[lm-sensors] ATK0110 on older Asus board

[lm-sensors] ATK0110 on older Asus board

[Karl Newman]

[-- Attachment #1.1: Type: text/plain, Size: 896 bytes --]

Hello,

I have a 4-year-old Asus A8N-VM CSM motherboard, which has a W83627EHG chip
for hardware monitoring. I recently upgraded to kernel 2.6.31 and ran into
the "Device or Resource busy" error when trying to load my existing
w83627ehf module. After some research, I discovered that I should use the
asus_atk0110 module instead. So, I updated my kernel config to build that
module and it will load with no problem, but "sensors" is ignoring it. When
I modprobe the module, it silently succeeds (it's listed in lsmod) but there
are no messages in dmesg or the syslog to indicate any activity from the
module, which makes me suspect that my board is not currently supported by
this module. I did confirm that it has the atk0110 chip (ls
/sys/bus/acpi/devices shows a ATK0110:00 entry). Am I doing something wrong,
or do you need a DSDT dump to support my motherboard, or...?

Thanks,

Karl Newman

[-- Attachment #1.2: Type: text/html, Size: 940 bytes --]

[-- Attachment #2: Type: text/plain, Size: 153 bytes --]

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Jean Delvare]

Hi Karl,

On Thu, 14 Jan 2010 06:56:34 -0800, Karl Newman wrote:
> I have a 4-year-old Asus A8N-VM CSM motherboard, which has a W83627EHG chip
> for hardware monitoring. I recently upgraded to kernel 2.6.31 and ran into
> the "Device or Resource busy" error when trying to load my existing
> w83627ehf module. After some research, I discovered that I should use the
> asus_atk0110 module instead. So, I updated my kernel config to build that
> module and it will load with no problem, but "sensors" is ignoring it. When
> I modprobe the module, it silently succeeds (it's listed in lsmod) but there
> are no messages in dmesg or the syslog to indicate any activity from the
> module, which makes me suspect that my board is not currently supported by
> this module. I did confirm that it has the atk0110 chip (ls
> /sys/bus/acpi/devices shows a ATK0110:00 entry). Am I doing something wrong,
> or do you need a DSDT dump to support my motherboard, or...?

Which version of lm-sensors are you running? "sensors -v" will tell
you. Minimum version for asus_atk0110 support is 3.1.0, as documented
on:
  http://www.lm-sensors.org/wiki/Devices

-- 
Jean Delvare
http://khali.linux-fr.org/wishlist.html

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Karl Newman]

[-- Attachment #1.1: Type: text/plain, Size: 1487 bytes --]

On Thu, Jan 14, 2010 at 7:19 AM, Jean Delvare <khali@linux-fr.org> wrote:

> Hi Karl,
>
> On Thu, 14 Jan 2010 06:56:34 -0800, Karl Newman wrote:
> > I have a 4-year-old Asus A8N-VM CSM motherboard, which has a W83627EHG
> chip
> > for hardware monitoring. I recently upgraded to kernel 2.6.31 and ran
> into
> > the "Device or Resource busy" error when trying to load my existing
> > w83627ehf module. After some research, I discovered that I should use the
> > asus_atk0110 module instead. So, I updated my kernel config to build that
> > module and it will load with no problem, but "sensors" is ignoring it.
> When
> > I modprobe the module, it silently succeeds (it's listed in lsmod) but
> there
> > are no messages in dmesg or the syslog to indicate any activity from the
> > module, which makes me suspect that my board is not currently supported
> by
> > this module. I did confirm that it has the atk0110 chip (ls
> > /sys/bus/acpi/devices shows a ATK0110:00 entry). Am I doing something
> wrong,
> > or do you need a DSDT dump to support my motherboard, or...?
>
> Which version of lm-sensors are you running? "sensors -v" will tell
> you. Minimum version for asus_atk0110 support is 3.1.0, as documented
> on:
>  http://www.lm-sensors.org/wiki/Devices
>
> --
> Jean Delvare
> http://khali.linux-fr.org/wishlist.html
>

Sorry, I forgot to mention that. I'm using 3.1.1 (sensors version 3.1.1 with
libsensors version 3.1.1), on Gentoo, along with kernel 2.6.31-gentoo-r6.

