[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [MiNT] virtual memory



On Tue, Apr 18, 2000 at 08:38:34PM -0500, Michael White wrote:

> Second, if we want virtual memory, we really can't afford to give
> each process an MMU table entry, as we do now.  If I recall, the table
> size would be in the megabytes per process.  Therefore, I'm going

Only if you use 256-byte pages. I would also suggect to have a page table
per process ans simply switch the root pointers on taskswitch (which is
expensive enough, since you have to clear the TLB).

> allocated from anywhere it TT RAM.  Note that I'm not mucking with ST RAM,
> due to all the DMA stuff.  ST RAM will have to remain as it is, more

Hm. You know that DMA is *not* impossible in TT-RAM? The TT SCSI controller
can do it, the Magnum/F030 RAM is dma-capable, and the Milan TT RAM also.

IMHO, the solution is to do DMA to physical addresses. The Milan SCSI driver
already does this (since virtual != physical addresses in Milan TT RAM),
however, modifications would be needed to the used virt_to_phys-function (in
the PCI bios) to support multiple address spaces.

A nice solution might be a new interface to the lowlevel drivers that passes
physical addresses instead of the current virtual ones.

cu
Michael
-- 
Michael Schwingen, Ahornstrasse 36, 52074 Aachen