mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Usama Arif <usama.arif@linux.dev>
To: "David Hildenbrand (Arm)" <david@kernel.org>,
	Andrew Morton <akpm@linux-foundation.org>,
	chrisl@kernel.org, kasong@tencent.com, ljs@kernel.org,
	ziy@nvidia.com, linux-mm@kvack.org
Cc: ying.huang@linux.alibaba.com, Baoquan He <baoquan.he@linux.dev>,
	willy@infradead.org, youngjun.park@lge.com, hannes@cmpxchg.org,
	riel@surriel.com, shakeel.butt@linux.dev, alex@ghiti.fr,
	kas@kernel.org, baohua@kernel.org, dev.jain@arm.com,
	baolin.wang@linux.alibaba.com, Nico Pache <nico.pache@linux.dev>,
	"Liam R. Howlett" <liam@infradead.org>,
	ryan.roberts@arm.com, Vlastimil Babka <vbabka@kernel.org>,
	lance.yang@linux.dev, linux-kernel@vger.kernel.org,
	nphamcs@gmail.com, shikemeng@huaweicloud.com, yosry@kernel.org,
	qi.zheng@linux.dev, luizcap@redhat.com, kernel-team@meta.com
Subject: Re: [RESEND v7 10/29] mm: make PMD migration-entry splitting explicit
Date: Thu, 24 Sep 2026 17:12:46 +0100	[thread overview]
Message-ID: <0f5d22fd-24f2-4235-92f5-3dc065e0f153@linux.dev> (raw)
In-Reply-To: <a8ff997a-2d09-44f4-85c3-322927f75d1d@kernel.org>



On 23/09/2026 12:04, David Hildenbrand (Arm) wrote:
> Having another look ...
> 
> 
>> +
>> +/*
>> + * Split a present PMD into PTE migration entries, for the rmap migration
>> + * walker.  
> 
> I assume pmd_is_device_private_entry() is also allowed?
> 
> "present or device-private PMDs"
> 

Yes, ack.

> 
>> Like split_huge_pmd_locked(), the caller must hold the PMD lock and
>> + * must already be inside an mmu_notifier invalidate range.
>> + */
>> +void split_pmd_to_migration_entries(struct vm_area_struct *vma,
>> +				    unsigned long address, pmd_t *pmd)
>> +{
>> +	VM_WARN_ON_ONCE(!IS_ALIGNED(address, HPAGE_PMD_SIZE));
>> +	if (pmd_trans_huge(*pmd) || pmd_is_valid_softleaf(*pmd))
> 
> pmd_is_valid_softleaf() would include migration entries and swap entries?
> 
> I'd assume that try_to_migrate_one() will never call us with something that is
> not a present or device-private PMD?
> 
> Shouldn't we instead just
> 
> 
> VM_WARN_ON_ONCE(!pmd_present(*pmd) && !pmd_is_device_private_entry(*pmd));
> __split_huge_pmd_locked(vma, pmd, address, true);
> 
> And not offer some weird fallback where we silently don't split?


Yes, that sounds much better. The function right now is:


/**
 * split_pmd_to_migration_entries() - Split a present or device private PMD into
 * PTE migration entries.
 * @vma: The VMA containing the PMD.
 * @address: The PMD-aligned address the PMD maps.
 * @pmd: A pointer to the leaf PMD entry.
 *
 * For the rmap migration walker, which only ever hands back those two entry
 * types.  Like split_huge_pmd_locked(), the caller must hold the PMD lock and
 * must already be inside an mmu_notifier invalidate range.
 */
void split_pmd_to_migration_entries(struct vm_area_struct *vma,
		unsigned long address, pmd_t *pmd)
{
	VM_WARN_ON_ONCE(!IS_ALIGNED(address, HPAGE_PMD_SIZE));
	VM_WARN_ON_ONCE(!pmd_present(*pmd) &&
			!pmd_is_device_private_entry(*pmd));
	__split_huge_pmd_locked(vma, pmd, address, true);
}

> 
>> +		__split_huge_pmd_locked(vma, pmd, address, true);
>>  }
> 


  reply	other threads:[~2026-09-24 16:12 UTC|newest]

Thread overview: 73+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-14 12:27 [RESEND v7 00/29] mm: PMD-level swap entries for anonymous THPs Usama Arif
2026-09-14 12:27 ` [RESEND v7 01/29] mm: rename pmd_to_softleaf_folio() to pmd_softleaf_to_folio() Usama Arif
2026-09-14 21:17   ` Barry Song
2026-09-15  3:06   ` Baoquan He
2026-09-15  3:34   ` Qi Zheng
2026-09-14 12:27 ` [RESEND v7 02/29] arm64: mm: add PMD swap-exclusive helpers Usama Arif
2026-09-18 21:11   ` David Hildenbrand (Arm)
2026-09-22 12:34     ` Usama Arif
2026-09-22 14:45       ` David Hildenbrand (Arm)
2026-09-22 15:03         ` Usama Arif
2026-09-14 12:27 ` [RESEND v7 03/29] loongarch: " Usama Arif
2026-09-18 21:18   ` David Hildenbrand (Arm)
2026-09-14 12:27 ` [RESEND v7 04/29] powerpc: " Usama Arif
2026-09-18 21:22   ` David Hildenbrand (Arm)
2026-09-22 12:48     ` Usama Arif
2026-09-14 12:27 ` [RESEND v7 05/29] riscv: " Usama Arif
2026-09-18 21:24   ` David Hildenbrand (Arm)
2026-09-14 12:27 ` [RESEND v7 06/29] s390: " Usama Arif
2026-09-18 21:41   ` David Hildenbrand (Arm)
2026-09-22 12:51     ` Usama Arif
2026-09-14 12:27 ` [RESEND v7 07/29] x86: " Usama Arif
2026-09-16 14:31   ` Kiryl Shutsemau
2026-09-22 12:52     ` Usama Arif
2026-09-18 21:43   ` David Hildenbrand (Arm)
2026-09-22 12:54     ` Usama Arif
2026-09-14 12:27 ` [RESEND v7 08/29] mm: recognize PMD swap entries in the softleaf layer Usama Arif
2026-09-16 14:47   ` Kiryl Shutsemau
2026-09-18 21:56   ` David Hildenbrand (Arm)
2026-09-22 13:08     ` Usama Arif
2026-09-22 14:48       ` David Hildenbrand (Arm)
2026-09-14 12:27 ` [RESEND v7 09/29] mm/debug_vm_pgtable: test PMD swap-exclusive helpers Usama Arif
2026-09-16 14:52   ` Kiryl Shutsemau
2026-09-18 21:57   ` David Hildenbrand (Arm)
2026-09-14 12:28 ` [RESEND v7 10/29] mm: make PMD migration-entry splitting explicit Usama Arif
2026-09-16 14:58   ` Kiryl Shutsemau
2026-09-18 22:10   ` David Hildenbrand (Arm)
2026-09-22 13:15     ` Usama Arif
2026-09-23 11:04   ` David Hildenbrand (Arm)
2026-09-24 16:12     ` Usama Arif [this message]
2026-09-14 12:28 ` [RESEND v7 11/29] mm: split PMD swap entries into PTE swap entries Usama Arif
2026-09-16 15:08   ` Kiryl Shutsemau
2026-09-22 11:46     ` David Hildenbrand (Arm)
2026-09-22 13:18       ` Usama Arif
2026-09-23 11:20   ` David Hildenbrand (Arm)
2026-09-24 17:34     ` Usama Arif
2026-09-24 19:57       ` David Hildenbrand (Arm)
2026-09-14 12:28 ` [RESEND v7 12/29] mm: handle PMD swap entries in fork path Usama Arif
2026-09-24 20:28   ` David Hildenbrand (Arm)
2026-09-25 11:16     ` Usama Arif
2026-09-14 12:28 ` [RESEND v7 13/29] mm: zswap: reject high-order swap cache allocations backed by zswap Usama Arif
2026-09-24 20:30   ` David Hildenbrand (Arm)
