[lm-sensors] Analogue Devices ADT7476ARQZ chip status

[lm-sensors] Analogue Devices ADT7476ARQZ chip status

[Reuben Farrelly]

Hi,

A while ago I opened a support ticket for a chip which I believe is currently 
unsupported:

http://www.lm-sensors.org/readticket.cgi?ticket 43

It was suggested that I post more information to the mailing list, so here goes :)

The board I have is an Intel D945PVSLKR, and it has two relevant i2c chips on 
it.  There is an SMC LPC47M182 chip, but also an Analogue Devices ADT7476ARQZ 
onboard near the CPU.  I suspect that the Analogue Devices chip is the more 
useful chip of the two (see ticket comment by Khali).

I have posted the rather lengthy output of sensors-detect up on my website at 
http://www.reub.net/files/kernel/

There's a data sheet about this chip here:

http://www.analog.com/en/prod/0%2C2877%2CADT7476%2C00.html

If this chip is in fact unsupported, what can I do to help with getting it 
supported and functional?

Thanks,
Reuben

[lm-sensors] Analogue Devices ADT7476ARQZ chip status

[Jean Delvare]

Hi Reuben,

> The board I have is an Intel D945PVSLKR, and it has two relevant i2c chips on 
> it.  There is an SMC LPC47M182 chip, but also an Analogue Devices ADT7476ARQZ 
> onboard near the CPU.  I suspect that the Analogue Devices chip is the more 
> useful chip of the two (see ticket comment by Khali).

It's "SMSC", and "Analog Devices". The LPC47M182 is not an "i2c chip".

> I have posted the rather lengthy output of sensors-detect up on my website at 
> http://www.reub.net/files/kernel/
> 
> There's a data sheet about this chip here:
> 
> http://www.analog.com/en/prod/0%2C2877%2CADT7476%2C00.html
> 
> If this chip is in fact unsupported, what can I do to help with getting it 
> supported and functional?

Please provide a dump of the Analog Devices chip:
i2cdump 0 0x2e b

-- 
Jean Delvare

[lm-sensors] Analogue Devices ADT7476ARQZ chip status

[Reuben Farrelly]

Hi again,

On 11/26/2005 10:52 PM, Jean Delvare wrote:
> Hi Reuben,
> 
>> The board I have is an Intel D945PVSLKR, and it has two relevant i2c chips on 
>> it.  There is an SMC LPC47M182 chip, but also an Analogue Devices ADT7476ARQZ 
>> onboard near the CPU.  I suspect that the Analogue Devices chip is the more 
>> useful chip of the two (see ticket comment by Khali).
> 
> It's "SMSC", and "Analog Devices". The LPC47M182 is not an "i2c chip".

Oops <embarassed look>

>> I have posted the rather lengthy output of sensors-detect up on my website at 
>> http://www.reub.net/files/kernel/
>>
>> There's a data sheet about this chip here:
>>
>> http://www.analog.com/en/prod/0%2C2877%2CADT7476%2C00.html
>>
>> If this chip is in fact unsupported, what can I do to help with getting it 
>> supported and functional?
> 
> Please provide a dump of the Analog Devices chip:
> i2cdump 0 0x2e b

[root@typhoon ~]# i2cdump 0 0x2e b
WARNING! This program can confuse your I2C bus, cause data loss and worse!
I will probe file /dev/i2c-0, address 0x2e, mode byte
Continue? [Y/n] y
      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 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
20: 77 6c c3 c5 be 36 29 27 62 0c 57 0d 46 36 ff ff    wl???6)'b?W?F6..
30: 33 4c 4c 00 00 00 00 00 ff ff ff 00 00 76 41 69    3LL..........vAi
40: 05 00 00 00 00 ff 00 ff 00 ff 00 ff 00 ff 81 7f    ?.............??
50: 81 7f 81 7f ff ff ff ff ff ff ff ff 02 22 22 7a    ????........?""z
60: 8a 80 e8 88 33 4c 4c 38 2d 2d 45 41 41 24 40 00    ????3LL8--EAA$@.
70: 00 00 00 00 00 00 1a 23 f0 00 00 55 01 00 00 00    ......?#?..U?...
80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
90: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
a0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
b0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
c0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
d0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
e0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
f0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
[root@typhoon ~]#


Reuben

[lm-sensors] Analogue Devices ADT7476ARQZ chip status

[Henrique de Moraes Holschuh]

On Sat, 26 Nov 2005, Jean Delvare wrote:
> > The board I have is an Intel D945PVSLKR, and it has two relevant i2c chips on 
> > it.  There is an SMC LPC47M182 chip, but also an Analogue Devices ADT7476ARQZ 
> > onboard near the CPU.  I suspect that the Analogue Devices chip is the more 
> > useful chip of the two (see ticket comment by Khali).
> 
> It's "SMSC", and "Analog Devices". The LPC47M182 is not an "i2c chip".

Still, the LPC47M182 is an ISA interface chip and it has sensors in it (two
tachometer inputs), so while it is not i2c material, it could be supported
by hwmon/lm-sensors.

I *really* doubt Intel is using tachometer inputs on the LPC47M182 if they
have a ADT7476ARQZ on-board, though.

-- 
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh

[lm-sensors] Analogue Devices ADT7476ARQZ chip status

[Jean Delvare]

Hi Henrique,

> > (...) The LPC47M182 is not an "i2c chip".
> 
> Still, the LPC47M182 is an ISA interface chip and it has sensors in it (two
> tachometer inputs), so while it is not i2c material, it could be supported
> by hwmon/lm-sensors.

True.

> I *really* doubt Intel is using tachometer inputs on the LPC47M182 if they
> have a ADT7476ARQZ on-board, though.

My point exactly, which is why I think we better don't waste time on
it. As I said in ticket #2043, the LPC47M182 is special in that the
hardware monitoring LDN (logical device number) depends on a
configuration bit. So even adding proper detection to sensors-detect
wouldn't be trivial.

-- 
Jean Delvare

[lm-sensors] Analogue Devices ADT7476ARQZ chip status

[Jean Delvare]

Hi Reuben,

> > Please provide a dump of the Analog Devices chip:
> > i2cdump 0 0x2e b
> 
> [root@typhoon ~]# i2cdump 0 0x2e b
> (...)
>       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 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
> 20: 77 6c c3 c5 be 36 29 27 62 0c 57 0d 46 36 ff ff    wl???6)'b?W?F6..
> 30: 33 4c 4c 00 00 00 00 00 ff ff ff 00 00 76 41 69    3LL..........vAi

This matches what the ADT7476 datasheet says. I've added support in
sensors-detect (in CVS). Please test if possible.

I've also added an entry in our "New Drivers" page.

My first impression is that the ADT7476 is partly compatible with the
ADT7467/ADT7468, which is itself partly compatible with the ADT7463,
which is an improved version of the ADM1027, which is itself mostly
compatible with the National Semiconductor LM85. We have a driver
(lm85) which covers the LM85, ADM1027 and ADT7463. If we want to
support the ADT7467/ADT7468 and ADT7476, I think we should start a new
driver. The lm85 driver already supports many chips, let's not add to
the mess.

-- 
Jean Delvare

[lm-sensors] Analogue Devices ADT7476ARQZ chip status

[Henrique de Moraes Holschuh]

On Sat, 26 Nov 2005, Jean Delvare wrote:
> My first impression is that the ADT7476 is partly compatible with the
> ADT7467/ADT7468, which is itself partly compatible with the ADT7463,
> which is an improved version of the ADM1027, which is itself mostly
> compatible with the National Semiconductor LM85. We have a driver
> (lm85) which covers the LM85, ADM1027 and ADT7463. If we want to
> support the ADT7467/ADT7468 and ADT7476, I think we should start a new
> driver. The lm85 driver already supports many chips, let's not add to
> the mess.

Well, that really depends on how the newer chips differ.  If they have
*extra* functionality only (as opposed to *different* functionality) over
whatever is already there for the amd1027 and emc6d102, it could be actually
better to support them on the lm85 driver.  Less duplicate code to keep
track of.

-- 
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh

[lm-sensors] Analogue Devices ADT7476ARQZ chip status

[Reuben Farrelly]

Hi Jean,

On 11/27/2005 12:59 AM, Jean Delvare wrote:
> Hi Reuben,
> 
>>> Please provide a dump of the Analog Devices chip:
>>> i2cdump 0 0x2e b
>> [root@typhoon ~]# i2cdump 0 0x2e b
>> (...)
>>       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 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
>> 20: 77 6c c3 c5 be 36 29 27 62 0c 57 0d 46 36 ff ff    wl???6)'b?W?F6..
>> 30: 33 4c 4c 00 00 00 00 00 ff ff ff 00 00 76 41 69    3LL..........vAi
> 
> This matches what the ADT7476 datasheet says. I've added support in
> sensors-detect (in CVS). Please test if possible.

sensors-detect in CVS works great:

Probing for `SMSC EMC6D100, EMC6D101 or EMC6D102'... Failed!
Probing for `Analog Devices ADT7467 or ADT7468'... Failed!
Probing for `Analog Devices ADT7476'... Success!
     (confidence 7, driver `to-be-written')
Probing for `National Semiconductor LM87'... Failed!

and of course finishes up with:

#----cut here----
# I2C adapter drivers
modprobe i2c-i801
# I2C chip drivers
# no driver for Analog Devices ADT7476 yet
modprobe eeprom
# sleep 2 # optional
/usr/local/bin/sensors -s # recommended
#----cut here----

This is good.  At least from an end user perspective now it looks like it is 
known to the system even if a driver hasn't yet been written.

Thank for the quick response :)

Reuben