slub: Fix calculation of cpu slabs

PATCH In review Li Zefan 4 participants 5 messages Sep 10, 2013 – Dec 29, 2013

Subsystems (2)

Memory management Slab allocator

lore

Share & tools

lore

Li Zefan

author

2013-09-10 03:43 UTC #0 lore raw

PATCH 1/1 slub: Fix calculation of cpu slabs

/sys/kernel/slab/:t-0000048 # cat cpu_slabs
231 N0=16 N1=215
/sys/kernel/slab/:t-0000048 # cat slabs
145 N0=36 N1=109

See, the number of slabs is smaller than that of cpu slabs.

The bug was introduced by commit 49e2258586b423684f03c278149ab46d8f8b6700 ("slub: per cpu cache for partial pages").

We should use page->pages instead of page->pobjects when calculating the number of cpu partial slabs. This also fixes the mapping of slabs and nodes.

As there's no variable storing the number of total/active objects in cpu partial slabs, and we don't have user interfaces requiring those statistics, I just add WARN_ON for those cases.

Cc <[email protected]> # 3.2+

Signed-off-by Li Zefan <[email protected]>

---
 mm/slub.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

mm/slub.c +7 -1

index e3ba1f2..6ea461d 100644
@@ -4300,7 +4300,13 @@ static ssize_t show_slab_objects(struct kmem_cache *s,
 
 			page = ACCESS_ONCE(c->partial);
 			if (page) {
-				x = page->pobjects;
+				node = page_to_nid(page);
+				if (flags & SO_TOTAL)
+					WARN_ON_ONCE(1);
+				else if (flags & SO_OBJECTS)
+					WARN_ON_ONCE(1);
+				else
+					x = page->pages;
 				total += x;
 				nodes[node] += x;
 			}

Signature · 7 lines

1.8.0.2

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to [email protected].  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>

Christoph Lameter

Acked-by on v1 1/1

2013-09-10 21:06 UTC #1 lore raw

Quoted 3 lines · Li Zefan

We should use page->pages instead of page->pobjects when calculating the number of cpu partial slabs. This also fixes the mapping of slabs and nodes.

True.

Quoted 3 lines

As there's no variable storing the number of total/active objects in cpu partial slabs, and we don't have user interfaces requiring those statistics, I just add WARN_ON for those cases.

Well that is not strictly required but it does not hurt either.

Acked-by Christoph Lameter <[email protected]>

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>

Li Zefan

author

2013-10-14 07:28 UTC #2 lore raw

Hi Pekka,

could you pick up this patch?

Quoted 8 lines · Christoph Lameter

Quoted 3 lines · Li Zefan
We should use page->pages instead of page->pobjects when calculating the number of cpu partial slabs. This also fixes the mapping of slabs and nodes.
True.

Quoted 9 lines

Quoted 3 lines
As there's no variable storing the number of total/active objects in cpu partial slabs, and we don't have user interfaces requiring those statistics, I just add WARN_ON for those cases.
Well that is not strictly required but it does not hurt either.
Acked-by Christoph Lameter <[email protected]>

Wanpeng Li

Reviewed-by on v1 1/1

2013-10-05 07:48 UTC #3 lore raw

In reply to Li Zefan

Quoted 14 lines · Li Zefan

/sys/kernel/slab/:t-0000048 # cat cpu_slabs
231 N0=16 N1=215
/sys/kernel/slab/:t-0000048 # cat slabs
145 N0=36 N1=109
See, the number of slabs is smaller than that of cpu slabs.
The bug was introduced by commit 49e2258586b423684f03c278149ab46d8f8b6700 ("slub: per cpu cache for partial pages").
We should use page->pages instead of page->pobjects when calculating the number of cpu partial slabs. This also fixes the mapping of slabs and nodes.

Quoted 6 lines

As there's no variable storing the number of total/active objects in cpu partial slabs, and we don't have user interfaces requiring those statistics, I just add WARN_ON for those cases.
Cc <[email protected]> # 3.2+
Signed-off-by Li Zefan <[email protected]>

Reviewed-by Wanpeng Li <[email protected]>

Quoted 31 lines

---
mm/slub.c | 8 +++++++-

1 file changed, 7 insertions(+), 1 deletion(-)

mm/slub.c +7 -1

index e3ba1f2..6ea461d 100644
@@ -4300,7 +4300,13 @@ static ssize_t show_slab_objects(struct kmem_cache *s,
 
 			page = ACCESS_ONCE(c->partial);
 			if (page) {
-				x = page->pobjects;
+				node = page_to_nid(page);
+				if (flags & SO_TOTAL)
+					WARN_ON_ONCE(1);
+				else if (flags & SO_OBJECTS)
+					WARN_ON_ONCE(1);
+				else
+					x = page->pages;
 				total += x;
 				nodes[node] += x;
 			}

Signature · 7 lines

1.8.0.2

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to [email protected].  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>

Pekka Enberg

2013-12-29 11:45 UTC #4 lore raw

Quoted 15 lines · Wanpeng Li

Quoted 14 lines · Li Zefan
/sys/kernel/slab/:t-0000048 # cat cpu_slabs
231 N0=16 N1=215
/sys/kernel/slab/:t-0000048 # cat slabs
145 N0=36 N1=109
See, the number of slabs is smaller than that of cpu slabs.
The bug was introduced by commit 49e2258586b423684f03c278149ab46d8f8b6700 ("slub: per cpu cache for partial pages").
We should use page->pages instead of page->pobjects when calculating the number of cpu partial slabs. This also fixes the mapping of slabs and nodes.

Quoted 8 lines

Quoted 6 lines
As there's no variable storing the number of total/active objects in cpu partial slabs, and we don't have user interfaces requiring those statistics, I just add WARN_ON for those cases.
Cc <[email protected]> # 3.2+
Signed-off-by Li Zefan <[email protected]>
Reviewed-by Wanpeng Li <[email protected]>

Applied and sorry for the slow response!