All of lore.kernel.org
 help / color / mirror / Atom feed
From: Karsten de Freese <karsten@maxi-dsl.de>
To: lm-sensors@vger.kernel.org
Subject: Re: [lm-sensors] looking for help with W83795ADG
Date: Sun, 20 Nov 2011 23:03:36 +0000	[thread overview]
Message-ID: <4EC98748.4030805@maxi-dsl.de> (raw)
In-Reply-To: <20101022135607.6996731s0m2gcr48@webmail.mnet-online.de>

hello Eddi,

a less-than-welcome update:

at least when the three modules i2c-piix4.ko, i2c-piix4-n36l.ko, 
w83795.ko are loaded,  a restart (sudo shutdown -r now) of my N36L 
fails, the system does not come up until I do a complete power cycle. 
I'll have to look more closely at this, once I have physical access to 
the box again.

regards -

Karsten

Am 20.11.2011 14:38, schrieb Karsten de Freese:
> hello Eddi,
>
> Success!
>
> Thanks for the detailed steps. This all works as described, at least 
> as far as the 'sensors' output is concerned. With the respective 
> modules, I get all the measurements you're listing! (See attached.)
>
> 'i2cdetect -l' will however not produce any output for whatsoever reason.
>
> The next thing I'd like to find out is how to load the modules 
> automatically in a clean / kernel-upgrade-proof way..
>
> thanks and regards -
>
> Karsten
>
>
> Am 18.11.2011 11:09, schrieb Eddi De Pieri:
>> I suggest you:
>>
>>
>> mkdir ~/piix4-n36l
>> cp 
>> [yourlinuxtree]/usr/src/linux-source-2.6.32/drivers/i2c/busses/i2c-piix4.c
>> ~/piix4-n36l/
>> cp ~/Downloads/Makefile ~/piix4-n36l/
>> cd  ~/piix4-n36l/
>>
>> patch -p1<  proliant-microserver-n36l.diff
>> make
>> you will have your new module inside ~/piix4-n36l/
>>
>> rmmod i2c-piix4
>> insmod ./i2c-piix4.ko
>> insmod ./i2c-piix4-n36l.ko
>>
>> insmod w83795 (you should already build separately)
>>
>> sensors
>>
>> regards Eddi.
>>
>> I hope an answer from Jean... I'd like if this patch may be inserted
>> in mainstream linux with the necessary adjustment..
>>
>> On Thu, Nov 17, 2011 at 11:39 PM, Karsten de 
>> Freese<karsten@maxi-dsl.de>  wrote:
>>> hello Eddi,
>>>
>>> I'm ready for two questions ;-)
>>>
>>> .diff / patch looks simple, and I also found i2c-amd756-s4882.c - 
>>> however:
>>> - the diff file actually describes the patch to be on top of 
>>> i2c-piix4.c?
>>> - if I'd compile the patched file, would that give me a standalone 
>>> driver?
>>>
>>> thanks for your help -
>>>
>>> Karsten
>>>
>>> On 17.11.2011 22:25, Karsten de Freese wrote:
>>>> hi Eddi,
>>>>
>>>> thanks very much - I had almost forgotten about the topic, but 
>>>> stumbled
>>>> across it again a few days ago, so that's very interesting.
>>>> I'm not yet familiar with .diff patches, and may welcome a little 
>>>> help.
>>>> However, let me try to find out on my own first..
>>>>
>>>> thanks&  regards -
>>>>
>>>> Karsten
>>>>
>>>> On 16.11.2011 17:18, Eddi De Pieri wrote:
>>>>> Finally I get W83795ADG working on my proliant microserver
>>>>>
>>>>> The patch is based on i2c-amd756-s4882.c and adapted for sb800 
>>>>> southbus.
>>>>>
>>>>> Karsten, please try it too...
>>>>>
>>>>> root@proliant:/usr/src/lm-sensors/eddi# i2cdetect -l
>>>>> i2c-0   smbus           SMBus piix4 adapter (SDA0)              SMBus
>>>>> adapter
>>>>> i2c-1   smbus           SMBus piix4 adapter (SDA2)              SMBus
>>>>> adapter
>>>>> i2c-2   smbus           SMBus piix4 adapter (SDA3)              SMBus
>>>>> adapter
>>>>> i2c-3   smbus           SMBus piix4 adapter (SDA4)              SMBus
>>>>> adapter
>>>>> root@proliant:/usr/src/lm-sensors/eddi#
>>>>>
>>>>> yes SDA1 is reserved... so i can't multiplex it
>>>>>
>>>>> root@proliant:/usr/src/lm-sensors/eddi# sensors
>>>>> k10temp-pci-00c3
>>>>> Adapter: PCI adapter
>>>>> temp1:       +24.5°C  (high = +70.0°C, crit = +100.0°C)
