[lm-sensors] hwmon: (adm1021) Clarify documentation regarding Xeon

[lm-sensors] hwmon: (adm1021) Clarify documentation regarding Xeon

[Jean Delvare]

Recent Xeon processor thermal sensors are supported by the coretemp
driver and not the adm1021 driver. Only one old generation of Xeon
processors (the first Netburst ones) are supported by the adm1021
driver.

Reported-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
---
 Documentation/hwmon/adm1021 |   36 +++++++++++++++++++-----------------
 drivers/hwmon/Kconfig       |    3 +--
 2 files changed, 20 insertions(+), 19 deletions(-)

--- linux-2.6.39-rc2.orig/Documentation/hwmon/adm1021	2010-08-02 00:11:14.000000000 +0200
+++ linux-2.6.39-rc2/Documentation/hwmon/adm1021	2011-04-08 10:11:13.000000000 +0200
@@ -14,10 +14,6 @@ Supported chips:
     Prefix: 'gl523sm'
     Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
     Datasheet:
-  * Intel Xeon Processor
-    Prefix: - any other - may require 'force_adm1021' parameter
-    Addresses scanned: none
-    Datasheet: Publicly available at Intel website
   * Maxim MAX1617
     Prefix: 'max1617'
     Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
@@ -91,21 +87,27 @@ will do no harm, but will return 'old' v
 ADM1021-clones do faster measurements, but there is really no good reason
 for that.
 
-Xeon support
-------------
 
-Some Xeon processors have real max1617, adm1021, or compatible chips
-within them, with two temperature sensors.
+Netburst-based Xeon support
+---------------------------
 
-Other Xeons have chips with only one sensor.
+Some Xeon processors based on the Netburst (early Pentium 4, from 2001 to
+2003) microarchitecture had real MAX1617, ADM1021, or compatible chips
+within them, with two temperature sensors. Other Xeon processors of this
+era (with 400 MHz FSB) had chips with only one temperature sensor.
 
-If you have a Xeon, and the adm1021 module loads, and both temperatures
-appear valid, then things are good.
+If you have such an old Xeon, and you get two valid temperatures when
+loading the adm1021 module, then things are good.
 
-If the adm1021 module doesn't load, you should try this:
-	modprobe adm1021 force_adm1021=BUS,ADDRESS
-	ADDRESS can only be 0x18, 0x1a, 0x29, 0x2b, 0x4c, or 0x4e.
+If nothing happens when loading the adm1021 module, are you are certain
+that your specific Xeon processor model includes compatible sensors, you
+will have to explicitly instantiate the sensor chips from user-space. See
+method 4 in Documentation/i2c/instantiating-devices. Possible slave
+addresses are 0x18, 0x1a, 0x29, 0x2b, 0x4c, or 0x4e. It is likely that
+only temp2 will be correct and temp1 will have to be ignored.
 
-If you have dual Xeons you may have appear to have two separate
-adm1021-compatible chips, or two single-temperature sensors, at distinct
-addresses.
+Previous generations of the Xeon processor (based on Pentium II/III)
+didn't have these sensors. Next generations of Xeon processors (533 MHz
+FSB and faster) lost them, until the Core-based generation which
+introduced integrated digital thermal sensors. These are supported by
+the coretemp driver.
--- linux-2.6.39-rc2.orig/drivers/hwmon/Kconfig	2011-04-07 18:51:51.000000000 +0200
+++ linux-2.6.39-rc2/drivers/hwmon/Kconfig	2011-04-08 09:36:50.000000000 +0200
@@ -110,8 +110,7 @@ config SENSORS_ADM1021
 	help
 	  If you say yes here you get support for Analog Devices ADM1021
 	  and ADM1023 sensor chips and clones: Maxim MAX1617 and MAX1617A,
-	  Genesys Logic GL523SM, National Semiconductor LM84, TI THMC10,
-	  and the XEON processor built-in sensor.
+	  Genesys Logic GL523SM, National Semiconductor LM84 and TI THMC10.
 
 	  This driver can also be built as a module.  If so, the module
 	  will be called adm1021.


