[sage-support] Sage 9.0 source build hyperelliptic_curves error

[Greg Marks]

Dear Sage Developers,

Compiling sage-9.0 from source on a Linux machine, at the "make
ptestlong" stage I got this error:

sage -t --long --warn-long 42.1 src/sage/schemes/hyperelliptic_curves/hyperelliptic_finite_field.py # 2 doctests failed

Re-running the above command produced this:

File "src/sage/schemes/hyperelliptic_curves/hyperelliptic_finite_field.py", line 481, in sage.schemes.hyperelliptic_curves.hyperelliptic_finite_field.HyperellipticCurve_finite_field.frobenius_polynomial_pari
Failed example:
C.frobenius_polynomial_pari()
Expected:
x^8 + 2*x^7 + 6*x^6 + 9*x^5 + 18*x^4 + 27*x^3 + 54*x^2 + 54*x + 81
Got:
x^8 + 731*x^7 + 6*x^6 - 720*x^5 + 18*x^4 - 2160*x^3 + 54*x^2 + 19737*x + 81
**********************************************************************
File "src/sage/schemes/hyperelliptic_curves/hyperelliptic_finite_field.py", line 561, in sage.schemes.hyperelliptic_curves.hyperelliptic_finite_field.HyperellipticCurve_finite_field.frobenius_polynomial
Failed example:
C.frobenius_polynomial()
Expected:
x^8 + 2*x^7 + 6*x^6 + 9*x^5 + 18*x^4 + 27*x^3 + 54*x^2 + 54*x + 81
Got:
x^8 + 731*x^7 + 6*x^6 - 720*x^5 + 18*x^4 - 2160*x^3 + 54*x^2 + 19737*x + 81
**********************************************************************
2 items had failures:
1 of 32 in sage.schemes.hyperelliptic_curves.hyperelliptic_finite_field.HyperellipticCurve_finite_field.frobenius_polynomial
1 of 26 in sage.schemes.hyperelliptic_curves.hyperelliptic_finite_field.HyperellipticCurve_finite_field.frobenius_polynomial_pari
[381 tests, 2 failures, 48.84 s]

Sincerely,
Greg Marks

------------------------------------------
| Greg Marks |
| Department of Mathematics and Statistics |
| St. Louis University |
| St. Louis, MO 63103-2006 |
| U.S.A. |
| |
| Phone: (314)977-7206 |
| PGP encryption public key ID: 0x53F269E8 |
| Web: https://gmarks.org |
------------------------------------------

[slelievre]

Seems related to

- Sage Trac ticket 28789
  Zeta functions of hyperelliptic curves giving wrong answers
  https://trac.sagemath.org/ticket/28789

[ksk...@gmail.com]

To summarize the ticket, this was a bug in Pari that was fixed by updating from 2.11.1 to 2.11.2. If it is still showing up on some architectures, that may have to go to upstream.

One way to test this would be to directly install Pari 2.11.2 on your machine and try this test directly:

? hyperellcharpoly(Mod(x^10 + x^9 + x^8 + x, 3))

%1 = x^8 + 2*x^7 + 6*x^6 + 9*x^5 + 18*x^4 + 27*x^3 + 54*x^2 + 54*x + 81

Kiran