linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: David Hildenbrand <david@redhat.com>
To: Usama Arif <usamaarif642@gmail.com>,
	sj@kernel.org, akpm@linux-foundation.org
Cc: damon@lists.linux.dev, linux-mm@kvack.org
Subject: Re: [PATCH v3 2/2] mm/damon: introduce DAMOS filter type hugepage
Date: Mon, 20 Jan 2025 20:25:54 +0100	[thread overview]
Message-ID: <68262ad8-a417-4183-af56-a3691002c3d1@redhat.com> (raw)
In-Reply-To: <f8e95ec8-f800-495d-a723-53edd0d487f5@gmail.com>

On 20.01.25 20:18, Usama Arif wrote:
> 
> 
> On 20/01/2025 18:57, David Hildenbrand wrote:
>> On 20.01.25 19:19, Usama Arif wrote:
>>> This is to gather statistics to check if memory regions of specific
>>> access tempratures are backed by hugepages. This includes both THPs
>>> and hugetlbfs.
>>> This filter can help to observe and prove the effectivenes of
>>> different schemes for shrinking/collapsing hugepages.
>>>
>>> Signed-off-by: Usama Arif <usamaarif642@gmail.com>
>>> ---
>>> v2 -> v3:
>>> - expose hugepage via sysfs even if the kernel is
>>>     built without hugepage support. DAMON will just
>>>     just return 0. (SJ Park)
>>>
>>> v1 -> v2:
>>> - Wrap DAMOS_FILTER_TYPE_HUGEPAGE case with
>>>     CONFIG_PGTABLE_HAS_HUGE_LEAVES (SJ Park)
>>> ---
>>>    include/linux/damon.h    | 2 ++
>>>    mm/damon/paddr.c         | 5 +++++
>>>    mm/damon/sysfs-schemes.c | 1 +
>>>    3 files changed, 8 insertions(+)
>>>
>>> diff --git a/include/linux/damon.h b/include/linux/damon.h
>>> index af525252b853..1d94d7d88b36 100644
>>> --- a/include/linux/damon.h
>>> +++ b/include/linux/damon.h
>>> @@ -326,6 +326,7 @@ struct damos_stat {
>>>     * @DAMOS_FILTER_TYPE_ANON:    Anonymous pages.
>>>     * @DAMOS_FILTER_TYPE_MEMCG:    Specific memcg's pages.
>>>     * @DAMOS_FILTER_TYPE_YOUNG:    Recently accessed pages.
>>> + * @DAMOS_FILTER_TYPE_HUGEPAGE:    Page is part of a hugepage.
>>>     * @DAMOS_FILTER_TYPE_ADDR:    Address range.
>>>     * @DAMOS_FILTER_TYPE_TARGET:    Data Access Monitoring target.
>>>     * @NR_DAMOS_FILTER_TYPES:    Number of filter types.
>>> @@ -345,6 +346,7 @@ enum damos_filter_type {
>>>        DAMOS_FILTER_TYPE_ANON,
>>>        DAMOS_FILTER_TYPE_MEMCG,
>>>        DAMOS_FILTER_TYPE_YOUNG,
>>> +    DAMOS_FILTER_TYPE_HUGEPAGE,
>>>        DAMOS_FILTER_TYPE_ADDR,
>>>        DAMOS_FILTER_TYPE_TARGET,
>>>        NR_DAMOS_FILTER_TYPES,
>>> diff --git a/mm/damon/paddr.c b/mm/damon/paddr.c
>>> index c0ccf4fade24..224308140441 100644
>>> --- a/mm/damon/paddr.c
>>> +++ b/mm/damon/paddr.c
>>> @@ -222,6 +222,11 @@ static bool damos_pa_filter_match(struct damos_filter *filter,
>>>            if (matched)
>>>                damon_folio_mkold(folio);
>>>            break;
>>> +#if defined(CONFIG_PGTABLE_HAS_HUGE_LEAVES)
>>> +    case DAMOS_FILTER_TYPE_HUGEPAGE:
>>> +        matched = folio_size(folio) == HPAGE_PMD_SIZE;
>>
>>
>> Can we directly embed in the name and the comments/docs that we are only talking about PMD size (both, THP and hugetlb)?
>>
>> DAMOS_FILTER_TYPE_PMD_HUGEPAGE or sth. like that.
>>
>>
> 
> I always think of gigantic page as PUD size, hugepage as PMD size (and not anything smaller), mTHP as smaller than PMD :)

And cont-PMD ... hugetlb, cont-PTE hugetlb ... are ? :)

> But I don't think thats the official term or standardized across the kernel.
> 

We didn't know better when we designed some of the old interfaces, 
that's why I am hoping that we can do better with new interfaces. Either 
spell out the PMD size, or have filters for size lists / size ranges.

