zeropain reports
Colin (478) 2433 posts |
One for BASIC or is it Alarm or maybe ShareFS with a smidgen of SharedCLibrary. My first non NULL pointer dereference.
|
Dave Higton (1515) 3526 posts |
It shows nicely a serious deficiency of ZeroPain: it can’t tell us what made the call into SharedCLibrary. Without that information, how can we find the real source of the problem? Is there any way that it could be able to tell us? Presumably it would mean a lot of stack unwinding. |
Steve Pampling (1551) 8170 posts |
I think during earlier discussion the consensus was that two things were at fault in this sort of case: Would I be wrong in thinking that having SharedCLib refuse to accept the rubbish would make it blindingly obvious which offender was sending it by causing the offender to simply not work? |
Jon Abbott (1421) 2651 posts |
Input validation would certainly help in locating many bugs that are currently hidden from sight. It’s not going to catch all of these issues though. |
Steve Pampling (1551) 8170 posts |
Never thought it would, but my fault elimination technique in work tends to work on fixing all visible faults and then looking to see what is revealed on the next layer. Remove a few trees and you get to see the wood… |