acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Ville-Pekka Vainio]

Hi,

I hope I'm on the right mailing list for this issue. The kernel (I'm
using Fedora's 2.6.37rc7) loads the acer-wmi module on a Lenovo Ideapad
S12 netbook and I think that's not necessary. Loading the module causes
problems with NetworkManager, the wireless network can't be enabled at
all. This seems to happen because acer-wmi adds a soft blocked wifi
killswitch, which can't be unblocked with rfkill.

Now that the kernel has the ideapad-laptop module, which seems to handle
the killswitches better, acer-wmi could probably be disabled with this
netbook. I know almost nothing about ACPI/WMI myself, which is why I'm
writing on the list instead of just fixing this.

-- 
Ville-Pekka Vainio

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Joey Lee]

Hi Vainio, 

於 二，2010-12-28 於 14:31 +0200，Ville-Pekka Vainio 提到：
> Hi,
> 
> I hope I'm on the right mailing list for this issue. The kernel (I'm
> using Fedora's 2.6.37rc7) loads the acer-wmi module on a Lenovo Ideapad
> S12 netbook and I think that's not necessary. Loading the module causes
> problems with NetworkManager, the wireless network can't be enabled at
> all. This seems to happen because acer-wmi adds a soft blocked wifi
> killswitch, which can't be unblocked with rfkill.
> 
> Now that the kernel has the ideapad-laptop module, which seems to handle
> the killswitches better, acer-wmi could probably be disabled with this
> netbook. I know almost nothing about ACPI/WMI myself, which is why I'm
> writing on the list instead of just fixing this.
> 

It's odd, because acer-wmi driver in 2.6.37rc7 only auto-probe when the
it detected the following wmi GUIDs:

86 MODULE_ALIAS("wmi:67C3371D-95A3-4C37-BB61-DD47B491DAAB");
87 MODULE_ALIAS("wmi:6AF4F258-B401-42fd-BE91-3D4AC2D7C0D3");

It's not make sense acer-wmi probe on Ideapad S12 netbook, could you
please help to attached the following information?

- modinfo acer-wmi > acer-wmi.log
- dmesg > dmesg.log

better attached:
- dmidecode > dmidecode.log
- acpidump > acpidump.dat


Thank's a lot!
Joey Lee

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Ville-Pekka Vainio]

I decided to upload the files to the web instead of attaching them to
the email since they are quite large.

ti, 2010-12-28 kello 08:36 -0700, Joey Lee kirjoitti:
> It's not make sense acer-wmi probe on Ideapad S12 netbook, could you
> please help to attached the following information?
> 
> - modinfo acer-wmi > acer-wmi.log

http://vpv.kapsi.fi/ideapad-s12/modinfo-acer-wmi.txt

> - dmesg > dmesg.log

http://vpv.kapsi.fi/ideapad-s12/dmesg-acer-wmi.txt

> - dmidecode > dmidecode.log

http://vpv.kapsi.fi/ideapad-s12/dmidecode-acer-wmi.txt

> - acpidump > acpidump.dat

http://vpv.kapsi.fi/ideapad-s12/acpidump.dat

The dmesg may not reveal much, but if I blacklist acer-wmi for booting
and then load it later manually, this is what appears
in /var/log/messages:

kernel: [ 4795.867540] acer-wmi: Acer Laptop ACPI-WMI Extras
NetworkManager[914]: <info> found WiFi radio killswitch rfkill7 (at /sys/devices/platform/acer-wmi/rfkill/rfkill7) (driver acer-wmi)
NetworkManager[914]: <info> WiFi now disabled by radio killswitch

Then, if I unload acer-wmi:

kernel: [ 4852.875590] acer-wmi: Acer Laptop WMI Extras unloaded
NetworkManager[914]: <info> radio
killswitch /sys/devices/platform/acer-wmi/rfkill/rfkill7 disappeared
NetworkManager[914]: <info> WiFi now enabled by radio killswitch
NetworkManager[914]: <info> (wlan0): bringing up device.
kernel: [ 4852.905625] wmi: Mapper unloaded

-- 
Ville-Pekka Vainio

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Joey Lee]

