On Thu, Oct 23, 2008 at 03:20:30PM +0200, Jean Delvare wrote: > On Thu, 23 Oct 2008 15:13:23 +0200, Manuel Lauss wrote: > > On Thu, Oct 23, 2008 at 02:57:10PM +0200, Jean Delvare wrote: > > > I don't know much about the SPI side of things so I would like someone > > > else to comment on that aspect of the patch (in particular the way > > > different device types are supported by the same driver.) David, can > > > you please take a look and let us know if you have any objection? > > > > I'd be more than happy to just use a number cast to void * for > > spi platform_data and get rid of this header. I'm not sure however > > whether this is acceptable style for kernel code. > > Err, no, that's not what I was suggesting. Abusing pointers to store > integers is bad, please don't do that. > > I was more curious about the need to have such arbitrary numbers to > differentiate between chip types. For i2c, we can just tell a driver to > support several chip names. But maybe the spi subsystem doesn't support > that. Ah, yes, the lack of an i2c_driver.id_table parameter was why I originally wrote a separate driver. I suppose I could add another struct spi_driver for the TMP121 and sort out chip types internally. Thanks! Manuel Lauss
