> You can read the contents of the eeproms more directly my looking in
> the directories /proc/sys/dev/sensors/*eeprom*.
Didn't know that (well, I read it in the eeprom module docs but forgot it
one minute later.) I just had a look at the files. I compared the data
before and after switching the two memory modules, and it is completely
different. It didn't helped me a bit understanding what could be wrong.
Where may I find docs to know which values are important?
I noticed an extra file: /proc/sys/dev/sensors/chips
It contains:
256 eeprom-i2c-0-50
257 eeprom-i2c-0-57
Needless to say I am very suprised by the value '257'. I think this may
be a clue for you, even if it won't help me for now.
I think that another interesting clue would be to understand why the
confidence for this slot was 1 instead of 8. I'll have a look at it,
although I guess you already have the answer (a wild guess from me would
be: "reading it doesn't give a usual memory size value, but it should
consiedring the address, so we set a confidence of 1 instead of 8".)
> It is possible that that slot is having issues. Although your
> computer sounds like it is running correctly?? If the eeprom(s) were
> unreadable, then at least that DIMM would not be used for RAM for the
> system since the DIMM contains the timings and such.
Indeed, the memory is perfectly detected and working.
As said before, I will try to have a look at the detection code and maybe
the sensors code itself, but probably won't find anything unless you tell
me where to look at and what to look for.
Thanks.
--
/~~ Jean "Khali" Delvare
-----\_ mail: delvare at ensicaen.ismra.fr
--------\ http://www.ensicaen.ismra.fr/~delvare/
---=ISMRA/- ____________________________________________________
