Received-SPF: pass (google.com: domain of linux-kernel+bounces-132415-linux.lists.archive=gmail.com@vger.kernel.org designates 139.178.88.99 as permitted sender) client-ip=139.178.88.99;
Precedence: bulk
MIME-Version: 1.0
References: <a25c9e14cd03907d5978b60546a69e6aa3fc2a7d.1712151833.git.baolin.wang@linux.alibaba.com>
In-Reply-To: <a25c9e14cd03907d5978b60546a69e6aa3fc2a7d.1712151833.git.baolin.wang@linux.alibaba.com>
From: Barry Song <21cnbao@gmail.com>
Date: Fri, 5 Apr 2024 17:53:24 +1300
Message-ID: <CAGsJ_4xvXT6Qy1JgP2khH0DKvTGhjsrexsSj8ddieM_XXn8bng@mail.gmail.com>
Subject: Re: [PATCH] mm: page_alloc: use the correct THP order for THP PCP
To: Baolin Wang <baolin.wang@linux.alibaba.com>
Cc: akpm@linux-foundation.org, mgorman@techsingularity.net, vbabka@suse.cz, 
	linux-mm@kvack.org, linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Thu, Apr 4, 2024 at 2:47=E2=80=AFAM Baolin Wang
<baolin.wang@linux.alibaba.com> wrote:
>
> Commit 44042b449872 ("mm/page_alloc: allow high-order pages to be stored
> on the per-cpu lists") extends the PCP allocator to store THP pages, and
> it determines whether to cache THP pags in PCP by comparing with pagebloc=
k_order.
> But the pageblock_order is not always equal to THP order, it might also
> be MAX_PAGE_ORDER, which could prevent PCP from caching THP pages.
>
> Therefore, using HPAGE_PMD_ORDER instead to determine the need for cachin=
g
> THP for PCP can fix this issue
>
> Fixes: 44042b449872 ("mm/page_alloc: allow high-order pages to be stored =
on the per-cpu lists")
> Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>

LGTM,
Reviewed-by: Barry Song <baohua@kernel.org>

In the context of using mTHP, perhaps there arises a need for PCP
allocation for frequently
requested mTHP orders. These orders typically exceed PAGE_ALLOC_COSTLY_ORDE=
R
but are smaller than HPAGE_PMD_ORDER.

> ---
>  mm/page_alloc.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 1beb56f75319..915f4ef070da 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -506,7 +506,7 @@ static inline unsigned int order_to_pindex(int migrat=
etype, int order)
>  {
>  #ifdef CONFIG_TRANSPARENT_HUGEPAGE
>         if (order > PAGE_ALLOC_COSTLY_ORDER) {
> -               VM_BUG_ON(order !=3D pageblock_order);
> +               VM_BUG_ON(order !=3D HPAGE_PMD_ORDER);
>                 return NR_LOWORDER_PCP_LISTS;
>         }
>  #else
> @@ -522,7 +522,7 @@ static inline int pindex_to_order(unsigned int pindex=
)
>
>  #ifdef CONFIG_TRANSPARENT_HUGEPAGE
>         if (pindex =3D=3D NR_LOWORDER_PCP_LISTS)
> -               order =3D pageblock_order;
> +               order =3D HPAGE_PMD_ORDER;
>  #else
>         VM_BUG_ON(order > PAGE_ALLOC_COSTLY_ORDER);
>  #endif
> @@ -535,7 +535,7 @@ static inline bool pcp_allowed_order(unsigned int ord=
er)
>         if (order <=3D PAGE_ALLOC_COSTLY_ORDER)
>                 return true;
>  #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> -       if (order =3D=3D pageblock_order)
> +       if (order =3D=3D HPAGE_PMD_ORDER)
>                 return true;
>  #endif
>         return false;
> --
> 2.39.3
>

Thanks
Barry