Changes between Version 45 and Version 51 of Ticket #8335


Ignore:
Timestamp:
06/18/13 14:32:09 (8 years ago)
Author:
jpflori
Comment:

There was some actual bug in the code which triggered the computation of the pseudo-Conway polynomials tree twice in the case where the extension degree was prime. First the way it should, and then using the same arguments as in the case where this degree is not prime which at some point triggered the computation of the power of modular integer with a small modulus and a huge exponent and PARI rants when you do that; just try

Mod(3,5)._pari_()**28172187218728127182718271821982918291829182918291

So the newly uploaded patch makes it so we only build the tree once as we always should have, and at least in Luca's example it prevents PARI rants to get on the screen.

Legend:

Unmodified
Added
Removed
Modified
  • Ticket #8335

    • Property Keywords days49 added
    • Property Reviewers changed from Jean-Pierre Flori to Jean-Pierre Flori, Luca De Feo
  • Ticket #8335 – Description

    v45 v51  
    22221. [attachment:trac_8335-pseudo_conway-5.10.b3.patch]
    23231. [attachment:trac_8335-finite_field_coerce-5.8.b0.patch]
    24 1. [attachment:trac_8335-fixes-5.8.b0.patch]
     241. [attachment:trac_8335-fixes-5.11.b1.patch]