[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [MiNT] Compiling a FreeMiNT kernel


Am Samstag, den 12.10.2013, 08:18 +1000 schrieb Miro Kropáček: 
>         > make M68K_ATARI_MINT_CROSS=yes all
> btw I use make CROSS=yes all the time :)

I suspect most most builds of FreeMiNT are infact cross compiled these
days, even if a native compile is still possible. 
>         Anyway you guys insisted on 2.95 for quite a time so which is
>         a version 4 gcc so much worse? 
> Did we? I use latest gcc for years ... there's only one serious
> advantage of the 2.95 over 4.x -- Sparemint's gdb has troubles to read
> symbols from the latter. But if you're only after compilation, gcc4 is
> always better choice.

For the time being, my cross gcc seems to work ok.
Any serious repacking or such would be based on official builds. I am
aware now that that might be a better approach.

>         in fsck or something. i guess I am still missing some support
>         libs.
>         cflib.h or such. I have included the logfile with this mail. 
> Yes, you need cflib for one or two of them.
>         - Did this compile all xfs and xdd drivers aswell? Or really
>         just the
>         main kernel? 
> Nope. But you can do this if you jump into 'sys' and either a) type
> just 'make' or b) type 'make xfs' or so, I think there's 'make help'
> available
>         - Where do I find the PRGs and all thos driver files? I plan
>         to
>         duplicate the VanillaMiNT efforts to learn more and maybe run
>         a basic
>         000 kernel in Hatari and later on my STE. 
> Unfortunately, it's quite chaotic. Kernels are in sys/.compile_<cpu>,
> drivers are in the source directories, so usually the best is to look
> for either executable files (with the x bit set) or for
> *.{xfs,xdd,km,xif,<and perhaps some others>}.

A thanks, with a bit of find magic I could trace it and it seems my
partial build did make some .xfs and .xdd files as well.

Now I just wonder why the compiled kernel .prgs are in hidden
directories while all drivers are directly insitu with their sources?

Is this a historical sorting thing or due to different subpackage
maintainers? I'm just curious....

Matthias Arndt <marndt@asmsoftware.de>
PGP-Key: http://www.final-memory.org/files/marndt.asc
Jabber: simonsunnyboy@atari-jabber.org