[lm-sensors] RE: w83792d driver for the 2.6 kernel

[lm-sensors] RE: w83792d driver for the 2.6 kernel

[Huang0]

Hi Fabien, Rudolf

> I saw that you were porting the w83792d driver to the 2.6 kernel. I
need it
> to
> monitor the temperature on the SuperMicro P8SCT motherboard that my
team is
> evaluating.

I have finished the porting of w83792d driver to the 2.6 kernel, and
submitted
it to the lm_sensors group. Rudolf, the member of lm_sensors group
helped me
to finish the documents and will submit it to the kernel term for me in
the
near future, I think.


> Could you tell me where I can get the source code for the 2.6 kernel ?
Also,
> when do you plan to submit it to the Kernel team ?

Hi Rudolf
Would you please give us the answer to the above questions?


Thanks
Best Regards
Chunhao


==============================================The privileged confidential information contained in this email is intended for use only by the addressees as indicated by the original sender of this email. If you are not the addressee indicated in this email or are not responsible for delivery of the email to such  a person, please kindly reply to the sender indicating this fact and delete all copies of it from your computer and network server immediately. Your cooperation is highly appreciated. It is advised that any unauthorized use of confidential information of Winbond is strictly prohibited; and any information in this email irrelevant to the official business of Winbond shall be deemed as neither given nor endorsed by Winbond.
==============================================If your computer is unable to decode Chinese font, please ignore the following message.It essentially repeats the statement in English given above.本信件內所含華邦電子的財產性機密性資訊, 僅授權原發信人指定之收信人取閱\之用. 假使您並非被指定之收信人或因任何原因在未經授權的情形之下收到本信件, 請您告知原發信人並立即將信件從電腦與網路伺服器中予以消除. 對於您的合作, 我們先此致謝. 特此提醒, 任何未經授權擅自使用華邦電子的機密資訊的行為是被嚴格禁止的. 信件與華邦電子營業無關之內容,不得視為華邦電子之立場或意見.

[lm-sensors] RE: w83792d driver for the 2.6 kernel

[Rudolf Marek]

Hi all,
>>to
>>monitor the temperature on the SuperMicro P8SCT motherboard that my
> 
> team is
> 
>>evaluating.

I have ready a patch for 2.6.x kernels (2.6.11, 2.6.12 older might work too) please see the attachment.
And also you need to use CVS version of lm_sensors.Download from here:
http://secure.netroedge.com/~lm78/archive/lm_sensors-daily.tar.gz
and compile with:
make user; make user_install

Please let us know how it works.

> I have finished the porting of w83792d driver to the 2.6 kernel, and
> submitted
> it to the lm_sensors group. Rudolf, the member of lm_sensors group
> helped me
> to finish the documents and will submit it to the kernel term for me in
> the
> near future, I think.

Yes I will work on it today. It is nearly ready and soon it will go to -mm- series of kernels.
There were some sysfs callbacks changes in -mm- and this needs to be incorporated to the driver.

Regards
Rudolf
-------------- next part --------------
A non-text attachment was scrubbed...
Name: patch_w83792d-2.6.12.gz
Type: application/x-gzip
Size: 14396 bytes
Desc: not available
Url : http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20050621/8dfe2e28/patch_w83792d-2.6.12-0001.gz

[lm-sensors] RE: w83792d driver for the 2.6 kernel

[fabien.pourias]

Hi,

Thanks a lot for the patch. It works pretty well (i.e. we see meaningful 
values when we run sensors -u) but the user tools are not up to date :
 - sensors-detect fails to detect the w83792d
 - the "sensors" output only includes the old w83627hf data

So we run sensors -u to get all the results. NOTE: we always use sensors -u 
since we need to postprocess the output and the "raw" results are usually 
better for that.

 sincerely,

    Fabien Pourias

[lm-sensors] RE: w83792d driver for the 2.6 kernel

[Jean Delvare]

Hi Fabien,

> Thanks a lot for the patch. It works pretty well (i.e. we see meaningful
> values when we run sensors -u) but the user tools are not up to date :
>  - sensors-detect fails to detect the w83792d
>  - the "sensors" output only includes the old w83627hf data

Did you update the lm_sensors user-space tools as Rudolf Marek
recommended? Support for the W83792D is in CVS only.

--
Jean Delvare

[lm-sensors] RE: w83792d driver for the 2.6 kernel

[fabien.pourias]

>=== Original Message From "Jean Delvare" <khali@linux-fr.org> ==>Hi Fabien,
>
>> Thanks a lot for the patch. It works pretty well (i.e. we see meaningful
>> values when we run sensors -u) but the user tools are not up to date :
>>  - sensors-detect fails to detect the w83792d
>>  - the "sensors" output only includes the old w83627hf data
>
>Did you update the lm_sensors user-space tools as Rudolf Marek
>recommended? Support for the W83792D is in CVS only.

I used the daily snapshot as recommanded by Rudolf and had to :
 - copy the etc/sensors.conf.eg to /etc/sensors.conf
 - run the sensors-detect from the lm_sensors distribution (the make 
user_install command did not install it)
 - run /usr/local/bin/sensors to make sure I did not run the Fedora version of 
sensors.

I also removed the Fedora version but the sensors-detect script still failed 
to detect the 83792d. Could it be because it is not properly installed ? What 
commands should I issue in addition to "make user; make user_install" ?

   thanks for your help,

    F.  Pourias

[lm-sensors] RE: w83792d driver for the 2.6 kernel

[Jean Delvare]

Hi Fabien,

> I used the daily snapshot as recommanded by Rudolf and had to :
>  - copy the etc/sensors.conf.eg to /etc/sensors.conf

This is true.

>  - run the sensors-detect from the lm_sensors distribution (the make
> user_install command did not install it)

This would be very surprising, please double-check. Remember that
everything is installed under /usr/local by default, and sensors-detect
is only usable by root so it's installed in /usr/local/sbin.

>  - run /usr/local/bin/sensors to make sure I did not run the Fedora
> version of sensors.

You should even uninstall the Fedora package entierly, else the new
sensors command may pick the old library and you won't get correct
results.

> I also removed the Fedora version but the sensors-detect script still
> failed to detect the 83792d. Could it be because it is not properly
> installed ? What commands should I issue in addition to "make user; make
> user_install" ?

These commands should be sufficient. Make sure that your kernel supports
the I2C Device Interface (i2c-dev driver), sensors-detect needs it (and
should have complained if it's not present). Also note that
sensors-detect cannot detect I2C chips for which you have a driver
currently loaded (because the probe would possibly confuse the driver).
If you can't figure it out, please send the complete output of
sensors-detect after unloading all i2c chip drivers.

--
Jean Delvare

[lm-sensors] RE: w83792d driver for the 2.6 kernel

[Huang0]

Hi Fabien, Jean

> >  - run the sensors-detect from the lm_sensors distribution (the make
> > user_install command did not install it)
> 
> This would be very surprising, please double-check. Remember that
> everything is installed under /usr/local by default, and
sensors-detect
> is only usable by root so it's installed in /usr/local/sbin.
> 
> >  - run /usr/local/bin/sensors to make sure I did not run the Fedora
> > version of sensors.
> 
> You should even uninstall the Fedora package entierly, else the new
> sensors command may pick the old library and you won't get correct
> results.
> 
> > I also removed the Fedora version but the sensors-detect script
still
> > failed to detect the 83792d. Could it be because it is not properly
> > installed ? What commands should I issue in addition to "make user;
make
> > user_install" ?

This is Chunhao, 
I'm always using the Redhat/Fedora Core to develop 792 driver, I'd like
to
introduce the problem I met before:
After I install Redhat/Fedora, there will be the lm_sensors related
programs
in the /usr/sbin/ directory, such as:
/usr/sbin/sensors
/usr/sbin/sensors-detect
/usr/sbin/i2cset
/usr/sbin/isaset

And after I install the latest lm_sensors version(CVS version),
the latest lm_sensors programs will be installed into /usr/local/sbin/
directory, if I use #sensors or #sensors-detect at this time, the
programs
will be the old one in /usr/sbin/ and can NOT find the 792 chip.

So, usually I will remove the above 4 programs in /usr/sbin/ directory
and make soft links from /usr/local/sbin/_program_ to /usr/sbin/ (this
soft link step may be omitted, you only need remove) then it will work.

