On Wed, Mar 18, 2009 at 10:52 PM, Misael Lopez <[email protected]> wrote: >>> Below are the my changes in kernel side. As mentioned in previous >>> mail, I'm using Switch GPIO driver (which uses Switch Class driver) to >>> generate the headset uevent. In the patch, it's also included the >>> board file (for OMAP 3430SDP board) although it's not part of standard >>> android kernel it may be useful for reference. >> >> Do you have a pointer to those drivers? > > Those drivers are provided along with the android kernel. You can find them > at: > > http://android.git.kernel.org/?p=kernel/common.git;a=tree;f=drivers/switch;h=c02cb39fc78a60aa4b046481596d7b184b39fa8a;hb=453dc2b690b52248ac6dcb068965f1a37d675cef
Ah, thanks, missed that. Perhaps such a generic "switch" name would be good to change, that's very vague. Also, why a special class? Why not use the input layer for stuff like this? And, why have a name attribute that should be obvious from the name of the struct device? Why the duplication? Also, putting 280 bytes on the stack isn't that nice in the switch_set_state() function, is it really needed? Why can't that be dynamic? And I don't think the "name" has to be duplicated here as it is already transmitted in the uevent. Hey code review on the android-kernel list, it's about time :) thanks, greg k-h --~--~---------~--~----~------------~-------~--~----~ unsubscribe: [email protected] website: http://groups.google.com/group/android-kernel -~----------~----~----~----~------~----~------~--~---
