Messages in this thread | | | Date | Mon, 20 Jul 1998 09:28:16 +1000 | From | Richard Gooch <> | Subject | Re: [PATCH] New phys_addr() syscall |
| |
Alan Cox writes: > > Alan Cox writes: > > > > People writing pure userspace network drivers may also find this handy > > > > if they want to DMA directly into userspace. > > > > > > It doesnt help. You have locking and copy/write issues to worry about for > > > the DMA case. This seems an extremely fringe and bizarre syscall to consider > > > for general use! > > > > What issues? If I pin down a set of pages, I can then get their > > physical address and tell the card to DMA its heart out. How exactly > > is this going to fail? > > DMA doesnt honour the MMU write protect bits for copy on write, or for > page locking. Nor does it set the dirty bit on page table entries.
OK, so it's not a safe and clean mechanism. On the other hand, if you want a fast and dirty mechanism, it should work fine. Something akin to the U-NET ATM drivers could use it. Note: I'm not saying that this is the way networking should be done! It's just something that I noticed you could do if you wanted.
Regards,
Richard....
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html
| |