Re: [PATCH v3 2/5] mm: Add batched versions of ptep_modify_prot_start/commit

[Dev Jain <dev.jain@arm.com>]

On 21/05/25 4:46 pm, Ryan Roberts wrote:
> On 19/05/2025 08:48, Dev Jain wrote:
>> Batch ptep_modify_prot_start/commit in preparation for optimizing mprotect.
>> Architecture can override these helpers; in case not, they are implemented
>> as a simple loop over the corresponding single pte helpers.
>>
>> Signed-off-by: Dev Jain <dev.jain@arm.com>
>> ---
>>   include/linux/pgtable.h | 75 +++++++++++++++++++++++++++++++++++++++++
>>   mm/mprotect.c           |  4 +--
>>   2 files changed, 77 insertions(+), 2 deletions(-)
>>
>> diff --git a/include/linux/pgtable.h b/include/linux/pgtable.h
>> index b50447ef1c92..e40ed57e034d 100644
>> --- a/include/linux/pgtable.h
>> +++ b/include/linux/pgtable.h
>> @@ -1333,6 +1333,81 @@ static inline void ptep_modify_prot_commit(struct vm_area_struct *vma,
>>   	__ptep_modify_prot_commit(vma, addr, ptep, pte);
>>   }
>>   #endif /* __HAVE_ARCH_PTEP_MODIFY_PROT_TRANSACTION */
>> +
>> +/**
>> + * modify_prot_start_ptes - Start a pte protection read-modify-write transaction
>> + * over a batch of ptes, which protects against asynchronous hardware modifications
> nit: This overflows the 80 char soft limit.
>
>> + * to the ptes. The intention is not to prevent the hardware from making pte
>> + * updates, but to prevent any updates it may make from being lost.
>> + * Please see the comment above ptep_modify_prot_start() for full description.
>> + *
>> + * @vma: The virtual memory area the pages are mapped into.
>> + * @addr: Address the first page is mapped at.
>> + * @ptep: Page table pointer for the first entry.
>> + * @nr: Number of entries.
>> + *
>> + * May be overridden by the architecture; otherwise, implemented as a simple
>> + * loop over ptep_modify_prot_start(), collecting the a/d bits of the mapped
>> + * folio.
> nit: "mapped folio" is a bit confusing given we are operating on ptes. Perhaps
> "collecting the a/d bits from each pte in the batch" is clearer.


Sure.


>
>> + *
>> + * Note that PTE bits in the PTE range besides the PFN can differ.
> nit: Perhaps "batch" would be more consistent than "range"?


Sure.


>
>> + *
>> + * Context: The caller holds the page table lock.  The PTEs map consecutive
>> + * pages that belong to the same folio.  The PTEs are all in the same PMD.
>> + */
>> +#ifndef modify_prot_start_ptes
>> +static inline pte_t modify_prot_start_ptes(struct vm_area_struct *vma,
>> +		unsigned long addr, pte_t *ptep, unsigned int nr)
> I thought David H suggested modify_prot_ptes_start() and
> modify_prot_ptes_commit(), which we settled on? I'm personally fine with either
> though.


No strong opinion, I'll do that.


>
>> +{
>> +	pte_t pte, tmp_pte;
>> +
>> +	pte = ptep_modify_prot_start(vma, addr, ptep);
>> +	while (--nr) {
> I thought we agreed to make the loop logic a bit more standard. I don't recall
> exactly what was finally agreed, but I would think something like this would be
> better:
>
> 	for (i = 1; i < nr; i++) {


Sure.


>
>> +		ptep++;
>> +		addr += PAGE_SIZE;
>> +		tmp_pte = ptep_modify_prot_start(vma, addr, ptep);
>> +		if (pte_dirty(tmp_pte))
>> +			pte = pte_mkdirty(pte);
>> +		if (pte_young(tmp_pte))
>> +			pte = pte_mkyoung(pte);
>> +	}
>> +	return pte;
>> +}
>> +#endif
>> +
>> +/**
>> + * modify_prot_commit_ptes - Commit an update to a batch of ptes, leaving any
>> + * hardware-controlled bits in the PTE unmodified.
>> + *
>> + * @vma: The virtual memory area the pages are mapped into.
>> + * @addr: Address the first page is mapped at.
>> + * @ptep: Page table pointer for the first entry.
> You've missed pte and old_pte params here.


