IBM PPC 405 and I2C

Mark Studebaker mds at
Tue Aug 5 00:03:08 CEST 2003

As Khali implied, it's unlikely that the drivers in our tree are up-to-date,
check w/ the PPC mailing lists.

Nobody is working on slave mode to our knowledge.
Sounds like you already found the relevant emails.

Jean Delvare wrote:
> > I am working with a powerpc 405 chip, which supports both master and
> > slave in the hardware.  The linux distribution is Montavista Hardhat.
> >
> > 1)  The i2c files i2c-algo-ibm_ocp.h, i2c-algo-ibm_ocp.c, and
> > i2c-adap-ibm_ocp.c were written by Montavista.  Unfortunately, they do
> > not support slave mode.  Has anyone outside Montavista added
> > functionality to this code?  Does Montavista provide updates often (it
> > appears the code with the Montavista distribution has significant
> > modifications over what is distributed with i2c, unfortunately I do
> > not have the code available to me now)?
> According to the CVS logs, there were no updates from Montavista since
> the initial commit back in July 2002. If there are changes that should
> make their way to our repository, please let us know.
> > 2)  Has anyone successfully insmoded the i2c-algo_ibm_ocp and
> > i2c-adap-ibm_ocp modules together?  Insmod complains about recursive
> > includes (which is true, the files call on functions in the other
> > file) when loading either module.
> This requires specific hardware we don't have, so it is unlikely any of
> us could try these modules. From my understanding, the adapter needs the
> algorithm, and not the other way around. That's the way it seems to work
> in our version of the driver. Maybe you could give it a try?
> > 3)  There have been several inquiries about using a host/bus in slave
> > mode. Has anyone done work on this since the postings?  Are their
> > plans to implement this at some point?  What would be required to add
> > slave functionality to the core i2c files?
> I don't think anyone worked seriously on this. Anyway, this is far over
> my knowledge, so I can't help.
> --
> Jean Delvare

More information about the lm-sensors mailing list