>>>>>
>>>>> w83795adg-i2c-1-2f
>>>>> Adapter: SMBus piix4 adapter (SDA2)
>>>>> in0:         +1.02 V  (min =  +0.00 V, max =  +2.05 V)
>>>>> in1:         +1.52 V  (min =  +0.00 V, max =  +2.05 V)
>>>>> in2:         +1.10 V  (min =  +0.00 V, max =  +2.05 V)
>>>>> in3:         +0.89 V  (min =  +0.00 V, max =  +2.05 V)
>>>>> in12:        +3.35 V  (min =  +0.00 V, max =  +6.14 V)
>>>>> in13:        +3.28 V  (min =  +0.00 V, max =  +6.14 V)
>>>>> fan1:        703 RPM  (min =  329 RPM)
>>>>> temp1:       +23.0°C  (high = +109.0°C, hyst = +109.0°C)
>>>>>                        (crit = +109.0°C, hyst = +109.0°C)  sensor =
>>>>> thermal diode
>>>>> temp2:       +33.2°C  (high = +105.0°C, hyst = +105.0°C)
>>>>>                        (crit = +105.0°C, hyst = +105.0°C)  sensor =
>>>>> thermal diode
>>>>> temp5:       +14.0°C  (high = +39.0°C, hyst = +39.0°C)
>>>>>                        (crit = +44.0°C, hyst = +44.0°C)  sensor =
>>>>> thermistor
>>>>> beep_enable:disabled
>>>>>
>>>>> jc42-i2c-0-18
>>>>> Adapter: SMBus piix4 adapter (SDA0)
>>>>> temp1:       +20.5°C  (low  =  +0.0°C, high =  +0.0°C)  ALARM
>>>>>                        (crit =  +0.0°C, hyst =  +0.0°C)  ALARM
>>>>>
>>>>>
>>>>> root@proliant:/usr/src/lm-sensors/eddi# i2cdetect -y 0
>>>>>       0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
>>>>> 00:          -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 10: -- -- -- -- -- -- -- -- UU -- -- -- -- -- -- --
>>>>> 20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 50: 50 -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 70: -- -- -- -- -- -- -- --
>>>>>
>>>>> root@proliant:/usr/src/lm-sensors/eddi# i2cdetect -y 1
>>>>>       0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
>>>>> 00:          -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- UU
>>>>> 30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 60: -- 61 -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 70: -- -- -- -- -- -- -- --
>>>>>
>>>>> root@proliant:/usr/src/lm-sensors/eddi# i2cdetect -y 2
>>>>>       0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
>>>>> 00:          -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 40: -- -- -- -- -- -- -- -- -- -- -- -- 4c -- -- --
>>>>> 50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 70: -- -- -- -- -- -- -- --
>>>>>
>>>>> root@proliant:/usr/src/lm-sensors/eddi# i2cdetect -y 3
>>>>>       0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
>>>>> 00:          -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
>>>>> 70: -- -- -- -- -- -- -- --
>>>>>
>>>>> pay attention that the msleep seems to be really needed...
>>>>>
>>>>> Regards, Eddi
>>>>>
>>>>> On Tue, Aug 16, 2011 at 9:26 AM, Eddi De 
>>>>> Pieri<eddi@depieri.net>    wrote:
>>>>>> Hi Jean...
>>>>>>
>>>>>> The bios show fan rpm and temperature. Since my own proliant 
>>>>>> microserver
>>>>>> lacks of ipmi card, I doubt the bios give me the results using 
>>>>>> ipmi call.
>>>>>>
>>>>>> The HP Proliant Microserver chipset (SB820M) have a multiplexed 
>>>>>> smbus
>>>>>> like your.
>>>>>>
>>>>>>
>>>>>> http://support.amd.com/it/Embedded_TechDocs/47283_sb820m_ds_pub_2.00.pdf 
>>>>>>
>>>>>>
>>>>>> Here is technical document from amd...
>>>>>>
>>>>>>
>>>>>> at p52/53 they say:
>>>>>>
>>>>>> (1) The SDA1 and SCL1 SMBus interface is dedicated for ASF 
>>>>>> devices only.
>>>>>> It should not be used to
>>>>>> connect to any other devices.
>>>>>> (2) There are only two SMBus controllers. The SCL1/SDA1 pair is
>>>>>> controlled by SMBus controller 1.
>>>>>>
>>>>>>
>>>>>> SCL0/SDA0, SCL2/SDA2, SCL3/SDA3 and SCL4/SDA4 are multiplexed 
>>>>>> pins that
>>>>>> are all controlled by