My bad.


>
>> + * @nr: Number of entries.
>> + *
>> + * May be overridden by the architecture; otherwise, implemented as a simple
>> + * loop over ptep_modify_prot_commit().
>> + *
>> + * Note that PTE bits in the PTE range besides the PFN can differ.
> How can it? All the applied bits other than the PFN will be exactly the same for
> the range because they all come from pte. I think this line can be dropped.


Copy pasted, then forgot to remove :)


>
>> + *
>> + * Context: The caller holds the page table lock.  The PTEs map consecutive
>> + * pages that belong to the same folio.  The PTEs are all in the same PMD.
> The middle sentance doesn't apply; the PTEs will all initially be none if using
> the default version of modify_prot_start_ptes(). I think that can be dropped.
> But I think you need to explain that this will be the case on exit.


Ah got it. "On exit, the set ptes will map the same folio."


>
>> + */
>> +#ifndef modify_prot_commit_ptes
>> +static inline void modify_prot_commit_ptes(struct vm_area_struct *vma, unsigned long addr,
>> +		pte_t *ptep, pte_t old_pte, pte_t pte, unsigned int nr)
>> +{
>> +	int i;
>> +
>> +	for (i = 0; i < nr; ++i) {
>> +		ptep_modify_prot_commit(vma, addr, ptep, old_pte, pte);
>> +		ptep++;
>> +		addr += PAGE_SIZE;
>> +		old_pte = pte_next_pfn(old_pte);
>> +		pte = pte_next_pfn(pte);
>> +	}
>> +}
>> +#endif
> I have some general concerns about the correctness of batching these functions.
> The support was originally added by Commit 1ea0704e0da6 ("mm: add a
> ptep_modify_prot transaction abstraction"), and the intent was to make it easier
> to defer the pte updates for XEN on x86.
>
> Your default implementations of the batched versions will match the number of
> ptep_modify_prot_start() calls with the same number of ptep_modify_prot_commit()
> calls, even if modify_prot_commit_ptes() is called incrementally for sub-batches
> of the batch used for modify_prot_start_ptes(). That's a requirement and you've
> met it. But in the batched case, there are 2 differences;
>
>    - You can now have multiple PTEs within a start-commit block at one time. I
> hope none of the specialized implementations care about that (i.e. XEN).
>
>    - when calling ptep_modify_prot_commit(), old_pte may not be exactly what
> ptep_modify_prot_start() returned for that pte. You have collected the A/D bits,
> and according to your docs "PTE bits in the PTE range besides the PFN can
> differ" when calling modify_prot_start_ptes() so R/W and other things could
> differ here.
>
> I'm not sure if these are problems in practice; they probably are not. But have
> you checked the XEN implementation (and any other specialized implementations)
> are definitely compatible with your batched semantics?
>
> Thanks,
> Ryan
>
>> +
>>   #endif /* CONFIG_MMU */
>>   
>>   /*
>> diff --git a/mm/mprotect.c b/mm/mprotect.c
>> index 1ee160ed0b14..124612ce3d24 100644
>> --- a/mm/mprotect.c
>> +++ b/mm/mprotect.c
>> @@ -188,7 +188,7 @@ static long change_pte_range(struct mmu_gather *tlb,
>>   						jiffies_to_msecs(jiffies));
>>   			}
>>   
>> -			oldpte = ptep_modify_prot_start(vma, addr, pte);
>> +			oldpte = modify_prot_start_ptes(vma, addr, pte, nr_ptes);
>>   			ptent = pte_modify(oldpte, newprot);
>>   
>>   			if (uffd_wp)
>> @@ -214,7 +214,7 @@ static long change_pte_range(struct mmu_gather *tlb,
>>   			    can_change_pte_writable(vma, addr, ptent))
>>   				ptent = pte_mkwrite(ptent, vma);
>>   
>> -			ptep_modify_prot_commit(vma, addr, pte, oldpte, ptent);
>> +			modify_prot_commit_ptes(vma, addr, pte, oldpte, ptent, nr_ptes);
>>   			if (pte_needs_flush(oldpte, ptent))
>>   				tlb_flush_pte_range(tlb, addr, PAGE_SIZE);
>>   			pages++;