1. Patchew
  2. linux
  3. [PATCH v4 0/4] Let unmap_hugepage_range() and several related functions to take folio instead of page
  4. View patch

[PATCH v4 2/4] mm/hugetlb: Refactor unmap_hugepage_range() to take folio instead of page

nifan.cxl@gmail.com posted 4 patches 7 months, 2 weeks ago
[PATCH v4 2/4] mm/hugetlb: Refactor unmap_hugepage_range() to take folio instead of page
Posted by nifan.cxl@gmail.com 7 months, 2 weeks ago 
From: Fan Ni <fan.ni@samsung.com>

The function unmap_hugepage_range() has two kinds of users:
1) unmap_ref_private(), which passes in the head page of a folio.  Since
   unmap_ref_private() already takes folio and there are no other uses
   of the folio struct in the function, it is natural for
   unmap_hugepage_range() to take folio also.
2) All other uses, which pass in NULL pointer.

In both cases, we can pass in folio. Refactor unmap_hugepage_range() to
take folio.

Signed-off-by: Fan Ni <fan.ni@samsung.com>
Reviewed-by: Muchun Song <muchun.song@linux.dev>
Reviewed-by: Sidhartha Kumar <sidhartha.kumar@oracle.com>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Acked-by: David Hildenbrand <david@redhat.com>
---
 include/linux/hugetlb.h | 4 ++--
 mm/hugetlb.c            | 7 ++++---
 2 files changed, 6 insertions(+), 5 deletions(-)

diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h
index 23ebf49c5d6a..f6d5f24e793c 100644
--- a/include/linux/hugetlb.h
+++ b/include/linux/hugetlb.h
@@ -129,8 +129,8 @@ int move_hugetlb_page_tables(struct vm_area_struct *vma,
 int copy_hugetlb_page_range(struct mm_struct *, struct mm_struct *,
 			    struct vm_area_struct *, struct vm_area_struct *);
 void unmap_hugepage_range(struct vm_area_struct *,
-			  unsigned long, unsigned long, struct page *,
-			  zap_flags_t);
+			  unsigned long start, unsigned long end,
+			  struct folio *, zap_flags_t);
 void __unmap_hugepage_range(struct mmu_gather *tlb,
 			  struct vm_area_struct *vma,
 			  unsigned long start, unsigned long end,
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 0c2b264a7ab8..c339ffe05556 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -6046,7 +6046,7 @@ void __hugetlb_zap_end(struct vm_area_struct *vma,
 }
 
 void unmap_hugepage_range(struct vm_area_struct *vma, unsigned long start,
-			  unsigned long end, struct page *ref_page,
+			  unsigned long end, struct folio *folio,
 			  zap_flags_t zap_flags)
 {
 	struct mmu_notifier_range range;
@@ -6058,7 +6058,8 @@ void unmap_hugepage_range(struct vm_area_struct *vma, unsigned long start,
 	mmu_notifier_invalidate_range_start(&range);
 	tlb_gather_mmu(&tlb, vma->vm_mm);
 
-	__unmap_hugepage_range(&tlb, vma, start, end, ref_page, zap_flags);
+	__unmap_hugepage_range(&tlb, vma, start, end,
+			       &folio->page, zap_flags);
 
 	mmu_notifier_invalidate_range_end(&range);
 	tlb_finish_mmu(&tlb);
@@ -6116,7 +6117,7 @@ static void unmap_ref_private(struct mm_struct *mm, struct vm_area_struct *vma,
 		if (!is_vma_resv_set(iter_vma, HPAGE_RESV_OWNER))
 			unmap_hugepage_range(iter_vma, address,
 					     address + huge_page_size(h),
-					     &folio->page, 0);
+					     folio, 0);
 	}
 	i_mmap_unlock_write(mapping);
 }
-- 
2.47.2
Re: [PATCH v4 2/4] mm/hugetlb: Refactor unmap_hugepage_range() to take folio instead of page
Posted by Vishal Moola (Oracle) 7 months, 2 weeks ago 
On Mon, May 05, 2025 at 11:22:42AM -0700, nifan.cxl@gmail.com wrote:
> From: Fan Ni <fan.ni@samsung.com>
> 
> The function unmap_hugepage_range() has two kinds of users:
> 1) unmap_ref_private(), which passes in the head page of a folio.  Since
>    unmap_ref_private() already takes folio and there are no other uses
>    of the folio struct in the function, it is natural for
>    unmap_hugepage_range() to take folio also.
> 2) All other uses, which pass in NULL pointer.
> 
> In both cases, we can pass in folio. Refactor unmap_hugepage_range() to
> take folio.

