Re: Patch to avoid SegFault in natClass.cc

[Bryce McKinlay <bryce at waitaki dot otago dot ac dot nz>]

Tom Tromey wrote:

> >>>>> "Bryce" == Bryce McKinlay <bryce@waitaki.otago.ac.nz> writes:
>
> Bryce> Ideally, I think the compiler should be ensuring classes are
> Bryce> initialized before calling these routines. This way many redundant
> Bryce> calls can be eliminated.
>
> It seems to me that there's a potential tradeoff between code bloat
> (lots of calls to class initialization in the object code) and
> performance.

I think that any added bloat will be minimal as long as the compiler is
smart about eliminating the redundant ones. For example, it needs to know
that it never needs to initialize the current class, or a superclass of the
current class.

> I think this bug is important enough to fix in 3.0.1.  Bryce, do you
> think it would be acceptable to check in Martin's patch on the branch?
> Do you agree that this is important enough?

Yes, I agree that the patch can go in as an interim fix. It can't break
anything. There is a performance impact here but it is probibly fairly
minimal.

regards

Bryce.