Re: which serious bugs exactly? (Re: if you care about debian on powerpc, please react ...)
On Sun, Nov 26, 2006 at 07:25:27PM +0100, Johannes Berg wrote:
> Hi,
>
> > During installation. How often does one install?
Holger, you should know i have installed that box maybe 4-5 times a day, and
stayed in the installer a long time in order to investigate various bugs,
including the infamous parted-doesn't-work-in-di one, which was trully
bewildering.
> Well, it does take a while :) Though when I installed mine, the
> installer didn't work at all yet because the kernel was too old so I
> built my own installer... I also occasionally hack the low level kernel
> code so I guess I simply don't count as a user in this discussion ;)
Indeed, but i guess that with etch now soon supporting 2.6.18 kernels, all
apple powerpc hardware will be supported. At least this is one benefit of
apple going over to intel :)
> > AFAIK they are incuded in the sid kernels (at least, maybe etch too) and you
> > can easily manually load them. Problem solved.
>
> Yes, I should have said "won't be loaded in the installed system". As a
> user I'd have to know which ones to load which is a support problem we
> wouldn't have if the installer just wrote the proper lines
> to /etc/modules. I understand it does now, but I don't understand why
Johannes, actually this is a different issue, the information has been sent to
maks and the initramfs-tools maintainers since a couple of month. The
therm_pm72 support was added, but to this date i don't think the various
windfarm stuff has been added, despite me regularly pinging the
initramfs-tools folk about this.
You can easily add the following modules (and in this order) to your
/etc/initramfs-tools/modules :
i2c-powermac
therm_pm72
windfarm_core
windfarm_cpufreq_clamp
windfarm_lm75_sensor
windfarm_max6690_sensor
windfarm_pid
windfarm_pm81
windfarm_pm91
windfarm_smu_sat
windfarm_pm112
windfarm_smu_sensors
> such a trivial change is generating this much noise (actually, I have my
> suspicions but that's beside the point). But then, I haven't understood
> the whole noise and probably don't want to.
>
> That said, the time being spent on this particular issue alone would
> suffice to fix the upstream kernel to auto-load the modules in question
> based on smu sysfs nodes or whatever... (and in fact, that's the way I
> work with(out) Debian most of the time; if something bothers me I fix it
> upstream.)
I doubt so, and as said it is an upstream matter, not a debian maintainer
matter. Two different set of skills.
Friendly,
Sven Luther
Reply to: