[PATCH v2 1/7] mm/compaction: split freepages without holding the zone lock

[PATCH v2 1/7] mm/compaction: split freepages without holding the zone lock

[js1304]

From: Joonsoo Kim <iamjoonsoo.kim@lge.com>

We don't need to split freepages with holding the zone lock. It will cause
more contention on zone lock so not desirable.

Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
---
 include/linux/mm.h |  1 -
 mm/compaction.c    | 42 ++++++++++++++++++++++++++++++------------
 mm/page_alloc.c    | 27 ---------------------------
 3 files changed, 30 insertions(+), 40 deletions(-)

diff --git a/include/linux/mm.h b/include/linux/mm.h
index a00ec81..1a1782c 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -537,7 +537,6 @@ void __put_page(struct page *page);
 void put_pages_list(struct list_head *pages);
 
 void split_page(struct page *page, unsigned int order);
-int split_free_page(struct page *page);
 
 /*
  * Compound pages have a destructor function.  Provide a
diff --git a/mm/compaction.c b/mm/compaction.c
index 1427366..8e013eb 100644
--- a/mm/compaction.c
+++ b/mm/compaction.c
@@ -65,13 +65,31 @@ static unsigned long release_freepages(struct list_head *freelist)
 
 static void map_pages(struct list_head *list)
 {
-	struct page *page;
+	unsigned int i, order, nr_pages;
+	struct page *page, *next;
+	LIST_HEAD(tmp_list);
+
+	list_for_each_entry_safe(page, next, list, lru) {
+		list_del(&page->lru);
+
+		order = page_private(page);
+		nr_pages = 1 << order;
+		set_page_private(page, 0);
+		set_page_refcounted(page);
+
+		arch_alloc_page(page, order);
+		kernel_map_pages(page, nr_pages, 1);
+		kasan_alloc_pages(page, order);
+		if (order)
+			split_page(page, order);
 
-	list_for_each_entry(page, list, lru) {
-		arch_alloc_page(page, 0);
-		kernel_map_pages(page, 1, 1);
-		kasan_alloc_pages(page, 0);
+		for (i = 0; i < nr_pages; i++) {
+			list_add(&page->lru, &tmp_list);
+			page++;
+		}
 	}
+
+	list_splice(&tmp_list, list);
 }
 
 static inline bool migrate_async_suitable(int migratetype)
@@ -368,12 +386,13 @@ static unsigned long isolate_freepages_block(struct compact_control *cc,
 	unsigned long flags = 0;
 	bool locked = false;
 	unsigned long blockpfn = *start_pfn;
+	unsigned int order;
 
 	cursor = pfn_to_page(blockpfn);
 
 	/* Isolate free pages. */
 	for (; blockpfn < end_pfn; blockpfn++, cursor++) {
-		int isolated, i;
+		int isolated;
 		struct page *page = cursor;
 
 		/*
@@ -439,13 +458,12 @@ static unsigned long isolate_freepages_block(struct compact_control *cc,
 				goto isolate_fail;
 		}
 
-		/* Found a free page, break it into order-0 pages */
-		isolated = split_free_page(page);
+		/* Found a free page, will break it into order-0 pages */
+		order = page_order(page);
+		isolated = __isolate_free_page(page, page_order(page));
+		set_page_private(page, order);
 		total_isolated += isolated;
-		for (i = 0; i < isolated; i++) {
-			list_add(&page->lru, freelist);
-			page++;
-		}
+		list_add_tail(&page->lru, freelist);
 
 		/* If a page was split, advance to the end of it */
 		if (isolated) {
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index d27e8b9..5134f46 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -2525,33 +2525,6 @@ int __isolate_free_page(struct page *page, unsigned int order)
 }
 
 /*
- * Similar to split_page except the page is already free. As this is only
- * being used for migration, the migratetype of the block also changes.
- * As this is called with interrupts disabled, the caller is responsible
- * for calling arch_alloc_page() and kernel_map_page() after interrupts
- * are enabled.
- *
- * Note: this is probably too low level an operation for use in drivers.
- * Please consult with lkml before using this in your driver.
- */
-int split_free_page(struct page *page)
-{
-	unsigned int order;
-	int nr_pages;
-
-	order = page_order(page);
-
-	nr_pages = __isolate_free_page(page, order);
-	if (!nr_pages)
-		return 0;
-
-	/* Split into individual pages */
-	set_page_refcounted(page);
-	split_page(page, order);
-	return nr_pages;
-}
-
-/*
  * Update NUMA hit/miss statistics
  *
  * Must be called with interrupts disabled.
-- 
1.9.1

[PATCH v2 2/7] mm/page_owner: initialize page owner without holding the zone lock

[js1304]

From: Joonsoo Kim <iamjoonsoo.kim@lge.com>

It's not necessary to initialized page_owner with holding the zone lock.
It would cause more contention on the zone lock although it's not
a big problem since it is just debug feature. But, it is better
than before so do it. This is also preparation step to use stackdepot
in page owner feature. Stackdepot allocates new pages when there is no
reserved space and holding the zone lock in this case will cause deadlock.

Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
---
 mm/compaction.c     | 3 +++
 mm/page_alloc.c     | 2 --
 mm/page_isolation.c | 9 ++++++---
 3 files changed, 9 insertions(+), 5 deletions(-)

diff --git a/mm/compaction.c b/mm/compaction.c
index 8e013eb..6043ef8 100644
--- a/mm/compaction.c
+++ b/mm/compaction.c
@@ -20,6 +20,7 @@
 #include <linux/kasan.h>
 #include <linux/kthread.h>
 #include <linux/freezer.h>
+#include <linux/page_owner.h>
 #include "internal.h"
 
 #ifdef CONFIG_COMPACTION
@@ -80,6 +81,8 @@ static void map_pages(struct list_head *list)
 		arch_alloc_page(page, order);
 		kernel_map_pages(page, nr_pages, 1);
 		kasan_alloc_pages(page, order);
+
+		set_page_owner(page, order, __GFP_MOVABLE);
 		if (order)
 			split_page(page, order);
 
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 5134f46..1b1ca57 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -2507,8 +2507,6 @@ int __isolate_free_page(struct page *page, unsigned int order)
 	zone->free_area[order].nr_free--;
 	rmv_page_order(page);
 
-	set_page_owner(page, order, __GFP_MOVABLE);
-
 	/* Set the pageblock if the isolated page is at least a pageblock */
 	if (order >= pageblock_order - 1) {
 		struct page *endpage = page + (1 << order) - 1;
diff --git a/mm/page_isolation.c b/mm/page_isolation.c
index 612122b..927f5ee 100644
--- a/mm/page_isolation.c
+++ b/mm/page_isolation.c
@@ -7,6 +7,7 @@
 #include <linux/pageblock-flags.h>
 #include <linux/memory.h>
 #include <linux/hugetlb.h>
+#include <linux/page_owner.h>
 #include "internal.h"
 
 #define CREATE_TRACE_POINTS
@@ -108,8 +109,6 @@ static void unset_migratetype_isolate(struct page *page, unsigned migratetype)
 			if (pfn_valid_within(page_to_pfn(buddy)) &&
 			    !is_migrate_isolate_page(buddy)) {
 				__isolate_free_page(page, order);
-				kernel_map_pages(page, (1 << order), 1);
-				set_page_refcounted(page);
 				isolated_page = page;
 			}
 		}
@@ -128,8 +127,12 @@ static void unset_migratetype_isolate(struct page *page, unsigned migratetype)
 	zone->nr_isolate_pageblock--;
 out:
 	spin_unlock_irqrestore(&zone->lock, flags);
-	if (isolated_page)
+	if (isolated_page) {
+		kernel_map_pages(page, (1 << order), 1);
+		set_page_refcounted(page);
+		set_page_owner(page, order, __GFP_MOVABLE);
 		__free_pages(isolated_page, order);
+	}
 }
 
 static inline struct page *
-- 
1.9.1

Re: [PATCH v2 2/7] mm/page_owner: initialize page owner without holding the zone lock

[Vlastimil Babka]

On 05/26/2016 04:37 AM, js1304@gmail.com wrote:
> From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>
> It's not necessary to initialized page_owner with holding the zone lock.
> It would cause more contention on the zone lock although it's not
> a big problem since it is just debug feature. But, it is better
> than before so do it. This is also preparation step to use stackdepot
> in page owner feature. Stackdepot allocates new pages when there is no
> reserved space and holding the zone lock in this case will cause deadlock.
>
> Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
> ---
>  mm/compaction.c     | 3 +++
>  mm/page_alloc.c     | 2 --
>  mm/page_isolation.c | 9 ++++++---
>  3 files changed, 9 insertions(+), 5 deletions(-)
>
> diff --git a/mm/compaction.c b/mm/compaction.c
> index 8e013eb..6043ef8 100644
> --- a/mm/compaction.c
> +++ b/mm/compaction.c
> @@ -20,6 +20,7 @@
>  #include <linux/kasan.h>
>  #include <linux/kthread.h>
>  #include <linux/freezer.h>
> +#include <linux/page_owner.h>
>  #include "internal.h"
>
>  #ifdef CONFIG_COMPACTION
> @@ -80,6 +81,8 @@ static void map_pages(struct list_head *list)
>  		arch_alloc_page(page, order);
>  		kernel_map_pages(page, nr_pages, 1);
>  		kasan_alloc_pages(page, order);
> +
> +		set_page_owner(page, order, __GFP_MOVABLE);
>  		if (order)
>  			split_page(page, order);
>
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 5134f46..1b1ca57 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -2507,8 +2507,6 @@ int __isolate_free_page(struct page *page, unsigned int order)
>  	zone->free_area[order].nr_free--;
>  	rmv_page_order(page);
>
> -	set_page_owner(page, order, __GFP_MOVABLE);
> -
>  	/* Set the pageblock if the isolated page is at least a pageblock */
>  	if (order >= pageblock_order - 1) {
>  		struct page *endpage = page + (1 << order) - 1;
> diff --git a/mm/page_isolation.c b/mm/page_isolation.c
> index 612122b..927f5ee 100644
> --- a/mm/page_isolation.c
> +++ b/mm/page_isolation.c
> @@ -7,6 +7,7 @@
>  #include <linux/pageblock-flags.h>
>  #include <linux/memory.h>
>  #include <linux/hugetlb.h>
> +#include <linux/page_owner.h>
>  #include "internal.h"
>
>  #define CREATE_TRACE_POINTS
> @@ -108,8 +109,6 @@ static void unset_migratetype_isolate(struct page *page, unsigned migratetype)
>  			if (pfn_valid_within(page_to_pfn(buddy)) &&
>  			    !is_migrate_isolate_page(buddy)) {
>  				__isolate_free_page(page, order);
> -				kernel_map_pages(page, (1 << order), 1);
> -				set_page_refcounted(page);
>  				isolated_page = page;
>  			}
>  		}
> @@ -128,8 +127,12 @@ static void unset_migratetype_isolate(struct page *page, unsigned migratetype)
>  	zone->nr_isolate_pageblock--;
>  out:
>  	spin_unlock_irqrestore(&zone->lock, flags);
> -	if (isolated_page)
> +	if (isolated_page) {
> +		kernel_map_pages(page, (1 << order), 1);

So why we don't need the other stuff done by e.g. map_pages()? For example 
arch_alloc_page() and kasan_alloc_pages(). Maybe kasan_free_pages() (called 
below via __free_pages() I assume) now doesn't check if the allocation part was 
done. But maybe it will start doing that?

See how the multiple places doing similar stuff is fragile? :(

> +		set_page_refcounted(page);
> +		set_page_owner(page, order, __GFP_MOVABLE);
>  		__free_pages(isolated_page, order);

This mixing of "isolated_page" and "page" is not a bug, but quite ugly. Can't 
isolated_page variable just be a bool?

> +	}
>  }
>
>  static inline struct page *
>

Re: [PATCH v2 2/7] mm/page_owner: initialize page owner without holding the zone lock

[Joonsoo Kim]

2016-06-03 19:23 GMT+09:00 Vlastimil Babka <vbabka@suse.cz>:
> On 05/26/2016 04:37 AM, js1304@gmail.com wrote:
>>
>> From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>>
>> It's not necessary to initialized page_owner with holding the zone lock.
>> It would cause more contention on the zone lock although it's not
>> a big problem since it is just debug feature. But, it is better
>> than before so do it. This is also preparation step to use stackdepot
>> in page owner feature. Stackdepot allocates new pages when there is no
>> reserved space and holding the zone lock in this case will cause deadlock.
>>
>> Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>> ---
>>  mm/compaction.c     | 3 +++
>>  mm/page_alloc.c     | 2 --
>>  mm/page_isolation.c | 9 ++++++---
>>  3 files changed, 9 insertions(+), 5 deletions(-)
>>
>> diff --git a/mm/compaction.c b/mm/compaction.c
>> index 8e013eb..6043ef8 100644
>> --- a/mm/compaction.c
>> +++ b/mm/compaction.c
>> @@ -20,6 +20,7 @@
>>  #include <linux/kasan.h>
>>  #include <linux/kthread.h>
>>  #include <linux/freezer.h>
>> +#include <linux/page_owner.h>
>>  #include "internal.h"
>>
>>  #ifdef CONFIG_COMPACTION
>> @@ -80,6 +81,8 @@ static void map_pages(struct list_head *list)
>>                 arch_alloc_page(page, order);
>>                 kernel_map_pages(page, nr_pages, 1);
>>                 kasan_alloc_pages(page, order);
>> +
>> +               set_page_owner(page, order, __GFP_MOVABLE);
>>                 if (order)
>>                         split_page(page, order);
>>
>> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
>> index 5134f46..1b1ca57 100644
>> --- a/mm/page_alloc.c
>> +++ b/mm/page_alloc.c
>> @@ -2507,8 +2507,6 @@ int __isolate_free_page(struct page *page, unsigned
>> int order)
>>         zone->free_area[order].nr_free--;
>>         rmv_page_order(page);
>>
>> -       set_page_owner(page, order, __GFP_MOVABLE);
>> -
>>         /* Set the pageblock if the isolated page is at least a pageblock
>> */
>>         if (order >= pageblock_order - 1) {
>>                 struct page *endpage = page + (1 << order) - 1;
>> diff --git a/mm/page_isolation.c b/mm/page_isolation.c
>> index 612122b..927f5ee 100644
>> --- a/mm/page_isolation.c
>> +++ b/mm/page_isolation.c
>> @@ -7,6 +7,7 @@
>>  #include <linux/pageblock-flags.h>
>>  #include <linux/memory.h>
>>  #include <linux/hugetlb.h>
>> +#include <linux/page_owner.h>
>>  #include "internal.h"
>>
>>  #define CREATE_TRACE_POINTS
>> @@ -108,8 +109,6 @@ static void unset_migratetype_isolate(struct page
>> *page, unsigned migratetype)
>>                         if (pfn_valid_within(page_to_pfn(buddy)) &&
>>                             !is_migrate_isolate_page(buddy)) {
>>                                 __isolate_free_page(page, order);
>> -                               kernel_map_pages(page, (1 << order), 1);
>> -                               set_page_refcounted(page);
>>                                 isolated_page = page;
>>                         }
>>                 }
>> @@ -128,8 +127,12 @@ static void unset_migratetype_isolate(struct page
>> *page, unsigned migratetype)
>>         zone->nr_isolate_pageblock--;
>>  out:
>>         spin_unlock_irqrestore(&zone->lock, flags);
>> -       if (isolated_page)
>> +       if (isolated_page) {
>> +               kernel_map_pages(page, (1 << order), 1);
>
>
> So why we don't need the other stuff done by e.g. map_pages()? For example
> arch_alloc_page() and kasan_alloc_pages(). Maybe kasan_free_pages() (called
> below via __free_pages() I assume) now doesn't check if the allocation part
> was done. But maybe it will start doing that?
>
> See how the multiple places doing similar stuff is fragile? :(

I answered it in reply of comment of patch 1.

>> +               set_page_refcounted(page);
>> +               set_page_owner(page, order, __GFP_MOVABLE);
>>                 __free_pages(isolated_page, order);
>
>
> This mixing of "isolated_page" and "page" is not a bug, but quite ugly.
> Can't isolated_page variable just be a bool?
>

Looks better. I will do it.

Thanks.

Re: [PATCH v2 2/7] mm/page_owner: initialize page owner without holding the zone lock

[Vlastimil Babka]

On 06/03/2016 02:47 PM, Joonsoo Kim wrote:
[...]

>>> @@ -128,8 +127,12 @@ static void unset_migratetype_isolate(struct page
>>> *page, unsigned migratetype)
>>>         zone->nr_isolate_pageblock--;
>>>  out:
>>>         spin_unlock_irqrestore(&zone->lock, flags);
>>> -       if (isolated_page)
>>> +       if (isolated_page) {
>>> +               kernel_map_pages(page, (1 << order), 1);
>>
>>
>> So why we don't need the other stuff done by e.g. map_pages()? For example
>> arch_alloc_page() and kasan_alloc_pages(). Maybe kasan_free_pages() (called
>> below via __free_pages() I assume) now doesn't check if the allocation part
>> was done. But maybe it will start doing that?
>>
>> See how the multiple places doing similar stuff is fragile? :(
>
> I answered it in reply of comment of patch 1.

Right.

Acked-by: Vlastimil Babka <vbabka@suse.cz>

>
>>> +               set_page_refcounted(page);
>>> +               set_page_owner(page, order, __GFP_MOVABLE);
>>>                 __free_pages(isolated_page, order);
>>
>>
>> This mixing of "isolated_page" and "page" is not a bug, but quite ugly.
>> Can't isolated_page variable just be a bool?
>>
>
> Looks better. I will do it.
>
> Thanks.
>

[PATCH v2 3/7] mm/page_owner: copy last_migrate_reason in copy_page_owner()

[js1304]

From: Joonsoo Kim <iamjoonsoo.kim@lge.com>

Currently, copy_page_owner() doesn't copy all the owner information.
It skips last_migrate_reason because copy_page_owner() is used for
migration and it will be properly set soon. But, following patch
will use copy_page_owner() and this skip will cause the problem that
allocated page has uninitialied last_migrate_reason. To prevent it,
this patch also copy last_migrate_reason in copy_page_owner().

Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
---
 mm/page_owner.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/mm/page_owner.c b/mm/page_owner.c
index c6cda3e..73e202f 100644
--- a/mm/page_owner.c
+++ b/mm/page_owner.c
@@ -118,6 +118,7 @@ void __copy_page_owner(struct page *oldpage, struct page *newpage)
 
 	new_ext->order = old_ext->order;
 	new_ext->gfp_mask = old_ext->gfp_mask;
+	new_ext->last_migrate_reason = old_ext->last_migrate_reason;
 	new_ext->nr_entries = old_ext->nr_entries;
 
 	for (i = 0; i < ARRAY_SIZE(new_ext->trace_entries); i++)
-- 
1.9.1

Re: [PATCH v2 3/7] mm/page_owner: copy last_migrate_reason in copy_page_owner()

[Vlastimil Babka]

On 05/26/2016 04:37 AM, js1304@gmail.com wrote:
> From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>
> Currently, copy_page_owner() doesn't copy all the owner information.
> It skips last_migrate_reason because copy_page_owner() is used for
> migration and it will be properly set soon. But, following patch
> will use copy_page_owner() and this skip will cause the problem that
> allocated page has uninitialied last_migrate_reason. To prevent it,
> this patch also copy last_migrate_reason in copy_page_owner().
>
> Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>

Acked-by: Vlastimil Babka <vbabka@suse.cz>

> ---
>  mm/page_owner.c | 1 +
>  1 file changed, 1 insertion(+)
>
> diff --git a/mm/page_owner.c b/mm/page_owner.c
> index c6cda3e..73e202f 100644
> --- a/mm/page_owner.c
> +++ b/mm/page_owner.c
> @@ -118,6 +118,7 @@ void __copy_page_owner(struct page *oldpage, struct page *newpage)
>
>  	new_ext->order = old_ext->order;
>  	new_ext->gfp_mask = old_ext->gfp_mask;
> +	new_ext->last_migrate_reason = old_ext->last_migrate_reason;
>  	new_ext->nr_entries = old_ext->nr_entries;
>
>  	for (i = 0; i < ARRAY_SIZE(new_ext->trace_entries); i++)
>

[PATCH v2 4/7] mm/page_owner: introduce split_page_owner and replace manual handling

[js1304]

From: Joonsoo Kim <iamjoonsoo.kim@lge.com>

split_page() calls set_page_owner() to set up page_owner to each pages.
But, it has a drawback that head page and the others have different
stacktrace because callsite of set_page_owner() is slightly differnt.
To avoid this problem, this patch copies head page's page_owner to
the others. It needs to introduce new function, split_page_owner() but
it also remove the other function, get_page_owner_gfp() so looks good
to do.

Acked-by: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
---
 include/linux/page_owner.h | 12 +++++-------
 mm/page_alloc.c            |  8 ++------
 mm/page_owner.c            | 14 +++++++-------
 3 files changed, 14 insertions(+), 20 deletions(-)

diff --git a/include/linux/page_owner.h b/include/linux/page_owner.h
index 46f1b93..30583ab 100644
--- a/include/linux/page_owner.h
+++ b/include/linux/page_owner.h
@@ -10,7 +10,7 @@ extern struct page_ext_operations page_owner_ops;
 extern void __reset_page_owner(struct page *page, unsigned int order);
 extern void __set_page_owner(struct page *page,
 			unsigned int order, gfp_t gfp_mask);
-extern gfp_t __get_page_owner_gfp(struct page *page);
+extern void __split_page_owner(struct page *page, unsigned int order);
 extern void __copy_page_owner(struct page *oldpage, struct page *newpage);
 extern void __set_page_owner_migrate_reason(struct page *page, int reason);
 extern void __dump_page_owner(struct page *page);
@@ -28,12 +28,10 @@ static inline void set_page_owner(struct page *page,
 		__set_page_owner(page, order, gfp_mask);
 }
 
-static inline gfp_t get_page_owner_gfp(struct page *page)
+static inline void split_page_owner(struct page *page, unsigned int order)
 {
 	if (static_branch_unlikely(&page_owner_inited))
-		return __get_page_owner_gfp(page);
-	else
-		return 0;
+		__split_page_owner(page, order);
 }
 static inline void copy_page_owner(struct page *oldpage, struct page *newpage)
 {
@@ -58,9 +56,9 @@ static inline void set_page_owner(struct page *page,
 			unsigned int order, gfp_t gfp_mask)
 {
 }
-static inline gfp_t get_page_owner_gfp(struct page *page)
+static inline void split_page_owner(struct page *page,
+			unsigned int order)
 {
-	return 0;
 }
 static inline void copy_page_owner(struct page *oldpage, struct page *newpage)
 {
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 1b1ca57..616ada9 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -2459,7 +2459,6 @@ void free_hot_cold_page_list(struct list_head *list, bool cold)
 void split_page(struct page *page, unsigned int order)
 {
 	int i;
-	gfp_t gfp_mask;
 
 	VM_BUG_ON_PAGE(PageCompound(page), page);
 	VM_BUG_ON_PAGE(!page_count(page), page);
@@ -2473,12 +2472,9 @@ void split_page(struct page *page, unsigned int order)
 		split_page(virt_to_page(page[0].shadow), order);
 #endif
 
-	gfp_mask = get_page_owner_gfp(page);
-	set_page_owner(page, 0, gfp_mask);
-	for (i = 1; i < (1 << order); i++) {
+	for (i = 1; i < (1 << order); i++)
 		set_page_refcounted(page + i);
-		set_page_owner(page + i, 0, gfp_mask);
-	}
+	split_page_owner(page, order);
 }
 EXPORT_SYMBOL_GPL(split_page);
 
diff --git a/mm/page_owner.c b/mm/page_owner.c
index 73e202f..499ad26 100644
--- a/mm/page_owner.c
+++ b/mm/page_owner.c
@@ -94,17 +94,17 @@ void __set_page_owner_migrate_reason(struct page *page, int reason)
 	page_ext->last_migrate_reason = reason;
 }
 
-gfp_t __get_page_owner_gfp(struct page *page)
+void __split_page_owner(struct page *page, unsigned int order)
 {
+	int i;
 	struct page_ext *page_ext = lookup_page_ext(page);
+
 	if (unlikely(!page_ext))
-		/*
-		 * The caller just returns 0 if no valid gfp
-		 * So return 0 here too.
-		 */
-		return 0;
+		return;
 
-	return page_ext->gfp_mask;
+	page_ext->order = 0;
+	for (i = 1; i < (1 << order); i++)
+		__copy_page_owner(page, page + i);
 }
 
 void __copy_page_owner(struct page *oldpage, struct page *newpage)
-- 
1.9.1

[PATCH v2 5/7] tools/vm/page_owner: increase temporary buffer size

[js1304]

From: Joonsoo Kim <iamjoonsoo.kim@lge.com>

Page owner will be changed to store more deep stacktrace
so current temporary buffer size isn't enough. Increase it.

Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
---
 tools/vm/page_owner_sort.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/tools/vm/page_owner_sort.c b/tools/vm/page_owner_sort.c
index 77147b4..f1c055f 100644
--- a/tools/vm/page_owner_sort.c
+++ b/tools/vm/page_owner_sort.c
@@ -79,12 +79,12 @@ static void add_list(char *buf, int len)
 	}
 }
 
-#define BUF_SIZE	1024
+#define BUF_SIZE	(128 * 1024)
 
 int main(int argc, char **argv)
 {
 	FILE *fin, *fout;
-	char buf[BUF_SIZE];
+	char *buf;
 	int ret, i, count;
 	struct block_list *list2;
 	struct stat st;
@@ -107,6 +107,11 @@ int main(int argc, char **argv)
 	max_size = st.st_size / 100; /* hack ... */
 
 	list = malloc(max_size * sizeof(*list));
+	buf = malloc(BUF_SIZE);
+	if (!list || !buf) {
+		printf("Out of memory\n");
+		exit(1);
+	}
 
 	for ( ; ; ) {
 		ret = read_block(buf, BUF_SIZE, fin);
-- 
1.9.1

[PATCH v2 6/7] mm/page_owner: use stackdepot to store stacktrace

[js1304]

From: Joonsoo Kim <iamjoonsoo.kim@lge.com>

Currently, we store each page's allocation stacktrace on corresponding
page_ext structure and it requires a lot of memory. This causes the problem
that memory tight system doesn't work well if page_owner is enabled.
Moreover, even with this large memory consumption, we cannot get full
stacktrace because we allocate memory at boot time and just maintain
8 stacktrace slots to balance memory consumption. We could increase it
to more but it would make system unusable or change system behaviour.

To solve the problem, this patch uses stackdepot to store stacktrace.
It obviously provides memory saving but there is a drawback that
stackdepot could fail.

stackdepot allocates memory at runtime so it could fail if system has
not enough memory. But, most of allocation stack are generated at very
early time and there are much memory at this time. So, failure would not
happen easily. And, one failure means that we miss just one page's
allocation stacktrace so it would not be a big problem. In this patch,
when memory allocation failure happens, we store special stracktrace
handle to the page that is failed to save stacktrace. With it, user
can guess memory usage properly even if failure happens.

Memory saving looks as following. (4GB memory system with page_owner)

static allocation:
92274688 bytes -> 25165824 bytes

dynamic allocation after kernel build:
0 bytes -> 327680 bytes

total:
92274688 bytes -> 25493504 bytes

72% reduction in total.

Note that implementation looks complex than someone would imagine because
there is recursion issue. stackdepot uses page allocator and page_owner
is called at page allocation. Using stackdepot in page_owner could re-call
page allcator and then page_owner. That is a recursion. To detect and
avoid it, whenever we obtain stacktrace, recursion is checked and
page_owner is set to dummy information if found. Dummy information means
that this page is allocated for page_owner feature itself
(such as stackdepot) and it's understandable behavior for user.

v2:
o calculate memory saving with including dynamic allocation
after kernel build
o change maximum stacktrace entry size due to possible stack overflow

Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
---
 include/linux/page_ext.h |   4 +-
 lib/Kconfig.debug        |   1 +
 mm/page_owner.c          | 138 ++++++++++++++++++++++++++++++++++++++++-------
 3 files changed, 122 insertions(+), 21 deletions(-)

diff --git a/include/linux/page_ext.h b/include/linux/page_ext.h
index e1fe7cf..03f2a3e 100644
--- a/include/linux/page_ext.h
+++ b/include/linux/page_ext.h
@@ -3,6 +3,7 @@
 
 #include <linux/types.h>
 #include <linux/stacktrace.h>
+#include <linux/stackdepot.h>
 
 struct pglist_data;
 struct page_ext_operations {
@@ -44,9 +45,8 @@ struct page_ext {
 #ifdef CONFIG_PAGE_OWNER
 	unsigned int order;
 	gfp_t gfp_mask;
-	unsigned int nr_entries;
 	int last_migrate_reason;
-	unsigned long trace_entries[8];
+	depot_stack_handle_t handle;
 #endif
 };
 
diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
index 930bf8e..fc37c66 100644
--- a/lib/Kconfig.debug
+++ b/lib/Kconfig.debug
@@ -248,6 +248,7 @@ config PAGE_OWNER
 	depends on DEBUG_KERNEL && STACKTRACE_SUPPORT
 	select DEBUG_FS
 	select STACKTRACE
+	select STACKDEPOT
 	select PAGE_EXTENSION
 	help
 	  This keeps track of what call chain is the owner of a page, may
diff --git a/mm/page_owner.c b/mm/page_owner.c
index 499ad26..587dcca 100644
--- a/mm/page_owner.c
+++ b/mm/page_owner.c
@@ -7,11 +7,18 @@
 #include <linux/page_owner.h>
 #include <linux/jump_label.h>
 #include <linux/migrate.h>
+#include <linux/stackdepot.h>
+
 #include "internal.h"
 
+#define PAGE_OWNER_STACK_DEPTH (16)
+
 static bool page_owner_disabled = true;
 DEFINE_STATIC_KEY_FALSE(page_owner_inited);
 
+static depot_stack_handle_t dummy_handle;
+static depot_stack_handle_t failure_handle;
+
 static void init_early_allocated_pages(void);
 
 static int early_page_owner_param(char *buf)
@@ -34,11 +41,41 @@ static bool need_page_owner(void)
 	return true;
 }
 
+static noinline void register_dummy_stack(void)
+{
+	unsigned long entries[4];
+	struct stack_trace dummy;
+
+	dummy.nr_entries = 0;
+	dummy.max_entries = ARRAY_SIZE(entries);
+	dummy.entries = &entries[0];
+	dummy.skip = 0;
+
+	save_stack_trace(&dummy);
+	dummy_handle = depot_save_stack(&dummy, GFP_KERNEL);
+}
+
+static noinline void register_failure_stack(void)
+{
+	unsigned long entries[4];
+	struct stack_trace failure;
+
+	failure.nr_entries = 0;
+	failure.max_entries = ARRAY_SIZE(entries);
+	failure.entries = &entries[0];
+	failure.skip = 0;
+
+	save_stack_trace(&failure);
+	failure_handle = depot_save_stack(&failure, GFP_KERNEL);
+}
+
 static void init_page_owner(void)
 {
 	if (page_owner_disabled)
 		return;
 
+	register_dummy_stack();
+	register_failure_stack();
 	static_branch_enable(&page_owner_inited);
 	init_early_allocated_pages();
 }
@@ -61,25 +98,66 @@ void __reset_page_owner(struct page *page, unsigned int order)
 	}
 }
 
-void __set_page_owner(struct page *page, unsigned int order, gfp_t gfp_mask)
+static inline bool check_recursive_alloc(struct stack_trace *trace,
+					unsigned long ip)
 {
-	struct page_ext *page_ext = lookup_page_ext(page);
+	int i, count;
+
+	if (!trace->nr_entries)
+		return false;
+
+	for (i = 0, count = 0; i < trace->nr_entries; i++) {
+		if (trace->entries[i] == ip && ++count == 2)
+			return true;
+	}
 
+	return false;
+}
+
+static noinline depot_stack_handle_t save_stack(gfp_t flags)
+{
+	unsigned long entries[PAGE_OWNER_STACK_DEPTH];
 	struct stack_trace trace = {
 		.nr_entries = 0,
-		.max_entries = ARRAY_SIZE(page_ext->trace_entries),
-		.entries = &page_ext->trace_entries[0],
-		.skip = 3,
+		.entries = entries,
+		.max_entries = PAGE_OWNER_STACK_DEPTH,
+		.skip = 0
 	};
+	depot_stack_handle_t handle;
+
+	save_stack_trace(&trace);
+	if (trace.nr_entries != 0 &&
+	    trace.entries[trace.nr_entries-1] == ULONG_MAX)
+		trace.nr_entries--;
+
+	/*
+	 * We need to check recursion here because our request to stackdepot
+	 * could trigger memory allocation to save new entry. New memory
+	 * allocation would reach here and call depot_save_stack() again
+	 * if we don't catch it. There is still not enough memory in stackdepot
+	 * so it would try to allocate memory again and loop forever.
+	 */
+	if (check_recursive_alloc(&trace, _RET_IP_))
+		return dummy_handle;
+
+	handle = depot_save_stack(&trace, flags);
+	if (!handle)
+		handle = failure_handle;
+
+	return handle;
+}
+
+noinline void __set_page_owner(struct page *page, unsigned int order,
+					gfp_t gfp_mask)
+{
+	struct page_ext *page_ext = lookup_page_ext(page);
 
 	if (unlikely(!page_ext))
 		return;
 
-	save_stack_trace(&trace);
-
+	page_ext->handle = save_stack(gfp_mask);
 	page_ext->order = order;
 	page_ext->gfp_mask = gfp_mask;
-	page_ext->nr_entries = trace.nr_entries;
 	page_ext->last_migrate_reason = -1;
 
 	__set_bit(PAGE_EXT_OWNER, &page_ext->flags);
@@ -111,7 +189,6 @@ void __copy_page_owner(struct page *oldpage, struct page *newpage)
 {
 	struct page_ext *old_ext = lookup_page_ext(oldpage);
 	struct page_ext *new_ext = lookup_page_ext(newpage);
-	int i;
 
 	if (unlikely(!old_ext || !new_ext))
 		return;
@@ -119,10 +196,7 @@ void __copy_page_owner(struct page *oldpage, struct page *newpage)
 	new_ext->order = old_ext->order;
 	new_ext->gfp_mask = old_ext->gfp_mask;
 	new_ext->last_migrate_reason = old_ext->last_migrate_reason;
-	new_ext->nr_entries = old_ext->nr_entries;
-
-	for (i = 0; i < ARRAY_SIZE(new_ext->trace_entries); i++)
-		new_ext->trace_entries[i] = old_ext->trace_entries[i];
+	new_ext->handle = old_ext->handle;
 
 	/*
 	 * We don't clear the bit on the oldpage as it's going to be freed
@@ -138,14 +212,18 @@ void __copy_page_owner(struct page *oldpage, struct page *newpage)
 
 static ssize_t
 print_page_owner(char __user *buf, size_t count, unsigned long pfn,
-		struct page *page, struct page_ext *page_ext)
+		struct page *page, struct page_ext *page_ext,
+		depot_stack_handle_t handle)
 {
 	int ret;
 	int pageblock_mt, page_mt;
 	char *kbuf;
+	unsigned long entries[PAGE_OWNER_STACK_DEPTH];
 	struct stack_trace trace = {
-		.nr_entries = page_ext->nr_entries,
-		.entries = &page_ext->trace_entries[0],
+		.nr_entries = 0,
+		.entries = entries,
+		.max_entries = PAGE_OWNER_STACK_DEPTH,
+		.skip = 0
 	};
 
 	kbuf = kmalloc(count, GFP_KERNEL);
@@ -174,6 +252,7 @@ print_page_owner(char __user *buf, size_t count, unsigned long pfn,
 	if (ret >= count)
 		goto err;
 
+	depot_fetch_stack(handle, &trace);
 	ret += snprint_stack_trace(kbuf + ret, count - ret, &trace, 0);
 	if (ret >= count)
 		goto err;
@@ -204,10 +283,14 @@ err:
 void __dump_page_owner(struct page *page)
 {
 	struct page_ext *page_ext = lookup_page_ext(page);
+	unsigned long entries[PAGE_OWNER_STACK_DEPTH];
 	struct stack_trace trace = {
-		.nr_entries = page_ext->nr_entries,
-		.entries = &page_ext->trace_entries[0],
+		.nr_entries = 0,
+		.entries = entries,
+		.max_entries = PAGE_OWNER_STACK_DEPTH,
+		.skip = 0
 	};
+	depot_stack_handle_t handle;
 	gfp_t gfp_mask = page_ext->gfp_mask;
 	int mt = gfpflags_to_migratetype(gfp_mask);
 
@@ -221,6 +304,13 @@ void __dump_page_owner(struct page *page)
 		return;
 	}
 
+	handle = READ_ONCE(page_ext->handle);
+	if (!handle) {
+		pr_alert("page_owner info is not active (free page?)\n");
+		return;
+	}
+
+	depot_fetch_stack(handle, &trace);
 	pr_alert("page allocated via order %u, migratetype %s, gfp_mask %#x(%pGg)\n",
 		 page_ext->order, migratetype_names[mt], gfp_mask, &gfp_mask);
 	print_stack_trace(&trace, 0);
@@ -236,6 +326,7 @@ read_page_owner(struct file *file, char __user *buf, size_t count, loff_t *ppos)
 	unsigned long pfn;
 	struct page *page;
 	struct page_ext *page_ext;
+	depot_stack_handle_t handle;
 
 	if (!static_branch_unlikely(&page_owner_inited))
 		return -EINVAL;
@@ -284,10 +375,19 @@ read_page_owner(struct file *file, char __user *buf, size_t count, loff_t *ppos)
 		if (!test_bit(PAGE_EXT_OWNER, &page_ext->flags))
 			continue;
 