Karl

[-- Attachment #1.2: Type: text/html, Size: 2095 bytes --]

[-- Attachment #2: Type: text/plain, Size: 153 bytes --]

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Jean Delvare]

Hi Karl,

On Thu, 14 Jan 2010 08:12:33 -0800, Karl Newman wrote:
> On Thu, Jan 14, 2010 at 7:19 AM, Jean Delvare <khali@linux-fr.org> wrote:
> > On Thu, 14 Jan 2010 06:56:34 -0800, Karl Newman wrote:
> > > I have a 4-year-old Asus A8N-VM CSM motherboard, which has a W83627EHG
> > > chip for hardware monitoring. I recently upgraded to kernel 2.6.31 and
> > > ran into the "Device or Resource busy" error when trying to load my
> > > existing w83627ehf module. After some research, I discovered that I
> > > should use the asus_atk0110 module instead. So, I updated my kernel
> > > config to build that module and it will load with no problem, but
> > > "sensors" is ignoring it. When I modprobe the module, it silently
> > > succeeds (it's listed in lsmod) but there are no messages in dmesg or
> > > the syslog to indicate any activity from the module, which makes me
> > > suspect that my board is not currently supported by this module. I did
> > > confirm that it has the atk0110 chip (ls /sys/bus/acpi/devices shows a
> > > ATK0110:00 entry). Am I doing something wrong, or do you need a DSDT
> > > dump to support my motherboard, or...?
> >
> > Which version of lm-sensors are you running? "sensors -v" will tell
> > you. Minimum version for asus_atk0110 support is 3.1.0, as documented
> > on:
> >  http://www.lm-sensors.org/wiki/Devices
> 
> Sorry, I forgot to mention that. I'm using 3.1.1 (sensors version 3.1.1 with
> libsensors version 3.1.1), on Gentoo, along with kernel 2.6.31-gentoo-r6.

OK, that would be a kernel-side problem then. I guess that you don't
have a /sys/class/hwmon/hwmon* directory created for the asus_atk0110?

I'm adding Luca to Cc, your best chance is probably to send him your
DSDT in private and let him investigate.

If an updated BIOS exists for your motherboard, you may have to use it.

-- 
Jean Delvare
http://khali.linux-fr.org/wishlist.html

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Luca Tettamanti]

On Fri, Jan 15, 2010 at 8:43 AM, Jean Delvare <khali@linux-fr.org> wrote:
> Hi Karl,
> On Thu, 14 Jan 2010 08:12:33 -0800, Karl Newman wrote:
>> On Thu, Jan 14, 2010 at 7:19 AM, Jean Delvare <khali@linux-fr.org> wrote:
>> > On Thu, 14 Jan 2010 06:56:34 -0800, Karl Newman wrote:
>> > > I have a 4-year-old Asus A8N-VM CSM motherboard, which has a W83627EHG
>> > > chip for hardware monitoring. I recently upgraded to kernel 2.6.31 and
>> > > ran into the "Device or Resource busy" error when trying to load my
>> > > existing w83627ehf module. After some research, I discovered that I
>> > > should use the asus_atk0110 module instead. So, I updated my kernel
>> > > config to build that module and it will load with no problem, but
>> > > "sensors" is ignoring it. When I modprobe the module, it silently
>> > > succeeds (it's listed in lsmod) but there are no messages in dmesg or
>> > > the syslog to indicate any activity from the module,

The module is very quiet unless an error is detected (or debug is enabled).
I guess I could add a printk to let the user known that the driver is
loaded and functional.

>> > > which makes me
>> > > suspect that my board is not currently supported by this module. I did
>> > > confirm that it has the atk0110 chip (ls /sys/bus/acpi/devices shows a
>> > > ATK0110:00 entry). Am I doing something wrong, or do you need a DSDT
>> > > dump to support my motherboard, or...?
>> >
>> > Which version of lm-sensors are you running? "sensors -v" will tell
>> > you. Minimum version for asus_atk0110 support is 3.1.0, as documented
>> > on:
>> >  http://www.lm-sensors.org/wiki/Devices
>>
>> Sorry, I forgot to mention that. I'm using 3.1.1 (sensors version 3.1.1 with
>> libsensors version 3.1.1), on Gentoo, along with kernel 2.6.31-gentoo-r6.
>
> OK, that would be a kernel-side problem then. I guess that you don't
> have a /sys/class/hwmon/hwmon* directory created for the asus_atk0110?
>
> I'm adding Luca to Cc, your best chance is probably to send him your
> DSDT in private and let him investigate.

