1. Patchew
  2. linux
  3. [PATCH v2 0/2] mm/vmalloc: free unused pages on vrealloc() shrink
  4. View patch

[PATCH v2 1/2] mm/vmalloc: extract vmalloc_free_pages() helper from vfree()

Shivam Kalra via B4 Relay posted 2 patches 1 month, 1 week ago
There is a newer version of this series
[PATCH v2 1/2] mm/vmalloc: extract vmalloc_free_pages() helper from vfree()
Posted by Shivam Kalra via B4 Relay 1 month, 1 week ago 
From: Shivam Kalra <shivamkalra98@zohomail.in>

Extract the page-freeing loop and NR_VMALLOC stat accounting from
vfree() into a reusable vmalloc_free_pages() helper. The helper operates
on a range [start, end) of pages from a vm_struct, making it suitable
for both full free (vfree) and partial free (upcoming vrealloc shrink).

No functional change.

Signed-off-by: Shivam Kalra <shivamkalra98@zohomail.in>
---
 mm/vmalloc.c | 42 ++++++++++++++++++++++++++++--------------
 1 file changed, 28 insertions(+), 14 deletions(-)

diff --git a/mm/vmalloc.c b/mm/vmalloc.c
index c607307c657a..e2aef0a79f2e 100644
--- a/mm/vmalloc.c
+++ b/mm/vmalloc.c
@@ -3416,6 +3416,32 @@ void vfree_atomic(const void *addr)
 		schedule_work(&p->wq);
 }
 
+/*
+ * vmalloc_free_pages - free a range of pages from a vmalloc allocation
+ * @vm: the vm_struct containing the pages
+ * @start: first page index to free (inclusive)
+ * @end: last page index to free (exclusive)
+ *
+ * Free pages [start, end) updating NR_VMALLOC stat accounting.
+ * Caller is responsible for unmapping (vunmap_range) and KASAN
+ * poisoning before calling this.
+ */
+static void vmalloc_free_pages(struct vm_struct *vm, unsigned int start,
+			       unsigned int end)
+{
+	unsigned int i;
+
+	for (i = start; i < end; i++) {
+		struct page *page = vm->pages[i];
+
+		BUG_ON(!page);
+		if (!(vm->flags & VM_MAP_PUT_PAGES))
+			mod_lruvec_page_state(page, NR_VMALLOC, -1);
+		__free_page(page);
+		cond_resched();
+	}
+}
+
 /**
  * vfree - Release memory allocated by vmalloc()
  * @addr:  Memory base address
@@ -3436,7 +3462,6 @@ void vfree_atomic(const void *addr)
 void vfree(const void *addr)
 {
 	struct vm_struct *vm;
-	int i;
 
 	if (unlikely(in_interrupt())) {
 		vfree_atomic(addr);
@@ -3459,19 +3484,8 @@ void vfree(const void *addr)
 
 	if (unlikely(vm->flags & VM_FLUSH_RESET_PERMS))
 		vm_reset_perms(vm);
-	for (i = 0; i < vm->nr_pages; i++) {
-		struct page *page = vm->pages[i];
-
-		BUG_ON(!page);
-		/*
-		 * High-order allocs for huge vmallocs are split, so
-		 * can be freed as an array of order-0 allocations
-		 */
-		if (!(vm->flags & VM_MAP_PUT_PAGES))
-			mod_lruvec_page_state(page, NR_VMALLOC, -1);
-		__free_page(page);
-		cond_resched();
-	}
+	if (vm->nr_pages)
+		vmalloc_free_pages(vm, 0, vm->nr_pages);
 	kvfree(vm->pages);
 	kfree(vm);
 }

