[PATCH 09/13] gup: Don't allow FOLL_LONGTERM pinning of FS DAX pages

[David Hildenbrand]

On 02.07.24 01:47, Alistair Popple wrote:
> 
> David Hildenbrand <david at redhat.com> writes:
> 
>> On 27.06.24 02:54, Alistair Popple wrote:
>>> Longterm pinning of FS DAX pages should already be disallowed by
>>> various pXX_devmap checks. However a future change will cause these
>>> checks to be invalid for FS DAX pages so make
>>> folio_is_longterm_pinnable() return false for FS DAX pages.
>>> Signed-off-by: Alistair Popple <apopple at nvidia.com>
>>> ---
>>>    include/linux/memremap.h | 11 +++++++++++
>>>    include/linux/mm.h       |  4 ++++
>>>    2 files changed, 15 insertions(+)
>>> diff --git a/include/linux/memremap.h b/include/linux/memremap.h
>>> index 6505713..19a448e 100644
>>> --- a/include/linux/memremap.h
>>> +++ b/include/linux/memremap.h
>>> @@ -193,6 +193,17 @@ static inline bool folio_is_device_coherent(const struct folio *folio)
>>>    	return is_device_coherent_page(&folio->page);
>>>    }
>>>    +static inline bool is_device_dax_page(const struct page *page)
>>> +{
>>> +	return is_zone_device_page(page) &&
>>> +		page_dev_pagemap(page)->type == MEMORY_DEVICE_FS_DAX;
>>> +}
>>> +
>>> +static inline bool folio_is_device_dax(const struct folio *folio)
>>> +{
>>> +	return is_device_dax_page(&folio->page);
>>> +}
>>> +
>>>    #ifdef CONFIG_ZONE_DEVICE
>>>    void zone_device_page_init(struct page *page);
>>>    void *memremap_pages(struct dev_pagemap *pgmap, int nid);
>>> diff --git a/include/linux/mm.h b/include/linux/mm.h
>>> index b84368b..4d1cdea 100644
>>> --- a/include/linux/mm.h
>>> +++ b/include/linux/mm.h
>>> @@ -2032,6 +2032,10 @@ static inline bool folio_is_longterm_pinnable(struct folio *folio)
>>>    	if (folio_is_device_coherent(folio))
>>>    		return false;
>>>    +	/* DAX must also always allow eviction. */
>>> +	if (folio_is_device_dax(folio))
>>> +		return false;
>>> +
>>>    	/* Otherwise, non-movable zone folios can be pinned. */
>>>    	return !folio_is_zone_movable(folio);
>>>    
>>
>> Why is the check in check_vma_flags() insufficient? GUP-fast maybe?
> 
> Right. This came up when I was changing the code for GUP-fast, but also
> they shouldn't be longterm pinnable and adding the case to
> folio_is_longterm_pinnable() is an excellent way of documenting that.

Makes sense, might be worth adding that (GUP-fast and check_vma_flags() 
covering GUP-slow) to the patch description.

-- 
Cheers,

David / dhildenb