I've never seen the DSDT of this board, so it's possible that it's
using a variation of the interface that is not handled by the driver.

L

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Karl Newman]

[-- Attachment #1.1: Type: text/plain, Size: 2505 bytes --]

On Thu, Jan 14, 2010 at 11:43 PM, Jean Delvare <khali@linux-fr.org> wrote:

> Hi Karl,
>
> On Thu, 14 Jan 2010 08:12:33 -0800, Karl Newman wrote:
> > On Thu, Jan 14, 2010 at 7:19 AM, Jean Delvare <khali@linux-fr.org>
> wrote:
> > > On Thu, 14 Jan 2010 06:56:34 -0800, Karl Newman wrote:
> > > > I have a 4-year-old Asus A8N-VM CSM motherboard, which has a
> W83627EHG
> > > > chip for hardware monitoring. I recently upgraded to kernel 2.6.31
> and
> > > > ran into the "Device or Resource busy" error when trying to load my
> > > > existing w83627ehf module. After some research, I discovered that I
> > > > should use the asus_atk0110 module instead. So, I updated my kernel
> > > > config to build that module and it will load with no problem, but
> > > > "sensors" is ignoring it. When I modprobe the module, it silently
> > > > succeeds (it's listed in lsmod) but there are no messages in dmesg or
> > > > the syslog to indicate any activity from the module, which makes me
> > > > suspect that my board is not currently supported by this module. I
> did
> > > > confirm that it has the atk0110 chip (ls /sys/bus/acpi/devices shows
> a
> > > > ATK0110:00 entry). Am I doing something wrong, or do you need a DSDT
> > > > dump to support my motherboard, or...?
> > >
> > > Which version of lm-sensors are you running? "sensors -v" will tell
> > > you. Minimum version for asus_atk0110 support is 3.1.0, as documented
> > > on:
> > >  http://www.lm-sensors.org/wiki/Devices
> >
> > Sorry, I forgot to mention that. I'm using 3.1.1 (sensors version 3.1.1
> with
> > libsensors version 3.1.1), on Gentoo, along with kernel 2.6.31-gentoo-r6.
>
> OK, that would be a kernel-side problem then. I guess that you don't
> have a /sys/class/hwmon/hwmon* directory created for the asus_atk0110?
>
> I'm adding Luca to Cc, your best chance is probably to send him your
> DSDT in private and let him investigate.
>
> If an updated BIOS exists for your motherboard, you may have to use it.
>
>
I have a /sys/class/hwmon/hwmon0 directory, but it's for the k8temp module.
If I unload k8temp and load asus_atk0110, there's no /sys/class/hwmon
directory. I can't use a new BIOS because the newest one breaks boot-time
USB support, so I'm using one version back. I had DSDT problems with this
board with early version of the BIOS (I had to load a custom table in the
kernel to fix the errors), but later BIOS versions fixed the problem. I'm
sending my DSDT to Luca in a separate mail. Thanks for your help!

Karl Newman

[-- Attachment #1.2: Type: text/html, Size: 3274 bytes --]

[-- Attachment #2: Type: text/plain, Size: 153 bytes --]

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Luca Tettamanti]

On Fri, Jan 15, 2010 at 5:52 PM, Karl Newman <siliconfiend@gmail.com> wrote:
> The reason I thought there would be some
> dmesg or syslog entries is that I saw something like that in the mailing
> list about the ATK0110 (it looked like it was enumerating devices it found).
> The email didn't say where that text came from, though...

They comes from the driver, but only when HWMON_DEBUG_CHIP is enabled.

> Thanks for looking into this. I've attached my DSDT (dumped via cp
> /proc/acpi/dsdt A8N-VM_CSM.DSDT).

