Skip to contentSkip to conversation

mm/page_alloc: simplify kmem cgroup charge/uncharge code

PATCH In review Hui Su 4 participants 7 messages Dec 7, 2020 – Dec 9, 2020 1 message not in the archive

Subsystems (2)
lore
Share & tools
lore

Hui Su

author
#0 lore raw

PATCH 1/1 mm/page_alloc: simplify kmem cgroup charge/uncharge code

Since the commit 60cd4bcd6238 ("memcg: localize memcg_kmem_enabled() check"), we have supplied the api which users don't have to explicitly check memcg_kmem_enabled().

---
 mm/page_alloc.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
mm/page_alloc.c +6 -6
index eaa227a479e4..dc990a899ded 100644
@@ -1214,8 +1214,8 @@ static __always_inline bool free_pages_prepare(struct page *page,
 		 * Do not let hwpoison pages hit pcplists/buddy
 		 * Untie memcg state and reset page's owner
 		 */
-		if (memcg_kmem_enabled() && PageKmemcg(page))
-			__memcg_kmem_uncharge_page(page, order);
+		if (PageKmemcg(page))
+			memcg_kmem_uncharge_page(page, order);
 		reset_page_owner(page, order);
 		return false;
 	}
@@ -1244,8 +1244,8 @@ static __always_inline bool free_pages_prepare(struct page *page,
 	}
 	if (PageMappingFlags(page))
 		page->mapping = NULL;
-	if (memcg_kmem_enabled() && PageKmemcg(page))
-		__memcg_kmem_uncharge_page(page, order);
+	if (PageKmemcg(page))
+		memcg_kmem_uncharge_page(page, order);
 	if (check_free)
 		bad += check_free_page(page);
 	if (bad)
@@ -4965,8 +4965,8 @@ __alloc_pages_nodemask(gfp_t gfp_mask, unsigned int order, int preferred_nid,
 	page = __alloc_pages_slowpath(alloc_mask, order, &ac);
 
 out:
-	if (memcg_kmem_enabled() && (gfp_mask & __GFP_ACCOUNT) && page &&
-	    unlikely(__memcg_kmem_charge_page(page, gfp_mask, order) != 0)) {
+	if ((gfp_mask & __GFP_ACCOUNT) && page &&
+	    unlikely(memcg_kmem_charge_page(page, gfp_mask, order) != 0)) {
 		__free_pages(page, order);
 		page = NULL;
 	}
Signature · 1 line
2.29.2
Reviewed-by on v1 1/1
#1 lore raw
Quoted 6 lines · Hui Su

Since the commit 60cd4bcd6238 ("memcg: localize memcg_kmem_enabled() check"), we have supplied the api which users don't have to explicitly check memcg_kmem_enabled().

Quoted 45 lines
---
 mm/page_alloc.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
mm/page_alloc.c +6 -7
index eaa227a479e4..dc990a899ded 100644
@@ -1214,8 +1214,8 @@ static __always_inline bool free_pages_prepare(struct page *page,
                 * Do not let hwpoison pages hit pcplists/buddy
                 * Untie memcg state and reset page's owner
                 */
-               if (memcg_kmem_enabled() && PageKmemcg(page))
-                       __memcg_kmem_uncharge_page(page, order);
+               if (PageKmemcg(page))
+                       memcg_kmem_uncharge_page(page, order);
                reset_page_owner(page, order);
                return false;
        }
@@ -1244,8 +1244,8 @@ static __always_inline bool free_pages_prepare(struct page *page,
        }
        if (PageMappingFlags(page))
                page->mapping = NULL;
-       if (memcg_kmem_enabled() && PageKmemcg(page))
-               __memcg_kmem_uncharge_page(page, order);
+       if (PageKmemcg(page))
+               memcg_kmem_uncharge_page(page, order);
        if (check_free)
                bad += check_free_page(page);
        if (bad)
@@ -4965,8 +4965,8 @@ __alloc_pages_nodemask(gfp_t gfp_mask, unsigned int order, int preferred_nid,
        page = __alloc_pages_slowpath(alloc_mask, order, &ac);
 
 out:
-       if (memcg_kmem_enabled() && (gfp_mask & __GFP_ACCOUNT) && page &&
-           unlikely(__memcg_kmem_charge_page(page, gfp_mask, order) != 0)) {
+       if ((gfp_mask & __GFP_ACCOUNT) && page &&
+           unlikely(memcg_kmem_charge_page(page, gfp_mask, order) != 0)) {
                __free_pages(page, order);
                page = NULL;
        }
--

