[i2c] "regd i2c-viapro bus driver"

[Jean Delvare]

Hi Savitha,

> I saw the patch in the following link. It looks like
> very minimal changes to i2c-viapro.c.

Indeed.

> how do i check if the i2c-viapro works for VT 8251
> after applying patch?

After loading i2c-viapro and i2c-dev, "i2cdetect -l" should list an
entry for the VIA south bridge.

i2c-0   unknown         SMBus Via Pro adapter at 5000           Algorithm unavailable

> Secondly if i am interfacing a i2c device to VT 8251
> southbridge, which ports of VT 8251 are used to
> communicate with the i2c device?

You don't really care. The implementation details belong to the
i2c-viapro driver. The ports used are motherboard-dependent.
"i2cdetect -l" will tell you, but this is for information only, again
you don't need to know. What you really care about is the i2c bus
number (which again "i2cdetect -l" will give you.)

> and also, If the i2c device generates interrupts to VT
> 8251, how it should be handled?

That's a complex question. First of all, the I2C protocol has no
provision for interrupts. The SMBus protocol does, in a form known as
"SMBus Alert". However, not all motherboards are wired to support it,
and more importantly, the Linux i2c-core doesn't support it at the
moment, and the i2c-viapro driver (and others) would need to be
converted from poll-based to interrupt-driven to use it if it were.
It's a long way to go.

-- 
Jean Delvare