All of lore.kernel.org
 help / color / mirror / Atom feed
From: Baolin Wang <baolin.wang@linux.alibaba.com>
To: Pedro Demarchi Gomes <pedrodemargomes@gmail.com>,
	Andrew Morton <akpm@linux-foundation.org>,
	David Hildenbrand <david@kernel.org>,
	Lorenzo Stoakes <ljs@kernel.org>, Zi Yan <ziy@nvidia.com>,
	"Liam R . Howlett" <liam@infradead.org>,
	Nico Pache <nico.pache@linux.dev>,
	Ryan Roberts <ryan.roberts@arm.com>, Dev Jain <dev.jain@arm.com>,
	Barry Song <baohua@kernel.org>, Lance Yang <lance.yang@linux.dev>,
	Usama Arif <usama.arif@linux.dev>
Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH mm-stable] mm/khugepaged: avoid unnecessary checking for swap entries when collapsing a mTHP
Date: Thu, 27 Aug 2026 16:06:12 +0800	[thread overview]
Message-ID: <1ec4ef97-47e4-48af-a8c8-4e50faf781f0@linux.alibaba.com> (raw)
In-Reply-To: <20260825192433.3185880-1-pedrodemargomes@gmail.com>



On 8/26/26 3:24 AM, Pedro Demarchi Gomes wrote:
> mthp_collapse() tries to swap in PTEs when collapsing a mTHP if there are any
> swap PTEs in the PMD range, even if none of those swap PTEs are
> actually part of the mTHP's range.

Are you sure? I wonder how you tested your patch? Because we never 
swapin PTEs for mTHP collapse, see the code in 
__collapse_huge_page_swapin():

/*
  * TODO: Support swapin without leading to further mTHP
  * collapses. Currently bringing in new pages via swapin may
  * cause a future higher order collapse on a rescan of the same
  * range.
  */
if (!is_pmd_order(order)) {
	count_mthp_stat(order, MTHP_STAT_COLLAPSE_EXCEED_SWAP);
	pte_unmap(pte);
	mmap_read_unlock(mm);
	result = SCAN_EXCEED_SWAP_PTE;
	goto out;
}

> 
> Track swap PTEs in cc->mthp_unmapped_ptes so that mthp_collapse() can
> tell whether the mTHP's own range contains a swap PTE, and skip the
> swapin check when it doesn't.
> 
> Signed-off-by: Pedro Demarchi Gomes <pedrodemargomes@gmail.com>
> ---
>   mm/khugepaged.c | 29 +++++++++++++++++++----------
>   1 file changed, 19 insertions(+), 10 deletions(-)
> 
> diff --git a/mm/khugepaged.c b/mm/khugepaged.c
> index 11ff98d55c76..38596172f7dc 100644
> --- a/mm/khugepaged.c
> +++ b/mm/khugepaged.c
> @@ -117,6 +117,9 @@ struct collapse_control {
>   
>   	/* Each bit represents a single occupied (!none/zero) page. */
>   	DECLARE_BITMAP(mthp_present_ptes, MAX_PTRS_PER_PTE);
> +
> +	/* Each bit represents a single not present and not none/zero pte. */
> +	DECLARE_BITMAP(mthp_unmapped_ptes, MAX_PTRS_PER_PTE);
>   };
>   
>   /**
> @@ -634,6 +637,7 @@ static void collapse_control_init_scan(struct collapse_control *cc)
>   	memset(cc->node_load, 0, sizeof(cc->node_load));
>   	nodes_clear(cc->alloc_nmask);
>   	bitmap_zero(cc->mthp_present_ptes, MAX_PTRS_PER_PTE);
> +	bitmap_zero(cc->mthp_unmapped_ptes, MAX_PTRS_PER_PTE);
>   }
>   
>   static void release_pte_folio(struct folio *folio)
> @@ -1291,7 +1295,7 @@ static enum scan_result alloc_charge_folio(struct folio **foliop, struct mm_stru
>    * Note that the VMA must be rechecked after grabbing the mmap_lock again.
>    */
>   static enum scan_result collapse_huge_page(struct mm_struct *mm, unsigned long start_addr,
> -		int referenced, int unmapped, struct collapse_control *cc,
> +		int referenced, bool swapin, struct collapse_control *cc,
>   		unsigned int order)
>   {
>   	const unsigned long pmd_addr = start_addr & HPAGE_PMD_MASK;
> @@ -1330,7 +1334,7 @@ static enum scan_result collapse_huge_page(struct mm_struct *mm, unsigned long s
>   		goto out_nolock;
>   	}
>   
> -	if (unmapped) {
> +	if (swapin) {
>   		/*
>   		 * __collapse_huge_page_swapin() will return with mmap_lock
>   		 * released when it fails. So we jump out_nolock directly in
> @@ -1502,10 +1506,10 @@ static unsigned int max_order_from_offset(unsigned int offset)
>    * mTHP.
>    */
>   static enum scan_result mthp_collapse(struct mm_struct *mm,
> -		unsigned long address, int referenced, int unmapped,
> +		unsigned long address, int referenced,
>   		struct collapse_control *cc, unsigned long enabled_orders)
>   {
> -	unsigned int nr_occupied_ptes, nr_ptes, max_ptes_none;
> +	unsigned int nr_occupied_ptes, nr_unmapped_ptes, nr_ptes, max_ptes_none;
>   	enum scan_result last_result = SCAN_FAIL;
>   	int collapsed = 0;
>   	bool alloc_failed = false;
> @@ -1522,21 +1526,25 @@ static enum scan_result mthp_collapse(struct mm_struct *mm,
>   		max_ptes_none = collapse_max_ptes_none(cc, NULL, order);
>   		nr_occupied_ptes = bitmap_weight_from(cc->mthp_present_ptes, offset,
>   						      offset + nr_ptes);
> +		nr_unmapped_ptes = bitmap_weight_from(cc->mthp_unmapped_ptes, offset,
> +						      offset + nr_ptes);
> +
>   
>   		/*
> -		 * Swap PTEs accepted during the scan are counted in @unmapped,
> -		 * not in the present-PTE bitmap. Account them for the PMD-order
> -		 * candidate.
> +		 * Swap PTEs accepted during the scan are counted in
> +		 * nr_unmapped_ptes, not in the present-PTE bitmap. Account
> +		 * them for the PMD-order candidate.
>   		 */
>   		if (is_pmd_order(order))
> -			nr_occupied_ptes += unmapped;
> +			nr_occupied_ptes += nr_unmapped_ptes;
>   
>   		if (nr_occupied_ptes >= nr_ptes - max_ptes_none) {
>   			enum scan_result ret;
> +			bool swapin = nr_unmapped_ptes > 0;
>   
>   			collapse_address = address + offset * PAGE_SIZE;
>   			ret = collapse_huge_page(mm, collapse_address, referenced,
> -						 unmapped, cc, order);
> +						 swapin, cc, order);
>   
>   			switch (ret) {
>   			/* Cases where we continue to next collapse candidate */
> @@ -1667,6 +1675,7 @@ static enum scan_result collapse_scan_pmd(struct mm_struct *mm,
>   			continue;
>   		}
>   		if (!pte_present(pteval)) {
> +			__set_bit(i, cc->mthp_unmapped_ptes);
>   			if (++unmapped > max_ptes_swap) {
>   				result = SCAN_EXCEED_SWAP_PTE;
>   				count_collapse_event(HPAGE_PMD_ORDER, THP_SCAN_EXCEED_SWAP_PTE,
> @@ -1785,7 +1794,7 @@ static enum scan_result collapse_scan_pmd(struct mm_struct *mm,
>   		/* collapse_huge_page() expects the lock to be dropped before calling */
>   		mmap_read_unlock(mm);
>   		result = mthp_collapse(mm, start_addr, referenced,
> -				       unmapped, cc, enabled_orders);
> +				       cc, enabled_orders);
>   		/* mmap_lock was released above, set lock_dropped */
>   		*lock_dropped = true;
>   	}



  parent reply	other threads:[~2026-08-27  8:06 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-08-25 19:24 [PATCH mm-stable] mm/khugepaged: avoid unnecessary checking for swap entries when collapsing a mTHP Pedro Demarchi Gomes
2026-08-26  4:43 ` Lance Yang
2026-08-27  8:06 ` Baolin Wang [this message]
2026-08-27 11:02   ` David Hildenbrand (Arm)
2026-08-27 12:11     ` Lance Yang
2026-08-27 14:07       ` Kiryl Shutsemau
2026-08-31 18:53         ` Pedro Demarchi Gomes
2026-09-02 10:08           ` Kiryl Shutsemau
2026-08-28  3:24       ` Pedro Demarchi Gomes

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=1ec4ef97-47e4-48af-a8c8-4e50faf781f0@linux.alibaba.com \
    --to=baolin.wang@linux.alibaba.com \
    --cc=akpm@linux-foundation.org \
    --cc=baohua@kernel.org \
    --cc=david@kernel.org \
    --cc=dev.jain@arm.com \
    --cc=lance.yang@linux.dev \
    --cc=liam@infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=ljs@kernel.org \
    --cc=nico.pache@linux.dev \
    --cc=pedrodemargomes@gmail.com \
    --cc=ryan.roberts@arm.com \
    --cc=usama.arif@linux.dev \
    --cc=ziy@nvidia.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.