Ugh. The AKT0110 node is totally empty, nonetheless the hwmon methods
are available.
The driver is failing because the method used to identify the mobo
(MBIF) is missing (and the printk isn't verbose enough).
I suppose I could relax the check and probe the interface even if it's
not possible to id the board.
I'll prepare a patch.

Luca

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Jean Delvare]

Hi Luca,

On Fri, 15 Jan 2010 18:11:53 +0100, Luca Tettamanti wrote:
> On Fri, Jan 15, 2010 at 5:52 PM, Karl Newman <siliconfiend@gmail.com> wrote:
> > The reason I thought there would be some
> > dmesg or syslog entries is that I saw something like that in the mailing
> > list about the ATK0110 (it looked like it was enumerating devices it found).
> > The email didn't say where that text came from, though...
> 
> They comes from the driver, but only when HWMON_DEBUG_CHIP is enabled.
> 
> > Thanks for looking into this. I've attached my DSDT (dumped via cp
> > /proc/acpi/dsdt A8N-VM_CSM.DSDT).
> 
> Ugh. The AKT0110 node is totally empty, nonetheless the hwmon methods
> are available.
> The driver is failing because the method used to identify the mobo
> (MBIF) is missing (and the printk isn't verbose enough).
> I suppose I could relax the check and probe the interface even if it's
> not possible to id the board.
> I'll prepare a patch.

Isn't this similar to the problem (and patch!) reported by Juan RP 3
days ago?

http://lists.lm-sensors.org/pipermail/lm-sensors/2010-January/027606.html

-- 
Jean Delvare

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Karl Newman]

[-- Attachment #1.1: Type: text/plain, Size: 1447 bytes --]

On Fri, Jan 15, 2010 at 9:30 AM, Jean Delvare <khali@linux-fr.org> wrote:

> Hi Luca,
>
> On Fri, 15 Jan 2010 18:11:53 +0100, Luca Tettamanti wrote:
> > On Fri, Jan 15, 2010 at 5:52 PM, Karl Newman <siliconfiend@gmail.com>
> wrote:
> > > The reason I thought there would be some
> > > dmesg or syslog entries is that I saw something like that in the
> mailing
> > > list about the ATK0110 (it looked like it was enumerating devices it
> found).
> > > The email didn't say where that text came from, though...
> >
> > They comes from the driver, but only when HWMON_DEBUG_CHIP is enabled.
> >
> > > Thanks for looking into this. I've attached my DSDT (dumped via cp
> > > /proc/acpi/dsdt A8N-VM_CSM.DSDT).
> >
> > Ugh. The AKT0110 node is totally empty, nonetheless the hwmon methods
> > are available.
> > The driver is failing because the method used to identify the mobo
> > (MBIF) is missing (and the printk isn't verbose enough).
> > I suppose I could relax the check and probe the interface even if it's
> > not possible to id the board.
> > I'll prepare a patch.
>
> Isn't this similar to the problem (and patch!) reported by Juan RP 3
> days ago?
>
> http://lists.lm-sensors.org/pipermail/lm-sensors/2010-January/027606.html
>
> --
> Jean Delvare
>

That patch worked great, even on 2.6.31-gentoo-r6 (offset -222 lines). Now
the module loads and 'sensors' shows all the connected devices. Thanks for
all your help!

Sincerely,

Karl Newman

[-- Attachment #1.2: Type: text/html, Size: 2152 bytes --]

[-- Attachment #2: Type: text/plain, Size: 153 bytes --]

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Luca Tettamanti]

On Fri, Jan 15, 2010 at 6:30 PM, Jean Delvare <khali@linux-fr.org> wrote:
> Hi Luca,
>
> On Fri, 15 Jan 2010 18:11:53 +0100, Luca Tettamanti wrote:
>> On Fri, Jan 15, 2010 at 5:52 PM, Karl Newman <siliconfiend@gmail.com> wrote:
>> > Thanks for looking into this. I've attached my DSDT (dumped via cp
>> > /proc/acpi/dsdt A8N-VM_CSM.DSDT).
>>
>> Ugh. The AKT0110 node is totally empty, nonetheless the hwmon methods
>> are available.
>> The driver is failing because the method used to identify the mobo
>> (MBIF) is missing (and the printk isn't verbose enough).
>> I suppose I could relax the check and probe the interface even if it's
>> not possible to id the board.
>> I'll prepare a patch.
>
> Isn't this similar to the problem (and patch!) reported by Juan RP 3
> days ago?
>
> http://lists.lm-sensors.org/pipermail/lm-sensors/2010-January/027606.html

Yes, same problem. I haven't been able to decode MBIF data, so ATM it's unused.
I'll send a patch in reply to this mail. Jean can you please push it
to Linus for .34?

