Re: [patch/rfc 2/4] pcf875x I2C GPIO expander driver

[David Brownell <david-b@pacbell.net>]

On Friday 30 November 2007, Jean Delvare wrote:
>
> So the user-space interface would be part of the generic GPIO
> infrastructure? I like the idea.

I thought that would make sense too! :)  Someone would need to
write the code though.  Having such a mechanism would provide
another "carrot" to migrate folk towards the gpiolib core.

I think adding a gpiochip primitive to mark a (potential) GPIO
as invalid would support the converse of /sys/kernel/debug/gpio.
Invalid GPIOs include pins set up for non-GPIO usage (like being
used for MMC or MII), or not wired up on a given board.  Pins
that were valid as GPIOs and not requested by a kernel driver
might reasonably be managed by userspace code.


> > > > +#include <linux/pcf857x.h>
> > > 
> > > I suspect that there will be many more such header files in the future.
> > > Would it make sense to move them to include/linux/gpio?
> > 
> > I was thinking more like <linux/i2c/...> myself.  There are many more
> > I2C chips than GPIO expanders.
> 
> But most i2c chip drivers don't need a header file. Or is this going to
> change with the new-style i2c drivers?

I expect it will become a lot more common.  Remember that legacy
I2C drivers *couldn't* get any board-specific config data; that's
been problematic, since it meant the drivers themselves ended up
with lots of board-specific cruft.   That prevented many drivers
from going upstream at all.  (As I mentioned about pcf8574 code,
although in that case the problem was worsened by lack of any
reusable kernel interface for such GPIO signals.)


> Along the same line, I am wondering if it would make sense to put the
> various GPIO drivers in drivers/gpio.

Could be.  Right now we have three "GPIO expander" drivers using
the new "gpiolib" framework:  pcf875x and pca9539 for I2C, and
mcp23s08 for SPI.  There are many more that *could* be used with
Linux boxes.  And there are other drivers/XYZ directories that
are (currently) that small.  Maybe gpiolib should go upstream
like that, and lib/gpiolib should be in drivers/gpio too...

However, keep in mind that lots of chips export a few GPIOs but
don't have that as their core functionality ... one example is
the drivers/i2c/chips/tps65010 driver.  So it'd never be the
case that GPIO drivers only live in that directory.

- Dave