Kevin Lawton wrote:
> > Just some little details to think about before you load
> > it into CVS:
> >
> > - I think you should remove user/iodev and user/gui
> > Perhaps use cvs -P next time ;)
> > - You should add guest/test to guest/Makefile.in and configure.in
> > Right now, that dir is neither built or cleaned from the main
> > Makefile, which is irritating. Also, IIRC there's currently
> > a binary in that dir !
> > - Adding docs/ to configure.in / main Makefile would also be
> > helpful.
> > - I haven't had the time yet to look in detail at the new code
> > (I'll do that real soon!) but what is kernel/crap.S ?
> > Not a very descriptive name ;)
>
> Yes, I can fix these for the CVS tree. Oops, you can
> send crap.S to /dev/null.
One more point I just noticed: I had removed all dependencies on
Linux-specific headers, e.g. on the u32 etc. types, from all files
except host-linux.c. You have apparently reverted those changes
(e.g. your monitor.c/h again contains u32 where I had already replaced
it with Bit32u ...), and reintroduced stuff like #include <linux/types.h>
in config.h.in and a host.h file.
Was this a deliberate change? I don't see what advantage this brings,
especially if we want to write code that runs also in the guest context
(where specific Linux services aren't available anyway).
Another CVS-specific problem: it is impossible to rename directories
in CVS, so that in order to rename the main directory to plex86 we'll
probably have to create a new repository anyway ...
Bye,
Ulrich
--
Ulrich Weigand,
IMMD 1, Universitaet Erlangen-Nuernberg,
Martensstr. 3, D-91058 Erlangen, Phone: +49 9131 85-27688