+		/*
+		 * Access to page_ext->handle isn't synchronous so we should
+		 * be careful to access it.
+		 */
+		handle = READ_ONCE(page_ext->handle);
+		if (!handle)
+			continue;
+
 		/* Record the next PFN to read in the file offset */
 		*ppos = (pfn - min_low_pfn) + 1;
 
-		return print_page_owner(buf, count, pfn, page, page_ext);
+		return print_page_owner(buf, count, pfn, page,
+				page_ext, handle);
 	}
 
 	return 0;
-- 
1.9.1

Re: [PATCH v2 6/7] mm/page_owner: use stackdepot to store stacktrace

[Michal Hocko]

On Thu 26-05-16 11:37:54, Joonsoo Kim wrote:
> From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
> 
> Currently, we store each page's allocation stacktrace on corresponding
> page_ext structure and it requires a lot of memory. This causes the problem
> that memory tight system doesn't work well if page_owner is enabled.
> Moreover, even with this large memory consumption, we cannot get full
> stacktrace because we allocate memory at boot time and just maintain
> 8 stacktrace slots to balance memory consumption. We could increase it
> to more but it would make system unusable or change system behaviour.
> 
> To solve the problem, this patch uses stackdepot to store stacktrace.
> It obviously provides memory saving but there is a drawback that
> stackdepot could fail.
> 
> stackdepot allocates memory at runtime so it could fail if system has
> not enough memory. But, most of allocation stack are generated at very
> early time and there are much memory at this time. So, failure would not
> happen easily. And, one failure means that we miss just one page's
> allocation stacktrace so it would not be a big problem. In this patch,
> when memory allocation failure happens, we store special stracktrace
> handle to the page that is failed to save stacktrace. With it, user
> can guess memory usage properly even if failure happens.
> 
> Memory saving looks as following. (4GB memory system with page_owner)

I still have troubles to understand your numbers

> static allocation:
> 92274688 bytes -> 25165824 bytes

I assume that the first numbers refers to the static allocation for the
given amount of memory while the second one is the dynamic after the
boot, right?

> dynamic allocation after kernel build:
> 0 bytes -> 327680 bytes

And this is the additional dynamic allocation after the kernel build.

> total:
> 92274688 bytes -> 25493504 bytes
> 
> 72% reduction in total.
> 
> Note that implementation looks complex than someone would imagine because
> there is recursion issue. stackdepot uses page allocator and page_owner
> is called at page allocation. Using stackdepot in page_owner could re-call
> page allcator and then page_owner. That is a recursion. To detect and
> avoid it, whenever we obtain stacktrace, recursion is checked and
> page_owner is set to dummy information if found. Dummy information means
> that this page is allocated for page_owner feature itself
> (such as stackdepot) and it's understandable behavior for user.
> 
> v2:
> o calculate memory saving with including dynamic allocation
> after kernel build
> o change maximum stacktrace entry size due to possible stack overflow
> 
> Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>

Other than the small remark below I haven't spotted anything wrong and
I like the approach.

Acked-by: Michal Hocko <mhocko@suse.com>

> ---
>  include/linux/page_ext.h |   4 +-
>  lib/Kconfig.debug        |   1 +
>  mm/page_owner.c          | 138 ++++++++++++++++++++++++++++++++++++++++-------
>  3 files changed, 122 insertions(+), 21 deletions(-)
> 
[...]
> @@ -7,11 +7,18 @@
>  #include <linux/page_owner.h>
>  #include <linux/jump_label.h>
>  #include <linux/migrate.h>
> +#include <linux/stackdepot.h>
> +
>  #include "internal.h"
>  

This is still 128B of the stack which is a lot in the allocation paths
so can we add something like

/*
 * TODO: teach PAGE_OWNER_STACK_DEPTH (__dump_page_owner and save_stack)
 * to use off stack temporal storage
 */
> +#define PAGE_OWNER_STACK_DEPTH (16)
-- 
Michal Hocko
SUSE Labs

Re: [PATCH v2 6/7] mm/page_owner: use stackdepot to store stacktrace

[Joonsoo Kim]

On Mon, Jun 06, 2016 at 03:56:04PM +0200, Michal Hocko wrote:
> On Thu 26-05-16 11:37:54, Joonsoo Kim wrote:
> > From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
> > 
> > Currently, we store each page's allocation stacktrace on corresponding
> > page_ext structure and it requires a lot of memory. This causes the problem
> > that memory tight system doesn't work well if page_owner is enabled.
> > Moreover, even with this large memory consumption, we cannot get full
> > stacktrace because we allocate memory at boot time and just maintain
> > 8 stacktrace slots to balance memory consumption. We could increase it
> > to more but it would make system unusable or change system behaviour.
> > 
> > To solve the problem, this patch uses stackdepot to store stacktrace.
> > It obviously provides memory saving but there is a drawback that
> > stackdepot could fail.
> > 
> > stackdepot allocates memory at runtime so it could fail if system has
> > not enough memory. But, most of allocation stack are generated at very
> > early time and there are much memory at this time. So, failure would not
> > happen easily. And, one failure means that we miss just one page's
> > allocation stacktrace so it would not be a big problem. In this patch,
> > when memory allocation failure happens, we store special stracktrace
> > handle to the page that is failed to save stacktrace. With it, user
> > can guess memory usage properly even if failure happens.
> > 
> > Memory saving looks as following. (4GB memory system with page_owner)
> 
> I still have troubles to understand your numbers
> 
> > static allocation:
> > 92274688 bytes -> 25165824 bytes
> 
> I assume that the first numbers refers to the static allocation for the
> given amount of memory while the second one is the dynamic after the
> boot, right?

No, first number refers to the static allocation before the patch and
second one is for after the patch.

> 
> > dynamic allocation after kernel build:
> > 0 bytes -> 327680 bytes
> 
> And this is the additional dynamic allocation after the kernel build.

This is the additional dynamic allocation after booting + the kernel
build. (before the patch -> after the patch)

> > total:
> > 92274688 bytes -> 25493504 bytes
> > 
> > 72% reduction in total.
> > 
> > Note that implementation looks complex than someone would imagine because
> > there is recursion issue. stackdepot uses page allocator and page_owner
> > is called at page allocation. Using stackdepot in page_owner could re-call
> > page allcator and then page_owner. That is a recursion. To detect and
> > avoid it, whenever we obtain stacktrace, recursion is checked and
> > page_owner is set to dummy information if found. Dummy information means
> > that this page is allocated for page_owner feature itself
> > (such as stackdepot) and it's understandable behavior for user.
> > 
> > v2:
> > o calculate memory saving with including dynamic allocation
> > after kernel build
> > o change maximum stacktrace entry size due to possible stack overflow
> > 
> > Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
> 
> Other than the small remark below I haven't spotted anything wrong and
> I like the approach.
> 
> Acked-by: Michal Hocko <mhocko@suse.com>

Thanks.

> > ---
> >  include/linux/page_ext.h |   4 +-
> >  lib/Kconfig.debug        |   1 +
> >  mm/page_owner.c          | 138 ++++++++++++++++++++++++++++++++++++++++-------
> >  3 files changed, 122 insertions(+), 21 deletions(-)
> > 
> [...]
> > @@ -7,11 +7,18 @@
> >  #include <linux/page_owner.h>
> >  #include <linux/jump_label.h>
> >  #include <linux/migrate.h>
> > +#include <linux/stackdepot.h>
> > +
> >  #include "internal.h"
> >  
> 
> This is still 128B of the stack which is a lot in the allocation paths
> so can we add something like
> 
> /*
>  * TODO: teach PAGE_OWNER_STACK_DEPTH (__dump_page_owner and save_stack)
>  * to use off stack temporal storage
>  */
> > +#define PAGE_OWNER_STACK_DEPTH (16)

Will add.

