Re: PATCH: support for firmwares sizes up to 128000 bytes

Arnaldo Carvalho de Melo (acme@conectiva.com.br)
Tue, 22 Dec 1998 15:17:54 -0200 (EDT)


On Tue, 22 Dec 1998, Alan Cox wrote:

> > > A 128K kmalloc isnt going to work
> >
> > Ok, I know that, that's why I'm using up to 128000 bytes, not more than
> > 131048 bytes :)
>
> The odds of even 128,000 working on a small machine except close to boot
> are low too.

ok

>
> > I also add that I'm beggining kernel programming, and I'm using the books
> > from Rubini and M. Beck et al, and both point outs the limit for kmalloc
> > at slightly less than 32 pages (128 KB on the Intel).
>
> I guessed you may be, thats why I pointed you at vmalloc. vmalloc memory
> isnt physically linear but is logically linear (that is DMA can see the
> different but the kernel cant). If you use vmalloc/vfree instead of
> kmalloc/kfree your 128000 will work fine andn reliably

Thanks for the explanation! I'll try and see if I manage (as I said, I'm a
beginner :) ) to redo the patches to use vmalloc/vfree.

- Arnaldo

"These are questions for action,
not speculation, which is idle." - Noam Chomsky

-
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/