Re: [hatari-devel] ST bus error results (was: New EmuTOS floppy code causes bus error on (Hatari-emulated) MegaST)

[Thomas Huth <th.huth@xxxxxxxxx>]

Am Sat, 28 Apr 2018 10:46:16 +0200
schrieb Christian Zietz <czietz@xxxxxxx>:

> Thomas Huth schrieb:
> 
> > Am Fri, 27 Apr 2018 21:23:24 +0200
> > schrieb Christian Zietz <czietz@xxxxxxx>:  
> 
> >> EmuTOS also crashes on real hardware, namely one of my two
> >> 1040STF. It runs on the other one. I think Hatari is wrong insofar
> >> as the difference is not between ST and MegaST but between Ricoh
> >> chipset (=no bus error for $FFFF860F) and IMP chipset (=bus error
> >> for $FFFF860F).  
> 
> > Oh, that's interesting. Could you please run the bus error testers
> > from Hatari (folder tests/buserror/) on both machines and mail the
> > results to the hatari-devel mailing list?  
> 
> Here they are, for a 1040STF with Ricoh chipset (-38) and a 1040STF
> with IMP chipset. You can see some differences between those two
> chipsets. Note that the IMP machine has been retrofitted with a
> Blitter which explains the differences around $FF8Axx.

Great, thanks for the files! The IMP file indeed matches the file that
has once been contributed by a Mega-ST user. So the difference is
really not ST vs. Mega-ST, but IMP vs. Ricoh! Would be interesting to
know whether Ricoh chips have still been used for Mega-STs, or if all
Megas were already shipped with an IMP chip...

> If you really
> need results without Blitter I'll have to see how to remove it
> temporarily. (I think this involves some soldering.)

Thanks for the offer, but I think that's not necessary right now.

 Thomas