[PATCH v4 5/7] dma-buf: system_heap: Allocate higher order pages if available

John Stultz john.stultz at linaro.org
Fri Oct 30 03:30:04 UTC 2020


On Thu, Oct 29, 2020 at 7:34 PM Hillf Danton <hdanton at sina.com> wrote:
> On Thu, 29 Oct 2020 12:34:51 -0700 John Stultz wrote:
> > As for your comment on HPAGE_PMD_ORDER (9 on arm64/arm) and
> > PAGE_ALLOC_COSTLY_ORDER(3), I'm not totally sure I understand your
> > question? Are you suggesting those values would be more natural orders
> > to choose from?
>
> The numbers, 9 and 3, are not magic themselves but under the mm diretory
> they draw more attentions than others do. Sometimes it would take two
> minutes for me to work out that HPAGE_PMD_ORDER does not mean 1MiB, on
> platforms like arm64 or not.

Yes, I can say it took me longer than two minutes to dig around and
work out HPAGE_PMD_ORDER for my last reply.  :)

Though I'm still a bit unsure if you are proposing something more than
just a comment to explain why order 8 and order 4 allocations are used
in my patch? Please let me know if so.

thanks
-john


More information about the dri-devel mailing list