Hi Vainio, 

於 二，2010-12-28 於 19:28 +0200，Ville-Pekka Vainio 提到：
> I decided to upload the files to the web instead of attaching them to
> the email since they are quite large.
> 
> ti, 2010-12-28 kello 08:36 -0700, Joey Lee kirjoitti:
> > It's not make sense acer-wmi probe on Ideapad S12 netbook, could you
> > please help to attached the following information?
> > 
> > - modinfo acer-wmi > acer-wmi.log
> 
> http://vpv.kapsi.fi/ideapad-s12/modinfo-acer-wmi.txt
> 
> > - dmesg > dmesg.log
> 
> http://vpv.kapsi.fi/ideapad-s12/dmesg-acer-wmi.txt
> 
> > - dmidecode > dmidecode.log
> 
> http://vpv.kapsi.fi/ideapad-s12/dmidecode-acer-wmi.txt
> 
> > - acpidump > acpidump.dat
> 
> http://vpv.kapsi.fi/ideapad-s12/acpidump.dat
> 

It's funny Lenovo S12's _WDG have GUID that matches with Acer AMW0.

Dump from Lenovo Ideapad S12's _WDG:

67C3371D-95A3-4C37-BB61-DD47B491DAAB:
        object_id: AB
        notify_id: 41
        reserved: 42
        instance_count: 1
        flags: 0x2 ACPI_WMI_METHOD


In acer-wmi code:

#define AMW0_GUID1              "67C3371D-95A3-4C37-BB61-DD47B491DAAB"

MODULE_ALIAS("wmi:67C3371D-95A3-4C37-BB61-DD47B491DAAB");	//match



Maybe Wistron is the ODM (or OEM) to manufacture Lenovo Ideapad S12 ?
Anyway, 
I added this Ideapad S12 to blacklist in acer-wmi driver because it must
supported by ideapad-laptop driver.

I also added Cc. to Carlos Corbacho, need his review.

Does is possible you help to test this patch?


Thank's a lot!
Joey Lee



From 194a1b499c36db057319712220af8ec7400030d7 Mon Sep 17 00:00:00 2001
From: Lee, Chun-Yi <jlee@novell.com>
Date: Wed, 29 Dec 2010 00:17:23 +0800
Subject: [PATCH] Add Lenovo Ideapad S12 to acer blacklist

The Lenovo Ideapad S12's _WDG have GUID that matches with AMW0, it causes
acer-wmi driver auto-probe when system boot but there already have
ideapad-laptop driver to support Ideapad serial.
So, add Lenovo Ideapad S12 to acer blacklist to disable it.

Signed-off-by: Lee, Chun-Yi <jlee@novell.com>
Cc: Carlos Corbacho <carlos@strangeworlds.co.uk>
---
 drivers/platform/x86/acer-wmi.c |   11 +++++++++++
 1 files changed, 11 insertions(+), 0 deletions(-)

diff --git a/drivers/platform/x86/acer-wmi.c b/drivers/platform/x86/acer-wmi.c
index 28288ad..387ac7c 100644
--- a/drivers/platform/x86/acer-wmi.c
+++ b/drivers/platform/x86/acer-wmi.c
@@ -310,6 +310,17 @@ static struct dmi_system_id __devinitdata acer_blacklist[] = {
 			DMI_MATCH(DMI_PRODUCT_NAME, "AOA150"),
 		},
 	},
+	/*
+	 * The Lenovo Ideapad S12 have GUID that matches with AMW0 - disable it.
+	 * There already have ideapad-laptop driver to support Ideapad serial.
+	 */
+	{
+		.ident = "Lenovo Ideapad S12",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "20021,2959"),
+		},
+	},
 	{}
 };
 
-- 
1.6.0.2

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Ville-Pekka Vainio]

Hi Joey and others,

ti, 2010-12-28 kello 21:30 -0700, Joey Lee kirjoitti:
> Does is possible you help to test this patch?

The patch works, thanks! acer-wmi is no longer loaded on the Ideapad
S12.

-- 
Ville-Pekka Vainio

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Joey Lee]

Hi Vainio, 

於 四，2010-12-30 於 00:12 +0200，Ville-Pekka Vainio 提到：
> Hi Joey and others,
> 
> ti, 2010-12-28 kello 21:30 -0700, Joey Lee kirjoitti:
> > Does is possible you help to test this patch?
> 
> The patch works, thanks! acer-wmi is no longer loaded on the Ideapad
> S12.
> 

Thank's for your help to test it, I will send out the patch again to
Carlos Corbacho and other platform driver experts for review.

Joey Lee

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Carlos Corbacho]

On Wednesday 29 December 2010 04:30:03 Joey Lee wrote:

[..]

> Maybe Wistron is the ODM (or OEM) to manufacture Lenovo Ideapad S12 ?
> Anyway,
> I added this Ideapad S12 to blacklist in acer-wmi driver because it must
> supported by ideapad-laptop driver.
> 
> I also added Cc. to Carlos Corbacho, need his review.

