|Bill Allombert on Tue, 26 Nov 2002 15:16:56 +0100|
[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]
Dear PARI developers, In a crazy attempt at testing I try to compute bnfinit(polredabs(x^4-9192452294206514*x^2+375140840302653653082195143217)) After 44h, 30mn, 15,550 ms of CPU time on a 1GHZ PIV and 128 Mb os stack I get ++++ cglob = 2134: new relation (need 2135) rel = 2^-5 3^-5 5^-18 9^-1 11^-15 12^-15 13^-1 14^-1 23^-1 24^-1 75^-1 77^-1 ++++ cglob = 2135: new relation (need 2135) rel = 1^-5 2^-13 3^-13 4^-5 5^-2 6^-2 8^-1 11^-6 12^-6 15^-1 18^-1 25^-1 26^-1 59^-1 61^-1 *** the PARI stack overflows ! current stack size: 128.0 Mbytes [hint] you can increase GP stack with allocatemem() Does this computation is outside the reach of PARI or does something go weird ? Well something go weird at least, since it had started to find a few number of relation but increased but never got maximal rank, so either there is a large subestimation of the number of required relation, or this field is catastrophic. I do not remember exactly why and how I build this polynomial, but it was supposed to be very tough for a certain computation, but not bnfinit as far as i can remember. Cheers, Bill.