On Wed, 3 Mar 1999, Jonathan Stone wrote:

> oops, missed this:
> 
> >I will assert (again) that mapping the device registers is an inherently
>                                                ^^^^^^^^^
> >bus-specific operation and should be done by the driver through the use of
> >bus_space*() routines rather than the VM system.
> 
> I'm talking about mapping device *memory*, not registers.  if an app
> is going to map the memory and treat it like memory-- e.g., to blit
> into it -- then it better behave like memory once it's mapped,
> (at both  the  bus_space and d_mmap level!)
> 
> Device registers are (or can be) another kettle of fish:
> memory barriers, &c.

I suppose I should have said device registers and DMA regions, since they
have similar issues with memory barriers, cache flushes, non-cacheability,
etc.

=========================================================================
Eduardo Horvath				eeh@one-o.com
	"I need to find a pithy new quote." -- me