2026-09-14 12:28 ` [RESEND v7 14/29] mm: swap in PMD swap entries as whole THPs during swapoff Usama Arif
2026-09-14 12:28 ` [RESEND v7 15/29] fs/proc: account PMD swap entries in smaps Usama Arif
2026-09-24 20:37   ` David Hildenbrand (Arm)
2026-09-25 13:28     ` Usama Arif
2026-09-14 12:28 ` [RESEND v7 16/29] mm: handle soft-dirty and uffd-wp on PMD swap entries Usama Arif
2026-09-14 12:28 ` [RESEND v7 17/29] mm/hmm: fault PMD swap entries on demand Usama Arif
2026-09-14 12:28 ` [RESEND v7 18/29] mm: free PMD swap entries in zap_huge_pmd() Usama Arif
2026-09-14 12:28 ` [RESEND v7 19/29] mm/madvise: free PMD swap entries with MADV_FREE Usama Arif
2026-09-14 12:28 ` [RESEND v7 20/29] mm/madvise: skip PMD swap entries for MADV_COLD and MADV_PAGEOUT Usama Arif
2026-09-14 12:28 ` [RESEND v7 21/29] mm/madvise: keep PMD swap entries whole for MADV_GUARD_INSTALL/REMOVE Usama Arif
2026-09-14 12:28 ` [RESEND v7 22/29] mm/mincore: report PMD swap-cache residency Usama Arif
2026-09-14 12:28 ` [RESEND v7 23/29] mm/khugepaged: treat PMD swap entries as mapped THPs Usama Arif
2026-09-14 12:28 ` [RESEND v7 24/29] mm: handle PMD swap entries in MADV_WILLNEED Usama Arif
2026-09-14 12:28 ` [RESEND v7 25/29] mm: handle PMD swap entries in UFFDIO_MOVE Usama Arif
2026-09-14 12:28 ` [RESEND v7 26/29] mm: don't PTE-batch a swap-in over a hardware-poisoned subpage Usama Arif
2026-09-14 12:28 ` [RESEND v7 27/29] mm: handle PMD swap entry faults on swap-in Usama Arif
2026-09-14 12:28 ` [RESEND v7 28/29] mm: install PMD swap entries on swap-out Usama Arif
2026-09-14 12:28 ` [RESEND v7 29/29] selftests/mm: add PMD swap entry tests Usama Arif
2026-09-15  3:32 ` [RESEND v7 00/29] mm: PMD-level swap entries for anonymous THPs Andrew Morton
2026-09-15 14:09   ` Usama Arif
2026-09-16  0:00     ` Andrew Morton
2026-09-16 10:22       ` Usama Arif

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=0f5d22fd-24f2-4235-92f5-3dc065e0f153@linux.dev \
    --to=usama.arif@linux.dev \
    --cc=akpm@linux-foundation.org \
    --cc=alex@ghiti.fr \
    --cc=baohua@kernel.org \
    --cc=baolin.wang@linux.alibaba.com \
    --cc=baoquan.he@linux.dev \
    --cc=chrisl@kernel.org \
    --cc=david@kernel.org \
    --cc=dev.jain@arm.com \
    --cc=hannes@cmpxchg.org \
    --cc=kas@kernel.org \
    --cc=kasong@tencent.com \
    --cc=kernel-team@meta.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=luizcap@redhat.com \
    --cc=nico.pache@linux.dev \
    --cc=nphamcs@gmail.com \
    --cc=qi.zheng@linux.dev \
    --cc=riel@surriel.com \
    --cc=ryan.roberts@arm.com \
    --cc=shakeel.butt@linux.dev \
    --cc=shikemeng@huaweicloud.com \
    --cc=vbabka@kernel.org \
    --cc=willy@infradead.org \
    --cc=ying.huang@linux.alibaba.com \
    --cc=yosry@kernel.org \
    --cc=youngjun.park@lge.com \
    --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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®