Re: [PATCH 3/4] hugetlb: Change the function prototype to take vma_area_struct as arg

Benjamin Herrenschmidt [off-list ref] writes:

On Thu, 2016-11-10 at 14:59 +0530, Aneesh Kumar K.V wrote:

quoted

This help us to find the hugetlb page size which we need ot use on some
archs like ppc64 for tlbflush. This also make the interface consistent
with other hugetlb functions

What about my requested simpler approach ?

Still working on the changes.

For normal (non-huge) pages, we already know the size.

For huge pages, can't we encode in the top SW bits of the PTE the
page size that we obtain from set_pte_at ?

That would be a lot less churn and avoid touching all these archs...
especially since the current DD1 workaround is horrible and I want
the fix to be backported, so something simpler and contained in
arch/powerpc feels more suitable.

My take as of now is even though the modification lines will be less, it
is going to be much more difficult to follow and backport. I will try to
do a patch to show the complexity and we can decide which approach is
simpler.

-aneesh