Re: [PATCH v3 02/11] KVM: guest_memfd: Handle final folio_put() of guest_memfd pages

[David Hildenbrand <david@redhat.com>]

On 20.02.25 12:28, Vlastimil Babka wrote:
> On 2/20/25 12:25, David Hildenbrand wrote:
>> On 11.02.25 13:11, Fuad Tabba wrote:
>>> Before transitioning a guest_memfd folio to unshared, thereby
>>> disallowing access by the host and allowing the hypervisor to
>>> transition its view of the guest page as private, we need to be
>>> sure that the host doesn't have any references to the folio.
>>>
>>> This patch introduces a new type for guest_memfd folios, which
>>> isn't activated in this series but is here as a placeholder and
>>> to facilitate the code in the next patch. This will be used in
>>> the future to register a callback that informs the guest_memfd
>>> subsystem when the last reference is dropped, therefore knowing
>>> that the host doesn't have any remaining references.
>>>
>>> Signed-off-by: Fuad Tabba <tabba@google.com>
>>> ---
>>
>> [...]
>>
>>>    static const char *page_type_name(unsigned int page_type)
>>> diff --git a/mm/swap.c b/mm/swap.c
>>> index 47bc1bb919cc..241880a46358 100644
>>> --- a/mm/swap.c
>>> +++ b/mm/swap.c
>>> @@ -38,6 +38,10 @@
>>>    #include <linux/local_lock.h>
>>>    #include <linux/buffer_head.h>
>>>    
>>> +#ifdef CONFIG_KVM_GMEM_SHARED_MEM
>>> +#include <linux/kvm_host.h>
>>> +#endif
>>> +
>>>    #include "internal.h"
>>>    
>>>    #define CREATE_TRACE_POINTS
>>> @@ -101,6 +105,11 @@ static void free_typed_folio(struct folio *folio)
>>>    	case PGTY_hugetlb:
>>>    		free_huge_folio(folio);
>>>    		return;
>>> +#endif
>>> +#ifdef CONFIG_KVM_GMEM_SHARED_MEM
>>> +	case PGTY_guestmem:
>>> +		kvm_gmem_handle_folio_put(folio);
>>> +		return;
>>
>> Hm, if KVM is built as a module, will that work? Or would we need the
> 
> Good catch, I guess not?
> 
>> core-mm guest_memfd shim that would always be compiled into the core and
>> decouple KVM from guest_memfd ("library")?
> 
> That could also help avoid exporting the mpol symbols in the NUMA mempolicy
> series?

Yes [1]! :)

[1] 
https://lore.kernel.org/linux-mm/9392618e-32de-4a86-9e1e-bcfeefe39181@redhat.com/

-- 
Cheers,

David / dhildenb