I suspect this is the same as all the Acer netbooks - they provide a dummy WMI 
interface which doesn't actually do anything. It would be nicer if there were 
a more generic way to do this, but this will do for now (I may try to play 
around with my Aspire One and see if I can figure out some way to detect the 
dummy interface, without needing to resort to DMI).

[..]

> From: Lee, Chun-Yi <jlee@novell.com>
> Date: Wed, 29 Dec 2010 00:17:23 +0800
> Subject: [PATCH] Add Lenovo Ideapad S12 to acer blacklist
> 
> The Lenovo Ideapad S12's _WDG have GUID that matches with AMW0, it causes
> acer-wmi driver auto-probe when system boot but there already have
> ideapad-laptop driver to support Ideapad serial.
> So, add Lenovo Ideapad S12 to acer blacklist to disable it.
> 
> Signed-off-by: Lee, Chun-Yi <jlee@novell.com>
> Cc: Carlos Corbacho <carlos@strangeworlds.co.uk>

Acked-by: Carlos Corbacho <carlos@strangeworlds.co.uk>
> ---
>  drivers/platform/x86/acer-wmi.c |   11 +++++++++++
>  1 files changed, 11 insertions(+), 0 deletions(-)
> 
> diff --git a/drivers/platform/x86/acer-wmi.c
> b/drivers/platform/x86/acer-wmi.c index 28288ad..387ac7c 100644
> --- a/drivers/platform/x86/acer-wmi.c
> +++ b/drivers/platform/x86/acer-wmi.c
> @@ -310,6 +310,17 @@ static struct dmi_system_id __devinitdata
> acer_blacklist[] = { DMI_MATCH(DMI_PRODUCT_NAME, "AOA150"),
>  		},
>  	},
> +	/*
> +	 * The Lenovo Ideapad S12 have GUID that matches with AMW0 - disable it.
> +	 * There already have ideapad-laptop driver to support Ideapad serial.
> +	 */
> +	{
> +		.ident = "Lenovo Ideapad S12",
> +		.matches = {
> +			DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
> +			DMI_MATCH(DMI_PRODUCT_NAME, "20021,2959"),
> +		},
> +	},
>  	{}
>  };

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Joey Lee]

Add Cc. David Woodhouse

Hi Carlos, 

於 四，2010-12-30 於 10:04 +0000，Carlos Corbacho 提到：
> On Wednesday 29 December 2010 04:30:03 Joey Lee wrote:
> 
> [..]
> 
> > Maybe Wistron is the ODM (or OEM) to manufacture Lenovo Ideapad S12 ?
> > Anyway,
> > I added this Ideapad S12 to blacklist in acer-wmi driver because it must
> > supported by ideapad-laptop driver.
> > 
> > I also added Cc. to Carlos Corbacho, need his review.
> 
> I suspect this is the same as all the Acer netbooks - they provide a dummy WMI 
> interface which doesn't actually do anything. It would be nicer if there were 
> a more generic way to do this, but this will do for now (I may try to play 
> around with my Aspire One and see if I can figure out some way to detect the 
> dummy interface, without needing to resort to DMI).
> 

The issue reporter Vainio help to dump the acpi informatoin and put
here:
> - acpidump > acpidump.dat
http://vpv.kapsi.fi/ideapad-s12/acpidump.dat

There have AMW0 declare in DSDT:

        Device (AMW0)
        {  
            Name (_HID, "*pnp0c14")
            Name (_UID, 0x00)
            Name (WLMP, 0x00)
            Name (WMID, 0x00)
            Name (B0ED, Buffer (0x04)

Per Vainio's report, the wlan rfkill cann't set to un-block when he used
acer-wmi driver, but ideapad-laptop driver handle the killswitches
better.

Sorry,
I have no ideapad S12 machine for double check. Need David Woodhouse or
Ville-Pekka Vainio's confirm the ideapad-laptop driver supports Ideapad
S12 netbook.

I just simply checked the ideapad-laptop source code, it handle the
killswitch state by access EC address, if it works, then acer-wmi don't
need probe and touch the same function. So, I thought still need add
S12's dmi information to blacklist in acer-wmi to avoid conflict.

