Re: [patch-2.3.51-pre2] misc cleanups

From: Thierry Vignaud (tvignaud@mandrakesoft.com)
Date: Fri Mar 10 2000 - 11:13:56 EST


Tigran Aivazian wrote:
>
> > Anyway, for final user, either they'll never see this because either
> > distro won't provide i686 compiled kernels or they'll provide kernels
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
> Hi
>
> If you check carefully you will notice that microcode driver does not
> require 686 compiled kernel - it does the family/model checks at runtime.
> Btw, it was your colleague (Jeff Garzik) who pointed this out and I
> accepted his advice gratefully.

this is right because a distribution can't afford to have kernel for a
lot of cpu generation because of disk space problem because you've to
provide a minimal (disk/cdrom/network but no sound/dri/... stuff) boot
kernel, and a UP and a SMP kernel for each generation (as some drivers
've problem in SMP mode but not in UP), ...

And i think (i didn't test it) that this driver won't made a panic on
i386.
So add stuff for next generation (and as this driver can be compiled as
a module, it won't eat some ram on i686- machines) is not a problem as
either it don't panic (eg PIII patches on PII) or it's the
responsability of the super user to test it (eg load the microcode
module).

But what i means is that putting in source some code that _will_ cause a
compilation error to happen is not a good stuff in the kernel. You have
at least to say it in the doc. There are at this moment enough
compilation problem with forgoten dependencies to add some explicit
ones...

see you.

-- 
www.linux-mandrake.com
somewhere between the playstation and the craystation
	Thierry

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Wed Mar 15 2000 - 21:00:18 EST