Luca

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Jean Delvare]

On Sun, 17 Jan 2010 22:58:02 +0100, Luca Tettamanti wrote:
> On Fri, Jan 15, 2010 at 6:30 PM, Jean Delvare <khali@linux-fr.org> wrote:
> > Hi Luca,
> >
> > On Fri, 15 Jan 2010 18:11:53 +0100, Luca Tettamanti wrote:
> >> Ugh. The AKT0110 node is totally empty, nonetheless the hwmon methods
> >> are available.
> >> The driver is failing because the method used to identify the mobo
> >> (MBIF) is missing (and the printk isn't verbose enough).
> >> I suppose I could relax the check and probe the interface even if it's
> >> not possible to id the board.
> >> I'll prepare a patch.
> >
> > Isn't this similar to the problem (and patch!) reported by Juan RP 3
> > days ago?
> >
> > http://lists.lm-sensors.org/pipermail/lm-sensors/2010-January/027606.html
> 
> Yes, same problem. I haven't been able to decode MBIF data, so ATM it's unused.
> I'll send a patch in reply to this mail. Jean can you please push it
> to Linus for .34?

I'll push whatever patch is needed to 2.6._33_.

-- 
Jean Delvare

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Luca Tettamanti]

On Mon, Jan 18, 2010 at 9:03 AM, Jean Delvare <khali@linux-fr.org> wrote:
> On Sun, 17 Jan 2010 22:58:02 +0100, Luca Tettamanti wrote:
>> Yes, same problem. I haven't been able to decode MBIF data, so ATM it's unused.
>> I'll send a patch in reply to this mail. Jean can you please push it
>> to Linus for .34?
>
> I'll push whatever patch is needed to 2.6._33_.

Of course :P

thanks,
Luca

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Luca Tettamanti]

On Wed, Feb 10, 2010 at 8:39 PM, frollic nilsson <frollic@gmail.com> wrote:
> Sorry to disturb you like this, but I stumbled across a quick fix I
> belive you wrote for the ATK0110 for an older Asus mobo -
> http://lists.lm-sensors.org/pipermail/lm-sensors/2010-January/027647.html
> I have the same issue, but in my case the /sys/class/hwmon doesn't get
> populated even if I apply the fix in the link.
>
> I wonder if you would have time to look into this matter ?

Sure. I need a copy of the DSDT table: /sys/firmware/acpi/tables/DSDT

Luca

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Luca Tettamanti]

On Thu, Feb 11, 2010 at 8:47 AM, frollic nilsson <frollic@gmail.com> wrote:
> On Wed, Feb 10, 2010 at 11:42 PM, Luca Tettamanti <kronos.it@gmail.com> wrote:
>> On Wed, Feb 10, 2010 at 8:39 PM, frollic nilsson <frollic@gmail.com> wrote:
>>> Sorry to disturb you like this, but I stumbled across a quick fix I
>>> belive you wrote for the ATK0110 for an older Asus mobo -
>>> http://lists.lm-sensors.org/pipermail/lm-sensors/2010-January/027647.html
>>> I have the same issue, but in my case the /sys/class/hwmon doesn't get
>>> populated even if I apply the fix in the link.
>>>
>>> I wonder if you would have time to look into this matter ?
>>
>> Sure. I need a copy of the DSDT table: /sys/firmware/acpi/tables/DSDT
[...]
> DSDT file attached.

Hum, everything seems in order; please try loading the module with
HWMON_DEBUG_CHIP enabled (or just add a #define DEBUG at the top of
the driver).

Luca
(keep the list in CC)

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Luca Tettamanti]

Do not strip the CC!

On Thu, Feb 11, 2010 at 04:36:53PM +0100, frollic nilsson wrote:
> Well, I tried adding the #define DEBUG in the drivers/hwmon/asus_atk0110.c and
> recompiling, but there wasn't any more output in the dmesg/messages.
> They don't even say the driver gets loaded, but it's there if I list
> it using lsmod.