Thanks.

Re: [PATCH v2 6/7] mm/page_owner: use stackdepot to store stacktrace

[Michal Hocko]

On Fri 17-06-16 16:25:26, Joonsoo Kim wrote:
> On Mon, Jun 06, 2016 at 03:56:04PM +0200, Michal Hocko wrote:
[...]
> > I still have troubles to understand your numbers
> > 
> > > static allocation:
> > > 92274688 bytes -> 25165824 bytes
> > 
> > I assume that the first numbers refers to the static allocation for the
> > given amount of memory while the second one is the dynamic after the
> > boot, right?
> 
> No, first number refers to the static allocation before the patch and
> second one is for after the patch.

I guess we are both talking about the same thing in different words. All
the allocations are static before the patch while all are dynamic after
the patch. Your boot example just shows how much dynamic memory gets
allocated during your boot. This will depend on the particular
configuration but it will at least give a picture what the savings might
be.
-- 
Michal Hocko
SUSE Labs

Re: [PATCH v2 6/7] mm/page_owner: use stackdepot to store stacktrace

[Joonsoo Kim]

On Fri, Jun 17, 2016 at 11:55:59AM +0200, Michal Hocko wrote:
> On Fri 17-06-16 16:25:26, Joonsoo Kim wrote:
> > On Mon, Jun 06, 2016 at 03:56:04PM +0200, Michal Hocko wrote:
> [...]
> > > I still have troubles to understand your numbers
> > > 
> > > > static allocation:
> > > > 92274688 bytes -> 25165824 bytes
> > > 
> > > I assume that the first numbers refers to the static allocation for the
> > > given amount of memory while the second one is the dynamic after the
> > > boot, right?
> > 
> > No, first number refers to the static allocation before the patch and
> > second one is for after the patch.
> 
> I guess we are both talking about the same thing in different words. All
> the allocations are static before the patch while all are dynamic after

Hmm... maybe no? After the patch, there is two parts, static and dynamic.
Page extension has following fields for page owner.

Before the patch
{
 unsigned int order;
 gfp_t gfp_mask;
 unsigned int nr_entries;
 int last_migrate_reason;
 unsigned long trace_entries[8];
}

After the patch
{
 unsigned int order;
 gfp_t gfp_mask;
 int last_migrate_reason;
 depot_stack_handle_t handle;
}

This structure should be allocated for each page even if the patch is
applied so I said it as static memory usage. There is an amount
difference since 'trace_entries[8]' field is changed to 'handle'
field.

Before the patch, stacktrace is stored to static allocated memory per
page. So, no dynamic usage.

After the patch, handle is returned by stackdepot and stackdepot
consumes some memory for it. I said it as dynamic.

Thanks.

> the patch. Your boot example just shows how much dynamic memory gets
> allocated during your boot. This will depend on the particular
> configuration but it will at least give a picture what the savings might
> be.
> -- 
> Michal Hocko
> SUSE Labs
> 
> --
> To unsubscribe, send a message with 'unsubscribe linux-mm' in
> the body to majordomo@kvack.org.  For more info on Linux MM,
> see: http://www.linux-mm.org/ .
> Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH v2 6/7] mm/page_owner: use stackdepot to store stacktrace

[Vlastimil Babka]

On 05/26/2016 04:37 AM, js1304@gmail.com wrote:
> From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>
> Currently, we store each page's allocation stacktrace on corresponding
> page_ext structure and it requires a lot of memory. This causes the problem
> that memory tight system doesn't work well if page_owner is enabled.
> Moreover, even with this large memory consumption, we cannot get full
> stacktrace because we allocate memory at boot time and just maintain
> 8 stacktrace slots to balance memory consumption. We could increase it
> to more but it would make system unusable or change system behaviour.
>
> To solve the problem, this patch uses stackdepot to store stacktrace.
> It obviously provides memory saving but there is a drawback that
> stackdepot could fail.
>
> stackdepot allocates memory at runtime so it could fail if system has
> not enough memory. But, most of allocation stack are generated at very
> early time and there are much memory at this time. So, failure would not
> happen easily. And, one failure means that we miss just one page's
> allocation stacktrace so it would not be a big problem. In this patch,
> when memory allocation failure happens, we store special stracktrace
> handle to the page that is failed to save stacktrace. With it, user
> can guess memory usage properly even if failure happens.
>
> Memory saving looks as following. (4GB memory system with page_owner)
>
> static allocation:
> 92274688 bytes -> 25165824 bytes
>
> dynamic allocation after kernel build:
> 0 bytes -> 327680 bytes
>
> total:
> 92274688 bytes -> 25493504 bytes
>
> 72% reduction in total.
>
> Note that implementation looks complex than someone would imagine because
> there is recursion issue. stackdepot uses page allocator and page_owner
> is called at page allocation. Using stackdepot in page_owner could re-call
> page allcator and then page_owner. That is a recursion. To detect and
> avoid it, whenever we obtain stacktrace, recursion is checked and
> page_owner is set to dummy information if found. Dummy information means
> that this page is allocated for page_owner feature itself
> (such as stackdepot) and it's understandable behavior for user.
>
> v2:
> o calculate memory saving with including dynamic allocation
> after kernel build
> o change maximum stacktrace entry size due to possible stack overflow
>
> Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>

I was surprised that there's no stack removal handling, and then found 
out that stackdepot doesn't support it (e.g. via refcount as one would 
expect). Hopefully the occupied memory doesn't grow indefinitely over 
time then...

Other than that,
Acked-by: Vlastimil Babka <vbabka@suse.cz>

Re: [PATCH v2 6/7] mm/page_owner: use stackdepot to store stacktrace

[Alexander Potapenko]

On Mon, Jun 6, 2016 at 4:51 PM, Vlastimil Babka <vbabka@suse.cz> wrote:
> On 05/26/2016 04:37 AM, js1304@gmail.com wrote:
>>
>> From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>>
>> Currently, we store each page's allocation stacktrace on corresponding
>> page_ext structure and it requires a lot of memory. This causes the
>> problem
>> that memory tight system doesn't work well if page_owner is enabled.
>> Moreover, even with this large memory consumption, we cannot get full
>> stacktrace because we allocate memory at boot time and just maintain
>> 8 stacktrace slots to balance memory consumption. We could increase it
>> to more but it would make system unusable or change system behaviour.
>>
>> To solve the problem, this patch uses stackdepot to store stacktrace.
>> It obviously provides memory saving but there is a drawback that
>> stackdepot could fail.
>>
>> stackdepot allocates memory at runtime so it could fail if system has
>> not enough memory. But, most of allocation stack are generated at very
>> early time and there are much memory at this time. So, failure would not
>> happen easily. And, one failure means that we miss just one page's
>> allocation stacktrace so it would not be a big problem. In this patch,
>> when memory allocation failure happens, we store special stracktrace
>> handle to the page that is failed to save stacktrace. With it, user
>> can guess memory usage properly even if failure happens.
>>
>> Memory saving looks as following. (4GB memory system with page_owner)
>>
>> static allocation:
>> 92274688 bytes -> 25165824 bytes
>>
>> dynamic allocation after kernel build:
>> 0 bytes -> 327680 bytes
>>
>> total:
>> 92274688 bytes -> 25493504 bytes
>>
>> 72% reduction in total.
>>
>> Note that implementation looks complex than someone would imagine because
>> there is recursion issue. stackdepot uses page allocator and page_owner
>> is called at page allocation. Using stackdepot in page_owner could re-call
>> page allcator and then page_owner. That is a recursion. To detect and
>> avoid it, whenever we obtain stacktrace, recursion is checked and
>> page_owner is set to dummy information if found. Dummy information means
>> that this page is allocated for page_owner feature itself
>> (such as stackdepot) and it's understandable behavior for user.
>>
>> v2:
>> o calculate memory saving with including dynamic allocation
>> after kernel build
>> o change maximum stacktrace entry size due to possible stack overflow
>>
>> Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>
>
> I was surprised that there's no stack removal handling, and then found out
> that stackdepot doesn't support it (e.g. via refcount as one would expect).
> Hopefully the occupied memory doesn't grow indefinitely over time then...