-- 
Jean Delvare

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

Re: [lm-sensors] hwmon: (adm1021) Clarify documentation regarding

[Guenter Roeck]

On Sat, Apr 09, 2011 at 12:31:56PM -0400, Jean Delvare wrote:
> Recent Xeon processor thermal sensors are supported by the coretemp
> driver and not the adm1021 driver. Only one old generation of Xeon
> processors (the first Netburst ones) are supported by the adm1021
> driver.
> 
> Reported-by: Darren Hart <dvhart@linux.intel.com>
> Signed-off-by: Jean Delvare <khali@linux-fr.org>

Nitpick below, otherwise

Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>

> ---
>  Documentation/hwmon/adm1021 |   36 +++++++++++++++++++-----------------
>  drivers/hwmon/Kconfig       |    3 +--
>  2 files changed, 20 insertions(+), 19 deletions(-)
> 
> --- linux-2.6.39-rc2.orig/Documentation/hwmon/adm1021	2010-08-02 00:11:14.000000000 +0200
> +++ linux-2.6.39-rc2/Documentation/hwmon/adm1021	2011-04-08 10:11:13.000000000 +0200
> @@ -14,10 +14,6 @@ Supported chips:
>      Prefix: 'gl523sm'
>      Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
>      Datasheet:
> -  * Intel Xeon Processor
> -    Prefix: - any other - may require 'force_adm1021' parameter
> -    Addresses scanned: none
> -    Datasheet: Publicly available at Intel website
>    * Maxim MAX1617
>      Prefix: 'max1617'
>      Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
> @@ -91,21 +87,27 @@ will do no harm, but will return 'old' v
>  ADM1021-clones do faster measurements, but there is really no good reason
>  for that.
>  
> -Xeon support
> -------------
>  
> -Some Xeon processors have real max1617, adm1021, or compatible chips
> -within them, with two temperature sensors.
> +Netburst-based Xeon support
> +---------------------------
>  
> -Other Xeons have chips with only one sensor.
> +Some Xeon processors based on the Netburst (early Pentium 4, from 2001 to
> +2003) microarchitecture had real MAX1617, ADM1021, or compatible chips
> +within them, with two temperature sensors. Other Xeon processors of this
> +era (with 400 MHz FSB) had chips with only one temperature sensor.
>  
> -If you have a Xeon, and the adm1021 module loads, and both temperatures
> -appear valid, then things are good.
> +If you have such an old Xeon, and you get two valid temperatures when
> +loading the adm1021 module, then things are good.
>  
> -If the adm1021 module doesn't load, you should try this:
> -	modprobe adm1021 force_adm1021=BUS,ADDRESS
> -	ADDRESS can only be 0x18, 0x1a, 0x29, 0x2b, 0x4c, or 0x4e.
> +If nothing happens when loading the adm1021 module, are you are certain

		are you are -> and you are

> +that your specific Xeon processor model includes compatible sensors, you
> +will have to explicitly instantiate the sensor chips from user-space. See
> +method 4 in Documentation/i2c/instantiating-devices. Possible slave
> +addresses are 0x18, 0x1a, 0x29, 0x2b, 0x4c, or 0x4e. It is likely that
> +only temp2 will be correct and temp1 will have to be ignored.
>  
> -If you have dual Xeons you may have appear to have two separate
> -adm1021-compatible chips, or two single-temperature sensors, at distinct
> -addresses.
> +Previous generations of the Xeon processor (based on Pentium II/III)
> +didn't have these sensors. Next generations of Xeon processors (533 MHz
> +FSB and faster) lost them, until the Core-based generation which
> +introduced integrated digital thermal sensors. These are supported by
> +the coretemp driver.
> --- linux-2.6.39-rc2.orig/drivers/hwmon/Kconfig	2011-04-07 18:51:51.000000000 +0200
> +++ linux-2.6.39-rc2/drivers/hwmon/Kconfig	2011-04-08 09:36:50.000000000 +0200
> @@ -110,8 +110,7 @@ config SENSORS_ADM1021
>  	help
>  	  If you say yes here you get support for Analog Devices ADM1021
>  	  and ADM1023 sensor chips and clones: Maxim MAX1617 and MAX1617A,
> -	  Genesys Logic GL523SM, National Semiconductor LM84, TI THMC10,
> -	  and the XEON processor built-in sensor.
> +	  Genesys Logic GL523SM, National Semiconductor LM84 and TI THMC10.
>  
>  	  This driver can also be built as a module.  If so, the module
>  	  will be called adm1021.
> 
> 
> -- 
> Jean Delvare
> 
> _______________________________________________
> lm-sensors mailing list
> lm-sensors@lm-sensors.org
> http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

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

Re: [lm-sensors] hwmon: (adm1021) Clarify documentation regarding

[Jean Delvare]

On Sat, 9 Apr 2011 13:39:09 -0700, Guenter Roeck wrote:
> On Sat, Apr 09, 2011 at 12:31:56PM -0400, Jean Delvare wrote:
> > Recent Xeon processor thermal sensors are supported by the coretemp
> > driver and not the adm1021 driver. Only one old generation of Xeon
> > processors (the first Netburst ones) are supported by the adm1021
> > driver.
> > 
> > Reported-by: Darren Hart <dvhart@linux.intel.com>
> > Signed-off-by: Jean Delvare <khali@linux-fr.org>
> 
> Nitpick below, otherwise
> 
> Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>

Thanks for the review :)

> > ---
> >  Documentation/hwmon/adm1021 |   36 +++++++++++++++++++-----------------
> >  drivers/hwmon/Kconfig       |    3 +--
> >  2 files changed, 20 insertions(+), 19 deletions(-)
> > 
> > --- linux-2.6.39-rc2.orig/Documentation/hwmon/adm1021	2010-08-02 00:11:14.000000000 +0200
> > +++ linux-2.6.39-rc2/Documentation/hwmon/adm1021	2011-04-08 10:11:13.000000000 +0200
> > @@ -14,10 +14,6 @@ Supported chips:
> >      Prefix: 'gl523sm'
> >      Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
> >      Datasheet:
> > -  * Intel Xeon Processor
> > -    Prefix: - any other - may require 'force_adm1021' parameter
> > -    Addresses scanned: none
> > -    Datasheet: Publicly available at Intel website
> >    * Maxim MAX1617
> >      Prefix: 'max1617'
> >      Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
> > @@ -91,21 +87,27 @@ will do no harm, but will return 'old' v
> >  ADM1021-clones do faster measurements, but there is really no good reason
> >  for that.
> >  
> > -Xeon support
> > -------------
> >  
> > -Some Xeon processors have real max1617, adm1021, or compatible chips
> > -within them, with two temperature sensors.
> > +Netburst-based Xeon support
> > +---------------------------
> >  
> > -Other Xeons have chips with only one sensor.
> > +Some Xeon processors based on the Netburst (early Pentium 4, from 2001 to
> > +2003) microarchitecture had real MAX1617, ADM1021, or compatible chips
> > +within them, with two temperature sensors. Other Xeon processors of this
> > +era (with 400 MHz FSB) had chips with only one temperature sensor.
> >  
> > -If you have a Xeon, and the adm1021 module loads, and both temperatures
> > -appear valid, then things are good.
> > +If you have such an old Xeon, and you get two valid temperatures when
> > +loading the adm1021 module, then things are good.
> >  
> > -If the adm1021 module doesn't load, you should try this:
> > -	modprobe adm1021 force_adm1021=BUS,ADDRESS
> > -	ADDRESS can only be 0x18, 0x1a, 0x29, 0x2b, 0x4c, or 0x4e.
> > +If nothing happens when loading the adm1021 module, are you are certain
> 
> 		are you are -> and you are

Good catch, thanks!

> 
> > +that your specific Xeon processor model includes compatible sensors, you
> > +will have to explicitly instantiate the sensor chips from user-space. See
> > +method 4 in Documentation/i2c/instantiating-devices. Possible slave
> > +addresses are 0x18, 0x1a, 0x29, 0x2b, 0x4c, or 0x4e. It is likely that
> > +only temp2 will be correct and temp1 will have to be ignored.
> >  
> > -If you have dual Xeons you may have appear to have two separate
> > -adm1021-compatible chips, or two single-temperature sensors, at distinct
> > -addresses.
> > +Previous generations of the Xeon processor (based on Pentium II/III)
> > +didn't have these sensors. Next generations of Xeon processors (533 MHz
> > +FSB and faster) lost them, until the Core-based generation which
> > +introduced integrated digital thermal sensors. These are supported by
> > +the coretemp driver.
> > --- linux-2.6.39-rc2.orig/drivers/hwmon/Kconfig	2011-04-07 18:51:51.000000000 +0200
> > +++ linux-2.6.39-rc2/drivers/hwmon/Kconfig	2011-04-08 09:36:50.000000000 +0200
> > @@ -110,8 +110,7 @@ config SENSORS_ADM1021
> >  	help
> >  	  If you say yes here you get support for Analog Devices ADM1021
> >  	  and ADM1023 sensor chips and clones: Maxim MAX1617 and MAX1617A,
> > -	  Genesys Logic GL523SM, National Semiconductor LM84, TI THMC10,
> > -	  and the XEON processor built-in sensor.
> > +	  Genesys Logic GL523SM, National Semiconductor LM84 and TI THMC10.
> >  
> >  	  This driver can also be built as a module.  If so, the module
> >  	  will be called adm1021.
> > 
> > 
> > -- 
> > Jean Delvare
> > 
> > _______________________________________________
> > lm-sensors mailing list
> > lm-sensors@lm-sensors.org
> > http://lists.lm-sensors.org/mailman/listinfo/lm-sensors


-- 
Jean Delvare

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

Re: [lm-sensors] hwmon: (adm1021) Clarify documentation regarding

[Darren Hart]

On 04/09/2011 01:39 PM, Guenter Roeck wrote:
> On Sat, Apr 09, 2011 at 12:31:56PM -0400, Jean Delvare wrote:
>> Recent Xeon processor thermal sensors are supported by the coretemp
>> driver and not the adm1021 driver. Only one old generation of Xeon
>> processors (the first Netburst ones) are supported by the adm1021
>> driver.

In addition to changing the kernel config help, is there a way to avoid
loading this driver for non netburst type CPUs? Someway to blacklist
later CPUs from using this driver? Seems this would make sensors-detect
and such behave better.

Thanks,

Darren

>>
>> Reported-by: Darren Hart <dvhart@linux.intel.com>
>> Signed-off-by: Jean Delvare <khali@linux-fr.org>
> 
> Nitpick below, otherwise
> 
> Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>
> 
>> ---
>>  Documentation/hwmon/adm1021 |   36 +++++++++++++++++++-----------------
>>  drivers/hwmon/Kconfig       |    3 +--
>>  2 files changed, 20 insertions(+), 19 deletions(-)
>>
>> --- linux-2.6.39-rc2.orig/Documentation/hwmon/adm1021	2010-08-02 00:11:14.000000000 +0200
>> +++ linux-2.6.39-rc2/Documentation/hwmon/adm1021	2011-04-08 10:11:13.000000000 +0200
>> @@ -14,10 +14,6 @@ Supported chips:
>>      Prefix: 'gl523sm'
>>      Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
>>      Datasheet:
>> -  * Intel Xeon Processor
>> -    Prefix: - any other - may require 'force_adm1021' parameter
>> -    Addresses scanned: none
>> -    Datasheet: Publicly available at Intel website
>>    * Maxim MAX1617
>>      Prefix: 'max1617'
>>      Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
>> @@ -91,21 +87,27 @@ will do no harm, but will return 'old' v
>>  ADM1021-clones do faster measurements, but there is really no good reason
>>  for that.
>>  
>> -Xeon support
>> -------------
>>  
>> -Some Xeon processors have real max1617, adm1021, or compatible chips
>> -within them, with two temperature sensors.
>> +Netburst-based Xeon support
>> +---------------------------
>>  
>> -Other Xeons have chips with only one sensor.
>> +Some Xeon processors based on the Netburst (early Pentium 4, from 2001 to
>> +2003) microarchitecture had real MAX1617, ADM1021, or compatible chips
>> +within them, with two temperature sensors. Other Xeon processors of this
>> +era (with 400 MHz FSB) had chips with only one temperature sensor.
>>  
>> -If you have a Xeon, and the adm1021 module loads, and both temperatures
>> -appear valid, then things are good.
>> +If you have such an old Xeon, and you get two valid temperatures when
>> +loading the adm1021 module, then things are good.
>>  
>> -If the adm1021 module doesn't load, you should try this:
>> -	modprobe adm1021 force_adm1021=BUS,ADDRESS
>> -	ADDRESS can only be 0x18, 0x1a, 0x29, 0x2b, 0x4c, or 0x4e.
>> +If nothing happens when loading the adm1021 module, are you are certain
> 
> 		are you are -> and you are
> 
>> +that your specific Xeon processor model includes compatible sensors, you
>> +will have to explicitly instantiate the sensor chips from user-space. See
>> +method 4 in Documentation/i2c/instantiating-devices. Possible slave
>> +addresses are 0x18, 0x1a, 0x29, 0x2b, 0x4c, or 0x4e. It is likely that
>> +only temp2 will be correct and temp1 will have to be ignored.
>>  
>> -If you have dual Xeons you may have appear to have two separate
>> -adm1021-compatible chips, or two single-temperature sensors, at distinct
>> -addresses.
>> +Previous generations of the Xeon processor (based on Pentium II/III)
>> +didn't have these sensors. Next generations of Xeon processors (533 MHz
>> +FSB and faster) lost them, until the Core-based generation which
>> +introduced integrated digital thermal sensors. These are supported by
>> +the coretemp driver.
>> --- linux-2.6.39-rc2.orig/drivers/hwmon/Kconfig	2011-04-07 18:51:51.000000000 +0200
>> +++ linux-2.6.39-rc2/drivers/hwmon/Kconfig	2011-04-08 09:36:50.000000000 +0200
>> @@ -110,8 +110,7 @@ config SENSORS_ADM1021
>>  	help
>>  	  If you say yes here you get support for Analog Devices ADM1021
>>  	  and ADM1023 sensor chips and clones: Maxim MAX1617 and MAX1617A,
>> -	  Genesys Logic GL523SM, National Semiconductor LM84, TI THMC10,
>> -	  and the XEON processor built-in sensor.
>> +	  Genesys Logic GL523SM, National Semiconductor LM84 and TI THMC10.
>>  
>>  	  This driver can also be built as a module.  If so, the module
>>  	  will be called adm1021.
>>
>>
>> -- 
>> Jean Delvare
>>
>> _______________________________________________
>> lm-sensors mailing list
>> lm-sensors@lm-sensors.org
>> http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

