Re: [TCLCORE] Making crashes more reliable

[Alexandre F. <ale...@gm...>]

On 2/1/10, Donald G Porter <don...@ni...> wrote:
> Alexandre Ferrieux wrote:
>
> >  (3) What about obj->typePtr itself ?
> >
>
>  The reform you are talking about does not apply to objPtr->typePtr.
>  That field must either hold a valid pointer to the Tcl_ObjType struct
>  that governs the current intrep, or NULL to indicate there is no current
>  intrep.  No other state is valid.

Of course: the point is precisely to get an unambiguously invalid
state, so that a SIGBUS is obtained as soon as use is attempted from a
stale reference.

>  Unless you talking exclusively about those Tcl_Obj structs that have
>  been returned to the free list.  For those I don't think I care.

This scheme applies to any kind of allocation, be it direct ckalloc,
or block allocation like Tcl_Obj. So the objects in the free list are
one case among many. And I believe all these discarded objects would
uniformly benefit from a reliable "reuse detector" costing very little
at free time, and nothing at valid use time.

-Alex