8736x (Was: 3736x)

Jean Delvare khali at linux-fr.org
Tue Apr 6 10:09:30 CEST 2004

According to Mark Studebaker:

> I verified that the 37365/66 fan monitoring/control is identical to
> the one supported by pc37360-fan.

Correct, except that it's 87360, not 37360.

> Therefore I propose that we add the voltage and temp sensor support
> for the 65/66 to that module, and we rename the module to pc3736x.

The "fan" suffix was mainly there to avoid a name conflict with another
pc87360 driver, not to reflect the fact that these chips were not
complete hardware monitoring chips. These Super I/0 chips are much more
than hardware monitoring chips and I thought that maybe there would be
other drivers for the other logical devices (as part of the Linux
kernel tree, or a different external project).

So the new driver name could have been pc37360-hwmon. But now that I
come to think a little more about it, it turns out that the other
logical devices most likely don't need a specific driver, so the suffix
is probably not necessary and we can just drop it.

As for the exact name, I don't much like the idea of having "x" in
driver names. It's not very informative about which values "x" can
hold. And for example, someone googl'ing for "pc8736x" will get much
fewer and less relevant answers than with "pc87360". Also, if other
chips with a different name happen to be compatible with these ones,
they won't fit in the name anyway.

The lm_sensors' project tends to use the first supported chip name as
the driver name, as far as I can tell, and I would suggest that we
stick to this. For this reason, I propose that we rename the driver

I would then add write support for the fan part and add the 365 and 366
to the list of supported chips. Do we agree that we can ignore
National's recommendation of not writing to the chip while monitoring
is enabled?

Jean Delvare

More information about the lm-sensors mailing list