Re: X11 and mc68060...

[Ruben van Staveren <ruben%luna.net@localhost>]

Hi,
> 
> 
> Your understanding of -m68060 is incomplete.
> 
> -m68060, as currently implemented, is like -m68040, with the additional 
>  omission of 68060 unfriendly integer instructions.
> 
> Both -m68040 and -m68060 create FP instructions that are NOT implemented on
> the 68881 and 68882 coprocessors.
Ok, the kernel uses soft-float

> 
> This is no problem for the kernel (which does not do any FP computations of
> its own), but WILL break many user programs.
> 
> Regards,
>       Ignatios Souvatzis
> 
> (who once made the mistake to compile his X server with -m68060, and needed
> a while ot understand why it crashed on the 68030 machine, on which he 
> occasionally boots NetBSD.)
> 

Hmm lets compile the Xserver (and related stuff) with soft-float too :)
(with severe performance loss on all cpu classes)
Or introduce cyberpatcher alike features in NetBSD *grin*


Regards,
        Ruben

-- 
.-------------------------------------------------------------------.
| Ruben van Staveren - http://kip.sateh.com/~ruben - A1260@50/48   _|_o
| Luna Internet - Systems & Software Engineer - NetBSD - AmigaOS_ //| #>
| ruben%luna.net.spam@localhost - Invasion Of Our Piracy - PGP 2.6.3ia    \X/ | 
4
`-------------------------------------------------------------------'