RE: Alpha: virt_to_bus/GFP_DMA problem

Jakma, Paul (Paul.Jakma@compaq.com)
Fri, 10 Dec 1999 17:09:55 -0000


> Hmmm this is quite neat, my posting was of course based on the
> assumption that the bus controller was simple ... doesn't change the
> fact that the ESS Solo1 is a pile of j**k ;-)
>

most (all i know of) of the EV5 and up (maybe pre-EV5 aswell) machines are
real nice. From the 21172/21174 doc's, it seems it works very much like CPU
page lookups. AFAIK The chipset stores a map table in a page in RAM, and has
a pinnable 8 entry TLB.

Very advanced stuff. Potential to solve any type of bus->phys mapping
problem. And would allow large memory on Alpha without needing zones, and
any bus device could DMA to/from anywhere in RAM without needing bounce
buffers. But we don't use it yet.

> This would be good for low end devices that don't do scatter/gather
> themselves - I seem to remember some of the frame grabber ones have
> this problem.
>

Even for scatter-gather capable hardware it would be a win. Chipset
scatter-gather would allow them to access any RAM. Irregardless of how much
memory is in the machine.


> Right, this should of course be combined with generic scatter/gather
> support in important places like the skbuffs (through the use of
> kiobufs). Some well designed hardware (like some modern network cards)
> already supports this .... and it would solve a lot of problems for
> the HIPPI and Fibre Channel drivers.
>

generic scatter-gather interface would be nice. Alpha would really benefit
from it.

> Jes
>

-paul jakma.

-
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.tux.org/lkml/