![]() | |||||||||||||
Messages in this thread |
Kyle Moffett said: > The USB SL811 host has a structure stuck in linux/usb_sl811.h As this > structure is kernel private and only used by a single piece of code, That's very wrong ... the structure in question is the interface between the board-specific setup and the driver (as it says in the comment). A better way to put the situation is that www.kernel.org doesn't currently bundle support for a board that wires that chip into its platform_bus. > this patch moves it to the header file from which it is used, deleting the > old one. Here's a thought experiment for you ... do that for EVERY driver on the platform_bus, and see how quickly you get a big mess. Board support packages on 2.4 tended to take that approach; good riddance, IMO. When you take that approach, then adding support for a new board means ** CHANGING EVERY DRIVER USED WITH THAT BOARD ** to know how the chip is set up on that particular board. IRQs, addressing, CPU and GPIO pin multiplexing, support logic ... yeech! Better to have a single arch/XYZ/mach-ZYX/board-ABC.c file that just sets up the relevant platform devices, and the platform_data those drivers use. The board support packages should be small, mostly using standard drivers ... not be invasive monsters. - Dave - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/ | ||||||||||||
| Last update: 2005-03-22 14:09 [from the cache] ©2003-2008 | |||||||||||||