Re: [TCLCORE] Making crashes more reliable

[Kevin K. <kev...@gm...>]

Alexandre Ferrieux wrote:
> Hi,
> 
> For what it's worth, Donal decided to ignore the suggestion in tclVar.c
> Then let's hope that the remainder is 100% bug-free; otherwise many
> hours will be wasted in investigating the next occurrence in a
> production, non-mem-debug system.
> 
> -Alex

Alex, let's not give up so quickly.

Donal, Alex is talking about laying a presumably-invalid background
in dead memory, for a case where NULL in live memory would have a
different meaning.

I don't care that much what's in dead memory. If Alex things that
0xDECAFBAD would be a bit pattern that makes his debugging easier,
I'm with Joe here. It doesn't matter for correct C code; and it's
highly likely to be in the same space as NULL in terms of segv's and
bus errors.