Re: 4MB pages and framebuffer access, x11perf results, 2.1.125

Linus Torvalds (torvalds@transmeta.com)
Thu, 29 Oct 1998 09:15:51 -0800 (PST)

On Thu, 29 Oct 1998, MOLNAR Ingo wrote:
>
> theres one more reason why it's a granularity type of thing: the vma's
> virtual address has to be 4M aligned too. Basically mmap() picks a generic
> 4M-size vma, and only the driver-specific function knows that this is
> 'incorrect'. We could hack this around by dropping the vma and
> reallocating it, but this is not quite robust.

This is partly why I'd like a separate interface, but on the other hand
it's also something that we can really easily do just with a normal mmap:
the "get_unmapped_area()" thing is very much designed to allow us to have
a policy on how to align things (right now it doesn't).

We can't drop the vma and re-allocate it, that would just break all
interfaces too much.

Linus

-
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/