Received-SPF: pass (google.com: domain of linux-kernel+bounces-141806-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45e3:2400::1 as permitted sender) client-ip=2604:1380:45e3:2400::1;
Precedence: bulk
MIME-Version: 1.0
References: <20240408042437.10951-1-ioworker0@gmail.com> <20240408042437.10951-3-ioworker0@gmail.com>
 <3cd1036d-3814-4a10-b6d2-099937ceabc8@arm.com>
In-Reply-To: <3cd1036d-3814-4a10-b6d2-099937ceabc8@arm.com>
From: Lance Yang <ioworker0@gmail.com>
Date: Fri, 12 Apr 2024 10:09:36 +0800
Message-ID: <CAK1f24m=nC=ecuOZesQvytnxJms2d-wONKT4_7GUfvz0PJz+oQ@mail.gmail.com>
Subject: Re: [PATCH v5 2/2] mm/arm64: override mkold_clean_ptes() batch helper
To: Ryan Roberts <ryan.roberts@arm.com>
Cc: akpm@linux-foundation.org, david@redhat.com, 21cnbao@gmail.com, 
	mhocko@suse.com, fengwei.yin@intel.com, zokeefe@google.com, 
	shy828301@gmail.com, xiehuan09@gmail.com, wangkefeng.wang@huawei.com, 
	songmuchun@bytedance.com, peterx@redhat.com, minchan@kernel.org, 
	linux-mm@kvack.org, linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Thu, Apr 11, 2024 at 9:17=E2=80=AFPM Ryan Roberts <ryan.roberts@arm.com>=
 wrote:
>
> On 08/04/2024 05:24, Lance Yang wrote:
> > The per-pte get_and_clear/modify/set approach would result in
> > unfolding/refolding for contpte mappings on arm64. So we need
> > to override mkold_clean_ptes() for arm64 to avoid it.
>
> IIRC, in the last version, I suggested copying the wrprotect_ptes() patte=
rn to
> correctly iterate over contpte blocks. I meant for you to take it as insp=
iration
> but looks like you have done a carbon copy, including lots of things that=
 are
> unneeded here. That's my fault for not being clear - sorry!

My bad. I must have misunderstood your intention.

>
>
> >
> > Suggested-by: David Hildenbrand <david@redhat.com>
> > Suggested-by: Barry Song <21cnbao@gmail.com>
> > Suggested-by: Ryan Roberts <ryan.roberts@arm.com>
> > Signed-off-by: Lance Yang <ioworker0@gmail.com>
> > ---
> >  arch/arm64/include/asm/pgtable.h | 55 ++++++++++++++++++++++++++++++++
> >  arch/arm64/mm/contpte.c          | 15 +++++++++
> >  2 files changed, 70 insertions(+)
> >
> > diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/=
pgtable.h
> > index 9fd8613b2db2..395754638a9a 100644
> > --- a/arch/arm64/include/asm/pgtable.h
> > +++ b/arch/arm64/include/asm/pgtable.h
> > @@ -1223,6 +1223,34 @@ static inline void __wrprotect_ptes(struct mm_st=
ruct *mm, unsigned long address,
> >               __ptep_set_wrprotect(mm, address, ptep);
> >  }
> >
> > +static inline void ___ptep_mkold_clean(struct mm_struct *mm, unsigned =
long addr,
> > +                                    pte_t *ptep, pte_t pte)
> > +{
> > +     pte_t old_pte;
> > +
> > +     do {
> > +             old_pte =3D pte;
> > +             pte =3D pte_mkclean(pte_mkold(pte));
> > +             pte_val(pte) =3D cmpxchg_relaxed(&pte_val(*ptep),
> > +                                            pte_val(old_pte), pte_val(=
pte));
> > +     } while (pte_val(pte) !=3D pte_val(old_pte));
> > +}
>
> Given you are clearing old and dirty, you have nothing to race against, s=
o you
> shouldn't need the cmpxchg loop here; just a get/modify/set should do? Of=
 course
> if you are setting one or the other, then you need the loop.

Got it.