2.29.2

Signature · 2 lines
Yours,
Muchun
#2 lore raw
In reply to Hui Su
Quoted 37 lines · Hui Su

Since the commit 60cd4bcd6238 ("memcg: localize memcg_kmem_enabled() check"), we have supplied the api which users don't have to explicitly check memcg_kmem_enabled().

---
 mm/page_alloc.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
mm/page_alloc.c +4 -4
index eaa227a479e4..dc990a899ded 100644
@@ -1214,8 +1214,8 @@ static __always_inline bool free_pages_prepare(struct page *page,
                 * Do not let hwpoison pages hit pcplists/buddy
                 * Untie memcg state and reset page's owner
                 */
-               if (memcg_kmem_enabled() && PageKmemcg(page))
-                       __memcg_kmem_uncharge_page(page, order);
+               if (PageKmemcg(page))
+                       memcg_kmem_uncharge_page(page, order);
                reset_page_owner(page, order);
                return false;
        }
@@ -1244,8 +1244,8 @@ static __always_inline bool free_pages_prepare(struct page *page,
        }
        if (PageMappingFlags(page))
                page->mapping = NULL;
-       if (memcg_kmem_enabled() && PageKmemcg(page))
-               __memcg_kmem_uncharge_page(page, order);
+       if (PageKmemcg(page))
+               memcg_kmem_uncharge_page(page, order);
        if (check_free)
                bad += check_free_page(page);
        if (bad)

mm/page_alloc.c __alloc_pages_nodemask(gfp_t gfp_mask, unsigned int order, int preferred_nid, · line 4972

Quoted 10 lines
mm/page_alloc.c +2 -2
@@ -4965,8 +4965,8 @@ __alloc_pages_nodemask(gfp_t gfp_mask, unsigned int order, int preferred_nid,
        page = __alloc_pages_slowpath(alloc_mask, order, &ac);
 
 out:
-       if (memcg_kmem_enabled() && (gfp_mask & __GFP_ACCOUNT) && page &&
-           unlikely(__memcg_kmem_charge_page(page, gfp_mask, order) != 0)) {
+       if ((gfp_mask & __GFP_ACCOUNT) && page &&
+           unlikely(memcg_kmem_charge_page(page, gfp_mask, order) != 0)) {
                __free_pages(page, order);
                page = NULL;
        }

The reason to keep __memcg_kmem_[un]charge_page functions is that they were called in the very hot path. Can you please check the performance impact of your change and if the generated code is actually same or different.

author
#3 lore raw
Quoted 6 lines · Shakeel Butt

On Mon, Dec 7, 2020 at 6:22 AM Hui Su <[email protected]> wrote:

The reason to keep __memcg_kmem_[un]charge_page functions is that they were called in the very hot path. Can you please check the performance impact of your change and if the generated code is actually same or different.

Hi, Shakeel:

I objdump the mm/page_alloc.o and comapre them, it change the assemble code indeed. In fact, it change some code order, which i personally think won't have impact on performance. And i ran the ltp mm and conatiner test, it seems nothing abnormal.

BUT i still want to check whether this change will have negative impact on perforance due to this change code was called in the very hot path like you said, AND saddly i did not find a way to quantify the impact on performance. Can you give me some suggestion about how to quantify the performance or some tool?

Thanks.

#5 lore raw
In reply to a message not in the archive

+Michal Hocko

Message starts at https://lkml.kernel.org/r/20201207142204.GA18516@rlk

Quoted 9 lines · Hui Su
Quoted 6 lines · Shakeel Butt

On Mon, Dec 7, 2020 at 6:22 AM Hui Su <[email protected]> wrote:

The reason to keep __memcg_kmem_[un]charge_page functions is that they were called in the very hot path. Can you please check the performance impact of your change and if the generated code is actually same or different.

Quoted 6 lines

Hi, Shakeel:

I objdump the mm/page_alloc.o and comapre them, it change the assemble code indeed. In fact, it change some code order, which i personally think won't have impact on performance. And i ran the ltp mm and conatiner test, it seems nothing abnormal.

Did you run the tests in a memcg? The change is behind a static key of kmem accounting which is enabled for subcontainers.

Quoted 7 lines

BUT i still want to check whether this change will have negative impact on perforance due to this change code was called in the very hot path like you said, AND saddly i did not find a way to quantify the impact on performance. Can you give me some suggestion about how to quantify the performance or some tool?

At least I think we can try with a simple page allocation in a loop i.e. alloc_page(GFP_KERNEL_ACCOUNT). I will think of any existing benchmark which exercises this code path.

Michal, do you have any suggestions?

reviewer
#6 lore raw
Quoted 32 lines · Shakeel Butt

+Michal Hocko

Message starts at https://lkml.kernel.org/r/20201207142204.GA18516@rlk

Quoted 9 lines · Hui Su
Quoted 6 lines · Shakeel Butt

On Mon, Dec 7, 2020 at 6:22 AM Hui Su <[email protected]> wrote:

The reason to keep __memcg_kmem_[un]charge_page functions is that they were called in the very hot path. Can you please check the performance impact of your change and if the generated code is actually same or different.

Quoted 6 lines

Hi, Shakeel:

I objdump the mm/page_alloc.o and comapre them, it change the assemble code indeed. In fact, it change some code order, which i personally think won't have impact on performance. And i ran the ltp mm and conatiner test, it seems nothing abnormal.

Did you run the tests in a memcg? The change is behind a static key of kmem accounting which is enabled for subcontainers.

Quoted 7 lines

BUT i still want to check whether this change will have negative impact on perforance due to this change code was called in the very hot path like you said, AND saddly i did not find a way to quantify the impact on performance. Can you give me some suggestion about how to quantify the performance or some tool?

Quoted 5 lines

At least I think we can try with a simple page allocation in a loop i.e. alloc_page(GFP_KERNEL_ACCOUNT). I will think of any existing benchmark which exercises this code path.

Michal, do you have any suggestions?

I have to say I do not see any big benefit from the patch and it alters a real hot path to check for the flag even in cases where kmem accounting is not enabled, unless I am misreading the code.

Signature · 2 lines
Michal Hocko
SUSE Labs
#7 lore raw
Quoted 38 lines · Michal Hocko
Quoted 36 lines · Shakeel Butt

+Michal Hocko

Message starts at https://lkml.kernel.org/r/20201207142204.GA18516@rlk

Quoted 9 lines · Hui Su
Quoted 6 lines · Shakeel Butt

On Mon, Dec 7, 2020 at 6:22 AM Hui Su <[email protected]> wrote:

The reason to keep __memcg_kmem_[un]charge_page functions is that they were called in the very hot path. Can you please check the performance impact of your change and if the generated code is actually same or different.

Quoted 6 lines

Hi, Shakeel:

I objdump the mm/page_alloc.o and comapre them, it change the assemble code indeed. In fact, it change some code order, which i personally think won't have impact on performance. And i ran the ltp mm and conatiner test, it seems nothing abnormal.

Did you run the tests in a memcg? The change is behind a static key of kmem accounting which is enabled for subcontainers.

Quoted 7 lines

BUT i still want to check whether this change will have negative impact on perforance due to this change code was called in the very hot path like you said, AND saddly i did not find a way to quantify the impact on performance. Can you give me some suggestion about how to quantify the performance or some tool?

At least I think we can try with a simple page allocation in a loop i.e. alloc_page(GFP_KERNEL_ACCOUNT). I will think of any existing benchmark which exercises this code path.

Quoted 6 lines
Quoted 1 line

Michal, do you have any suggestions?

I have to say I do not see any big benefit from the patch and it alters a real hot path to check for the flag even in cases where kmem accounting is not enabled, unless I am misreading the code.

Yes you are right unless the super intelligent compiler re-arranges the checks and puts the static key check at front to optimize the non-kmem-accounting mode.