... I can investigate if nobody else is able to (since I wrote the Makefile for ppc), but I won't be able to get to it for a week or so. I've been very busy. ...
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 :...
ggnfs@yahoogroups.com
May 2, 2005 10:55 am
1061
Hi Tom. ... I would guess that the memalign() function does not exist under FreeBSD or requires some header file or library that GGNFS is not specifying. One...
... Ah, good to know. I did some Googling and it appears memalign() does not exist under FreeBSD, and most implementations fake it by making it a stub for...
... If you mean papers, here are some links: For block Lanczos, http://www.crypto-world.com/FactorPapers.html has Montgomery's paper in PostScript format under...
... Whew. :) ... Looks like the types ulong and ushort need to be defined for BSD. You can edit src/lasieve4/asm/lasieve-asm.h and modify the first line of #if...
... Ok. I guess we'll have to put in an #if !defined(__FreeBSD__) && !defined(__OpenBSD__) && !defined(__NetBSD__)/#endif around those #include <malloc.h>...
Hi Raghu, As I recall, the obvious idea works rather well: Suppose the vectors in question are u and v. (1) Compute the projection of u onto v, say: P(u:v) =...
Hello again, I just realized that my old junkie lattice sieving code is still in GGNFS - although I stopped long before I started making decent progress on the...
Hello, All! I've faced following bug when trying to run procrels on some collectes relations: i build factorbase with lprb=lpra=30 (same setting was used for...
... I can't say that I know what is wrong, but lpbX=30 is *Huge*! For the 202 digit SNFS 100^99+99^100 we can easily get by using lpbX=28. I would think you...
... Hi! In def-par.txt I followed the hints from Greg and limit lvpa/lbpr to 27. The result is excellent. I seldom now need more than two passes with reduction...
It might be a integer overflow, or atleast it looks that way without looking at the code. you might have to check to see if it is unsigned instead of signed. ...
Is there a way to parallelize GGNFS? A friend an I are using it for a school project, and we thought we saw somewhere that there was a way to do it, but we...
Hello, fpgiv! ... Well, running distributed factorisation with ggnfs is not very complicated. (1) Polynomial selection is easy to parallelize: use pol51* tools...
Hello All! Some more details are become available for makefb/procrels problem. Anton traced bug to completePartialRelFact() function (in rels.c). It returns...
Hi Andrei, This seems to be another facet of the known signed/unsigned issues in the code. The -191 occurrs when it tries to verify the factorization of a...
Hello all, Brian wanted me to find out how many people are using VC to compile GGNFS on a regular basis. Each time there are significant changes to the code,...
Hello, ggnfsmonico! Thanks a lot for your answer! ... It seems that I've found cause for this error. It occurs because 'a' member of relation_t structure is...
Hello Chris, Thursday, May 5, 2005, 6:38:28 PM, you wrote: g> will probably still be a while before I get to that. My suggestion in g> the meantime is just to...
Hello Anton! g>> will probably still be a while before I get to that. My suggestion in g>> the meantime is just to drop the LPBA/LPBR values to 29 and g>>...
... Did you add -D__ppc__ to the compile line? This would not be automatically defined in linux. jasonp ... This message was sent using BOO.net's Webmail. ...
Jason Papadopulos
jasonp@...
May 7, 2005 2:15 pm
1088
... I think you have an incompatibility. -march=k8 is used for Pentium processors and __ppc__ is for PowerPC processors. If you are building on a PPC, you...
I have updated the source to build on a PPC G5 with OS X 10.4. I have not tested it yet. You won't need the entirety of the source (which I have zipped)....
I got the following output from GGNFS checkMat() did not like something about the matrix: This is probably a sign that something has gone horribly wrong in the...
It is nice to see that you have solved your problem. There must be another solution that will allow use to use the ASM code as well. Unfortunately I do not...
Having problems with message search?
Fill out this form to ensure your group is one of the first to be migrated to the new message search system.
