Discussion related to cado-nfs

[Emmanuel Thomé <emmanuel.thome@gmail.com>]

Hi,

Just to complement Pierrick's answer. Revamping lingen is a
significant amount of work. But as Pierrick just said, this annoys us
very much, so let's hope that we do get time to work on it.

Meanwhile, we managed to fix an outstanding bug with an openmp version
of lingen, so you should observe some speedup with the git version.

E.

On Mon, Oct 27, 2014 at 5:36 PM, Pierrick Gaudry
<pierrick.gaudry@loria.fr> wrote:
> Hi Pierpaolo!
>
> Very good question, indeed!
>
> But easy answer: nobody implemented a parallel version yet :-(
>
> In fact, there are many issues with the lingen binary that was written
> ages ago... This is unfortunately not a 5-minute change.
>
> This actually bothers us very much (maybe more than it bothers you) and
> is likely to be fixed in the next few months.
>
> Regards,
> Pierrick
>
> On Mon, Oct 27, 2014 at 05:28:22PM +0100, Pierpaolo Santucci wrote:
>> Hello everyone!
>>
>> Using CADO-NFS 2.1.1 linear algebra without MPI with the command:
>>
>> ./bwc.pl :complete thr=8 mn=64 nullspace=left interval=100
>> matrix=/tmp/matrices/100.bin wdir=/tmp/wdir_8_2.1.1/100 interleaving=0
>> shuffled_product=1
>>
>> the lingen phase use only one CPU. It seems as if lingen was sequential.
>>
>> Why?
>>
>> Thanking you in advance.
>> Regards,
>>
>> Pierpaolo Santucci
>> _______________________________________________
>> Cado-nfs-discuss mailing list
>> Cado-nfs-discuss@lists.gforge.inria.fr
>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/cado-nfs-discuss
> _______________________________________________
> Cado-nfs-discuss mailing list
> Cado-nfs-discuss@lists.gforge.inria.fr
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/cado-nfs-discuss