On 10/25/10 01:45, Ben Dooks wrote: > On Mon, Oct 18, 2010 at 03:51:49PM -0400, Mike Frysinger wrote: >> On Mon, Oct 18, 2010 at 10:33, Jean Delvare wrote: >>> Why do we have set_irq_type() if we're not supposed to call it? I am >>> not claiming to be an expert in the area, but it seems totally >>> reasonable to me that the same piece of code instantiating an I2C >>> device is also responsible for setting its IRQ type. >> >> but we're back to the same issue mentioned earlier -- you cant have a >> single kernel build with modules supporting multiple drivers >> simultaneously. we like to ship development boards with a single >> kernel build on it with many modules. then people can pick the addon >> boards they wish to prototype with at runtime by plugging in the card >> and loading the module. > > I also dislike set_irq_type() as it doesn't check whether there is anyone > registered with the interrupt, which means that you could set the irq > type of someone else's irq. > > I wonder if we should pass a struct resource instead, in case there > are multiple interrupt sources, as well as having it registered with > the right resource systems. > Either works as far as I am concerned. Having seen a large set of drivers using the flags option (posted to linux-iio yesterday) I'm definitely convinced some means of allowing devices to match what the board config asks for is useful. I personally prefer the struct resource option as I have multiple drivers in IIO which have two interrupts and this is the only reason some of them use platform data. Thanks, Jonathan -- To unsubscribe from this list: send the line "unsubscribe linux-i2c" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