>
> > +
> > +static inline void __ptep_mkold_clean(struct mm_struct *mm, unsigned l=
ong addr,
> > +                                   pte_t *ptep)
> > +{
> > +     ___ptep_mkold_clean(mm, addr, ptep, __ptep_get(ptep));
> > +}
>
> I don't see a need for this intermediate function.
>
> > +
> > +static inline void __mkold_clean_ptes(struct mm_struct *mm, unsigned l=
ong addr,
> > +                                   pte_t *ptep, unsigned int nr)
> > +{
> > +     unsigned int i;
> > +
> > +     for (i =3D 0; i < nr; i++, addr +=3D PAGE_SIZE, ptep++)
>
> It would probably be good to use the for() loop pattern used by the gener=
ic
> impls here too.

Got it.

>
> > +             __ptep_mkold_clean(mm, addr, ptep);
> > +}
> > +
> >  #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> >  #define __HAVE_ARCH_PMDP_SET_WRPROTECT
> >  static inline void pmdp_set_wrprotect(struct mm_struct *mm,
> > @@ -1379,6 +1407,8 @@ extern void contpte_wrprotect_ptes(struct mm_stru=
ct *mm, unsigned long addr,
> >  extern int contpte_ptep_set_access_flags(struct vm_area_struct *vma,
> >                               unsigned long addr, pte_t *ptep,
> >                               pte_t entry, int dirty);
> > +extern void contpte_mkold_clean_ptes(struct mm_struct *mm, unsigned lo=
ng addr,
> > +                             pte_t *ptep, unsigned int nr);
> >
> >  static __always_inline void contpte_try_fold(struct mm_struct *mm,
> >                               unsigned long addr, pte_t *ptep, pte_t pt=
e)
> > @@ -1603,6 +1633,30 @@ static inline int ptep_set_access_flags(struct v=
m_area_struct *vma,
> >       return contpte_ptep_set_access_flags(vma, addr, ptep, entry, dirt=
y);
> >  }
> >
> > +#define mkold_clean_ptes mkold_clean_ptes
> > +static inline void mkold_clean_ptes(struct mm_struct *mm, unsigned lon=
g addr,
> > +                                 pte_t *ptep, unsigned int nr)
> > +{
> > +     if (likely(nr =3D=3D 1)) {
> > +             /*
> > +              * Optimization: mkold_clean_ptes() can only be called fo=
r present
> > +              * ptes so we only need to check contig bit as condition =
for unfold,
> > +              * and we can remove the contig bit from the pte we read =
to avoid
> > +              * re-reading. This speeds up madvise(MADV_FREE) which is=
 sensitive
> > +              * for order-0 folios. Equivalent to contpte_try_unfold()=
.
> > +              */
>
> Is this true? Do you have data that shows the cost? If not, I'd prefer to=
 avoid
> the optimization and do it the more standard way:
>
> contpte_try_unfold(mm, addr, ptep, __ptep_get(ptep));
>
> > +             pte_t orig_pte =3D __ptep_get(ptep);
> > +
> > +             if (unlikely(pte_cont(orig_pte))) {
> > +                     __contpte_try_unfold(mm, addr, ptep, orig_pte);
> > +                     orig_pte =3D pte_mknoncont(orig_pte);
> > +             }
> > +             ___ptep_mkold_clean(mm, addr, ptep, orig_pte);
> > +     } else {
> > +             contpte_mkold_clean_ptes(mm, addr, ptep, nr);
> > +     }
>
> ...but I don't think you should ever need to unfold in the first place. E=
ven if
> it's folded and you are trying to clear access/dirty for a single pte, yo=
u can
> just clear the whole block. See existing comment in
> contpte_ptep_test_and_clear_young().

Thanks for pointing that out.

