Message-ID: <91c4e7e6-e4c2-4ff0-8b13-7b3ff138e98e@redhat.com>
Date: Wed, 30 Apr 2025 23:25:56 +0200
From: David Hildenbrand <david@...hat.com>
To: Petr Vaněk <arkamar@...as.cz>
Cc: linux-kernel@...r.kernel.org, Andrew Morton <akpm@...ux-foundation.org>,
 Ryan Roberts <ryan.roberts@....com>, linux-mm@...ck.org,
 stable@...r.kernel.org
Subject: Re: [PATCH 1/1] mm: Fix folio_pte_batch() overcount with zero PTEs

On 30.04.25 18:00, Petr Vaněk wrote:
> On Wed, Apr 30, 2025 at 04:37:21PM +0200, David Hildenbrand wrote:
>> On 30.04.25 13:52, Petr Vaněk wrote:
>>> On Tue, Apr 29, 2025 at 08:56:03PM +0200, David Hildenbrand wrote:
>>>> On 29.04.25 20:33, Petr Vaněk wrote:
>>>>> On Tue, Apr 29, 2025 at 05:45:53PM +0200, David Hildenbrand wrote:
>>>>>> On 29.04.25 16:52, David Hildenbrand wrote:
>>>>>>> On 29.04.25 16:45, Petr Vaněk wrote:
>>>>>>>> On Tue, Apr 29, 2025 at 04:29:30PM +0200, David Hildenbrand wrote:
>>>>>>>>> On 29.04.25 16:22, Petr Vaněk wrote:
>>>>>>>>>> folio_pte_batch() could overcount the number of contiguous PTEs when
>>>>>>>>>> pte_advance_pfn() returns a zero-valued PTE and the following PTE in
>>>>>>>>>> memory also happens to be zero. The loop doesn't break in such a case
>>>>>>>>>> because pte_same() returns true, and the batch size is advanced by one
>>>>>>>>>> more than it should be.
>>>>>>>>>>
>>>>>>>>>> To fix this, bail out early if a non-present PTE is encountered,
>>>>>>>>>> preventing the invalid comparison.
>>>>>>>>>>
>>>>>>>>>> This issue started to appear after commit 10ebac4f95e7 ("mm/memory:
>>>>>>>>>> optimize unmap/zap with PTE-mapped THP") and was discovered via git
>>>>>>>>>> bisect.
>>>>>>>>>>
>>>>>>>>>> Fixes: 10ebac4f95e7 ("mm/memory: optimize unmap/zap with PTE-mapped THP")
>>>>>>>>>> Cc: stable@...r.kernel.org
>>>>>>>>>> Signed-off-by: Petr Vaněk <arkamar@...as.cz>
>>>>>>>>>> ---
>>>>>>>>>>        mm/internal.h | 2 ++
>>>>>>>>>>        1 file changed, 2 insertions(+)
>>>>>>>>>>
>>>>>>>>>> diff --git a/mm/internal.h b/mm/internal.h
>>>>>>>>>> index e9695baa5922..c181fe2bac9d 100644
>>>>>>>>>> --- a/mm/internal.h
>>>>>>>>>> +++ b/mm/internal.h
>>>>>>>>>> @@ -279,6 +279,8 @@ static inline int folio_pte_batch(struct folio *folio, unsigned long addr,
>>>>>>>>>>        			dirty = !!pte_dirty(pte);
>>>>>>>>>>        		pte = __pte_batch_clear_ignored(pte, flags);
>>>>>>>>>>        
>>>>>>>>>> +		if (!pte_present(pte))
>>>>>>>>>> +			break;
>>>>>>>>>>        		if (!pte_same(pte, expected_pte))
>>>>>>>>>>        			break;
>>>>>>>>>
>>>>>>>>> How could pte_same() suddenly match on a present and non-present PTE.
>>>>>>>>
>>>>>>>> In the problematic case pte.pte == 0 and expected_pte.pte == 0 as well.
>>>>>>>> pte_same() returns a.pte == b.pte -> 0 == 0. Both are non-present PTEs.
>>>>>>>
>>>>>>> Observe that folio_pte_batch() was called *with a present pte*.
>>>>>>>
>>>>>>> do_zap_pte_range()
>>>>>>> 	if (pte_present(ptent))
>>>>>>> 		zap_present_ptes()
>>>>>>> 			folio_pte_batch()
>>>>>>>
>>>>>>> How can we end up with an expected_pte that is !present, if it is based
>>>>>>> on the provided pte that *is present* and we only used pte_advance_pfn()
>>>>>>> to advance the pfn?
>>>>>>
>>>>>> I've been staring at the code for too long and don't see the issue.
>>>>>>
>>>>>> We even have
>>>>>>
>>>>>> VM_WARN_ON_FOLIO(!pte_present(pte), folio);
>>>>>>
>>>>>> So the initial pteval we got is present.
>>>>>>
>>>>>> I don't see how
>>>>>>
>>>>>> 	nr = pte_batch_hint(start_ptep, pte);
>>>>>> 	expected_pte = __pte_batch_clear_ignored(pte_advance_pfn(pte, nr), flags);
>>>>>>
>>>>>> would suddenly result in !pte_present(expected_pte).
>>>>>
>>>>> The issue is not happening in __pte_batch_clear_ignored but later in
>>>>> following line:
>>>>>
>>>>>      expected_pte = pte_advance_pfn(expected_pte, nr);
>>>>>
>>>>> The issue seems to be in __pte function which converts PTE value to
>>>>> pte_t in pte_advance_pfn, because warnings disappears when I change the
>>>>> line to
>>>>>
>>>>>      expected_pte = (pte_t){ .pte = pte_val(expected_pte) + (nr << PFN_PTE_SHIFT) };
>>>>>
>>>>> The kernel probably uses __pte function from
>>>>> arch/x86/include/asm/paravirt.h because it is configured with
>>>>> CONFIG_PARAVIRT=y:
>>>>>
>>>>>      static inline pte_t __pte(pteval_t val)
>>>>>      {
>>>>>      	return (pte_t) { PVOP_ALT_CALLEE1(pteval_t, mmu.make_pte, val,
>>>>>      					  "mov %%rdi, %%rax", ALT_NOT_XEN) };
>>>>>      }
>>>>>
>>>>> I guess it might cause this weird magic, but I need more time to
>>>>> understand what it does :)
>>>
>>> I understand it slightly more. __pte() uses xen_make_pte(), which calls
>>> pte_pfn_to_mfn(), however, mfn for this pfn contains INVALID_P2M_ENTRY
>>> value, therefore the pte_pfn_to_mfn() returns 0, see [1].
>>>
>>> I guess that the mfn was invalidated by xen-balloon driver?
>>>
>>> [1] https://web.git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/arch/x86/xen/mmu_pv.c?h=v6.15-rc4#n408
>>>
>>>> What XEN does with basic primitives that convert between pteval and
>>>> pte_t is beyond horrible.
>>>>
>>>> How come set_ptes() that uses pte_next_pfn()->pte_advance_pfn() does not
>>>> run into this?
>>>
>>> I don't know, but I guess it is somehow related to pfn->mfn translation.
>>>
>>>> Is it only a problem if we exceed a certain pfn?
>>>
>>> No, it is a problem if the corresponding mft to given pfn is invalid.
>>>
>>> I am not sure if my original patch is a good fix.
>>
>> No :)
>>
>> Maybe it would be
>>> better to have some sort of native_pte_advance_pfn() which will use
>>> native_make_pte() rather than __pte(). Or do you think the issue is in
>>> Xen part?
>>
>> I think what's happening is that -- under XEN only -- we might get garbage when
>> calling pte_advance_pfn() and the next PFN would no longer fall into the folio. And
>> the current code cannot deal with that XEN garbage.
>>
>> But still not 100% sure.
>>
>> The following is completely untested, could you give that a try?
> 
> Yes, it solves the issue for me.

Cool!

> 
> However, maybe it would be better to solve it with the following patch.
> The pte_pfn_to_mfn() actually returns the same value for non-present
> PTEs. I suggest to return original PTE if the mfn is INVALID_P2M_ENTRY,
> rather than empty non-present PTE, but the _PAGE_PRESENT bit will be set
> to zero. Thus, we will not loose information about original pfn but it
> will be clear that the page is not present.
> 
>  From e84781f9ec4fb7275d5e7629cf7e222466caf759 Mon Sep 17 00:00:00 2001
> From: =?UTF-8?q?Petr=20Van=C4=9Bk?= <arkamar@...as.cz>
> Date: Wed, 30 Apr 2025 17:08:41 +0200
> Subject: [PATCH] x86/mm: Reset pte _PAGE_PRESENT bit for invalid mft
> MIME-Version: 1.0
> Content-Type: text/plain; charset=UTF-8
> Content-Transfer-Encoding: 8bit
> 
> Signed-off-by: Petr Vaněk <arkamar@...as.cz>
> ---
>   arch/x86/xen/mmu_pv.c | 9 +++------
>   1 file changed, 3 insertions(+), 6 deletions(-)
> 
> diff --git a/arch/x86/xen/mmu_pv.c b/arch/x86/xen/mmu_pv.c
> index 38971c6dcd4b..92a6a9af0c65 100644
> --- a/arch/x86/xen/mmu_pv.c
> +++ b/arch/x86/xen/mmu_pv.c
> @@ -392,28 +392,25 @@ static pteval_t pte_mfn_to_pfn(pteval_t val)
>   static pteval_t pte_pfn_to_mfn(pteval_t val)
>   {
>   	if (val & _PAGE_PRESENT) {
>   		unsigned long pfn = (val & PTE_PFN_MASK) >> PAGE_SHIFT;
>   		pteval_t flags = val & PTE_FLAGS_MASK;
>   		unsigned long mfn;
>   
>   		mfn = __pfn_to_mfn(pfn);
>   
>   		/*
> -		 * If there's no mfn for the pfn, then just create an
> -		 * empty non-present pte.  Unfortunately this loses
> -		 * information about the original pfn, so
> -		 * pte_mfn_to_pfn is asymmetric.
> +		 * If there's no mfn for the pfn, then just reset present pte bit.
>   		 */
>   		if (unlikely(mfn == INVALID_P2M_ENTRY)) {
> -			mfn = 0;
> -			flags = 0;
> +			mfn = pfn;
> +			flags &= ~_PAGE_PRESENT;
>   		} else
>   			mfn &= ~(FOREIGN_FRAME_BIT | IDENTITY_FRAME_BIT);
>   		val = ((pteval_t)mfn << PAGE_SHIFT) | flags;
>   	}
>   
>   	return val;
>   }
>   
>   __visible pteval_t xen_pte_val(pte_t pte)
>   {