Ok, this means that ACPI core is not even calling the add method.
I guess that I can happen if _STA reports that the device is not functional.

Please apply this patch:

diff --git a/drivers/hwmon/asus_atk0110.c b/drivers/hwmon/asus_atk0110.c
index 028284f..a9e07e4 100644
--- a/drivers/hwmon/asus_atk0110.c
+++ b/drivers/hwmon/asus_atk0110.c
@@ -1405,6 +1405,14 @@ static int atk_remove(struct acpi_device *device, int type)
 static int __init atk0110_init(void)
 {
 	int ret;
+	acpi_status status;
+	u64 val;
+
+	status = acpi_evaluate_integer(NULL, "\\AIDI", NULL, &val);
+	if (status != AE_OK)
+		printk("Failed to eval AIDI: %s\n", acpi_format_exception(status));
+	else
+		printk("AIDI = %llu\n", val);
 
 	ret = acpi_bus_register_driver(&atk_driver);
 	if (ret)

Luca

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[frollic nilsson]

sorry, missed the reply to all button :\

with the extra code inserted, I get a AIDI = 1 in the logs

On Thu, Feb 11, 2010 at 5:25 PM, Luca Tettamanti <kronos.it@gmail.com> wrote:
> Do not strip the CC!
>
> On Thu, Feb 11, 2010 at 04:36:53PM +0100, frollic nilsson wrote:
>> Well, I tried adding the #define DEBUG in the drivers/hwmon/asus_atk0110.c and
>> recompiling, but there wasn't any more output in the dmesg/messages.
>> They don't even say the driver gets loaded, but it's there if I list
>> it using lsmod.
>
> Ok, this means that ACPI core is not even calling the add method.
> I guess that I can happen if _STA reports that the device is not functional.
>
> Please apply this patch:
>
> diff --git a/drivers/hwmon/asus_atk0110.c b/drivers/hwmon/asus_atk0110.c
> index 028284f..a9e07e4 100644
> --- a/drivers/hwmon/asus_atk0110.c
> +++ b/drivers/hwmon/asus_atk0110.c
> @@ -1405,6 +1405,14 @@ static int atk_remove(struct acpi_device *device, int type)
>  static int __init atk0110_init(void)
>  {
>        int ret;
> +       acpi_status status;
> +       u64 val;
> +
> +       status = acpi_evaluate_integer(NULL, "\\AIDI", NULL, &val);
> +       if (status != AE_OK)
> +               printk("Failed to eval AIDI: %s\n", acpi_format_exception(status));
> +       else
> +               printk("AIDI = %llu\n", val);
>
>        ret = acpi_bus_register_driver(&atk_driver);
>        if (ret)
>
> Luca
>

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Luca Tettamanti]

On Thu, Feb 11, 2010 at 9:10 PM, frollic nilsson <frollic@gmail.com> wrote:
> sorry, missed the reply to all button :\

While we are at it: don't top post :P

> On Thu, Feb 11, 2010 at 5:25 PM, Luca Tettamanti <kronos.it@gmail.com> wrote:
>> On Thu, Feb 11, 2010 at 04:36:53PM +0100, frollic nilsson wrote:
>>
>> Ok, this means that ACPI core is not even calling the add method.
>> I guess that I can happen if _STA reports that the device is not functional.
>
> with the extra code inserted, I get a AIDI = 1 in the logs

As I suspected the device is disabled by the BIOS (AI-whatever is the
commercial name of the hwmon features, I guess AIDI = AI DIsable).
You should try and see if there's a relevant option in the BIOS setup
screen (btw, is there a monitoring page? does it work?)