>
> So this ends up as something like:
>
> static inline void clear_young_dirty_ptes(struct mm_struct *mm,
>                         unsigned long addr, pte_t *ptep, unsigned int nr,
>                         bool clear_young, bool clear_dirty)
> {
>         if (likely(nr =3D=3D 1 && !pte_cont(__ptep_get(ptep))))
>                 clear_young_dirty_ptes(mm, addr, ptep, nr,
>                                         clear_young, clear_dirty);
>         else
>                 contpte_clear_young_dirty_ptes(mm, addr, ptep, nr,
>                                         clear_young, clear_dirty);
> }

Nice. I'll make sure to follow this approach.

>
>
> > +}
> > +
> >  #else /* CONFIG_ARM64_CONTPTE */
> >
> >  #define ptep_get                             __ptep_get
> > @@ -1622,6 +1676,7 @@ static inline int ptep_set_access_flags(struct vm=
_area_struct *vma,
> >  #define wrprotect_ptes                               __wrprotect_ptes
> >  #define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS
> >  #define ptep_set_access_flags                        __ptep_set_access=
_flags
> > +#define mkold_clean_ptes                     __mkold_clean_ptes
> >
> >  #endif /* CONFIG_ARM64_CONTPTE */
> >
> > diff --git a/arch/arm64/mm/contpte.c b/arch/arm64/mm/contpte.c
> > index 1b64b4c3f8bf..dbff9c5e9eff 100644
> > --- a/arch/arm64/mm/contpte.c
> > +++ b/arch/arm64/mm/contpte.c
> > @@ -361,6 +361,21 @@ void contpte_wrprotect_ptes(struct mm_struct *mm, =
unsigned long addr,
> >  }
> >  EXPORT_SYMBOL_GPL(contpte_wrprotect_ptes);
> >
> > +void contpte_mkold_clean_ptes(struct mm_struct *mm, unsigned long addr=
,
> > +                           pte_t *ptep, unsigned int nr)
> > +{
> > +     /*
> > +      * If clearing the young and dirty bits for an entire contig rang=
e, we can
> > +      * avoid unfolding. Just set old/clean and wait for the later mmu=
_gather
> > +      * flush to invalidate the tlb. If it's a partial range though, w=
e need to
> > +      * unfold.
> > +      */
>
> nit: Please reflow comments like this to 80 cols.
>
> We can avoid unfolding in all cases. See existing comment in
> contpte_ptep_test_and_clear_young(). Suggest something like this (unteste=
d):
>
> void clear_young_dirty_ptes(struct mm_struct *mm, unsigned long addr,
>                             pte_t *ptep, unsigned int nr,
>                             bool clear_young, bool clear_dirty)
> {
>         /*
>          * We can safely clear access/dirty without needing to unfold fro=
m the
>          * architectures perspective, even when contpte is set. If the ra=
nge
>          * starts or ends midway through a contpte block, we can just exp=
and to
>          * include the full contpte block. While this is not exactly what=
 the
>          * core-mm asked for, it tracks access/dirty per folio, not per p=
age.
>          * And since we only create a contpte block when it is covered by=
 a
>          * single folio, we can get away with clearing access/dirty for t=
he
>          * whole block.
>          */
>
>         unsigned int start =3D addr;
>         unsigned int end =3D start + nr;
>
>         if (pte_cont(__ptep_get(ptep + nr - 1)))
>                 end =3D ALIGN(end, CONT_PTE_SIZE);
>
>         if (pte_cont(__ptep_get(ptep))) {
>                 start =3D ALIGN_DOWN(start, CONT_PTE_SIZE);
>                 ptep =3D contpte_align_down(ptep);
>         }
>
>         __clear_young_dirty_ptes(mm, start, ptep, end - start,
>                                  clear_young, clear_dirty);
> }

Nice. Thanks a lot for your help!

Thanks,
Lance

>
> Thanks,
> Ryan
>
> > +
> > +     contpte_try_unfold_partial(mm, addr, ptep, nr);
> > +     __mkold_clean_ptes(mm, addr, ptep, nr);
> > +}
> > +EXPORT_SYMBOL_GPL(contpte_mkold_clean_ptes);
> > +
> >  int contpte_ptep_set_access_flags(struct vm_area_struct *vma,
> >                                       unsigned long addr, pte_t *ptep,
> >                                       pte_t entry, int dirty)
>