If it still can NOT work, then maybe there is some other reason, we need
more help of lm_sensors :-(



Thanks
Best Regards
Chunhao


==============================================The privileged confidential information contained in this email is intended for use only by the addressees as indicated by the original sender of this email. If you are not the addressee indicated in this email or are not responsible for delivery of the email to such  a person, please kindly reply to the sender indicating this fact and delete all copies of it from your computer and network server immediately. Your cooperation is highly appreciated. It is advised that any unauthorized use of confidential information of Winbond is strictly prohibited; and any information in this email irrelevant to the official business of Winbond shall be deemed as neither given nor endorsed by Winbond.
==============================================If your computer is unable to decode Chinese font, please ignore the following message.It essentially repeats the statement in English given above.本信件內所含華邦電子的財產性機密性資訊, 僅授權原發信人指定之收信人取閱\之用. 假使您並非被指定之收信人或因任何原因在未經授權的情形之下收到本信件, 請您告知原發信人並立即將信件從電腦與網路伺服器中予以消除. 對於您的合作, 我們先此致謝. 特此提醒, 任何未經授權擅自使用華邦電子的機密資訊的行為是被嚴格禁止的. 信件與華邦電子營業無關之內容,不得視為華邦電子之立場或意見.

[lm-sensors] RE: w83792d driver for the 2.6 kernel

[Huang0]

Hi Fabien


> I also removed the Fedora version but the sensors-detect script still
failed
> to detect the 83792d. Could it be because it is not properly installed
? What
> commands should I issue in addition to "make user; make user_install"
?

I check the motherboard SuperMicro P8SCT
http://www.supermicro.com/products/motherboard/P4/E7221/P8SCT.cfm
and it's user manual:
http://www.supermicro.com/manuals/motherboard/E7221/MNL-0776.pdf
and do NOT find w83792d chip but w83627hf chip, Are you sure that your
motherboard
contains the 792 chip?

Thanks
Best Regards

Chunhao


==============================================The privileged confidential information contained in this email is intended for use only by the addressees as indicated by the original sender of this email. If you are not the addressee indicated in this email or are not responsible for delivery of the email to such  a person, please kindly reply to the sender indicating this fact and delete all copies of it from your computer and network server immediately. Your cooperation is highly appreciated. It is advised that any unauthorized use of confidential information of Winbond is strictly prohibited; and any information in this email irrelevant to the official business of Winbond shall be deemed as neither given nor endorsed by Winbond.
==============================================If your computer is unable to decode Chinese font, please ignore the following message.It essentially repeats the statement in English given above.本信件內所含華邦電子的財產性機密性資訊, 僅授權原發信人指定之收信人取閱\之用. 假使您並非被指定之收信人或因任何原因在未經授權的情形之下收到本信件, 請您告知原發信人並立即將信件從電腦與網路伺服器中予以消除. 對於您的合作, 我們先此致謝. 特此提醒, 任何未經授權擅自使用華邦電子的機密資訊的行為是被嚴格禁止的. 信件與華邦電子營業無關之內容,不得視為華邦電子之立場或意見.

[lm-sensors] RE: w83792d driver for the 2.6 kernel

[P.O. Gaillard]

Huang0@Winbond.com.tw wrote:

> Hi Fabien
>
> I check the motherboard SuperMicro P8SCT
> http://www.supermicro.com/products/motherboard/P4/E7221/P8SCT.cfm
> and it's user manual:
> http://www.supermicro.com/manuals/motherboard/E7221/MNL-0776.pdf
> and do NOT find w83792d chip but w83627hf chip, Are you sure that your
> motherboard
> contains the 792 chip?
> 
Indeed, Ken Wo from Supermicro Nederland was kind enough to tell us about it and 
we saw that :
  - we can load the w83792d module on the P8SCT
  - doing sensors -u shows results from both chips and the ones from the w83792d 
are the ones that seem right (i.e. 40 Celsius instead of -1 Celsius)

So, it's pretty clear the there is a w83792d somewhere on the motherboard and 
that the temperature sensors and fans are connected to it.

	thank you for writing the driver which allows us to choose this motherboard for 
our project.

	sincerely,

	P.O. Gaillard

[lm-sensors] RE: w83792d driver for the 2.6 kernel

[Rudolf Marek]

Hi all

Please can you modprobe the bus driver

modprobe i2c-i801
modprobe i2c-dev
i2cdetect 0

And send us output of i2cdetect command?

Somewhere in region 0x20 - 0x2f should be one number. Use it in next command

i2cdump 0 0x2f

0x2f is an example. Most likely it will be 0x2f or 0x2d or 0x2e

Are you sure you ran  /usr/local/sbin/sensors-detect

and not /usr/sbin/sensors-detect

Thanks,
regards

Rudolf.

[lm-sensors] RE: w83792d driver for the 2.6 kernel

[fabien.pourias]

>=== Original Message From lm-sensors@lm-sensors.org ==>Hi all
>
>Please can you modprobe the bus driver
>
>modprobe i2c-i801
>modprobe i2c-dev
>i2cdetect 0
>
>And send us output of i2cdetect command?

see attach output_i2cdetect.txt file.

>Somewhere in region 0x20 - 0x2f should be one number. Use it in next command
>
>i2cdump 0 0x2f
>

see attach output_i2cdump.txt file.

>0x2f is an example. Most likely it will be 0x2f or 0x2d or 0x2e
>
>Are you sure you ran  /usr/local/sbin/sensors-detect
>
>and not /usr/sbin/sensors-detect
>

Yes i am sure because there is no binary sensors-detect in /usr/sbin.

Thank,
Fabien POURIAS.

>Thanks,
>regards
>
>Rudolf.

-------------- next part --------------
No size specified (using byte-data access)
WARNING! This program can confuse your I2C bus, cause data loss and worse!
I will probe file /dev/i2c-0, address 0x2f, mode byte
Continue? [Y/n] 
     0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f    0123456789abcdef
00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
10: 00 00 64 64 93 00 00 78 00 ff 00 00 04 00 e0 b3    ..dd?..x....?.??
20: a8 ff d2 bf 1f c4 d4 28 32 ff ff c6 94 b5 83 e2    ?.?????(2..?????
30: b9 ce a8 2f 13 d9 b1 cc a7 4b 46 ed ed ed 1f 01    ???/?????KF?????
40: 03 00 00 df 7f ff ff 33 2f 13 73 87 43 ff 80 5c    ?..??..3/?s?C.?\
50: ff ff ff ff ff ff ff ff 7a 10 ff 33 33 01 05 7f    ........z?.33???
60: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff    ................
70: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff    ................
80: 01 8f 01 8f 15 00 00 00 11 11 ff ff 3c 3c 06 06    ?????...??..<<??
90: 00 00 00 01 80 ff 00 00 11 ff 3c 03 df 01 01 ff    ...??...?.<????.
a0: 01 01 01 e0 8f 8f 8f ff 3a c2 10 c3 ff ff 00 00    ???????.:???....
b0: cb c8 ff ff cc a7 cc a7 33 ff ff ed ed ed ff ff    ??..????3..???..
c0: 25 80 00 4b 00 50 00 ff cf 80 00 4b 00 50 00 ff    %?.K.P..??.K.P..
d0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff    ................
e0: 8b 8b 8b 28 3c 50 28 3c 50 28 3c 50 ff ff ff ff    ???(<P(<P(<P....
f0: ff ff 00 ff 00 01 00 00 ff 02 ff 00 00 00 00 ff    .....?...?......
-------------- next part --------------
WARNING! This program can confuse your I2C bus, cause data loss and worse!
I will probe file /dev/i2c-0.
I will probe address range 0x03-0x77.
Continue? [Y/n] 
     0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
00:          XX XX XX XX XX 08 XX XX XX XX XX XX XX 
10: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX 
20: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX 2f 
30: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX 
40: XX XX XX XX 44 XX XX XX XX XX XX 4b XX XX XX 4f 
50: 50 XX XX XX 54 XX XX XX XX XX XX XX XX XX XX XX 
60: 60 61 XX XX XX XX XX XX XX 69 XX XX XX XX XX XX 
70: XX XX XX XX XX XX XX XX