> [..]
> 
> > From: Lee, Chun-Yi <jlee@novell.com>
> > Date: Wed, 29 Dec 2010 00:17:23 +0800
> > Subject: [PATCH] Add Lenovo Ideapad S12 to acer blacklist
> > 
> > The Lenovo Ideapad S12's _WDG have GUID that matches with AMW0, it causes
> > acer-wmi driver auto-probe when system boot but there already have
> > ideapad-laptop driver to support Ideapad serial.
> > So, add Lenovo Ideapad S12 to acer blacklist to disable it.
> > 
> > Signed-off-by: Lee, Chun-Yi <jlee@novell.com>
> > Cc: Carlos Corbacho <carlos@strangeworlds.co.uk>
> 
> Acked-by: Carlos Corbacho <carlos@strangeworlds.co.uk>

Thank's a lot for your acked.

Joey Lee

> > ---
> >  drivers/platform/x86/acer-wmi.c |   11 +++++++++++
> >  1 files changed, 11 insertions(+), 0 deletions(-)
> > 
> > diff --git a/drivers/platform/x86/acer-wmi.c
> > b/drivers/platform/x86/acer-wmi.c index 28288ad..387ac7c 100644
> > --- a/drivers/platform/x86/acer-wmi.c
> > +++ b/drivers/platform/x86/acer-wmi.c
> > @@ -310,6 +310,17 @@ static struct dmi_system_id __devinitdata
> > acer_blacklist[] = { DMI_MATCH(DMI_PRODUCT_NAME, "AOA150"),
> >  		},
> >  	},
> > +	/*
> > +	 * The Lenovo Ideapad S12 have GUID that matches with AMW0 - disable it.
> > +	 * There already have ideapad-laptop driver to support Ideapad serial.
> > +	 */
> > +	{
> > +		.ident = "Lenovo Ideapad S12",
> > +		.matches = {
> > +			DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
> > +			DMI_MATCH(DMI_PRODUCT_NAME, "20021,2959"),
> > +		},
> > +	},
> >  	{}
> >  };

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Matthew Garrett]

It looks like this GUID is present in a surprisingly wide range of 
devices (eg, http://pastebin.com/8BY0Dg64 from a Vaio...), which means 
that it's probably not a safe assumption that reading from the EC will 
give us reasonable results. Do the Acer drivers definitely do that, 
rather than there being some SMI that returns the data?

-- 
Matthew Garrett | mjg59@srcf.ucam.org

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Joey Lee]

Hi Carlos, 

Sorry bother you for this old issue.

於 一，2011-01-03 於 20:49 +0000，Matthew Garrett 提到：
> It looks like this GUID is present in a surprisingly wide range of 
> devices (eg, http://pastebin.com/8BY0Dg64 from a Vaio...), which means 
> that it's probably not a safe assumption that reading from the EC will 
> give us reasonable results. Do the Acer drivers definitely do that, 
> rather than there being some SMI that returns the data?
> 

The AMW0 in acer-wmi not just used by acer machines:

#define AMW0_GUID1              "67C3371D-95A3-4C37-BB61-DD47B491DAAB"

It was also included in Lenovo Ideapad S12 and Vaio machines' DSDT. For
Lenovo Ideapad, that will be better handle by ideapad-laptop driver. 

Could we just direct add those model to the blacklist in acer-wmi
driver? or we direct check dmi in acer-wmi for only support Acer
machines that were included AMW0 in DSDT?

Appreciate if you can give any suggestion.


Thank's a lot!
Joey Lee

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Carlos Corbacho]

On Wednesday 02 Mar 2011 08:53:55 Joey Lee wrote:
> 於 一，2011-01-03 於 20:49 +0000，Matthew Garrett 提到：
> 
> > It looks like this GUID is present in a surprisingly wide range of
> > devices (eg, http://pastebin.com/8BY0Dg64 from a Vaio...), which means
> > that it's probably not a safe assumption that reading from the EC will
> > give us reasonable results. Do the Acer drivers definitely do that,
> > rather than there being some SMI that returns the data?
> 
> The AMW0 in acer-wmi not just used by acer machines:
> 
> #define AMW0_GUID1              "67C3371D-95A3-4C37-BB61-DD47B491DAAB"
> 
> It was also included in Lenovo Ideapad S12 and Vaio machines' DSDT. For
> Lenovo Ideapad, that will be better handle by ideapad-laptop driver.

They might not be - Matthew's original question still stands. The old AMW0 was 
just a funny wrapper for direct BIOS calls - maybe we just need a different 
call here?

What does Acer do - are they poking the EC directly, or is the call to AMW0 
just different for these machines?