-- 
Darren Hart
Intel Open Source Technology Center
Yocto Project - Linux Kernel

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

Re: [lm-sensors] hwmon: (adm1021) Clarify documentation regarding

[Jean Delvare]

Hi Darren,

On Sat, 09 Apr 2011 16:15:45 -0700, Darren Hart wrote:
> On 04/09/2011 01:39 PM, Guenter Roeck wrote:
> > On Sat, Apr 09, 2011 at 12:31:56PM -0400, Jean Delvare wrote:
> >> Recent Xeon processor thermal sensors are supported by the coretemp
> >> driver and not the adm1021 driver. Only one old generation of Xeon
> >> processors (the first Netburst ones) are supported by the adm1021
> >> driver.
> 
> In addition to changing the kernel config help, is there a way to avoid
> loading this driver for non netburst type CPUs? Someway to blacklist
> later CPUs from using this driver? Seems this would make sensors-detect
> and such behave better.

This is all the other way around. The adm1021 driver doesn't auto-load
on PC systems. The way it usually goes is that sensors-detect will
detect a compatible device, and it will set the adm1021 driver to be
loaded when the lm_sensors service is started. The adm1021 driver, in
turn, tries to detect compatible devices, and bind to them if found.

So the problem boils down to two related issues:
* The sensors-detect script seemingly misdetected a device on your
  system.
* The adm1021 got loaded and misdetected the same device.
Which is expected as in general sensors-detect has the same detection
code as hwmon drivers on the kernel side.

That being said, this can only be fixed if you provide the full output
of sensors-detect on your system after unloading all hwmon drivers
(stopping the lm_sensors service sometimes do, but not always, depends
on the distribution) and a dump of the misdetected device. This will
confirm the misdetection and hopefully offer a way to avoid it.

