getting new hardware (arbor M1526) buttons to work w/ acpi

getting new hardware (arbor M1526) buttons to work w/ acpi

[Yedidia Klein]

Hello list,

I've a kiosk computer named arbor m1526
(http://www.arbor.com.tw/products/detail.aspx?Product_Name=M1526.)
this hardware has some functions key (F1-F8) that are connected to
Embedded Controller.

The company do not provide Linux support to these buttons but provided
information on addresses as follows:
EC RAM Address 0x0501

Bit0 = 1 indicative F1 Button be pressed
Bit1 = 1 indicative F2 Button be pressed
Bit2 = 1 indicative F3 Button be pressed
Bit3 = 1 indicative F4 Button be pressed
Bit4 = 1 indicative F5 Button be pressed
Bit5 = 1 indicative F6 Button be pressed
Bit6 = 1 indicative F7 Button be pressed
Bit7 = 1 indicative F8 Button be pressed
P.S. If read finish please clear EC ram bit flag.

I would like to write a new acpi patch/file to the linux kernel for
supporting these buttons.
seems to me that the right place is writing a new file at drivers/platform/x86

I'm looking for some good doc. or template for writing a new hardware
acpi support.

any help will be appreciate,

tnx,

--Y

Re: getting new hardware (arbor M1526) buttons to work w/ acpi

[Zhang Rui]

That's interesting.

The first question would be how to catch the hotkey events.
is this hotkey event routed via ACPI?

can you attach the acpidump of this machine please?

thanks,
rui

On Thu, 2010-08-26 at 13:56 +0800, Yedidia Klein wrote:
> Hello list,
> 
> I've a kiosk computer named arbor m1526
> (http://www.arbor.com.tw/products/detail.aspx?Product_Name=M1526.)
> this hardware has some functions key (F1-F8) that are connected to
> Embedded Controller.
> 
> The company do not provide Linux support to these buttons but provided
> information on addresses as follows:
> EC RAM Address 0x0501
> 
> Bit0 = 1 indicative F1 Button be pressed
> Bit1 = 1 indicative F2 Button be pressed
> Bit2 = 1 indicative F3 Button be pressed
> Bit3 = 1 indicative F4 Button be pressed
> Bit4 = 1 indicative F5 Button be pressed
> Bit5 = 1 indicative F6 Button be pressed
> Bit6 = 1 indicative F7 Button be pressed
> Bit7 = 1 indicative F8 Button be pressed
> P.S. If read finish please clear EC ram bit flag.
> 
> I would like to write a new acpi patch/file to the linux kernel for
> supporting these buttons.
> seems to me that the right place is writing a new file at drivers/platform/x86
> 
> I'm looking for some good doc. or template for writing a new hardware
> acpi support.
> 
> any help will be appreciate,
> 
> tnx,
> 
> --Y
> --
> To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: getting new hardware (arbor M1526) buttons to work w/ acpi

[Zhang Rui]

On Thu, 2010-08-26 at 17:17 +0800, Yedidia Klein wrote:
> here is my acpidump output.
> 
hmmm, please attach the output of "cat /proc/interrupts" and
"grep . /sys/firmware/acpi/interrupts/*", both before and after pressing
the hotkey.

thanks,
rui

> tnx,
> --Y
> 
> 
> On Thu, Aug 26, 2010 at 10:20 AM, Zhang Rui <rui.zhang@intel.com> wrote:
> >
> > That's interesting.
> >
> > The first question would be how to catch the hotkey events.
> > is this hotkey event routed via ACPI?
> >
> > can you attach the acpidump of this machine please?
> >
> > thanks,
> > rui
> >
> > On Thu, 2010-08-26 at 13:56 +0800, Yedidia Klein wrote:
> > > Hello list,
> > >
> > > I've a kiosk computer named arbor m1526
> > > (http://www.arbor.com.tw/products/detail.aspx?Product_Name=M1526.)
> > > this hardware has some functions key (F1-F8) that are connected to
> > > Embedded Controller.
> > >
> > > The company do not provide Linux support to these buttons but provided
> > > information on addresses as follows:
> > > EC RAM Address 0x0501
> > >
> > > Bit0 = 1 indicative F1 Button be pressed
> > > Bit1 = 1 indicative F2 Button be pressed
> > > Bit2 = 1 indicative F3 Button be pressed
> > > Bit3 = 1 indicative F4 Button be pressed
> > > Bit4 = 1 indicative F5 Button be pressed
> > > Bit5 = 1 indicative F6 Button be pressed
> > > Bit6 = 1 indicative F7 Button be pressed
> > > Bit7 = 1 indicative F8 Button be pressed
> > > P.S. If read finish please clear EC ram bit flag.
> > >
> > > I would like to write a new acpi patch/file to the linux kernel for
> > > supporting these buttons.
> > > seems to me that the right place is writing a new file at drivers/platform/x86
> > >
> > > I'm looking for some good doc. or template for writing a new hardware
> > > acpi support.
> > >
> > > any help will be appreciate,
> > >
> > > tnx,
> > >
> > > --Y
> > > --
> > > To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
> > > the body of a message to majordomo@vger.kernel.org
> > > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> >
> >

Re: getting new hardware (arbor M1526) buttons to work w/ acpi

[Thomas Renninger]

On Friday 27 August 2010 03:30:11 Zhang Rui wrote:
> On Thu, 2010-08-26 at 17:17 +0800, Yedidia Klein wrote:
> > here is my acpidump output.
> > 
> hmmm, please attach the output of "cat /proc/interrupts" and
> "grep . /sys/firmware/acpi/interrupts/*", both before and after pressing
> the hotkey.
Yep.
You could also try to load the wmi driver with:
debug_event=1
e.g. adding this line:
options wmi debug_event=1
in
/etc/modprobe.d/99-local.conf or /etc/modprobe.conf.local
(the first is the new the latter the older interface, depends on your
userspace versions)
Then hit the button and check syslog.

It might also be served by a Windows spec called "hot start" for which
I sent a driver which would still need some polishing some weeks ago.
The device for these is:
PNP0C32

Please add me to CC if you open bug at bugzilla.kernel.org. I won't
have time to code, but I can help a bit.

      Thomas

Fwd: getting new hardware (arbor M1526) buttons to work w/ acpi

[Yedidia Klein]

ok, found that wmi is compiled into my kernel and not built as module.
while the grep for PNP0C32 returned nothing - so I assume that it's not that.

is there any way to read these addresses directly via C and not via ACPI ?

tnx a lot,

--Y


2010/9/13 Thomas Renninger <trenn@suse.de>
>
> On Monday 13 September 2010 16:46:32 Yedidia Klein wrote:
> > hi,
> These are only two possibilities/interfaces how the buttons may be
> driven.
>
> > what is the wmi driver ? there is not any module in this name on my distro
> > (ubuntu 10.4)
> Oh, there should, they probably compiled this driver:
> wmi.ko
> ?
> A debug facility (debug_events=1 module parameter) was added
> recently and may not exist there yet.
>
> > the  PNP0C32 sounds interesting - what should I patch to enable it ?
> mkdir /tmp/acpi
> cd /tmp/acpi
> acpidump >acpidump
> acpixtract -a acpidump
> iasl -d DSDT.dat
> iasl -d -e DSDT.dat SSDT*.dat
> grep PNP0C32 *.dsl
>
> Does the last command (grep) give you any output that
> it found the PNP0C32 string in your ACPI BIOS tables
> (which were dumped from memory with the commands before).
> If yes, the button might be served by this interface.
>
>    Thomas
>
> >
> > tnx,
> >
> > --Y
> >
> >
> > On Mon, Sep 6, 2010 at 1:43 PM, Thomas Renninger <trenn@suse.de> wrote:
> >
> > > On Friday 27 August 2010 03:30:11 Zhang Rui wrote:
> > > > On Thu, 2010-08-26 at 17:17 +0800, Yedidia Klein wrote:
> > > > > here is my acpidump output.
> > > > >
> > > > hmmm, please attach the output of "cat /proc/interrupts" and
> > > > "grep . /sys/firmware/acpi/interrupts/*", both before and after pressing
> > > > the hotkey.
> > > Yep.
> > > You could also try to load the wmi driver with:
> > > debug_event=1
> > > e.g. adding this line:
> > > options wmi debug_event=1
> > > in
> > > /etc/modprobe.d/99-local.conf or /etc/modprobe.conf.local
> > > (the first is the new the latter the older interface, depends on your
> > > userspace versions)
> > > Then hit the button and check syslog.
> > >
> > > It might also be served by a Windows spec called "hot start" for which
> > > I sent a driver which would still need some polishing some weeks ago.
> > > The device for these is:
> > > PNP0C32
> > >
> > > Please add me to CC if you open bug at bugzilla.kernel.org. I won't
> > > have time to code, but I can help a bit.
> > >
> > >      Thomas
> > >
> >
>
--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: getting new hardware (arbor M1526) buttons to work w/ acpi

[Thomas Renninger]

On Monday 13 September 2010 20:11:55 Yedidia Klein wrote:
> ok, found that wmi is compiled into my kernel and not built as module.
> while the grep for PNP0C32 returned nothing - so I assume that it;s
> not that.
> 
> is there any way to read these addresses directly via C and not via
> ACPI ?
If the button is ACPI driven you have to look at ACPI code sooner or 
later.
Have you tried showkey yet?

Try:
sleep 1;showkey
and
sleep 1;showkey -s
(the sleep is only that you can escape with CTRL-c).
If you get output if you hit the keys, then forget about ACPI and all
I said. There is a lot documentation about this out there.
Best you learn about:
/etc/X11/Xmodmap
the input layer, xev is a nice tool for these, etc.

If you don't see something with showkey, it's likely that it's ACPI 
driven.
Do:
rmmod battery
rmmod thermal
to avoid other ACPI interrupts
Then:
watch -n1 cat /proc/interrupts
Press the buttons, is the ACPI irq (normally 9) increasing when you
hit any of these?
If yes, it's likely ACPI driven.

Next step would to find out which GPE is fired on each press:
VALID="";for x in /sys/firmware/acpi/interrupts/gpe*;do if cat $x|grep 
enabled;then VALID="$VALID $x";fi;done;watch -n1 cat $VALID

modprobe battery
modprobe thermal
afterwards...

          Thomas

Re : getting new hardware (arbor M1526) buttons to work w/ acpi

[Yedidia Klein]

Hi,

nothing shows, nothing changes.. not showkey/xev output
and not any change on ACPI interrupts or /proc

seems that these buttons do not exist....

any idea - what can be done  ?

tnx,

--Y


2010/9/13 Thomas Renninger <trenn@suse.de>
>
> On Monday 13 September 2010 20:11:55 Yedidia Klein wrote:
> > ok, found that wmi is compiled into my kernel and not built as module.
> > while the grep for PNP0C32 returned nothing - so I assume that it;s
> > not that.
> >
> > is there any way to read these addresses directly via C and not via
> > ACPI ?
> If the button is ACPI driven you have to look at ACPI code sooner or
> later.
> Have you tried showkey yet?
>
> Try:
> sleep 1;showkey
> and
> sleep 1;showkey -s
> (the sleep is only that you can escape with CTRL-c).
> If you get output if you hit the keys, then forget about ACPI and all
> I said. There is a lot documentation about this out there.
> Best you learn about:
> /etc/X11/Xmodmap
> the input layer, xev is a nice tool for these, etc.
>
> If you don't see something with showkey, it's likely that it's ACPI
> driven.
> Do:
> rmmod battery
> rmmod thermal
> to avoid other ACPI interrupts
> Then:
> watch -n1 cat /proc/interrupts
> Press the buttons, is the ACPI irq (normally 9) increasing when you
> hit any of these?
> If yes, it's likely ACPI driven.
>
> Next step would to find out which GPE is fired on each press:
> VALID="";for x in /sys/firmware/acpi/interrupts/gpe*;do if cat $x|grep
> enabled;then VALID="$VALID $x";fi;done;watch -n1 cat $VALID
>
> modprobe battery
> modprobe thermal
> afterwards...
>
>          Thomas
--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html