Re: [PATCH v3 14/16] mm: support GUP-triggered unsharing of anonymous pages

From: Vlastimil Babka <vbabka@suse.cz>
To: David Hildenbrand <david@redhat.com>, linux-kernel@vger.kernel.org
Cc: Andrew Morton <akpm@linux-foundation.org>,
	Hugh Dickins <hughd@google.com>,
	Linus Torvalds <torvalds@linux-foundation.org>,
	David Rientjes <rientjes@google.com>,
	Shakeel Butt <shakeelb@google.com>,
	John Hubbard <jhubbard@nvidia.com>,
	Jason Gunthorpe <jgg@nvidia.com>,
	Mike Kravetz <mike.kravetz@oracle.com>,
	Mike Rapoport <rppt@linux.ibm.com>,
	Yang Shi <shy828301@gmail.com>,
	"Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>,
	Matthew Wilcox <willy@infradead.org>,
	Jann Horn <jannh@google.com>, Michal Hocko <mhocko@kernel.org>,
	Nadav Amit <namit@vmware.com>, Rik van Riel <riel@surriel.com>,
	Roman Gushchin <guro@fb.com>,
	Andrea Arcangeli <aarcange@redhat.com>,
	Peter Xu <peterx@redhat.com>, Donald Dutile <ddutile@redhat.com>,
	Christoph Hellwig <hch@lst.de>, Oleg Nesterov <oleg@redhat.com>,
	Jan Kara <jack@suse.cz>, Liang Zhang <zhangliang5@huawei.com>,
	Pedro Gomes <pedrodemargomes@gmail.com>,
	Oded Gabbay <oded.gabbay@gmail.com>,
	linux-mm@kvack.org
Subject: Re: [PATCH v3 14/16] mm: support GUP-triggered unsharing of anonymous pages
Date: Tue, 19 Apr 2022 18:31:49 +0200	[thread overview]
Message-ID: <3cc33611-a430-1ce4-bb30-7dc62a530cb8@suse.cz> (raw)
In-Reply-To: <c003964a-9fa1-b17c-158a-8ccbf4ff8656@redhat.com>

On 4/19/22 18:29, David Hildenbrand wrote:
>>> @@ -4515,8 +4550,11 @@ static inline vm_fault_t create_huge_pmd(struct vm_fault *vmf)
>>>  /* `inline' is required to avoid gcc 4.1.2 build error */
>>>  static inline vm_fault_t wp_huge_pmd(struct vm_fault *vmf)
>>>  {
>>> +	const bool unshare = vmf->flags & FAULT_FLAG_UNSHARE;
>>> +
>>>  	if (vma_is_anonymous(vmf->vma)) {
>>> -		if (userfaultfd_huge_pmd_wp(vmf->vma, vmf->orig_pmd))
>>> +		if (unlikely(unshare) &&
>> 
>> Is this condition flipped, should it be "likely(!unshare)"? As the similar
>> code in do_wp_page() does.
> 
> Good catch, this should affect uffd-wp on THP -- it wouldn't trigger as expected. Thanks a lot for finding that!

Yay, glad I was right this time.

>> 
>>> +		    userfaultfd_huge_pmd_wp(vmf->vma, vmf->orig_pmd))
>>>  			return handle_userfault(vmf, VM_UFFD_WP);
>>>  		return do_huge_pmd_wp_page(vmf);
>>>  	}
>>> @@ -4651,10 +4689,11 @@ static vm_fault_t handle_pte_fault(struct vm_fault *vmf)
>>>  		update_mmu_tlb(vmf->vma, vmf->address, vmf->pte);
>>>  		goto unlock;
>>>  	}
>>> -	if (vmf->flags & FAULT_FLAG_WRITE) {
>>> +	if (vmf->flags & (FAULT_FLAG_WRITE|FAULT_FLAG_UNSHARE)) {
>>>  		if (!pte_write(entry))
>>>  			return do_wp_page(vmf);
>>> -		entry = pte_mkdirty(entry);
>>> +		else if (likely(vmf->flags & FAULT_FLAG_WRITE))
>>> +			entry = pte_mkdirty(entry);
>>>  	}
>>>  	entry = pte_mkyoung(entry);
>>>  	if (ptep_set_access_flags(vmf->vma, vmf->address, vmf->pte, entry,
>> 
> 
> 
> So the following on top, right?

Looks good!

> diff --git a/mm/memory.c b/mm/memory.c
> index 8b3cb73f5e44..4584c7e87a70 100644
> --- a/mm/memory.c
> +++ b/mm/memory.c
> @@ -3137,7 +3137,7 @@ static vm_fault_t wp_page_copy(struct vm_fault *vmf)
>                         free_swap_cache(old_page);
>                 put_page(old_page);
>         }
> -       return page_copied && !unshare ? VM_FAULT_WRITE : 0;
> +       return (page_copied && !unshare) ? VM_FAULT_WRITE : 0;
>  oom_free_new:
>         put_page(new_page);
>  oom:
> @@ -4604,7 +4604,7 @@ static inline vm_fault_t wp_huge_pmd(struct vm_fault *vmf)
>         const bool unshare = vmf->flags & FAULT_FLAG_UNSHARE;
>  
>         if (vma_is_anonymous(vmf->vma)) {
> -               if (unlikely(unshare) &&
> +               if (likely(!unshare) &&
>                     userfaultfd_huge_pmd_wp(vmf->vma, vmf->orig_pmd))
>                         return handle_userfault(vmf, VM_UFFD_WP);
>                 return do_huge_pmd_wp_page(vmf);
> 
> 