[ltp] DMA hotswap 0.2.0 vs. ibm-acpi-0.12a

[honey]

  This message is in MIME format.  The first part should be readable text,
  while the remaining parts are likely unreadable without MIME-aware tools.

--8323328-113700565-1141001249=:32746
Content-Type: TEXT/PLAIN; CHARSET=ISO-8859-1; FORMAT=flowed
Content-Transfer-Encoding: 8BIT
Content-ID: <Pine.LNX.4.64.0602270049291.32746@localhost.localdomain>

On Sun, 26 Feb 2006, André Wyrwa wrote:

> Joe Christy wrote:
>> 	Aha! That explains it.
>>
>> 	Since I've never been able to convince FC4 to honor any parameters to
>> ibm-acpi in /etc/modprobe.conf, I modprobe -r it and then reload it w/
>> its parameters at the end of rc.local. Looks like a Fedora bug. Sigh.
>
> If you load ibm-acpi at bootup through something the like of
> /etc/modules.autoload.d/ (Gentoo) or /etc/modules (Debian), try giving
> the parameters in the same line.
>
> If it's some power management setup script, try figuring how ibm-acpi
> get's loaded. I guess it's loaded via insmod, which doesn't respect
> things in /etc/modprobe.conf .
>
> André.

The reason ibm_acpi doesn't load with modprobe.conf params in Fedora is
this devilish section in /etc/rc.d/rc.sysinit:

# Initialiaze ACPI bits
if [ -d /proc/acpi ]; then
     for module in /lib/modules/$unamer/kernel/drivers/acpi/* ; do
        insmod $module >/dev/null 2>&1
     done
fi

- it basically tries to load every possible acpi module early on.  And
it uses insmod instead of modprobe, it doesn't honour modprobe.conf.

I never investigated further why it does this, or asked Fedora devels
if such an undiscriminating method of loading acpi modules can be
more finely tuned, or modprobed maybe (but maybe it's too early in
the boot).  But I'm pretty sure that's why, after a lot of hunting,
and I do the same as the original poster - unload and the modprobe
it again in rc.local.  Messy.

Honey
--8323328-113700565-1141001249=:32746--