make check results

Craig P Prescott prescott@phys.ufl.edu
Sun Oct 15 15:32:00 GMT 2000


Hi;

Brian Gough wrote:
> Recommend trying -O0, the effects of -O2 confuse the debugger
> (e.g. the variable 'data' has probably been optimized away, hence the
> apparent NULL pointer).

Nod.  I've rebuilt the cvs version with -O0 on mips-sgi-irix6.5 and
rerun the make check.  This solved the memory fault crashes in the
block/, vector/, and matrix/ tests, which are now all PASSes.

The remaining FAILs are in the miser and plain tests, and are
identical to the FAILs previously reported with -O2.

Cheers,
Craig
---
Craig Prescott                                (352) 846-3145 (Phone)
2011 New Physics Building                     (352) 392-8863 (Fax)
University of Florida, Gainesville, FL 32611  prescott@phys.ufl.edu


More information about the Gsl-discuss mailing list