That might do as well.


I assume the following would also work? (removing the early ==1 check)

It has the general benefit of removing the pte_pfn() call
from the loop body, which is why I like that fix. (almost looks like a cleanup)


 From 75948778b586d4759a480bf412fd4682067b12ea Mon Sep 17 00:00:00 2001
From: David Hildenbrand <david@...hat.com>
Date: Wed, 30 Apr 2025 16:35:12 +0200
Subject: [PATCH] tmp

Signed-off-by: David Hildenbrand <david@...hat.com>
---
  mm/internal.h | 27 +++++++++++----------------
  1 file changed, 11 insertions(+), 16 deletions(-)

diff --git a/mm/internal.h b/mm/internal.h
index e9695baa59226..25a29872c634b 100644
--- a/mm/internal.h
+++ b/mm/internal.h
@@ -248,11 +248,9 @@ static inline int folio_pte_batch(struct folio *folio, unsigned long addr,
  		pte_t *start_ptep, pte_t pte, int max_nr, fpb_t flags,
  		bool *any_writable, bool *any_young, bool *any_dirty)
  {
-	unsigned long folio_end_pfn = folio_pfn(folio) + folio_nr_pages(folio);
-	const pte_t *end_ptep = start_ptep + max_nr;
  	pte_t expected_pte, *ptep;
  	bool writable, young, dirty;
-	int nr;
+	int nr, cur_nr;
  
  	if (any_writable)
  		*any_writable = false;
@@ -265,11 +263,15 @@ static inline int folio_pte_batch(struct folio *folio, unsigned long addr,
  	VM_WARN_ON_FOLIO(!folio_test_large(folio) || max_nr < 1, folio);
  	VM_WARN_ON_FOLIO(page_folio(pfn_to_page(pte_pfn(pte))) != folio, folio);
  
+	/* Limit max_nr to the actual remaining PFNs in the folio we could batch. */
+	max_nr = min_t(unsigned long, max_nr,
+		       folio_pfn(folio) + folio_nr_pages(folio) - pte_pfn(pte));
+
  	nr = pte_batch_hint(start_ptep, pte);
  	expected_pte = __pte_batch_clear_ignored(pte_advance_pfn(pte, nr), flags);
  	ptep = start_ptep + nr;
  
-	while (ptep < end_ptep) {
+	while (nr < max_nr) {
  		pte = ptep_get(ptep);
  		if (any_writable)
  			writable = !!pte_write(pte);
@@ -282,14 +284,6 @@ static inline int folio_pte_batch(struct folio *folio, unsigned long addr,
  		if (!pte_same(pte, expected_pte))
  			break;
  
-		/*
-		 * Stop immediately once we reached the end of the folio. In
-		 * corner cases the next PFN might fall into a different
-		 * folio.
-		 */
-		if (pte_pfn(pte) >= folio_end_pfn)
-			break;
-
  		if (any_writable)
  			*any_writable |= writable;
  		if (any_young)
@@ -297,12 +291,13 @@ static inline int folio_pte_batch(struct folio *folio, unsigned long addr,
  		if (any_dirty)
  			*any_dirty |= dirty;
  
-		nr = pte_batch_hint(ptep, pte);
-		expected_pte = pte_advance_pfn(expected_pte, nr);
-		ptep += nr;
+		cur_nr = pte_batch_hint(ptep, pte);
+		expected_pte = pte_advance_pfn(expected_pte, cur_nr);
+		ptep += cur_nr;
+		nr += cur_nr;
  	}
  
-	return min(ptep - start_ptep, max_nr);
+	return min(nr, max_nr);
  }
  
  /**
-- 
2.49.0


-- 
Cheers,

David / dhildenb

Powered by blists - more mailing lists