linux-kernel - Re: [PATCH v2 1/3] Revert "mm/ksm: convert break_ksm() from walk_page_range_vma() to folio

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <42byvvz55omaszu6ep3g7n2dj5z7mfxy5h3zbc3xjdnslemkpp@kvdzrjz423mb>
Date: Wed, 5 Nov 2025 10:28:51 -0300
From: Pedro Demarchi Gomes <pedrodemargomes@...il.com>
To: "David Hildenbrand (Red Hat)" <david@...nel.org>
Cc: David Hildenbrand <david@...hat.com>, 
	Andrew Morton <akpm@...ux-foundation.org>, Xu Xin <xu.xin16@....com.cn>, 
	Chengming Zhou <chengming.zhou@...ux.dev>, linux-mm@...ck.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 1/3] Revert "mm/ksm: convert break_ksm() from
 walk_page_range_vma() to folio_walk"

On Mon, Nov 03, 2025 at 06:00:08PM +0100, David Hildenbrand (Red Hat) wrote:
> On 31.10.25 18:46, Pedro Demarchi Gomes wrote:
> > This reverts commit e317a8d8b4f600fc7ec9725e26417030ee594f52 and changes
> > function break_ksm_pmd_entry() to use folios.
> > 
> > This reverts break_ksm() to use walk_page_range_vma() instead of
> > folio_walk_start().
> > This will make it easier to later modify break_ksm() to perform a proper
> > range walk.
> > 
> > Suggested-by: David Hildenbrand <david@...hat.com>
> > Signed-off-by: Pedro Demarchi Gomes <pedrodemargomes@...il.com>
> > ---
> >   mm/ksm.c | 63 ++++++++++++++++++++++++++++++++++++++++++--------------
> >   1 file changed, 47 insertions(+), 16 deletions(-)
> > 
> > diff --git a/mm/ksm.c b/mm/ksm.c
> > index 4f672f4f2140..922d2936e206 100644
> > --- a/mm/ksm.c
> > +++ b/mm/ksm.c
> > @@ -607,6 +607,47 @@ static inline bool ksm_test_exit(struct mm_struct *mm)
> >   	return atomic_read(&mm->mm_users) == 0;
> >   }
> > +static int break_ksm_pmd_entry(pmd_t *pmd, unsigned long addr, unsigned long next,
> > +			struct mm_walk *walk)
> > +{
> > +	struct folio *folio = NULL;
> > +	spinlock_t *ptl;
> > +	pte_t *pte;
> > +	pte_t ptent;
> > +	int ret;
> > +
> > +	pte = pte_offset_map_lock(walk->mm, pmd, addr, &ptl);
> > +	if (!pte)
> > +		return 0;
> > +	ptent = ptep_get(pte);
> > +	if (pte_present(ptent)) {
> > +		folio = vm_normal_folio(walk->vma, addr, ptent);
> > +	} else if (!pte_none(ptent)) {
> > +		swp_entry_t entry = pte_to_swp_entry(ptent);
> > +
> > +		/*
> > +		 * As KSM pages remain KSM pages until freed, no need to wait
> > +		 * here for migration to end.
> > +		 */
> > +		if (is_migration_entry(entry))
> > +			folio = pfn_swap_entry_folio(entry);
> > +	}
> > +	/* return 1 if the page is an normal ksm page or KSM-placed zero page */
> > +	ret = (folio && folio_test_ksm(folio)) || is_ksm_zero_pte(ptent);
> 
> Staring again, we should really call is_ksm_zero_pte() only if we know the
> folio is present.
> 
> It's not super dangerous in the old code (because we would only look at
> present an migration entries), but now you are making it possible to call it
> on even more non-present ptes.
> 

IIUC vm_normal_folio will return NULL in the case of a ksm zero pte, so
we can not do
	found = folio && (folio_test_ksm(folio) || is_ksm_zero_pte(pte))
because it will always be false for a ksm zero pte.
So we should do 
	found = (folio && folio_test_ksm(folio)) || (pte_present(ptent) 
		&& is_ksm_zero_pte(ptent));
since if the pte is present and is a zero pte we can guarantee that
the folio is present.

Sorry if I am missing something.


> -- 
> Cheers
> 
> David
>