[ltp] [patch 11/11] ibm-acpi: implement fan watchdog command

Henrique de Moraes Holschuh linux-thinkpad@linux-thinkpad.org
Mon, 9 Oct 2006 12:02:24 -0300

On Mon, 09 Oct 2006, Shem Multinymous wrote:
> On 10/9/06, Henrique de Moraes Holschuh <hmh@hmh.eng.br> wrote:
> >The watchdog is programmed using echo "watchdog <number>" > fan, where 
> >number
> >is the number of seconds to wait before doing an "enable", and zero 
> >disables
> >the watchdog.
> Excellent approach; implementation awaited.

I am studying workqueues and timers to see how to better implement the
watchdog, and then I will send a patch 12 of 11 to this thread implementing
it :)

> I can't actually test the patches at the moment, but they look very
> good other than those minor comments.

They're tested (yes, even the intermediate ones) on a T43, but I didn't test
it against braindead userspace code.

> I expect acpi-devel to express some discomfort over the continued of a
> machine-specific procfs interface, as moving stuff to a
> generic-looking sysfs interface is now in vogue.

I will port the entire ibm-acpi myself to sysfs if I have to, but *only*
after I get a working tree (i.e. something I can actually run on my kernel
:-) ) of the rest of the ACPI code using sysfs.  I understand this is being
worked on upstream, and ibm-acpi will follow that with eagerness.  I am
*itching* to trash-dump the entire procfs interface like you wouldn't

I will remember to make that very clear when I send the patchset to

  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh