[Octave-bug-tracker] [bug #40812] glpk should avoid integer overflow when Octave uses 64-bit indexing

[John W. Eaton]

URL:
  <http://savannah.gnu.org/bugs/?40812>

                 Summary: glpk should avoid integer overflow when Octave uses
64-bit indexing
                 Project: GNU Octave
            Submitted by: jwe
            Submitted on: Wed 04 Dec 2013 07:56:30 PM GMT
                Category: Octave Function
                Severity: 3 - Normal
                Priority: 5 - Normal
              Item Group: Other
                  Status: None
             Assigned to: None
         Originator Name: jwe
        Originator Email: 
             Open/Closed: Open
         Discussion Lock: Any
                 Release: dev
        Operating System: Any

    _______________________________________________________

Details:

GLPK uses integer variables for problem dimensions, but Octave is storing
octave_idx_type values in those variables.   This can cause trouble if Octave
is compiled with 64-bit indexing and someone passes large arrays to glpk.  The
changes required should be similar to what was done in this changeset

http://hg.savannah.gnu.org/hgweb/octave/rev/ac74b0c4c564

for similar problems with Qhull.

It would probably be best to generalize the "octave_qhull_dims_ok" function
and avoid duplicating it in multiple files.






    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?40812>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/