Re: 64bit HPUX binary buggy

[Igor Schein <igor@txc.com>]

On Wed, Jun 12, 2002 at 05:48:07PM +0200, Karim BELABAS wrote:
> On Wed, 12 Jun 2002, Igor Schein wrote:
> 
> > 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.
> [...]
> > 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
> [...]
> > -adding vector = 2155177552
> > +adding vector = -4611686016274382136
> 
> Argh. Both are ridiculous. Had forgotten about this dreadful hack (using a

OK, I'm including the complete log of bnfinit(x^2+1) at \g9 on HPUX.
It's identical compared to that on Solaris 64bit, except that on
Solaris it goes on after outputing

#### Looking for 12 relations (small norms)

while on HPUX it stops.

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
   debug = 9
     1 factor of degree   1
...tried prime   2 (1   factor). Time = 0
     1 factor of degree   2
...tried prime   3 (1   factor). Time = 0
bound for resultant: 2^3
Time resultant mod 27457 (bound 2^14, stable = 0): 0
Time resultant mod 27479 (bound 2^29, stable = 1): 0
Time disc. factorisation: 0
Treating p^k = 2^2
  dedek: gcd has degree 0
initial parameters p=2,
  f=x^2 + 1
  entering Dedekind Basis with parameters p=2
  f = x^2 + 1,
  alpha = x
  new order: [2, 0; 0, 2]
Result for prime 2 is:
[1, 0; 0, 1]
Time round4: 0
Time matrix M: 0
get_red_G: starting LLL, prec = 4 (4 + 0)
k = K2
Time LLL basis: 0
Time matrix M: 0
Time inverse mod 27449 (stable=0): 0
Time inverse mod 27457 (stable=1): 0
Time ZM_inv done: 0
Time mult. table: 0
Time inverse mod 27449 (stable=0): 0
Time inverse mod 27457 (stable=1): 0
Time ZM_inv done: 0
Time matrices: 0
entering fincke_pohst
Entering gauss with inexact=1
Solving the triangular system
final LLL: prec = 5
k = K2
Time inverse mod 27449 (stable=0): 0
Time inverse mod 27457 (stable=1): 0
Time ZM_inv done: 0
entering smallvectors
smallvectors looking for norm <= 2
q = [1.99999999999999999999999999999999999999999999999999999999, 0.E-58; 0, 1.99999999999999999999999999999999999999999999999999999999]
x[2] = 1
1 2 
x[2] = 0
1 1 leaving fincke_pohst
Time initalg & rootsof1: 10
N = 2, R1 = 0, R2 = 1, RU = 1
D = -4

*** Bach constant: 0.300000
LIMC = 20, LIMC2 = 20
 2Time factmod: 0
Time simple primedec: 0
 3Time frobenius: 0
Time kernel: 0
Time factmod: 0
Time simple primedec: 0
 5Time frobenius: 0
Time kernel: 0
Time factmod: 0
bound for resultant: 2^3
Time resultant mod 27457 (bound 2^14, stable = 0): 0
Time resultant mod 27479 (bound 2^29, stable = 1): 0
bound for resultant: 2^3
Time resultant mod 27457 (bound 2^14, stable = 0): 0
Time resultant mod 27479 (bound 2^29, stable = 1): 0
Time simple primedec: 0
 7Time frobenius: 0
Time kernel: 0
Time factmod: 0
Time simple primedec: 0
 11Time frobenius: 0
Time kernel: 0
Time factmod: 0
Time simple primedec: 0
 13Time frobenius: 0
Time kernel: 0
Time factmod: 0
bound for resultant: 2^6
Time resultant mod 27457 (bound 2^14, stable = 0): 0
Time resultant mod 27479 (bound 2^29, stable = 1): 0
bound for resultant: 2^6
Time resultant mod 27457 (bound 2^14, stable = 0): 0
Time resultant mod 27479 (bound 2^29, stable = 1): 0
Time simple primedec: 0
 17Time frobenius: 0
Time kernel: 0
Time factmod: 0
bound for resultant: 2^5
Time resultant mod 27457 (bound 2^14, stable = 0): 0
Time resultant mod 27479 (bound 2^29, stable = 1): 0
bound for resultant: 2^5
Time resultant mod 27457 (bound 2^14, stable = 0): 0
Time resultant mod 27479 (bound 2^29, stable = 1): 0
Time simple primedec: 0
 19Time frobenius: 0
Time kernel: 0
Time factmod: 0
Time simple primedec: 0

########## FACTORBASE ##########

KC2=7, KC=7, KCZ=4, KCZ2=4, MAXRELSUP=28
++ idealbase[1] = [[2, [1, 1]~, 2, 1, [1, 1]~]]++ idealbase[2] = [[5, [-2, 1]~, 1, 1, [2, 1]~], [5, [2, 1]~, 1, 1, [-2, 1]~]]++ idealbase[3] = [[13, [-5, 1]~, 1, 1, [5, 1]~], [13, [5, 1]~, 1, 1, [-5, 1]~]]++ idealbase[4] = [[17, [-4, 1]~, 1, 1, [4, 1]~], [17, [4, 1]~, 1, 1, [-4, 1]~]]Time factor base: 0

***** IDEALS IN FACTORBASE *****

no 1 = [2, [1, 1]~, 2, 1, [1, 1]~]
no 2 = [5, [-2, 1]~, 1, 1, [2, 1]~]
no 3 = [5, [2, 1]~, 1, 1, [-2, 1]~]
no 4 = [13, [-5, 1]~, 1, 1, [5, 1]~]
no 5 = [13, [5, 1]~, 1, 1, [-5, 1]~]
no 6 = [17, [-4, 1]~, 1, 1, [4, 1]~]
no 7 = [17, [4, 1]~, 1, 1, [-4, 1]~]

***** IDEALS IN SUB FACTORBASE *****

[[5, [-2, 1]~, 1, 1, [2, 1]~], [13, [-5, 1]~, 1, 1, [5, 1]~], [17, [-4, 1]~, 1, 1, [4, 1]~]]~

***** INITIAL PERMUTATION *****

vperm = Vecsmall([2, 4, 6, 1, 3, 5, 7])

Time sub factorbase (3 elements): 0
relsup = 5, ss = 4, KCZ = 4, KC = 7, KCCO = 12
After trivial relations, cglob = 4
adding vector = Vecsmall([2, 0, 0, 0, 0, 0])
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 = Vecsmall([0, 1, 1, 0, 0, 0])
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 = Vecsmall([0, 0, 0, 1, 1, 0])
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 = Vecsmall([0, 0, 0, 0, 0, 1])
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()
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\