Re: [PATCH v1 5/7] mm/pagewalk: add walk_page_range_vma()

On 05.10.22 22:42, Peter Xu wrote:

On Fri, Sep 30, 2022 at 04:19:29PM +0200, David Hildenbrand wrote:

...

+int walk_page_range_vma(struct vm_area_struct *vma, unsigned long start,

• 	unsigned long end, const struct mm_walk_ops *ops,
  

• 	void *private)
  

+{

• struct mm_walk walk = {

• .ops		= ops,
  

• .mm		= vma->vm_mm,
  

• .vma		= vma,
  

• .private	= private,
  

• };
• int err;
• if (start >= end || !walk.mm)

• return -EINVAL;
  

• if (start < vma->vm_start || end > vma->vm_end)

• return -EINVAL;
  

• mmap_assert_locked(walk.mm);
• err = walk_page_test(start, end, &walk);

According to test_walk():

• @test_walk: caller specific callback function to determine whether

• 	we walk over the current vma or not. Returning 0 means
  

• 	"do page table walk over the current vma", returning
  

• 	a negative value means "abort current page table walk
  

• 	right now" and returning 1 means "skip the current vma"
  
  

Since this helper has vma passed in, not sure whether this is needed at all?

I kept it because walk_page_vma() similarly has it -- walk_page_vma() walks the whole VMA range.

I do agree that it's kind of weird to have it like that. All users of walk_page_vma() don't use it, so we can just get rid of it there as well. Might make the walk slightly faster.

walk_page_vma_range() sounds slightly better to me as it does look more like an extension of walk_page_vma(), rather than sister version of walk_page_range_novma() (which works for "no backing VMA" case). But no strong opinion.

I matched that to walk_page_range_novma(). Now we have

walk_page_range walk_page_vma walk_page_range_novma walk_page_range_vma