pmevent patch and PM

pmevent patch and PM

[John Fremlin]

Hi PM maintainers! ARM and PPC ppl, who is your PM maintainer?

"Grover, Andrew" <andrew.grover@intel.com> writes:

> ACPI has by far the richest set of capabilities. It is a superset of
> APM.  Therefore a combined APM/ACPI interface is going to look a lot
> like an ACPI interface.
>
> IMHO an abstracted interface at this point is overengineering. Maybe
> later it will make sense, though.

A simple event interface is presented with the pmevent patch. It is
very extensible. The individual PM system driver suggests an action to
be taken and identifies the event exactly. Any number of userspace
readers can therefore apply fine grained policy to specific events and
simultaneously handle unknown events reasonably well.

The pmevent patch is available here

        http://john.snoop.dk/programs/linux/offbutton

There are some badly done example bindings for APM (that part of the
patch needs to be cleaned up).  Will your PM driver be able to map
events well onto it? I hope it should remove the need for an PM driver
specific event interfaces.

Please take a look and comment.

[...]

--

	http://ape.n3.net

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: pmevent patch and PM

[Erik Mouw]

On Wed, May 02, 2001 at 06:59:32PM +0100, John Fremlin wrote:
> Hi PM maintainers! ARM and PPC ppl, who is your PM maintainer?

We (= ARM people) have a couple of them, there is no real "PM
maintainer". We do clock switching (on SA11x0 and XScale), CPU core
voltage switching, and other things. Note that most ARM systems don't
have things like ACPI BIOSes that do funky things, everything is
controlled from the OS. Russell usually coordinates things.


Erik

--
J.A.K. (Erik) Mouw, Information and Communication Theory Group, Department
of Electrical Engineering, Faculty of Information Technology and Systems,
Delft University of Technology, PO BOX 5031,  2600 GA Delft, The Netherlands
Phone: +31-15-2783635  Fax: +31-15-2781843  Email: J.A.K.Mouw@its.tudelft.nl
WWW: http://www-ict.its.tudelft.nl/~erik/

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: pmevent patch and PM

[Russell King - ARM Linux]

On Wed, May 02, 2001 at 08:57:46PM +0200, Erik Mouw wrote:
> maintainer". We do clock switching (on SA11x0 and XScale), CPU core
> voltage switching, and other things. Note that most ARM systems don't
> have things like ACPI BIOSes that do funky things, everything is
> controlled from the OS. Russell usually coordinates things.

I've been specifically discussing the CPU clock switching issues
with Erik, and we have the bones of a kernel-internal generic
interface that should be suitable for both ARM and Intel SpeedStep
(thanks to Andy Grover for his input to this).

By "clock switching" here, I'm talking about the ability for the OS
to reprogram the clock rate at which the CPU core is clocked, and/or
surrounding devices.  For instance, on some of the StrongARM cores,
you need to adjust the SDRAM timing parameters either before or after
the clock change).

The interface doesn't have anything to drive it yet - its literally
an interface where you can say "please set my CPU clock speed to
xxx MHz".
   _____
  |_____| ------------------------------------------------- ---+---+-
  |   |        Russell King       linux@arm.linux.org.uk      --- ---
  | | | |            http://www.arm.linux.org.uk/            /  /  |
  | +-+-+                                                     --- -+-
  /   |               THE developer of ARM Linux              |+| /|\
 /  | | |                                                     ---  |
    +-+-+ -------------------------------------------------  /\\\  |

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: pmevent patch and PM

[Dave Jones]

On Wed, 2 May 2001, Russell King - ARM Linux wrote:

> I've been specifically discussing the CPU clock switching issues
> with Erik, and we have the bones of a kernel-internal generic
> interface that should be suitable for both ARM and Intel SpeedStep

Interesting, Powertweak [http://powertweak.sourceforge.net] has had
this in development for a while (slowed down somewhat as I've been
busy with studies). So far we have basic AMD PowerNOW! support, and
have some pending stuff to be merged (VIA Cyrix & Transmeta support).

> The interface doesn't have anything to drive it yet - its literally
> an interface where you can say "please set my CPU clock speed to
> xxx MHz".

The code we have so far is in userspace. The clockspeed frobbing
will be moved to kernelspace as soon as my exams are over, and I
get some time to work on this.

The policies for adjusting the clockspeed sysctl have yet to be written,
but should slot into the framework we have quite easily.

Do you have a ptr to what you have done so far ?
Doing the kernel side part as platform independantly as possible
would make a lot of sense, and shouldn't be particularly difficult.

regards,

Dave.

--
| Dave Jones.        http://www.suse.de/~davej
| SuSE Labs

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/