Messages in this thread | | | Date | Tue, 7 Aug 2012 16:44:58 +0100 | From | Russell King <> | Subject | Re: [PATCH 0/5] mfd: replace IORESOURCE_IO by IORESOURCE_MEM |
| |
On Tue, Aug 07, 2012 at 05:22:45PM +0200, Geert Uytterhoeven wrote: > And as Arnd pointed out, if resources will be used for various new buses, > "IORESOURCE_FOO" or "IORESOURCE_OTHER" is a bit vague. > What about conflicts where one driver means i2c addresses and another > one means gpio addresses? The resource system will reject them?
I changed the subsequent patch to use IORESOURCE_REG - at least that better describes what it's for. Maybe IORESOURCE_REGRANGE would be better (so it can be used for any register range resource on any bus type) ?
However, one issue that I hope has already been addressed is what space the ranges are in, and how does a sub-driver get to know that. To put it another way, how does a sub-driver get to know about the 'base' for these register ranges. I hope that problem has been thought about in MFD land _before_ the approach of passing around register ranges through resources was allowed to happen.
-- Russell King Linux kernel 2.6 ARM Linux - http://www.arm.linux.org.uk/ maintainer of:
| |