Stefan Roese sr@denx.de wrote:
On Tuesday 01 September 2009 10:57:52 Haavard Skinnemoen wrote:
Well, usually we run with IC on and with DC off, usually because quite a number of drivers and other code do not use proper I/O accessors yet, and/or because it's easier and nobody really cares. For example on PowerPC, adding support for the data cache usually gives only a minimal performance boost. This may be different on other architectures.
Ok, so the code is broken and nobody else cares?
I wouldn't put it like this. The CFI driver assumes that the FLASH mapping is not cached. This makes perfect sense in my point of view.
Yes, and it can do that safely because it specifically asks for an uncached address from map_physmem(). That part didn't actually change with the patch in question -- the problem is that the address returned by map_physmem() is now exposed through external interfaces -- it's not just an internal implementation detail anymore.
I suppose I could disable the DC (which is a bit complicated, but possible), but that would just add to the already high cost (in terms of both code size and performance) of using common code (i.e. the CFI driver), so I'm leaning towards a custom flash driver instead.
On some 440 platforms we configure the FLASH cached upon powerup, and disable the caching after relocating to SDRAM. So when the CFI driver is started, the FLASH is not cached any more, but we have the cache speedup upon relocation.
That's what I want to do as well! And I can actually do that as long as the flash subsystem uses map_physmem() consistently.
Haavard