Discussion related to cado-nfs

[Pierrick Gaudry <pierrick.gaudry@loria.fr>]

Hello,

I could reproduce your problem.

There are indeed bugs (in at least 2 places) that are triggered by your
example. One in freerel that produces wrong relations, and one in
filter_galois that crashes. In both cases, I suspect that the local
behavior of x^4+4*x^2+1 at p=3 is not handled correctly.

This won't be fixed right now, but will probably be this summer.

For now, there is a workaround:
- disable Galois (you already did that);
- disable free relations by adding the following in the parameter file
    tasks.sieve.freerel.pmax=1

With this, I could have the Logarithms Reconstruction finish without
error.

Regards,
Pierrick


On Sat, May 18, 2019 at 08:22:18PM +0800, Xs. X. wrote:
> Hello everyone,
> 
> I'm solving a DLP problem in GF(p^2) with the polynomial built by myself,
> and the program failed at reconstructlog-dl.
> 
> The command is:
> 
> ./cado-nfs.py -dlp 170141183460469231731687303715884123283 -ell
> 42535295865117307932921825928971030821 -gfpext 2
> 
> The output is:
> 
> Info:Logarithms Reconstruction: Starting
> Warning:Command: Process with PID 6202 finished with return code 1
> Error:Logarithms Reconstruction: Program run on server failed with exit
> code 1
> Error:Logarithms Reconstruction: Command line was:
> /home/xz/git/cado-nfs/build/W/filter/reconstructlog-dl -ell
> 42535295865117307932921825928971030821 -mt 4 -log
> /tmp/cado.l7xi344p/p2dd40.bwc/K.sols0-1.0.txt -out
> /tmp/cado.l7xi344p/p2dd40.dlog -renumber
> /tmp/cado.l7xi344p/p2dd40.renumber.gz -poly /tmp/cado.l7xi344p/p2dd40.poly
> -purged /tmp/cado.l7xi344p/p2dd40.purged.gz -ideals
> /tmp/cado.l7xi344p/p2dd40.ideal -relsdel
> /tmp/cado.l7xi344p/p2dd40.relsdel.gz -nrels 448640 -nsm '1,0' >
> /tmp/cado.l7xi344p/p2dd40.reconstructlog-dl.stdout.1 2>
> /tmp/cado.l7xi344p/p2dd40.reconstructlog-dl.stderr.1
> Error:Logarithms Reconstruction: Stderr output (last 10 lines only) follow
> (stored in file /tmp/cado.l7xi344p/p2dd40.reconstructlog-dl.stderr.1):
> Error:Logarithms Reconstruction:        antebuffer set to
> /home/xz/git/cado-nfs/build/W/utils/antebuffer
> Error:Logarithms Reconstruction:        Warning, log is zero for h = a
> Error:Logarithms Reconstruction:        Warning, log is zero for h = 11
> Error:Logarithms Reconstruction:        Warning, log is zero for h = 36
> Error:Logarithms Reconstruction:        Warning, log is zero for h = 373
> Error:Logarithms Reconstruction:        Warning, log is zero for h = 1000
> Error:Logarithms Reconstruction:        Error, no unknown log in rel 107479
> and sum of log is not zero, sum is: 24350889115762306525772643466974011408
> 
> 
> It seems that there is a bug in DLP for GF(p^2) like bug#21707.
> The git version is efbd21713
> The polynomial I build is:
> # poly0 = x^4+4*x^2+1
> # poly1 = 7575543440937986073*x^2+7440663049042884025*x+7575543440937986073
> (By the way, I build the polynomial with the Conjugation method. I hope I'm
> not misunderstanding this method)
> 
> Regards
> 
> Chen Yongyan



> _______________________________________________
> Cado-nfs-discuss mailing list
> Cado-nfs-discuss@lists.gforge.inria.fr
> https://lists.gforge.inria.fr/mailman/listinfo/cado-nfs-discuss