Patches to acpi_hp

[Michael Gmelin]

On Wed, 14 Sep 2016 04:07:35 +0000
Alexey Dokuchaev <danfe at FreeBSD.org> wrote:

> On Tue, Jun 30, 2009 at 07:32:09PM +0200, Michael Gmelin wrote:
> > Can you send me the output you get when kldloading acpi_hp + the
> > output of cat /dev/wmistat0.  It's possible that your laptop
> > doesn't support CMI. ...
> > it seems like your model doesn't support this feature, sorry.  
> 
> Hi Michael,
> 
> What exactly indicates proper CMI support?  I have a problem with
> acpi_hp(4) on ProBook 645 G1 (infamous ``driver bug: Unable to set
> devclass (class: ppc devname: (unknown))'' error), but it looks like
> all GUIDs listed inside the `acpi_hp.c' are there:
> 
>   $ cat /dev/wmistat0 | egrep '95F24279|5FB7F034|2D114B49'
>   {5FB7F034-2C63-45E9-BE91-3D44E2C707E4}   1 NO   WMAA NO  NO    AA
>   {95F24279-4D7B-4334-9387-ACCDC67EF61C}   1 NO   NO   NO  0x80+ -
>   {2D114B49-2DFB-4130-B8FE-4A3C09E75133} 247 NO   NO   NO  NO    BC
> 
> Despite that "driver bug" message, /dev/hpcmi gets created, but
> reading it yields just a bunch (247 of them, like the number above)
> of blank lines, and issuing various sysctl(8) calls also does not
> work properly.
> 
>   $ sysctl dev.acpi_hp.0.hdd_temperature
>   dev.acpi_hp.0.hdd_temperature: 4		<-- ??? what units
> are these? # sysctl dev.acpi_hp.0.wlan_enabled=1
>   dev.acpi_hp.0.wlan_enabled: 0
>   sysctl: dev.acpi_hp.0.wlan_enabled=1: Interrupted system call
> 

Unfortunately my memory on this is pretty spotty at this point, so I
would probably start at reading this again:

http://h20331.www2.hp.com/Hpsub/downloads/cmi_whitepaper.pdf

I still have the laptop around I used to write it, so I could test
patches if necessary. Your best bet is probably to check if Linux
supports your model and if yes, figure out what's wrong with our driver.

-michael

-- 
Michael Gmelin