> Could we just direct add those model to the blacklist in acer-wmi
> driver? or we direct check dmi in acer-wmi for only support Acer
> machines that were included AMW0 in DSDT?

Interesting. If we can't autodetect whether it's a dummy interface or not, 
then, we should change acer-wmi to whitelist what it does support - Acer, 
Fujistu Siemens and Medion - rather than trying to blacklist every vendor that 
crops up.

-Carlos

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Ike Panhc]

Hi,

Here[1] is the output of wmidump for ideapad s10-3 and s12-via and there is no
WMI device of ideapad B550 and Y530.

[1] http://pastebin.ubuntu.com/574793/

For now, ideapad-laptop handles the VPC2004 devices that can be found in most of
the ideapads (maybe all, but I can not sure). In my opinion, I do not think to
handle WMI in ideapad-laptop is reasonable unless we have an extra function on
ideapads and we can only control it with WMI driver.

On 03/03/2011 05:09 AM, Carlos Corbacho wrote:
> On Wednesday 02 Mar 2011 08:53:55 Joey Lee wrote:
>> 於 一，2011-01-03 於 20:49 +0000，Matthew Garrett 提到：
>>
>>> It looks like this GUID is present in a surprisingly wide range of
>>> devices (eg, http://pastebin.com/8BY0Dg64 from a Vaio...), which means
>>> that it's probably not a safe assumption that reading from the EC will
>>> give us reasonable results. Do the Acer drivers definitely do that,
>>> rather than there being some SMI that returns the data?
>>
>> The AMW0 in acer-wmi not just used by acer machines:
>>
>> #define AMW0_GUID1              "67C3371D-95A3-4C37-BB61-DD47B491DAAB"
>>
>> It was also included in Lenovo Ideapad S12 and Vaio machines' DSDT. For
>> Lenovo Ideapad, that will be better handle by ideapad-laptop driver.
> 
> They might not be - Matthew's original question still stands. The old AMW0 was 
> just a funny wrapper for direct BIOS calls - maybe we just need a different 
> call here?
> 
> What does Acer do - are they poking the EC directly, or is the call to AMW0 
> just different for these machines?
> 
>> Could we just direct add those model to the blacklist in acer-wmi
>> driver? or we direct check dmi in acer-wmi for only support Acer
>> machines that were included AMW0 in DSDT?
> 
> Interesting. If we can't autodetect whether it's a dummy interface or not, 
> then, we should change acer-wmi to whitelist what it does support - Acer, 
> Fujistu Siemens and Medion - rather than trying to blacklist every vendor that 
> crops up.
> 
> -Carlos
> --
> To unsubscribe from this list: send the line "unsubscribe platform-driver-x86" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Joey Lee]

Hi Ike, 

於 四，2011-03-03 於 11:17 +0800，Ike Panhc 提到：
> Hi,
> 
> Here[1] is the output of wmidump for ideapad s10-3 and s12-via and there is no
> WMI device of ideapad B550 and Y530.
> 
> [1] http://pastebin.ubuntu.com/574793/
> 
> For now, ideapad-laptop handles the VPC2004 devices that can be found in most of
> the ideapads (maybe all, but I can not sure). In my opinion, I do not think to
> handle WMI in ideapad-laptop is reasonable unless we have an extra function on
> ideapads and we can only control it with WMI driver.
> 

Yes, I just simply check the Ideapad S12's DSDT that was dumped by
Ville-Pekka Vaino, it also declare VPC2004 devices in EC block.

And, 
per Vaino's mail, he said ideapad-laptop handle the killswitches better
on Ideapad S12.

Did you have any plan to add the module alias for all known machines
that were support VPC2004 devices?