>>>>>> SMBus controller 0, and only 1 pair of those pins can be active 
>>>>>> at any
>>>>>> time.
>>>>>>
>>>>>> I can't understand how to enable 2, 3 and 4 muxed bus..
>>>>>>
>>>>>>
>>>>>> If you consolidate a solution for i2c-i801.c I hope we can be 
>>>>>> replicated
>>>>>> for i2c-piix4.c too..
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Hi Karsten,
>>>>>>
>>>>>> On Fri, 22 Oct 2010 16:04:10 +0200, karsten at maxi-dsl.de wrote:
>>>>>>> hi Jean,
>>>>>>>> The w83795 driver uses the same detection mechanism used in
>>>>>>>> sensors-detect. So if sensors-detect doesn't see your chip, I 
>>>>>>>> am not
>>>>>>>> surprised that the driver didn't either.
>>>>>>>> The full output of sensors-detect is needed to diagnose your 
>>>>>>>> problem
>>>>>>>> further. It could be that your SMBus controller isn't 
>>>>>>>> supported, or
>>>>>>>> maybe you are supposed to access the chip through IPMI.
>>>>>>> IPMI is probably the key..
>>>>>>>    From the sensors-detect output:
>>>>>>> "Warning: the required module ipmisensors is not currently 
>>>>>>> installed
>>>>>>> on your system. If it is built into the kernel then it's OK.
>>>>>>> Otherwise, checkhttp://www.lm-sensors.org/wiki/Devices  for
>>>>>>> driver availability" (full ouput attached)
>>>>>>>
>>>>>>> I did see the output before, but didn't really understand the
>>>>>>> implications (IPMI seems to be an alternative way of accessing the
>>>>>>> monitoring chips?)
>>>>>> Yes. The strength of IPMI access is that it can be used remotely 
>>>>>> too,
>>>>>> for example from a BMC plugged into your system.
>>>>>>
>>>>>>
>>>>>>
>>>>>>> The suggested ipmisensors driver however does not seem to be 
>>>>>>> available,
>>>>>>> right?
>>>>>>> (http://lm-sensors.org/wiki/Devices  says "(2004-12-12)  Port to 
>>>>>>> Linux
>>>>>>> 2.6 in progress by Yani Ioannou")
>>>>>> That project is essentially dead. At the time being, the best way to
>>>>>> retrieve IPMI-based sensor values is using "ipmitool sensor". The
>>>>>> drawback is that this isn't integrated with libsensors, so the 
>>>>>> usually
>>>>>>
>>>>>>
>>>>>> monitoring applications won't work.
>>>>>>
>>>>>> If you have a BMC or are otherwise already using IPMI on your 
>>>>>> machine,
>>>>>> just use ipmitool and you're done.
>>>>>>
>>>>>> If not, then it might be worth trying ipmitool once, and then try
>>>>>>
>>>>>>
>>>>>> sensors-detect again. There is a small chance that, once ipmitool
>>>>>> managed to access the monitoring device, sensors-detect will see 
>>>>>> it. My
>>>>>> own mainboard which has a W83795ADG chip has a multiplexed SMBus, 
>>>>>> and I
>>>>>> can only see the W83795ADG in sensors-detect when the multiplexer is
>>>>>>
>>>>>>
>>>>>> set properly. Otherwise I would see the memory modules, which _do_
>>>>>> appear in your sensors-detect output. So maybe your board has a
>>>>>> multiplexed SMBus too.
>>>>>>
>>>>>> If the trick doesn't work, you may want to try the new jc42 driver
>>>>>>
>>>>>>
>>>>>> which can monitor the temperature of some DDR3 memory modules. I
>>>>>> suspect your memory modules have such sensors. A more recent 
>>>>>> version of
>>>>>> sensors-detect would confirm that:
>>>>>>    http://dl.lm-sensors.org/lm-sensors/files/sensors-detect
>>>>>>
>>>>>>
>>>>>>
>>>>>> I also see that sensors-detect suggested to use the k10temp driver,
>>>>>> which your kernel doesn't have. You can try the standalone flavor of
>>>>>> the drivers which I maintain at:
>>>>>>    http://khali.linux-fr.org/devel/misc/k10temp/
>>>>>>
>>>>>>
>>>>>> The accuracy of the CPU internal sensors is limited, but that's 
>>>>>> still
>>>>>> better than nothing.
>>>>>>
>>>>>> -- 
>>>>>> 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

  parent reply	other threads:[~2011-11-20 23:03 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-10-22 11:56 [lm-sensors] looking for help with W83795ADG karsten
2010-10-22 13:06 ` Jean Delvare
2010-10-22 14:04 ` karsten
2010-10-22 20:51 ` Karsten de Freese
2010-10-23 16:05 ` Jean Delvare
2010-10-25  7:30 ` Jean Delvare
2010-10-25  7:41 ` Jean Delvare
2010-10-26 18:46 ` Karsten de Freese
2010-10-26 19:06 ` Jean Delvare
2010-10-26 20:09 ` Karsten de Freese
2010-10-26 20:30 ` Jean Delvare
2010-10-26 21:51 ` Guenter Roeck
2010-10-26 22:32 ` Karsten de Freese
2011-08-16  7:26 ` Eddi De Pieri
2011-11-16 16:18 ` Eddi De Pieri
2011-11-17 21:25 ` Karsten de Freese
2011-11-17 22:39 ` Karsten de Freese
2011-11-18 10:09 ` Eddi De Pieri
2011-11-20 13:38 ` Karsten de Freese
2011-11-20 23:03 ` Karsten de Freese [this message]
2011-11-21 10:49 ` Eddi De Pieri
2011-11-21 16:09 ` Karsten de Freese
2011-11-25 10:22 ` Eddi De Pieri
2011-11-25 10:45 ` Guenter Roeck
2011-12-04 21:18 ` Karsten de Freese
2013-04-09 13:23 ` Marcin Nowak
2013-04-09 15:19 ` Jean Delvare
2013-04-09 15:22 ` Marcin Nowak

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=4EC98748.4030805@maxi-dsl.de \
    --to=karsten@maxi-dsl.de \
    --cc=lm-sensors@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.