Re: [PATCH v16 11/25] mm: pagewalk: Add p4d_entry() and pgd_entry()

[Steven Price]

On 12/12/2019 14:04, Thomas HellstrÃm (VMware) wrote:
> On 12/12/19 2:15 PM, Steven Price wrote:
> > On 12/12/2019 11:33, Thomas HellstrÃm (VMware) wrote:
> > > On 12/12/19 12:23 PM, Thomas HellstrÃm (VMware) wrote:
> > > > On 12/6/19 2:53 PM, Steven Price wrote:
> > > > > pgd_entry() and pud_entry() were removed by commit 0b1fbfe50006c410
> > > > > ("mm/pagewalk: remove pgd_entry() and pud_entry()") because there were
> > > > > no users. We're about to add users so reintroduce them, along with
> > > > > p4d_entry() as we now have 5 levels of tables.
> > > > >
> > > > > Note that commit a00cc7d9dd93d66a ("mm, x86: add support for
> > > > > PUD-sized transparent hugepages") already re-added pud_entry() but with
> > > > > different semantics to the other callbacks. Since there have never
> > > > > been upstream users of this, revert the semantics back to match the
> > > > > other callbacks. This means pud_entry() is called for all entries, not
> > > > > just transparent huge pages.
> >
> > When I wrote that there were no upstream users, which sadly shows how
> > long ago that was :(
> >
> > > > Actually, there are two users of pud_entry(), in hmm.c and since 5.5rc1 also mapping_dirty_helpers.c. The latter one is unproblematic and requires no attention but the one in hmm.c is probably largely untested, and seems to assume it was called outside of the spinlock.
> > > >
> > > > The problem with the current patch is that the hmm pud_entry will traverse also pmds, so that will be done twice now.
> > > >
> > > > In another thread we were discussing a means of rerunning the level (in case of a race), or continuing after a level, based on the return value after the callback. The change was fairly invasive,
> > > Hmm. Forgot to remove the above text that appears twice. :(. The correct one is inline below.
> > >
> > > > > Tested-by: Zong Li <zong.li@xxxxxxxxxx>
> > > > > Signed-off-by: Steven Price <steven.price@xxxxxxx>
> > > > > ---
> > > > > Â include/linux/pagewalk.h | 19 +++++++++++++------
> > > > > Â mm/pagewalk.cÂÂÂÂÂÂÂÂÂÂÂ | 27 ++++++++++++++++-----------
> > > > > Â 2 files changed, 29 insertions(+), 17 deletions(-)
> > > > >
> > > > > diff --git a/include/linux/pagewalk.h b/include/linux/pagewalk.h
> > > > > index 6ec82e92c87f..06790f23957f 100644
> > > > > --- a/include/linux/pagewalk.h
> > > > > +++ b/include/linux/pagewalk.h
> > > > > @@ -8,15 +8,15 @@ struct mm_walk;
> > > > > Â Â /**
> > > > > ÂÂ * mm_walk_ops - callbacks for walk_page_range
> > > > > - * @pud_entry:ÂÂÂÂÂÂÂ if set, called for each non-empty PUD (2nd-level) entry
> > > > > - *ÂÂÂÂÂÂÂÂÂÂÂ this handler should only handle pud_trans_huge() puds.
> > > > > - *ÂÂÂÂÂÂÂÂÂÂÂ the pmd_entry or pte_entry callbacks will be used for
> > > > > - *ÂÂÂÂÂÂÂÂÂÂÂ regular PUDs.
> > > > > - * @pmd_entry:ÂÂÂÂÂÂÂ if set, called for each non-empty PMD (3rd-level) entry
> > > > > + * @pgd_entry:ÂÂÂÂÂÂÂ if set, called for each non-empty PGD (top-level) entry
> > > > > + * @p4d_entry:ÂÂÂÂÂÂÂ if set, called for each non-empty P4D entry
> > > > > + * @pud_entry:ÂÂÂÂÂÂÂ if set, called for each non-empty PUD entry
> > > > > + * @pmd_entry:ÂÂÂÂÂÂÂ if set, called for each non-empty PMD entry
> > > > > ÂÂ *ÂÂÂÂÂÂÂÂÂÂÂ this handler is required to be able to handle
> > > > >  * pmd_trans_huge() pmds. They may simply choose to
> > > > > ÂÂ *ÂÂÂÂÂÂÂÂÂÂÂ split_huge_page() instead of handling it explicitly.
> > > > > - * @pte_entry:ÂÂÂÂÂÂÂ if set, called for each non-empty PTE (4th-level) entry
> > > > > + * @pte_entry:ÂÂÂÂÂÂÂ if set, called for each non-empty PTE (lowest-level)
> > > > > + *ÂÂÂÂÂÂÂÂÂÂÂ entry
> > > > > ÂÂ * @pte_hole:ÂÂÂÂÂÂÂ if set, called for each hole at all levels
> > > > > ÂÂ * @hugetlb_entry:ÂÂÂ if set, called for each hugetlb entry
> > > > > ÂÂ * @test_walk:ÂÂÂÂÂÂÂ caller specific callback function to determine whether
> > > > > @@ -27,8 +27,15 @@ struct mm_walk;
> > > > > ÂÂ * @pre_vma:ÂÂÂÂÂÂÂÂÂÂÂ if set, called before starting walk on a non-null vma.
> > > > > ÂÂ * @post_vma:ÂÂÂÂÂÂÂÂÂÂ if set, called after a walk on a non-null vma, provided
> > > > > ÂÂ *ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ that @pre_vma and the vma walk succeeded.
> > > > > + *
> > > > > + * p?d_entry callbacks are called even if those levels are folded on a
> > > > > + * particular architecture/configuration.
> > > > > ÂÂ */
> > > > > Â struct mm_walk_ops {
> > > > > +ÂÂÂ int (*pgd_entry)(pgd_t *pgd, unsigned long addr,
> > > > > +ÂÂÂÂÂÂÂÂÂÂÂÂ unsigned long next, struct mm_walk *walk);
> > > > > +ÂÂÂ int (*p4d_entry)(p4d_t *p4d, unsigned long addr,
> > > > > +ÂÂÂÂÂÂÂÂÂÂÂÂ unsigned long next, struct mm_walk *walk);
> > > > > ÂÂÂÂÂ int (*pud_entry)(pud_t *pud, unsigned long addr,
> > > > > ÂÂÂÂÂÂÂÂÂÂÂÂÂÂ unsigned long next, struct mm_walk *walk);
> > > > > ÂÂÂÂÂ int (*pmd_entry)(pmd_t *pmd, unsigned long addr,
> > > > > diff --git a/mm/pagewalk.c b/mm/pagewalk.c
> > > > > index ea0b9e606ad1..c089786e7a7f 100644
> > > > > --- a/mm/pagewalk.c
> > > > > +++ b/mm/pagewalk.c
> > > > > @@ -94,15 +94,9 @@ static int walk_pud_range(p4d_t *p4d, unsigned long addr, unsigned long end,
> > > > > ÂÂÂÂÂÂÂÂÂ }
> > > > > Â ÂÂÂÂÂÂÂÂÂ if (ops->pud_entry) {
> > > > > -ÂÂÂÂÂÂÂÂÂÂÂ spinlock_t *ptl = pud_trans_huge_lock(pud, walk->vma);
> > > > > -
> > > > > -ÂÂÂÂÂÂÂÂÂÂÂ if (ptl) {
> > > > > -ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ err = ops->pud_entry(pud, addr, next, walk);
> > > > > -ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ spin_unlock(ptl);
> > > > > -ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ if (err)
> > > > > -ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ break;
> > > > > -ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ continue;
> > > > > -ÂÂÂÂÂÂÂÂÂÂÂ }
> > > > > +ÂÂÂÂÂÂÂÂÂÂÂ err = ops->pud_entry(pud, addr, next, walk);
> > > > > +ÂÂÂÂÂÂÂÂÂÂÂ if (err)
> > > > > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ break;
> > > >
> > > > Actually, there are two current users of pud_entry(), in hmm.c and since 5.5rc1 also mapping_dirty_helpers.c. The latter one is unproblematic and requires no attention but the one in hmm.c is probably largely untested, and seems to assume it was called outside of the spinlock.
> >
> > Thanks for pointing that out, I guess the simplest fix would be to
> > squash in something like the below which should restore the old
> > behaviour for hmm.c without affecting others.
> >
> > Steve 
>
> I'm not fully sure that the old behaviour is the correct one, but definitely hmm's pud_entry needs some fixing.
> I'm more concerned with the pagewalk code. With your patch it actually splits all huge puds present in the page-table
> on each page walk which is not what we want.