-- 
2.43.0
Re: [PATCH v2 1/2] mm/vmalloc: extract vmalloc_free_pages() helper from vfree()
Posted by Uladzislau Rezki 1 month ago 
On Wed, Mar 04, 2026 at 08:23:15PM +0530, Shivam Kalra via B4 Relay wrote:
> From: Shivam Kalra <shivamkalra98@zohomail.in>
> 
> Extract the page-freeing loop and NR_VMALLOC stat accounting from
> vfree() into a reusable vmalloc_free_pages() helper. The helper operates
> on a range [start, end) of pages from a vm_struct, making it suitable
> for both full free (vfree) and partial free (upcoming vrealloc shrink).
> 
> No functional change.
> 
> Signed-off-by: Shivam Kalra <shivamkalra98@zohomail.in>
> ---
>  mm/vmalloc.c | 42 ++++++++++++++++++++++++++++--------------
>  1 file changed, 28 insertions(+), 14 deletions(-)
> 
> diff --git a/mm/vmalloc.c b/mm/vmalloc.c
> index c607307c657a..e2aef0a79f2e 100644
> --- a/mm/vmalloc.c
> +++ b/mm/vmalloc.c
> @@ -3416,6 +3416,32 @@ void vfree_atomic(const void *addr)
>  		schedule_work(&p->wq);
>  }
>  
> +/*
> + * vmalloc_free_pages - free a range of pages from a vmalloc allocation
> + * @vm: the vm_struct containing the pages
> + * @start: first page index to free (inclusive)
> + * @end: last page index to free (exclusive)
> + *
> + * Free pages [start, end) updating NR_VMALLOC stat accounting.
> + * Caller is responsible for unmapping (vunmap_range) and KASAN
> + * poisoning before calling this.
> + */
> +static void vmalloc_free_pages(struct vm_struct *vm, unsigned int start,
> +			       unsigned int end)
> +{
> +	unsigned int i;
> +
> +	for (i = start; i < end; i++) {
> +		struct page *page = vm->pages[i];
> +
> +		BUG_ON(!page);
> +		if (!(vm->flags & VM_MAP_PUT_PAGES))
> +			mod_lruvec_page_state(page, NR_VMALLOC, -1);
> +		__free_page(page);
> +		cond_resched();
> +	}
> +}
> +
>  /**
>   * vfree - Release memory allocated by vmalloc()
>   * @addr:  Memory base address
> @@ -3436,7 +3462,6 @@ void vfree_atomic(const void *addr)
>  void vfree(const void *addr)
>  {
>  	struct vm_struct *vm;
> -	int i;
>  
>  	if (unlikely(in_interrupt())) {
>  		vfree_atomic(addr);
> @@ -3459,19 +3484,8 @@ void vfree(const void *addr)
>  
>  	if (unlikely(vm->flags & VM_FLUSH_RESET_PERMS))
>  		vm_reset_perms(vm);
> -	for (i = 0; i < vm->nr_pages; i++) {
> -		struct page *page = vm->pages[i];
> -
> -		BUG_ON(!page);
> -		/*
> -		 * High-order allocs for huge vmallocs are split, so
> -		 * can be freed as an array of order-0 allocations
> -		 */
> -		if (!(vm->flags & VM_MAP_PUT_PAGES))
> -			mod_lruvec_page_state(page, NR_VMALLOC, -1);
> -		__free_page(page);
> -		cond_resched();
> -	}
> +	if (vm->nr_pages)
> +		vmalloc_free_pages(vm, 0, vm->nr_pages);
>  	kvfree(vm->pages);
>
probably we should move "if (vm->nr_pages)" condition inside the
vmalloc_free_pages(). 

I think, the function name should also be renamed to something like
vm_area_free_pages() so we align with vm_area_alloc_pages() helper.

--
Uladzislau Rezki
Re: [PATCH v2 1/2] mm/vmalloc: extract vmalloc_free_pages() helper from vfree()
Posted by Alice Ryhl 1 month, 1 week ago 
On Wed, Mar 04, 2026 at 08:23:15PM +0530, Shivam Kalra via B4 Relay wrote:
> From: Shivam Kalra <shivamkalra98@zohomail.in>
> 
> Extract the page-freeing loop and NR_VMALLOC stat accounting from
> vfree() into a reusable vmalloc_free_pages() helper. The helper operates
> on a range [start, end) of pages from a vm_struct, making it suitable
> for both full free (vfree) and partial free (upcoming vrealloc shrink).
> 
> No functional change.
> 
> Signed-off-by: Shivam Kalra <shivamkalra98@zohomail.in>

> -		/*
> -		 * High-order allocs for huge vmallocs are split, so
> -		 * can be freed as an array of order-0 allocations
> -		 */
> -		if (!(vm->flags & VM_MAP_PUT_PAGES))
> -			mod_lruvec_page_state(page, NR_VMALLOC, -1);

Any reason this comment was dropped?

Alice
Re: [PATCH v2 1/2] mm/vmalloc: extract vmalloc_free_pages() helper from vfree()
Posted by Shivam Kalra 1 month, 1 week ago 
On 05/03/26 03:14, Alice Ryhl wrote:
> On Wed, Mar 04, 2026 at 08:23:15PM +0530, Shivam Kalra via B4 Relay wrote:
>> From: Shivam Kalra <shivamkalra98@zohomail.in>
>>
>> Extract the page-freeing loop and NR_VMALLOC stat accounting from
>> vfree() into a reusable vmalloc_free_pages() helper. The helper operates
>> on a range [start, end) of pages from a vm_struct, making it suitable
>> for both full free (vfree) and partial free (upcoming vrealloc shrink).
>>
>> No functional change.
>>
>> Signed-off-by: Shivam Kalra <shivamkalra98@zohomail.in>
> 
>> -		/*
>> -		 * High-order allocs for huge vmallocs are split, so
>> -		 * can be freed as an array of order-0 allocations
>> -		 */
>> -		if (!(vm->flags & VM_MAP_PUT_PAGES))
>> -			mod_lruvec_page_state(page, NR_VMALLOC, -1);
> 
> Any reason this comment was dropped?
> 
> Alice
Great catch. This regressed during the rebase. I will add this back in
the next version. However, I'll wait for more feedback before the next
version.

Patchew 2.1-devel-ea86937

© 2016 - 2026 Red Hat, Inc.