Luca

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Luca Tettamanti]

On Mon, Feb 15, 2010 at 4:31 PM, frollic nilsson <frollic@gmail.com> wrote:
> On Fri, Feb 12, 2010 at 11:05 AM, Luca Tettamanti <kronos.it@gmail.com> wrote:
>> On Thu, Feb 11, 2010 at 9:10 PM, frollic nilsson <frollic@gmail.com> wrote:
>>> On Thu, Feb 11, 2010 at 5:25 PM, Luca Tettamanti <kronos.it@gmail.com> wrote:
>>>> On Thu, Feb 11, 2010 at 04:36:53PM +0100, frollic nilsson wrote:
>>>>
>>>> Ok, this means that ACPI core is not even calling the add method.
>>>> I guess that I can happen if _STA reports that the device is not functional.
>>>
>>> with the extra code inserted, I get a AIDI = 1 in the logs
>>
>> As I suspected the device is disabled by the BIOS (AI-whatever is the
>> commercial name of the hwmon features, I guess AIDI = AI DIsable).
>> You should try and see if there's a relevant option in the BIOS setup
>> screen (btw, is there a monitoring page? does it work?)
>
> Yeah, I got the cpu temp and fan RPMs displayed in BIOS.

Nothing to enable the hwmon or "AI-something"?
What about the voltages? Do you see them? Is "Vcore" is spelled all upper case?

> Parts from the messages log file (I grep:ed for "ACPI"), I restarted
> the systenm three days ago:

> Feb 12 22:41:28 atlantis kernel: ACPI: I/O resource w83627ehf
> [0x295-0x296] conflicts with ACPI region HWRE [0x290-0x299]
> Feb 12 22:41:28 atlantis kernel: ACPI: If an ACPI driver is available
> for this device, you should use it instead of the native driver
> Feb 12 22:51:12 atlantis kernel: ACPI: I/O resource w83627ehf
> [0x295-0x296] conflicts with ACPI region HWRE [0x290-0x299]
> Feb 12 22:51:12 atlantis kernel: ACPI: If an ACPI driver is available
> for this device, you should use it instead of the native driver
> Feb 12 22:51:43 atlantis kernel: ACPI: I/O resource w83627ehf
> [0x295-0x296] conflicts with ACPI region HWRE [0x290-0x299]
> Feb 12 22:51:43 atlantis kernel: ACPI: If an ACPI driver is available
> for this device, you should use it instead of the native driver

The resources are reserved even if the ATK0110 device itself is disabled, weird.
This looks like a BIOS bug to me: the device is explicitly disabled,
but ACPI still clams the ioports for the monitoring chip.
I'm not sure how to handle this mobo, but clearly it's safe to use
acpi_enforce_resources=lax and load the native driver.

Luca

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[frollic nilsson]

On Mon, Feb 15, 2010 at 5:08 PM, Luca Tettamanti <kronos.it@gmail.com> wrote:
> On Mon, Feb 15, 2010 at 4:31 PM, frollic nilsson <frollic@gmail.com> wrote:
>> On Fri, Feb 12, 2010 at 11:05 AM, Luca Tettamanti <kronos.it@gmail.com> wrote:
>>> On Thu, Feb 11, 2010 at 9:10 PM, frollic nilsson <frollic@gmail.com> wrote:
>>>> On Thu, Feb 11, 2010 at 5:25 PM, Luca Tettamanti <kronos.it@gmail.com> wrote:
>>>>> On Thu, Feb 11, 2010 at 04:36:53PM +0100, frollic nilsson wrote:
>>>>>
>>>>> Ok, this means that ACPI core is not even calling the add method.
>>>>> I guess that I can happen if _STA reports that the device is not functional.
>>>>
>>>> with the extra code inserted, I get a AIDI = 1 in the logs
>>>
>>> As I suspected the device is disabled by the BIOS (AI-whatever is the
>>> commercial name of the hwmon features, I guess AIDI = AI DIsable).
>>> You should try and see if there's a relevant option in the BIOS setup
>>> screen (btw, is there a monitoring page? does it work?)
>>
>> Yeah, I got the cpu temp and fan RPMs displayed in BIOS.
>
> Nothing to enable the hwmon or "AI-something"?
> What about the voltages? Do you see them? Is "Vcore" is spelled all upper case?
>
>> Parts from the messages log file (I grep:ed for "ACPI"), I restarted
>> the systenm three days ago:
>
>> Feb 12 22:41:28 atlantis kernel: ACPI: I/O resource w83627ehf
>> [0x295-0x296] conflicts with ACPI region HWRE [0x290-0x299]
>> Feb 12 22:41:28 atlantis kernel: ACPI: If an ACPI driver is available
>> for this device, you should use it instead of the native driver
>> Feb 12 22:51:12 atlantis kernel: ACPI: I/O resource w83627ehf
>> [0x295-0x296] conflicts with ACPI region HWRE [0x290-0x299]
>> Feb 12 22:51:12 atlantis kernel: ACPI: If an ACPI driver is available
>> for this device, you should use it instead of the native driver
>> Feb 12 22:51:43 atlantis kernel: ACPI: I/O resource w83627ehf
>> [0x295-0x296] conflicts with ACPI region HWRE [0x290-0x299]
>> Feb 12 22:51:43 atlantis kernel: ACPI: If an ACPI driver is available
>> for this device, you should use it instead of the native driver
>
> The resources are reserved even if the ATK0110 device itself is disabled, weird.
> This looks like a BIOS bug to me: the device is explicitly disabled,
> but ACPI still clams the ioports for the monitoring chip.
> I'm not sure how to handle this mobo, but clearly it's safe to use
> acpi_enforce_resources=lax and load the native driver.
>
> Luca
>

Well, tried the lax as well got the same "error"

w83627ehf: Found W83627EHG chip at 0x290
ACPI: I/O resource w83627ehf [0x295-0x296] conflicts with ACPI region
HWRE [0x290-0x299]
ACPI: This conflict may cause random problems and system instability
ACPI: If an ACPI driver is available for this device, you should use
it instead of the native driver

:D

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ATK0110 on older Asus board

[Luca Tettamanti]

On Mon, Feb 22, 2010 at 9:19 PM, frollic nilsson <frollic@gmail.com> wrote:
> On Mon, Feb 15, 2010 at 5:08 PM, Luca Tettamanti <kronos.it@gmail.com> wrote:
>> On Mon, Feb 15, 2010 at 4:31 PM, frollic nilsson <frollic@gmail.com> wrote:
>>> On Fri, Feb 12, 2010 at 11:05 AM, Luca Tettamanti <kronos.it@gmail.com> wrote:
>>>> On Thu, Feb 11, 2010 at 9:10 PM, frollic nilsson <frollic@gmail.com> wrote:
>>>>> On Thu, Feb 11, 2010 at 5:25 PM, Luca Tettamanti <kronos.it@gmail.com> wrote:
>>>>>> On Thu, Feb 11, 2010 at 04:36:53PM +0100, frollic nilsson wrote:
>>>>>>
>>>>>> Ok, this means that ACPI core is not even calling the add method.
>>>>>> I guess that I can happen if _STA reports that the device is not functional.
>>>>>
>>>>> with the extra code inserted, I get a AIDI = 1 in the logs
>>>>
>>>> As I suspected the device is disabled by the BIOS (AI-whatever is the
>>>> commercial name of the hwmon features, I guess AIDI = AI DIsable).
>>>> You should try and see if there's a relevant option in the BIOS setup
>>>> screen (btw, is there a monitoring page? does it work?)
>>>
>>> Yeah, I got the cpu temp and fan RPMs displayed in BIOS.
>>
>> Nothing to enable the hwmon or "AI-something"?
>> What about the voltages? Do you see them? Is "Vcore" is spelled all upper case?

This might sound a strange question but in other case allowed me to
understand whether the ACPI interface was actually used by the BIOS or
not.

>> The resources are reserved even if the ATK0110 device itself is disabled, weird.
>> This looks like a BIOS bug to me: the device is explicitly disabled,
>> but ACPI still clams the ioports for the monitoring chip.
>> I'm not sure how to handle this mobo, but clearly it's safe to use
>> acpi_enforce_resources=lax and load the native driver.
>
> Well, tried the lax as well got the same "error"
>
> w83627ehf: Found W83627EHG chip at 0x290
> ACPI: I/O resource w83627ehf [0x295-0x296] conflicts with ACPI region
> HWRE [0x290-0x299]
> ACPI: This conflict may cause random problems and system instability
> ACPI: If an ACPI driver is available for this device, you should use
> it instead of the native driver

Yes, this is expected. The kernel still warns you that the situation
is potentially unsafe, but with "lax" checking it will allow you to
load the native driver.

Luca

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors