On Thu, 01 Apr 1999 16:00:57 -0500, Brian Gerst
<bgerst@quark.vpplus.com> said:
>> > How about a dmamalloc() function?
>>
>> In the kernel, use GFP_DMA for dma-capable, physically contiguous memory
>> (only for ISA dma right now, as that is limited to 16MB; PCI can always
>> dma all of physical memory right now, as we don't support more than 4G
>> physical ram on Intel). In user space, use valloc() to get page-aligned
>> memory.
> That's a really bad assumption. There WILL eventually be a time when
> this is no longer true, so something should be done now to head this
> off.
Don't worry, I'm well aware of this: I've already started specing out
support for 64G physical memory on Intel. _Right now_, the only special
case we support is ISA dma. For the future, we will need something a
lot more powerful, but that is planned anyway.
> Adding a GFP_DMAPCI or whatever, even if it only currently maps to
> another value, will make life a whole lot easier when we start seeing
> systems with >4G memory with 32-bit PCI cards. And what about other
> arches besides Intel?
I haven't seen much interest from people wanting >4G support on other
32-bit architectures, and on 64-bit platforms it's a different game
altogether.
--Stephen
-
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/