[PATCH 1/2] mm/mmap: rename __vma_unlink_common() to __vma_unlink()

[PATCH 1/2] mm/mmap: rename __vma_unlink_common() to __vma_unlink()

[Wei Yang]

__vma_unlink_common() and __vma_unlink() are counterparts. Since there is
not function named __vma_unlink(), let's rename it to __vma_unlink() to
make the code more self-explain and easy for audience to understand.

Otherwise we may expect there are several variants of vma_unlink and
__vma_unlink_common() is used by them.

Signed-off-by: Wei Yang <richard.weiyang@linux.alibaba.com>
---
 mm/mmap.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/mm/mmap.c b/mm/mmap.c
index dcdab2675a21..ec780925da94 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -677,7 +677,7 @@ static void __insert_vm_struct(struct mm_struct *mm, struct vm_area_struct *vma)
 	mm->map_count++;
 }
 
-static __always_inline void __vma_unlink_common(struct mm_struct *mm,
+static __always_inline void __vma_unlink(struct mm_struct *mm,
 						struct vm_area_struct *vma,
 						struct vm_area_struct *ignore)
 {
@@ -859,7 +859,7 @@ int __vma_adjust(struct vm_area_struct *vma, unsigned long start,
 		 * us to remove next before dropping the locks.
 		 */
 		if (remove_next != 3)
-			__vma_unlink_common(mm, next, next);
+			__vma_unlink(mm, next, next);
 		else
 			/*
 			 * vma is not before next if they've been
@@ -870,7 +870,7 @@ int __vma_adjust(struct vm_area_struct *vma, unsigned long start,
 			 * "next" (which is stored in post-swap()
 			 * "vma").
 			 */
-			__vma_unlink_common(mm, next, vma);
+			__vma_unlink(mm, next, vma);
 		if (file)
 			__remove_shared_vm_struct(next, file, mapping);
 	} else if (insert) {
-- 
2.20.1 (Apple Git-117)

[PATCH 2/2] mm/mmap: leverage vma_rb_erase_ignore() to implement vma_rb_erase()

[Wei Yang]

These two functions share the same logic except ignore a different vma.

Let's reuse the code.

Signed-off-by: Wei Yang <richard.weiyang@linux.alibaba.com>
---
 mm/mmap.c | 16 +++++++---------
 1 file changed, 7 insertions(+), 9 deletions(-)

diff --git a/mm/mmap.c b/mm/mmap.c
index ec780925da94..90b1298d4222 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -474,8 +474,12 @@ static __always_inline void vma_rb_erase_ignore(struct vm_area_struct *vma,
 {
 	/*
 	 * All rb_subtree_gap values must be consistent prior to erase,
-	 * with the possible exception of the "next" vma being erased if
-	 * next->vm_start was reduced.
+	 * with the possible exception of
+	 *
+	 * a. the "next" vma being erased if next->vm_start was reduced in
+	 *    __vma_adjust() -> __vma_unlink()
+	 * b. the vma being erased in detach_vmas_to_be_unmapped() ->
+	 *    vma_rb_erase()
 	 */
 	validate_mm_rb(root, ignore);
 
@@ -485,13 +489,7 @@ static __always_inline void vma_rb_erase_ignore(struct vm_area_struct *vma,
 static __always_inline void vma_rb_erase(struct vm_area_struct *vma,
 					 struct rb_root *root)
 {
-	/*
-	 * All rb_subtree_gap values must be consistent prior to erase,
-	 * with the possible exception of the vma being erased.
-	 */
-	validate_mm_rb(root, vma);
-
-	__vma_rb_erase(vma, root);
+	vma_rb_erase_ignore(vma, root, vma);
 }
 
 /*
-- 
2.20.1 (Apple Git-117)

Re: [PATCH 1/2] mm/mmap: rename __vma_unlink_common() to __vma_unlink()

[Andrew Morton]

On Mon, 10 Aug 2020 07:20:56 +0800 Wei Yang <richard.weiyang@linux.alibaba.com> wrote:

> __vma_unlink_common() and __vma_unlink() are counterparts. Since there is

I assume you meant "__vma_link()" here?

> not function named __vma_unlink(), let's rename it to __vma_unlink() to
> make the code more self-explain and easy for audience to understand.
> 
> Otherwise we may expect there are several variants of vma_unlink and
> __vma_unlink_common() is used by them.

Re: [PATCH 1/2] mm/mmap: rename __vma_unlink_common() to __vma_unlink()

[Wei Yang]

On Wed, Aug 19, 2020 at 01:12:10PM -0700, Andrew Morton wrote:
>On Mon, 10 Aug 2020 07:20:56 +0800 Wei Yang <richard.weiyang@linux.alibaba.com> wrote:
>
>> __vma_unlink_common() and __vma_unlink() are counterparts. Since there is
>
>I assume you meant "__vma_link()" here?
>

Oops, my fault. You are right.

Do you prefer a v2, or you would like to fix it online?

>> not function named __vma_unlink(), let's rename it to __vma_unlink() to
>> make the code more self-explain and easy for audience to understand.
>> 
>> Otherwise we may expect there are several variants of vma_unlink and
>> __vma_unlink_common() is used by them.

-- 
Wei Yang
Help you, Help me