It looks like unmap_ref_private() is the only caller that cares about
passing a particular folio to unmap_hugepage_range(). Is there any
reason we shouldn't drop the folio argument and call
__unmap_hugepage_range() directly? 

> Signed-off-by: Fan Ni <fan.ni@samsung.com>
> Reviewed-by: Muchun Song <muchun.song@linux.dev>
> Reviewed-by: Sidhartha Kumar <sidhartha.kumar@oracle.com>
> Reviewed-by: Oscar Salvador <osalvador@suse.de>
> Acked-by: David Hildenbrand <david@redhat.com>
> ---
>  include/linux/hugetlb.h | 4 ++--
>  mm/hugetlb.c            | 7 ++++---
>  2 files changed, 6 insertions(+), 5 deletions(-)
> 
> diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h
> index 23ebf49c5d6a..f6d5f24e793c 100644
> --- a/include/linux/hugetlb.h
> +++ b/include/linux/hugetlb.h
> @@ -129,8 +129,8 @@ int move_hugetlb_page_tables(struct vm_area_struct *vma,
>  int copy_hugetlb_page_range(struct mm_struct *, struct mm_struct *,
>  			    struct vm_area_struct *, struct vm_area_struct *);
>  void unmap_hugepage_range(struct vm_area_struct *,
> -			  unsigned long, unsigned long, struct page *,
> -			  zap_flags_t);
> +			  unsigned long start, unsigned long end,
> +			  struct folio *, zap_flags_t);
>  void __unmap_hugepage_range(struct mmu_gather *tlb,
>  			  struct vm_area_struct *vma,
>  			  unsigned long start, unsigned long end,
> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> index 0c2b264a7ab8..c339ffe05556 100644
> --- a/mm/hugetlb.c
> +++ b/mm/hugetlb.c
> @@ -6046,7 +6046,7 @@ void __hugetlb_zap_end(struct vm_area_struct *vma,
>  }
>  
>  void unmap_hugepage_range(struct vm_area_struct *vma, unsigned long start,
> -			  unsigned long end, struct page *ref_page,
> +			  unsigned long end, struct folio *folio,
>  			  zap_flags_t zap_flags)
>  {
>  	struct mmu_notifier_range range;
> @@ -6058,7 +6058,8 @@ void unmap_hugepage_range(struct vm_area_struct *vma, unsigned long start,
>  	mmu_notifier_invalidate_range_start(&range);
>  	tlb_gather_mmu(&tlb, vma->vm_mm);
>  
> -	__unmap_hugepage_range(&tlb, vma, start, end, ref_page, zap_flags);
> +	__unmap_hugepage_range(&tlb, vma, start, end,
> +			       &folio->page, zap_flags);
>  
>  	mmu_notifier_invalidate_range_end(&range);
>  	tlb_finish_mmu(&tlb);
> @@ -6116,7 +6117,7 @@ static void unmap_ref_private(struct mm_struct *mm, struct vm_area_struct *vma,
>  		if (!is_vma_resv_set(iter_vma, HPAGE_RESV_OWNER))
>  			unmap_hugepage_range(iter_vma, address,
>  					     address + huge_page_size(h),
> -					     &folio->page, 0);
> +					     folio, 0);
>  	}
>  	i_mmap_unlock_write(mapping);
>  }
> -- 
> 2.47.2
>
Re: [PATCH v4 2/4] mm/hugetlb: Refactor unmap_hugepage_range() to take folio instead of page
Posted by Andrew Morton 6 months, 3 weeks ago 
On Mon, 5 May 2025 12:59:20 -0700 "Vishal Moola (Oracle)" <vishal.moola@gmail.com> wrote:

> On Mon, May 05, 2025 at 11:22:42AM -0700, nifan.cxl@gmail.com wrote:
> > From: Fan Ni <fan.ni@samsung.com>
> > 
> > The function unmap_hugepage_range() has two kinds of users:
> > 1) unmap_ref_private(), which passes in the head page of a folio.  Since
> >    unmap_ref_private() already takes folio and there are no other uses
> >    of the folio struct in the function, it is natural for
> >    unmap_hugepage_range() to take folio also.
> > 2) All other uses, which pass in NULL pointer.
> > 
> > In both cases, we can pass in folio. Refactor unmap_hugepage_range() to
> > take folio.
> 
> It looks like unmap_ref_private() is the only caller that cares about
> passing a particular folio to unmap_hugepage_range(). Is there any
> reason we shouldn't drop the folio argument and call
> __unmap_hugepage_range() directly? 

afaict there was no response to this review comment.

I'll proceed with the patchset, but please let's not lose sight of this.

Patchew 2.1-devel-b67e4c2

© 2016 - 2025 Red Hat, Inc.