Re: [MiNT] mintlib release

[Alan Hourihane <alanh@fairlite.demon.co.uk>]

On Sun, 2008-04-27 at 18:16 +0200, Frank Naumann wrote:
> Hello!
> 
> > The biggest issue is that no one really understands what happens :-(
> >
> > At first, MiKRO experienced crashes (before main ?) when running programs 
> > using nearly all the available TPA memory. So he added a nice message, and he 
> > increased the size of the preliminary stack.
> 
> With preliminary stack you mean the stack before the Mshrink()? This 
> happened under FreeMiNT? How can this reproduced? 
> Inside mint.cnf, was there TPA_INITIALMEM set and to which value (and how 
> many RAM does the machine have).
> 
> > Keith and Alan reverted MiKRO's patch, and everything went OK again.
> 
> This only affect crt0.s?
> 
> > Memory management is critical, especially in a multitasking non-protected 
> > environment... So it is probably better to do what you suggest - reverting to 
> > the original preliminary stack (while keeping the nice message ?).
> 
> This question I would answer with: go back to the existing behaviour. This 
> was used since ages inside mintlib and didn't made problems until now.

I think we should go back to the existing behaviour before MiKRO's
patch.

Then we need a repeatable case where we can really debug things.

Alan.