Igor Schein on Wed, 12 Jun 2002 11:07:23 -0400 |
[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]
Re: 64bit HPUX binary buggy |
On Wed, Jun 12, 2002 at 04:16:06PM +0200, Karim BELABAS wrote: > On Wed, 12 Jun 2002, Igor Schein wrote: > > > On Wed, Jun 12, 2002 at 02:01:59PM +0200, Karim BELABAS wrote: > > > On Tue, 11 Jun 2002, Igor Schein wrote: > > > > I compiled a 64bit gp binary on HPUX 11, but it's badly broken - > > > > a command as simple as bnfinit(x^2-2) keeps doubling the stack. > > > > I'd like to debug it, but I need some pointers on where to set > > > > breakpoints, what to look for, etc. > [...] > > nfinit(x^2-2) and other simpler commands work fine and yield the same > > result as on other 64bit architectures. It's specifically bnfinit() > > that is completely shut down. All diffs in make bench are related to > > bnfinit() not working. > > Nice... OK, set \g2, run the broken binary, and diff the log with the correct > one (other 64bit). Where do they diverge ? (if possible try \g3, etc. but you > will get an enormous amount of useless information from LLL, try to pinpoint > the precise point where it breaks down). > > Also bnfinit(x^2+1) would not be broken, by any chance ? (much easier to > debug, no units). yes, bnfinit(x^2+1) is broken. I tried it at \g9. Below the relevant diff portion, As you can see, the problem starts on "adding vector" line \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ --- solaris-output Wed Jun 12 10:52:02 2002 +++ hpux-output Wed Jun 12 10:51:40 2002 @@ -22,7 +22,7 @@ get_red_G: starting LLL, prec = 4 (4 + 0) k = K2 Time LLL basis: 0 -Time matrix M: 0 +Time matrix M: 10 Time inverse mod 27449 (stable=0): 0 Time inverse mod 27457 (stable=1): 0 Time ZM_inv done: 0 @@ -124,196 +124,32 @@ vperm = Vecsmall([2, 4, 6, 1, 3, 5, 7]) -Time sub factorbase (3 elements): 10 +Time sub factorbase (3 elements): 0 relsup = 5, ss = 4, KCZ = 4, KC = 7, KCCO = 12 After trivial relations, cglob = 4 -adding vector = 2155177512 +adding vector = -4611686016274382176 vector in new basis = [2, 0, 0, 0, 0, 0, 0]~ list = Vecsmall([0, 0, 0, 0, 0, 0, 0]) base change matrix = [1, 0, 0, 0, 0, 0, 0; 0, 1, 0, 0, 0, 0, 0; 0, 0, 1, 0, 0, 0, 0; 0, 0, 0, 1, 0, 0, 0; 0, 0, 0, 0, 1, 0, 0; 0, 0, 0, 0, 0, 1, 0; 0, 0, 0, 0, 0, 0, 1] -adding vector = 2155177552 +adding vector = -4611686016274382136 vector in new basis = [0, 1, 1, 0, 0, 0, 0]~ list = Vecsmall([1, 0, 0, 0, 0, 0, 0]) base change matrix = [1/2, 0, 0, 0, 0, 0, 0; 0, 1, 0, 0, 0, 0, 0; 0, 0, 1, 0, 0, 0, 0; 0, 0, 0, 1, 0, 0, 0; 0, 0, 0, 0, 1, 0, 0; 0, 0, 0, 0, 0, 1, 0; 0, 0, 0, 0, 0, 0, 1] -adding vector = 2155177592 +adding vector = -4611686016274382096 vector in new basis = [0, 0, 0, 1, 1, 0, 0]~ list = Vecsmall([1, 1, 0, 0, 0, 0, 0]) base change matrix = [1/2, 0, 0, 0, 0, 0, 0; 0, 1, 0, 0, 0, 0, 0; 0, -1, 1, 0, 0, 0, 0; 0, 0, 0, 1, 0, 0, 0; 0, 0, 0, 0, 1, 0, 0; 0, 0, 0, 0, 0, 1, 0; 0, 0, 0, 0, 0, 0, 1] -adding vector = 2155177632 +adding vector = -4611686016274382056 vector in new basis = [0, 0, 0, 0, 0, 1, 1]~ list = Vecsmall([1, 1, 0, 1, 0, 0, 0]) base change matrix = [1/2, 0, 0, 0, 0, 0, 0; 0, 1, 0, 0, 0, 0, 0; 0, -1, 1, 0, 0, 0, 0; 0, 0, 0, 1, 0, 0, 0; 0, 0, 0, -1, 1, 0, 0; 0, 0, 0, 0, 0, 1, 0; 0, 0, 0, 0, 0, 0, 1] #### Looking for 12 relations (small norms) + *** the PARI stack overflows ! + current stack size: 8000000 (7.629 Mbytes) + [hint] you can increase GP stack with allocatemem() -*** Ideal no 7: [17, [4, 1]~, 1, 1, [-4, 1]~] -k = K2 (7) -k = K2 -v[1]=34 v[2]=34 \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\