Thank's
Joey Lee

> On 03/03/2011 05:09 AM, Carlos Corbacho wrote:
> > On Wednesday 02 Mar 2011 08:53:55 Joey Lee wrote:
> >> 於 一，2011-01-03 於 20:49 +0000，Matthew Garrett 提到：
> >>
> >>> It looks like this GUID is present in a surprisingly wide range of
> >>> devices (eg, http://pastebin.com/8BY0Dg64 from a Vaio...), which means
> >>> that it's probably not a safe assumption that reading from the EC will
> >>> give us reasonable results. Do the Acer drivers definitely do that,
> >>> rather than there being some SMI that returns the data?
> >>
> >> The AMW0 in acer-wmi not just used by acer machines:
> >>
> >> #define AMW0_GUID1              "67C3371D-95A3-4C37-BB61-DD47B491DAAB"
> >>
> >> It was also included in Lenovo Ideapad S12 and Vaio machines' DSDT. For
> >> Lenovo Ideapad, that will be better handle by ideapad-laptop driver.
> > 
> > They might not be - Matthew's original question still stands. The old AMW0 was 
> > just a funny wrapper for direct BIOS calls - maybe we just need a different 
> > call here?
> > 
> > What does Acer do - are they poking the EC directly, or is the call to AMW0 
> > just different for these machines?
> > 
> >> Could we just direct add those model to the blacklist in acer-wmi
> >> driver? or we direct check dmi in acer-wmi for only support Acer
> >> machines that were included AMW0 in DSDT?
> > 
> > Interesting. If we can't autodetect whether it's a dummy interface or not, 
> > then, we should change acer-wmi to whitelist what it does support - Acer, 
> > Fujistu Siemens and Medion - rather than trying to blacklist every vendor that 
> > crops up.
> > 
> > -Carlos
> > --
> > To unsubscribe from this list: send the line "unsubscribe platform-driver-x86" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 
> --
> To unsubscribe from this list: send the line "unsubscribe platform-driver-x86" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Ike Panhc]

On 03/03/2011 02:50 PM, Joey Lee wrote:
> Hi Ike, 
> 
> 
> Yes, I just simply check the Ideapad S12's DSDT that was dumped by
> Ville-Pekka Vaino, it also declare VPC2004 devices in EC block.
> 
> And, 
> per Vaino's mail, he said ideapad-laptop handle the killswitches better
> on Ideapad S12.
> 
> Did you have any plan to add the module alias for all known machines
> that were support VPC2004 devices?

I believe its already done. The ideapad-laptop driver will bind on
ACPI:VPC2004 device.

In drivers/platform/x86/ideapad-laptop.c

static const struct acpi_device_id ideapad_device_ids[] = {
        { "VPC2004", 0},
        { "", 0},
};
MODULE_DEVICE_TABLE(acpi, ideapad_device_ids);

> 
> 
> Thank's
> Joey Lee
> 

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Joey Lee]

Hi Carios, 

於 三，2011-03-02 於 21:09 +0000，Carlos Corbacho 提到：
> On Wednesday 02 Mar 2011 08:53:55 Joey Lee wrote:
> > 於 一，2011-01-03 於 20:49 +0000，Matthew Garrett 提到：
> > 
> > > It looks like this GUID is present in a surprisingly wide range of
> > > devices (eg, http://pastebin.com/8BY0Dg64 from a Vaio...), which means
> > > that it's probably not a safe assumption that reading from the EC will
> > > give us reasonable results. Do the Acer drivers definitely do that,
> > > rather than there being some SMI that returns the data?
> > 
> > The AMW0 in acer-wmi not just used by acer machines:
> > 
> > #define AMW0_GUID1              "67C3371D-95A3-4C37-BB61-DD47B491DAAB"
> > 
> > It was also included in Lenovo Ideapad S12 and Vaio machines' DSDT. For
> > Lenovo Ideapad, that will be better handle by ideapad-laptop driver.
> 

The original issue was reported by Ville-Pekka Vaino, acer-wmi driver
auto-probe on Ideapad S12 machine by detection AWM0. But, acer-wmi adds
a soft blocked wifi killswitch, which can't be unblocked with rfkill, it
causes problems with NetworkManager wireless network can't be enabled at
all.

> They might not be - Matthew's original question still stands. The old AMW0 was 
> just a funny wrapper for direct BIOS calls - maybe we just need a different 
> call here?
> 

Understood your point.

In acer-wmi driver, by the default ACER_CAP_WIRELESS = 0, looks this
case not work for Ideapad S12.
If we want to support S12 in acer-wmi for maintain the wlan rfkill, then
we need find out the new EC pattern to read the wireless state and add
it to acer-wmi.

We need Vaino help to run acer_ec.pl to capture the EC register status.

> What does Acer do - are they poking the EC directly, or is the call to AMW0 
> just different for these machines?
> 

You are right, AMW0 just a wrapper the detail wlan rfkill state sync by
direct access EC register, we will need to know how the S12's EC
registers state changed.

On the other hand,
Vaino said the ideapad-laptop driver works fine on his S12 machine to
handle killswitches.

Per Per Ike's information, ideapad-laptop driver support S12 machine by
VPC2004 devices and I also simply check Ideapad S12's DSDT that was
dumped by Ville-Pekka Vaino, it declare VPC2004 devices in EC block.

The ideapad-laptop also maintain a wlan killswitch state by access
VPC2004 devices, but this driver have no alias and doesn't auto-probe.

