Re: Isn't the IDE-Doubler machine independent?

[Ilpo Ruotsalainen <lonewolf%iki.fi@localhost>]

On Tue Jan 14 2003 at 03:24:47 -0800, Ignatios Souvatzis wrote:
> \begin[devil]{advocate}
> 
> You will note that those IDE doublers (at least the one I know) are 
> machine independent - that is, they should work with any old-style IDE
> machine! Shouldn't that code be in sys/dev/ic/wdc.c ? ;-)
> 
> \end{advocate}
> 
> I guess it's ok to keep it confined to Amiga, unless until we find a way to
> run-time detect that eveil hack. Opinions?

I haven't _ever_ heard of that horrible hack used _any_where else than
amiga, and I _only_ created the support code so people who are used to
using them with AmigaOS could have them work in a better OS too - I
wouldn't want this horror to spread elsewhere ;)

(Really. Running without ability to reset - ARGH.)

PS. My tests last night show that the IDEfix commercial thing that comes
with the commercial IDE-doubler HW can somehow recover the interface
from almost any state on bootup. Hmm. Best theory currently is that it
uses some non-IDE-specific way to reset the whole Gayle chip before
trying to use it, but I'm not really sure, the code seems to be
generated by a C compiler (with very high optimizations) and isn't very
readable after disassembly. And I haven't been ably to find *any* docs
at all for the Gayle chip :(

-- 
Ilpo Ruotsalainen - <lonewolf%iki.fi@localhost> - http://www.iki.fi/lonewolf/