> I can send a v4, or maybe Andrew could apply the diff below?
> 
> diff --git a/include/linux/damon.h b/include/linux/damon.h
> index 1d94d7d88b36..261c0741dd0f 100644
> --- a/include/linux/damon.h
> +++ b/include/linux/damon.h
> @@ -326,7 +326,7 @@ struct damos_stat {
>    * @DAMOS_FILTER_TYPE_ANON:    Anonymous pages.
>    * @DAMOS_FILTER_TYPE_MEMCG:   Specific memcg's pages.
>    * @DAMOS_FILTER_TYPE_YOUNG:   Recently accessed pages.
> - * @DAMOS_FILTER_TYPE_HUGEPAGE:        Page is part of a hugepage.
> + * @DAMOS_FILTER_TYPE_PMD_HUGEPAGE:    Page is part of a hugepage (THP/hugetlb).
>    * @DAMOS_FILTER_TYPE_ADDR:    Address range.
>    * @DAMOS_FILTER_TYPE_TARGET:  Data Access Monitoring target.
>    * @NR_DAMOS_FILTER_TYPES:     Number of filter types.
> @@ -346,7 +346,7 @@ enum damos_filter_type {
>          DAMOS_FILTER_TYPE_ANON,
>          DAMOS_FILTER_TYPE_MEMCG,
>          DAMOS_FILTER_TYPE_YOUNG,
> -       DAMOS_FILTER_TYPE_HUGEPAGE,
> +       DAMOS_FILTER_TYPE_PMD_HUGEPAGE,
>          DAMOS_FILTER_TYPE_ADDR,
>          DAMOS_FILTER_TYPE_TARGET,
>          NR_DAMOS_FILTER_TYPES,
> diff --git a/mm/damon/paddr.c b/mm/damon/paddr.c
> index 224308140441..e374ea952308 100644
> --- a/mm/damon/paddr.c
> +++ b/mm/damon/paddr.c
> @@ -223,7 +223,7 @@ static bool damos_pa_filter_match(struct damos_filter *filter,
>                          damon_folio_mkold(folio);
>                  break;
>   #if defined(CONFIG_PGTABLE_HAS_HUGE_LEAVES)
> -       case DAMOS_FILTER_TYPE_HUGEPAGE:
> +       case DAMOS_FILTER_TYPE_PMD_HUGEPAGE:
>                  matched = folio_size(folio) == HPAGE_PMD_SIZE;
>                  break;
>   #endif
> 


-- 
Cheers,

David / dhildenb



      reply	other threads:[~2025-01-20 19:26 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-01-20 18:19 [PATCH v3 1/2] mm/damon: have damon_get_folio return folio even for tail pages Usama Arif
2025-01-20 18:19 ` [PATCH v3 2/2] mm/damon: introduce DAMOS filter type hugepage Usama Arif
2025-01-20 18:33   ` SeongJae Park
2025-01-20 18:57   ` David Hildenbrand
2025-01-20 19:16     ` SeongJae Park
2025-01-20 19:23       ` David Hildenbrand
2025-01-20 19:30         ` SeongJae Park
2025-01-20 19:58           ` Usama Arif
2025-01-20 20:03             ` David Hildenbrand
2025-01-21 17:52               ` SeongJae Park
2025-01-21 18:39                 ` David Hildenbrand
2025-01-21 20:05                   ` SeongJae Park
2025-01-20 20:12             ` SeongJae Park
2025-01-20 20:26               ` Usama Arif
2025-01-20 20:48                 ` SeongJae Park
2025-01-20 19:18     ` Usama Arif
2025-01-20 19:25       ` David Hildenbrand [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=68262ad8-a417-4183-af56-a3691002c3d1@redhat.com \
    --to=david@redhat.com \
    --cc=akpm@linux-foundation.org \
    --cc=damon@lists.linux.dev \
    --cc=linux-mm@kvack.org \
    --cc=sj@kernel.org \
    --cc=usamaarif642@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox