sensors vs 2.6, sensors 0 kernel 1

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Gene,

You really want to contact the mailing-list to report problems. More
readers, more chance to get your problem solved.

> I'm currently running the 2.6.1-rc1-mm1 kernel, and have followed the 
> instruction in lm_sensors-2.8.2, doing only a make user and make 
> user-install, and nothing in the ic2-2.8.2 dir.

As a general rule, 2.6.0 is best supported by lm_sensors-2.8.2 and
2.6.1-rc1 and next are better supported by lm_sensors2-CVS (or will be
soon).

> Everything required for a 2.4 kernel to read all my hardware is turned
> on in the 2.6 config.

As modules or into the kernel?

> ISTR being told that everything has been moved from the /proc virtual 
> dir to the /sys virtual dir, but while a bunch of stuffs from the 
> bttv and gforce2 card show up there, none of the hardware stuff for 
> temps, voltages and fans seems to have survived the trip.

What are you using for the geforce2 card? Rivatv? Is that something
present in the kernel or some additional patch?

> 
> Here is a tree of that dir:
> [root at coyote i2c]# pwd
> /sys/bus/i2c
> [root at coyote i2c]# tree
> .
> |-- devices
> |   |-- 0-0050 -> 
> ../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0050
> |   |-- 0-0051 -> 
> ../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0051
> |   |-- 0-0052 -> 
> ../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0052
> |   |-- 0-0053 -> 
> ../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0053
> |   |-- 0-0054 -> 
> ../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0054
> |   |-- 0-0055 -> 
> ../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0055
> |   |-- 0-0056 -> 
> ../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0056
> |   |-- 0-0057 -> 
> ../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0057
> |   |-- 0-0061 -> 
> ../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0061
> |   |-- 1-0050 -> 
> ../../../devices/pci0000:00/0000:00:11.0/i2c-1/1-0050
> |   `-- 1-0051 -> 
> ../../../devices/pci0000:00/0000:00:11.0/i2c-1/1-0051
> `-- drivers
>     |-- dev_driver
>     |-- eeprom
>     |   |-- 0-0050 -> 
> ../../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0050
>     |   |-- 0-0051 -> 
> ../../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0051
>     |   |-- 0-0052 -> 
> ../../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0052
>     |   |-- 0-0053 -> 
> ../../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0053
>     |   |-- 0-0054 -> 
> ../../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0054
>     |   |-- 0-0055 -> 
> ../../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0055
>     |   |-- 0-0056 -> 
> ../../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0056
>     |   |-- 0-0057 -> 
> ../../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0057
>     |   |-- 1-0050 -> 
> ../../../../devices/pci0000:00/0000:00:11.0/i2c-1/1-0050
>     |   `-- 1-0051 -> 
> ../../../../devices/pci0000:00/0000:00:11.0/i2c-1/1-0051
>     |-- generic i2c audio drive
>     |-- i2c TV tuner driver
>     |   `-- 0-0061 -> 
> ../../../../devices/pci0000:00/0000:00:09.0/i2c-0/0-0061
>     |-- i2c msp3400 driver
>     |-- i2c tda7432 driver
>     |-- i2c tda9875 driver
>     |-- i2c tda9887 driver
>     |-- i2c_adapter
>     |-- tv card mixer drive
>     `-- w83781d
> 
> 35 directories, 0 files

I suspect that tree doesn't follow symlinks. What if you "cd
/sys/bus/i2c/devices/1-0050" and list the files in there?

> I've just this week upgraded my atk, pango, glib and gtk+- to the 
> latest versions, running into version hell for a while, and gkrellm 
> is now 2.1.24.  Everybody who will bother to answer the questions 
> says it "just works"  except that it doesn't.  The only thing related 
> to i2c that shows up in dmesg is this:
> 
> Linux version 2.6.1-rc1-mm1 (root at coyote.coyote.den) (gcc version 3.2 
> 20020903 (Red Hat Linux 8.0 3.2-7)) #1 Fri Jan 2 19:25:06 EST 2004
> [...]
> i2c /dev entries driver

I'd expect messages like "1-0050 registered" too (can't tell you the
exact message, I'm not under 2.6). Don't you see them?

> running sensors-detect, it appears its still looking in the /proc
> dir, and bails out when it cannot find it:
>  We are now going to do the adapter probings. Some adapters may hang 
>  halfway through; we can't really help that. Also, some chips will be
>  double detected; we choose the one with the highest confidence value
>  in that case.
>  If you found that the adapter hung after probing a certain address, 
>  you can specify that address to remain unprobed. That often includes
>  address 0x69 (clock chip).
> Couldn't open /proc/bus/i2c?!? at /usr/local/sbin/sensors-detect line 
> 3184, <STDIN> line 9.
> [root at coyote i2c]#

Strange. It is supposed to work and was tested to, too. What does "type
-a i2cdetect" return?

> Now, I've just wander through /usr/local/sbin/sensors-detect, 
> changeing all the /proc to /sys, and the script at least runs and 
> generates a sample lm_sensors to be copied in /etc/init.d.
> I copied that, and rebooted again, but there isn't anything in the 
> logs or available to gkrellm from the w83781d device.
> 
> Am I doing something wrong yet?

No, it seems correct. I suspect that the w83781d driver doesn't detect
you chip correctly, but I can't tell why. Please provide the output of
"lsmod" so that we can take a look at your modules stack (if you used
modules). If you did not use module, please try to do so (for the i2c
stuff at least) and see it it is any better. I admit that we all mostly
use modules because it makes debugging much easier, so the in-kernel
method isn't too well tested.

-- 
Jean Delvare
http://www.ensicaen.ismra.fr/~delvare/



[Index of Archives]     [Linux Kernel]     [Linux Hardware Monitoring]     [Linux USB Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [Yosemite Backpacking]

  Powered by Linux