On Fri, 27 Jul 2018 10:13:42 +0200
Janosch Frank <frankja@xxxxxxxxxxxxx> wrote:
> On 27.07.2018 09:46, Christian Borntraeger wrote:
> > This is with
> >
> > # CONFIG_TRANSPARENT_HUGEPAGE is not set
> > # CONFIG_HUGETLBFS is not set
> >
> > so we need to fence this.
> > With hugetlbfs enabled this compiles (and also works fine).
>
> We can either ifdef this three times or create a pmd_alloc_map in pgtable.h.
>
> I think that's Martin's call to make.
Hmm, we have a similar sequence in pgtable.c:test_and_clear_guest_dirty
pgd = pgd_offset(mm, addr);
p4d = p4d_alloc(mm, pgd, addr);
if (!p4d)
return false;
pud = pud_alloc(mm, p4d, addr);
if (!pud)
return false;
pmd = pmd_alloc(mm, pud, addr);
if (!pmd)
return false;
Why not refactor this sequence into a separate function local to pgtable.c?
That way we can avoid using huge_pte_alloc in set_guest_storage_key.
--
blue skies,
Martin.
"Reality continues to ruin my life." - Calvin.
--
To unsubscribe from this list: send the line "unsubscribe linux-s390" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