-- 
Jean Delvare

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

Re: [lm-sensors] hwmon: (adm1021) Clarify documentation regarding

[Darren Hart]

[-- Attachment #1: Type: text/plain, Size: 2340 bytes --]



On 04/10/2011 12:08 AM, Jean Delvare wrote:
> Hi Darren,
> 
> On Sat, 09 Apr 2011 16:15:45 -0700, Darren Hart wrote:
>> On 04/09/2011 01:39 PM, Guenter Roeck wrote:
>>> On Sat, Apr 09, 2011 at 12:31:56PM -0400, Jean Delvare wrote:
>>>> Recent Xeon processor thermal sensors are supported by the coretemp
>>>> driver and not the adm1021 driver. Only one old generation of Xeon
>>>> processors (the first Netburst ones) are supported by the adm1021
>>>> driver.
>>
>> In addition to changing the kernel config help, is there a way to avoid
>> loading this driver for non netburst type CPUs? Someway to blacklist
>> later CPUs from using this driver? Seems this would make sensors-detect
>> and such behave better.
> 
> This is all the other way around. The adm1021 driver doesn't auto-load
> on PC systems. The way it usually goes is that sensors-detect will
> detect a compatible device, and it will set the adm1021 driver to be
> loaded when the lm_sensors service is started. The adm1021 driver, in
> turn, tries to detect compatible devices, and bind to them if found.
> 
> So the problem boils down to two related issues:
> * The sensors-detect script seemingly misdetected a device on your
>   system.
> * The adm1021 got loaded and misdetected the same device.
> Which is expected as in general sensors-detect has the same detection
> code as hwmon drivers on the kernel side.
> 
> That being said, this can only be fixed if you provide the full output
> of sensors-detect on your system after unloading all hwmon drivers
> (stopping the lm_sensors service sometimes do, but not always, depends
> on the distribution) and a dump of the misdetected device. This will
> confirm the misdetection and hopefully offer a way to avoid it.

This system is really throwing me for a loop. After returning from ELC I
powered it back up and have been completely unable to get sensors to
read anything from w83795 or ipmi-si even. Running sensors-detect
(without any hwmon modules loaded) finds only ipmi-si and w83627ehf, but
sensors only reports on w83627ehf).

I have attached the typescript of the sensors-detect session, but it
isn't very interesting. If I stumble across another odd set modules
detected, I'll report back to the list.

/me scratches head...

-- 
Darren Hart
Intel Open Source Technology Center
Yocto Project - Linux Kernel

[-- Attachment #2: typescript --]
[-- Type: application/octet-stream, Size: 7137 bytes --]

[-- Attachment #3: 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] hwmon: (adm1021) Clarify documentation regarding

[Jean Delvare]

On Thu, 14 Apr 2011 21:57:11 -0700, Darren Hart wrote:
> On 04/10/2011 12:08 AM, Jean Delvare wrote:
> > This is all the other way around. The adm1021 driver doesn't auto-load
> > on PC systems. The way it usually goes is that sensors-detect will
> > detect a compatible device, and it will set the adm1021 driver to be
> > loaded when the lm_sensors service is started. The adm1021 driver, in
> > turn, tries to detect compatible devices, and bind to them if found.
> > 
> > So the problem boils down to two related issues:
> > * The sensors-detect script seemingly misdetected a device on your
> >   system.
> > * The adm1021 got loaded and misdetected the same device.
> > Which is expected as in general sensors-detect has the same detection
> > code as hwmon drivers on the kernel side.
> > 
> > That being said, this can only be fixed if you provide the full output
> > of sensors-detect on your system after unloading all hwmon drivers
> > (stopping the lm_sensors service sometimes do, but not always, depends
> > on the distribution) and a dump of the misdetected device. This will
> > confirm the misdetection and hopefully offer a way to avoid it.
> 
> This system is really throwing me for a loop. After returning from ELC I
> powered it back up and have been completely unable to get sensors to
> read anything from w83795 or ipmi-si even.

Try a complete power-off including removal from the mains power plug
for 1 minute.

> Running sensors-detect
> (without any hwmon modules loaded) finds only ipmi-si and w83627ehf, but
> sensors only reports on w83627ehf).
> 
> I have attached the typescript of the sensors-detect session, but it
> isn't very interesting. If I stumble across another odd set modules
> detected, I'll report back to the list.

From the sensors-detect log:
> FATAL: Module i2c_dev not found.
> Failed to load module i2c-dev.

You'll have to check where this module has gone. If this is a
self-built kernel, check that you have:
CONFIG_I2C_CHARDEV=m

-- 
Jean Delvare

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