Should we still maintain a wlan killswitch by acer-wmi when
ideapad-laptop also maintain the same wlan killswitch by itself?

> > Could we just direct add those model to the blacklist in acer-wmi
> > driver? or we direct check dmi in acer-wmi for only support Acer
> > machines that were included AMW0 in DSDT?
> 
> Interesting. If we can't autodetect whether it's a dummy interface or not, 
> then, we should change acer-wmi to whitelist what it does support - Acer, 
> Fujistu Siemens and Medion - rather than trying to blacklist every vendor that 
> crops up.
> 
> -Carlos

I thought there have 2 choices:

- In acer-wmi driver, maybe we can walk through acpi device to check if
there have VPC2004 devices in DSDT, then we direct ignore this machine's
AWM0 supported.

or
- We still try to support Ideapad S12 in acer-wmi, we maintain a wlan
killswitch even it duplicate with ideapad-laptop's wlan killswitch.
We need Vaino help to run acer_ec.pl.

Do you think which one is better for this case?


Thank's a lot!
Joey Lee

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Joey Lee]

於 四，2011-03-03 於 15:11 +0800，Ike Panhc 提到：
> On 03/03/2011 02:50 PM, Joey Lee wrote:
> > Hi Ike, 
> > 
> > 
> > Yes, I just simply check the Ideapad S12's DSDT that was dumped by
> > Ville-Pekka Vaino, it also declare VPC2004 devices in EC block.
> > 
> > And, 
> > per Vaino's mail, he said ideapad-laptop handle the killswitches better
> > on Ideapad S12.
> > 
> > Did you have any plan to add the module alias for all known machines
> > that were support VPC2004 devices?
> 
> I believe its already done. The ideapad-laptop driver will bind on
> ACPI:VPC2004 device.
> 
> In drivers/platform/x86/ideapad-laptop.c
> 
> static const struct acpi_device_id ideapad_device_ids[] = {
>         { "VPC2004", 0},
>         { "", 0},
> };
> MODULE_DEVICE_TABLE(acpi, ideapad_device_ids);
> 

OH! yes, you are right. I lost it!


Thank's a lot!
Joey Lee

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Matthew Garrett]

Looking at this, it seems that the affected machines have AMW0_GUID1 but 
none of the WMID GUIDs. By the looks of things this is just a 
semi-standardised way to trigger an SMM trap and communicate with the 
firmware. This makes things tricker to track down (because once we're in 
SMM we have no way of knowing what the firwmare's doing), but 
potentially there's some method to identify the machine here. Carlos, do 
we have any idea what the inputs and outputs for this method mean other 
than what's in AMW0_set_capabilities()?

-- 
Matthew Garrett | mjg59@srcf.ucam.org

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Joey Lee]

於 五，2011-03-04 於 23:59 +0000，Matthew Garrett 提到：
> Looking at this, it seems that the affected machines have AMW0_GUID1 but 
> none of the WMID GUIDs. By the looks of things this is just a 
> semi-standardised way to trigger an SMM trap and communicate with the 
> firmware. This makes things tricker to track down (because once we're in 
> SMM we have no way of knowing what the firwmare's doing), but 
> potentially there's some method to identify the machine here. Carlos, do 
> we have any idea what the inputs and outputs for this method mean other 
> than what's in AMW0_set_capabilities()?
> 

Lenovo S12 have AMW0_GUID2 like Fujitsu Siemens Amilo Li2732, direct
assign wireless capability when detected AMW0_GUID2 in acer-wmi for
Fujitsu Li2732, but the same EC cmmand doesn't work for Lenovo S12.


Thank's
Joey Lee

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Ozan Çağlayan]

03-03-2011 09:23 tarihinde, Joey Lee yazdı:
> The original issue was reported by Ville-Pekka Vaino, acer-wmi driver
> auto-probe on Ideapad S12 machine by detection AWM0. But, acer-wmi adds
> a soft blocked wifi killswitch, which can't be unblocked with rfkill, it
> causes problems with NetworkManager wireless network can't be enabled at
> all.

This also happens with Lenovo 3000 V200 and Lenovo b560. I'm adding them 
to blacklist with the Ideapad S12
to fix users' no-wlan-at-all complains. Maybe you can just blacklist 
them for stable kernel releases and try to find a proper solution after.

Thanks,
Ozan Caglayan

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Joey Lee]

於 四，2011-03-10 於 08:23 +0200，Ozan Çağlayan 提到：
> 03-03-2011 09:23 tarihinde, Joey Lee yazdı:
> > The original issue was reported by Ville-Pekka Vaino, acer-wmi driver
> > auto-probe on Ideapad S12 machine by detection AWM0. But, acer-wmi adds
> > a soft blocked wifi killswitch, which can't be unblocked with rfkill, it
> > causes problems with NetworkManager wireless network can't be enabled at
> > all.
> 
> This also happens with Lenovo 3000 V200 and Lenovo b560. I'm adding them 
> to blacklist with the Ideapad S12
> to fix users' no-wlan-at-all complains. Maybe you can just blacklist 
> them for stable kernel releases and try to find a proper solution after.
> 
> Thanks,
> Ozan Caglayan
> 
> 

My idea is direct ignore all machines that have VPC2004 acpi device in
DSDT, because ideapad-laptop driver talks with BIOS though VPC2004.

Did you see VPC2004 device declare in Lenovo 3000 V200 and b560's DSDT?


Thank's a lot!
Joey Lee

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Ozan Çağlayan]

10-03-2011 08:36 tarihinde, Joey Lee yazdı:
> My idea is direct ignore all machines that have VPC2004 acpi device in
> DSDT, because ideapad-laptop driver talks with BIOS though VPC2004.
>
> Did you see VPC2004 device declare in Lenovo 3000 V200 and b560's DSDT?

Unfortunately I can't see any VPC2004 references in 3000 v200's dsdt 
which is at:

http://cekirdek.pardus.org.tr/~ozan/ivir/lenovo3000v200_dsdt.dsl

I don't have the other one's acpidump but I'll try to get it from the user.

Thanks.

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Joey Lee]

於 四，2011-03-10 於 08:58 +0200，Ozan Çağlayan 提到：
> 10-03-2011 08:36 tarihinde, Joey Lee yazdı:
> > My idea is direct ignore all machines that have VPC2004 acpi device in
> > DSDT, because ideapad-laptop driver talks with BIOS though VPC2004.
> >
> > Did you see VPC2004 device declare in Lenovo 3000 V200 and b560's DSDT?
> 
> Unfortunately I can't see any VPC2004 references in 3000 v200's dsdt 
> which is at:
> 
> http://cekirdek.pardus.org.tr/~ozan/ivir/lenovo3000v200_dsdt.dsl
> 
> I don't have the other one's acpidump but I'll try to get it from the user.
> 
> Thanks.
> 

OK, there have no VPC2004 and VPCW, VPCR method in your DSDT, that means
ideapad-laptop doesn't support 3000 V200 and b560.

If we are lucky, we can find out which EC register mapping to wifi key's
status.

If you want...
Could you please help to try Carlos's acer_ec.pl tool?

http://code.google.com/p/aceracpi/wiki/EmbeddedController

Download the acer_ec.pl file then put to your Lenovo machine.
Run:
	watch -n 1 perl acer_ec.pl regs

Start to press wifi Fn key on Lenovo machines' keyboard, then monitor
the result from acer_ec. Please press wifi Fn key a couple of times,
fine out which register changed when you press Fn key.

Please send out the register number and value when wifi on/off.
e.g.
	E0 04		128 (on)	144 (off)


Thank's a lot!
Joey Lee

Re: acer-wmi is unnecessarily loaded on a Lenovo Ideapad S12

[Ozan Çağlayan]

10-03-2011 09:45 tarihinde, Joey Lee yazdı:
> OK, there have no VPC2004 and VPCW, VPCR method in your DSDT, that means
> ideapad-laptop doesn't support 3000 V200 and b560.
>
> If we are lucky, we can find out which EC register mapping to wifi key's
> status.
>
> If you want...
> Could you please help to try Carlos's acer_ec.pl tool?
>
> http://code.google.com/p/aceracpi/wiki/EmbeddedController
>
> Download the acer_ec.pl file then put to your Lenovo machine.
> Run:
> 	watch -n 1 perl acer_ec.pl regs
>
> Start to press wifi Fn key on Lenovo machines' keyboard, then monitor
> the result from acer_ec. Please press wifi Fn key a couple of times,
> fine out which register changed when you press Fn key.
>
> Please send out the register number and value when wifi on/off.
> e.g.

OK will tell the users to try this if they can. BTW, note that I didn't 
send the dsdt of b560, it may contain
the ACPI device that you wanted to know.

Also, the user didn't complain about a non-functional Wifi Fn key. If 
everything works OK for him, should we bother about anything at all?

Thanks