[PATCH] mm/mlock: Use vma_lookup() instead of find_vma()

[PATCH] mm/mlock: Use vma_lookup() instead of find_vma()

[Miaohe Lin]

Using vma_lookup() verifies the start is contained in the found vma. This
results in easier to read the code.

Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
---
 mm/mlock.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/mm/mlock.c b/mm/mlock.c
index efd2dd2943de..fefa9644d0f9 100644
--- a/mm/mlock.c
+++ b/mm/mlock.c
@@ -453,8 +453,8 @@ static int apply_vma_lock_flags(unsigned long start, size_t len,
 		return -EINVAL;
 	if (end == start)
 		return 0;
-	vma = find_vma(current->mm, start);
-	if (!vma || vma->vm_start > start)
+	vma = vma_lookup(current->mm, start);
+	if (!vma)
 		return -ENOMEM;
 
 	prev = vma->vm_prev;
-- 
2.23.0

Re: [PATCH] mm/mlock: Use vma_lookup() instead of find_vma()

[Matthew Wilcox]

On Mon, Mar 14, 2022 at 11:17:28PM +0800, Miaohe Lin wrote:
> Using vma_lookup() verifies the start is contained in the found vma. This
> results in easier to read the code.

This conflicts with the maple tree.

Re: [PATCH] mm/mlock: Use vma_lookup() instead of find_vma()

[Miaohe Lin]

On 2022/3/15 0:16, Matthew Wilcox wrote:
> On Mon, Mar 14, 2022 at 11:17:28PM +0800, Miaohe Lin wrote:
>> Using vma_lookup() verifies the start is contained in the found vma. This
>> results in easier to read the code.
> 
> This conflicts with the maple tree.

IIUC, maple tree is the implementation detail of the vma and this patch
should not be conflict with it. But there might be some trival conflict.
Am I supposed to delay this patch until the maple tree's work is done?

Thanks.

> 
> .
> 

Re: [PATCH] mm/mlock: Use vma_lookup() instead of find_vma()

[Matthew Wilcox]

On Tue, Mar 15, 2022 at 08:24:04PM +0800, Miaohe Lin wrote:
> On 2022/3/15 0:16, Matthew Wilcox wrote:
> > On Mon, Mar 14, 2022 at 11:17:28PM +0800, Miaohe Lin wrote:
> >> Using vma_lookup() verifies the start is contained in the found vma. This
> >> results in easier to read the code.
> > 
> > This conflicts with the maple tree.
> 
> IIUC, maple tree is the implementation detail of the vma and this patch
> should not be conflict with it. But there might be some trival conflict.
> Am I supposed to delay this patch until the maple tree's work is done?

This function is rewritten as part of the maple tree patchset.
There's no point in merging it now, or later.

Re: [PATCH] mm/mlock: Use vma_lookup() instead of find_vma()

[Miaohe Lin]

On 2022/3/15 20:39, Matthew Wilcox wrote:
> On Tue, Mar 15, 2022 at 08:24:04PM +0800, Miaohe Lin wrote:
>> On 2022/3/15 0:16, Matthew Wilcox wrote:
>>> On Mon, Mar 14, 2022 at 11:17:28PM +0800, Miaohe Lin wrote:
>>>> Using vma_lookup() verifies the start is contained in the found vma. This
>>>> results in easier to read the code.
>>>
>>> This conflicts with the maple tree.
>>
>> IIUC, maple tree is the implementation detail of the vma and this patch
>> should not be conflict with it. But there might be some trival conflict.
>> Am I supposed to delay this patch until the maple tree's work is done?
> 
> This function is rewritten as part of the maple tree patchset.
> There's no point in merging it now, or later.
> 

I see. Thanks for explanation.

> .
>