Re: problem: linux and (really) fast I/O

Andrea Arcangeli (arcangeli@mbox.queen.it)
Mon, 8 Jun 1998 21:57:39 +0200 (CEST)


On Mon, 8 Jun 1998, Davide Rossetti wrote:

> Here I really need to someway lock mmap-ped pages and to get the
> 'real' address to be passed to the PCI device. Moreover I'm limited
> by the 4096 page size as the maximum DMA transaction size, unless i

Remeber that you can always force the kernel to use only a small part of
the ram in the system. For example in a 128Mbyte machine, you can run the
kernel in 64Mbyte di ram and use the remaining 64Mbyte of ram only as
contiguous DMA buffer area. This is the way that Rubini used in its device
driver. This won' t help you in transferring chunk > PAGESIZE of an
mmapped file via DMA, but sure will help you avoiding memory all
fragmentation problems if you need that ;-).

Andrea[s] Arcangeli

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu