error handling

Jeroen Demeyer jdemeyer at cage.ugent.be
Tue Nov 11 10:21:01 UTC 2014


On 2014-11-10 20:16, Victor Shoup wrote:
> I guess systems like Sage and others that build ob GMP (and NTL)
> would benefit from better error handling.  Time to move (slowly) into the
> 21st century...
Sage already has some hacks to deal with memory allocation errors (using 
set_memory_functions()). Other error conditions, e.g. overflow in 
mpz_pow_ui(), cannot be dealt with easily on the Sage side.


More information about the gmp-discuss mailing list