647 (composite) which would be...
1624
Max
relf6
Nov 5, 2005 3:35 am
... No! ULL_NO_UL indicates that '(unsigned) long long' type is different from '(unsigned) long' type. And that is *not* the case for x86_64. I've just...
1625
Max
relf6
Nov 5, 2005 3:44 am
... btw, 'ulong' type is carefully avoided in generic builds. Could you tell me why it is required for ppc/ppc64 ? I don't see it from the sources. ... OK....
1626
Mark Rodenkirch
mgrogue
Nov 5, 2005 5:58 am
... ulong does not appear to be needed anymore, so I think you can get rid of it. GGNFS_BIGENDIAN was the important one to define --Mark...
1627
Mark Rodenkirch
mgrogue
Nov 5, 2005 3:13 pm
It would be nice in a future release to make it easier to configure the output. I have a few examples. In the sievers, it would be nice to use \r instead of...
1628
Tyler Cadigan
t_cadigan
Nov 5, 2005 5:30 pm
which is now: 3120474947812619256123711293686439459 x 34869743428852458306924098704070378276533 which finishes a step. so it becomes: ...
1629
ggnfs@yahoogroups.com
Nov 5, 2005 7:11 pm
Hello, This email message is a notification to let you know that a file has been uploaded to the Files area of the ggnfs group. File :...
1630
Mark Rodenkirch
mgrogue
Nov 5, 2005 7:11 pm
... 1859212913155513398621420379 * 129419315461443556294410443443007372148709704112200650408263286521761020 09000569834552328618007938639131 I'm using msieve...
1631
ggnfs@yahoogroups.com
Nov 5, 2005 7:11 pm
Hello, This email message is a notification to let you know that a file has been uploaded to the Files area of the ggnfs group. File :...
1632
trilliwig
Nov 7, 2005 5:18 pm
... Hm, yes, it might be good to change those printf()s to printTmp()s. And since the information printed is cumulative statistics it makes sense to only show...
1633
Mark Rodenkirch
mgrogue
Nov 7, 2005 11:17 pm
... That's OK. I was just throwing it out there. It would be nice to see which step without all of the details. I can also look into the code to see what is...
1634
wavesky17
Nov 8, 2005 1:37 am
... Thanks! I do that as you say,I add a Makefile entry that looks like ... but still have errors,see below: # cd newggnfs # ls CVS Makefile...
1635
Mark Rodenkirch
mgrogue
Nov 8, 2005 1:52 am
... The C131 for HP 973 is factored as 1859212913155513398621420379 * 255634182504155602875632238625820041237 * ...
1636
Mark Rodenkirch
mgrogue
Nov 8, 2005 1:58 am
I am working on a couple of optimizations to pol51opt.c. Brief testing indicates a 10% improvement on PPC. I am still in the process of testing these...
1637
Max
relf6
Nov 8, 2005 8:27 pm
http://mathworld.wolfram.com/news/2005-11-08/rsa-640/ Way to go!...
1638
trilliwig
Nov 8, 2005 9:31 pm
... Hm, ok, open up src/Makefile and change the first line ALLOPT=-march=$(ARCH) -pipe to ALLOPT=-mcpu=$(ARCH) -pipe It looks like alpha is another of those...
1639
Mark Rodenkirch
mgrogue
Nov 8, 2005 10:10 pm
... Use -mtune instead of -mcpu since -mcpu is deprecated. --Mark...
1640
Max
relf6
Nov 8, 2005 10:38 pm
... ggnfs sources contain both assembler-optimized and generic (pure C/C++) versions. What was suggested to you is compilation of a generic version. The only...
1641
Max
relf6
Nov 8, 2005 10:49 pm
... It seems that the line above is cut at some point (it should continue with "-funroll-loops ..."). Could you please show it completely for better...
1642
trilliwig
Nov 9, 2005 12:25 pm
... -mtune does something different than -mcpu on Alpha. And -mcpu is only deprecated for x86/x86-64. ...
1643
Mark Rodenkirch
mgrogue
Nov 9, 2005 12:59 pm
... -mcpu is also deprecated on PPC. This link is the first time I see an actual difference between -mtune and -mcpu. The man pages in OS X for gcc imply...
1644
tlc4tvm
Nov 9, 2005 5:04 pm
I'm starting to push the limits a little to far for msieve c108, using a dual athlon XPBarton 2.4G with windows. I've downloaded, "ggnfs-0.73.4 Athlon.zip",...
1645
Torbjörn Alm
torbjornalm
Nov 9, 2005 7:39 pm
... First of all, do not try to use that old version,it is still full of bugs. The very best version is the snapshots on www.sourceforge.net. look for ggnfs. ...
