18 Apr
2003
18 Apr
'03
11:47 p.m.
On Fri, 2003-04-18 at 08:48, Martijn Sipkema wrote:
[...]
Well whatever the problem was, it is not happening anymore..
I really don't know what the problem is and to be honest, I don't think
anyone else on the list could tell me unless they were here to see the
problem happen and analyze it when it happens.
It may have been something I did in my code. I was building a somewhat
complicated memory management system for LADSPA plugins and the problem
started happening.
Once I fixed the code where it worked properly the problem went away.
I just did not manage to find out what exactly was causing the problem
or why.
I have no idea but I am pretty sure I did something with unitialized
memory, and than later fixed it without realizing it . I am pretty sure
I it was a pointer problem.
I can not run gdb on my program, it won't work. The code is too hard to
debug by just looking at it. The only way I can debug is to put hundreds
of printf() into the code. That was how I found log( blah ) was making
NAN as well. I have learned to live without a debugger, but it sucks,
because I really need it.
You always need a virtual destructor in the class
if you do inhertance.
Otherwise upon destruction of the derived class, the baseclass destructor
will not be called.
I doubt that is the problem. You should use virtual destructors when
deleting using a pointer so the correct destructor is called but when
you know the pointer is of the correct type or you allocate the instance
on the stack then the correct destructor will be called and I think the
baseclass destructor(s) will be called also even if they are not
virtual.
--ms