Good catch - yes that's certainly not ideal.

> One idea would be to add a new member to struct_mm_walk:
>
> enum page_walk_ret_action {
>  ÂÂÂÂACTION_SUBTREE = 0,
>  ÂÂÂÂACTION_CONTINUE = 1,
>  ÂÂÂÂACTION_AGAIN = 2 /* Only for levels that thave p?d_unstable */
> };
>
> struct mm_walk {
>  ÂÂÂÂ...
>  ÂÂÂÂenum page_walk_ret_action action; /* or perhaps as an enum */
> };
>
>
> if (ops->pud_entry) {
>  ÂÂÂÂwalk->action = ACTION_SUBTREE;
>  ÂÂÂÂ...
>  ÂÂÂÂ...
>  ÂÂÂÂ...
>  ÂÂÂÂif (walk->action == ACTION_AGAIN)Â /* Callback tried to split huge entry, but failed */
>  ÂÂÂÂÂÂÂ goto again;
>  ÂÂÂÂelse if (walk->action == ACTION_CONTINUE) /* Done with this subtree. Probably huge entry handled. */
>  ÂÂÂÂÂÂÂ continue;
>  ÂÂÂÂ/* ACTION_SUBTREE falls through */
> }

I'll have a go at implementing the above - this might also allow removing the test_p?d() callbacks as they can simply return ACTION_CONTINUE.

Steve

> we discussed something similar before on linux-mm, but the idea then was to redefine
> the positive return value of the callback to the action, but that meant changing those existing callbacks that relied on
> a positive return value. The above would be helpful also for pmd_entry.
>
> /Thomas
>
>
>
>
>
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel@xxxxxxxxxxxxxxxxxxx
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel