[i2c] [PATCH] Is review of AT91 patch pending?
hskinnemoen at atmel.com
Wed Nov 7 23:53:51 CET 2007
On Wed, 7 Nov 2007 12:35:17 -0800
David Brownell <david-b at pacbell.net> wrote:
> > Users that know what they are doing can adjust this
> > value themselves, or should it be in Kconfig?
> One suggestion to Haavard (maintainer of i2c-gpio) is to
> switch the platform data so it can use ndelay() ... that
> way it can support higher clock rates (e.g. 1Mbit Fm+).
Yeah, but it would involve changing i2c_algo_bit, and all the other
drivers that depend on it, since the i2c-gpio driver simply passes the
udelay parameter along without looking much at it.
> I tend to think Kconfig here would be confusing overkill.
Yeah. Why not add a module parameter while we're at it? ;-)
The reasoning behind the platform_data interface is that whoever writes
the board code knows what speed the i2c bus is supposed to run at. If
you're adding more i2c devices than the board had originally, or trying
to push the limits of the existing ones, you should also know how to
modify the board code.
But if you have a very extensible board, you could of course add
board-specific Kconfig options to control some parameters. Or just
gather a few #defines near the top of the main board support file to
make it clear what parameters are configurable.
More information about the i2c