The existing use case (allocation/deallocation stacks for KASAN
reports) doesn't require reference counts. Introducing those would
have added unwanted contention and increase memory usage.
The amount of memory used by the stack depot is bounded above, and
should be theoretically enough for everyone (as noted in another
thread, the number of unique allocation/deallocation stacks is way
less than 30k).
> Other than that,
> Acked-by: Vlastimil Babka <vbabka@suse.cz>
>



-- 
Alexander Potapenko
Software Engineer

Google Germany GmbH
Erika-Mann-Straße, 33
80636 München

Geschäftsführer: Matthew Scott Sucherman, Paul Terence Manicle
Registergericht und -nummer: Hamburg, HRB 86891
Sitz der Gesellschaft: Hamburg

[PATCH v2 7/7] mm/page_alloc: introduce post allocation processing on page allocator

[js1304]

From: Joonsoo Kim <iamjoonsoo.kim@lge.com>

This patch is motivated from Hugh and Vlastimil's concern [1].

There are two ways to get freepage from the allocator. One is using
normal memory allocation API and the other is __isolate_free_page() which
is internally used for compaction and pageblock isolation. Later usage is
rather tricky since it doesn't do whole post allocation processing
done by normal API.

One problematic thing I already know is that poisoned page would not be
checked if it is allocated by __isolate_free_page(). Perhaps, there would
be more.

We could add more debug logic for allocated page in the future and this
separation would cause more problem. I'd like to fix this situation
at this time. Solution is simple. This patch commonize some logic
for newly allocated page and uses it on all sites. This will solve
the problem.

[1] http://marc.info/?i=alpine.LSU.2.11.1604270029350.7066%40eggly.anvils%3E

Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
---
 mm/compaction.c     |  8 +-------
 mm/internal.h       |  2 ++
 mm/page_alloc.c     | 22 +++++++++++++---------
 mm/page_isolation.c |  4 +---
 4 files changed, 17 insertions(+), 19 deletions(-)

diff --git a/mm/compaction.c b/mm/compaction.c
index 6043ef8..e15d350 100644
--- a/mm/compaction.c
+++ b/mm/compaction.c
@@ -75,14 +75,8 @@ static void map_pages(struct list_head *list)
 
 		order = page_private(page);
 		nr_pages = 1 << order;
-		set_page_private(page, 0);
-		set_page_refcounted(page);
 
-		arch_alloc_page(page, order);
-		kernel_map_pages(page, nr_pages, 1);
-		kasan_alloc_pages(page, order);
-
-		set_page_owner(page, order, __GFP_MOVABLE);
+		post_alloc_hook(page, order, __GFP_MOVABLE);
 		if (order)
 			split_page(page, order);
 
diff --git a/mm/internal.h b/mm/internal.h
index b6ead95..420bbe3 100644
--- a/mm/internal.h
+++ b/mm/internal.h
@@ -153,6 +153,8 @@ extern int __isolate_free_page(struct page *page, unsigned int order);
 extern void __free_pages_bootmem(struct page *page, unsigned long pfn,
 					unsigned int order);
 extern void prep_compound_page(struct page *page, unsigned int order);
+extern void post_alloc_hook(struct page *page, unsigned int order,
+					gfp_t gfp_flags);
 extern int user_min_free_kbytes;
 
 #if defined CONFIG_COMPACTION || defined CONFIG_CMA
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 616ada9..baa5999 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -1722,6 +1722,18 @@ static bool check_new_pages(struct page *page, unsigned int order)
 	return false;
 }
 
+void post_alloc_hook(struct page *page, unsigned int order, gfp_t gfp_flags)
+{
+	set_page_private(page, 0);
+	set_page_refcounted(page);
+
+	arch_alloc_page(page, order);
+	kernel_map_pages(page, 1 << order, 1);
+	kernel_poison_pages(page, 1 << order, 1);
+	kasan_alloc_pages(page, order);
+	set_page_owner(page, order, gfp_flags);
+}
+
 static void prep_new_page(struct page *page, unsigned int order, gfp_t gfp_flags,
 							unsigned int alloc_flags)
 {
@@ -1734,13 +1746,7 @@ static void prep_new_page(struct page *page, unsigned int order, gfp_t gfp_flags
 			poisoned &= page_is_poisoned(p);
 	}
 
-	set_page_private(page, 0);
-	set_page_refcounted(page);
-
-	arch_alloc_page(page, order);
-	kernel_map_pages(page, 1 << order, 1);
-	kernel_poison_pages(page, 1 << order, 1);
-	kasan_alloc_pages(page, order);
+	post_alloc_hook(page, order, gfp_flags);
 
 	if (!free_pages_prezeroed(poisoned) && (gfp_flags & __GFP_ZERO))
 		for (i = 0; i < (1 << order); i++)
@@ -1749,8 +1755,6 @@ static void prep_new_page(struct page *page, unsigned int order, gfp_t gfp_flags
 	if (order && (gfp_flags & __GFP_COMP))
 		prep_compound_page(page, order);
 
-	set_page_owner(page, order, gfp_flags);
-
 	/*
 	 * page is set pfmemalloc when ALLOC_NO_WATERMARKS was necessary to
 	 * allocate the page. The expectation is that the caller is taking
diff --git a/mm/page_isolation.c b/mm/page_isolation.c
index 927f5ee..4639163 100644
--- a/mm/page_isolation.c
+++ b/mm/page_isolation.c
@@ -128,9 +128,7 @@ static void unset_migratetype_isolate(struct page *page, unsigned migratetype)
 out:
 	spin_unlock_irqrestore(&zone->lock, flags);
 	if (isolated_page) {
-		kernel_map_pages(page, (1 << order), 1);
-		set_page_refcounted(page);
-		set_page_owner(page, order, __GFP_MOVABLE);
+		post_alloc_hook(page, order, __GFP_MOVABLE);
 		__free_pages(isolated_page, order);
 	}
 }
-- 
1.9.1

Re: [PATCH v2 7/7] mm/page_alloc: introduce post allocation processing on page allocator

[Vlastimil Babka]

On 05/26/2016 04:37 AM, js1304@gmail.com wrote:
> From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>
> This patch is motivated from Hugh and Vlastimil's concern [1].
>
> There are two ways to get freepage from the allocator. One is using
> normal memory allocation API and the other is __isolate_free_page() which
> is internally used for compaction and pageblock isolation. Later usage is
> rather tricky since it doesn't do whole post allocation processing
> done by normal API.
>
> One problematic thing I already know is that poisoned page would not be
> checked if it is allocated by __isolate_free_page(). Perhaps, there would
> be more.
>
> We could add more debug logic for allocated page in the future and this
> separation would cause more problem. I'd like to fix this situation
> at this time. Solution is simple. This patch commonize some logic
> for newly allocated page and uses it on all sites. This will solve
> the problem.
>
> [1] http://marc.info/?i=alpine.LSU.2.11.1604270029350.7066%40eggly.anvils%3E
>
> Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>

Yes that's much better. Hopefully introducing a function call into 
prep_new_page() (or can compiler still inline it there?) doesn't impact 
the fast paths though.

Acked-by: Vlastimil Babka <vbabka@suse.cz>

> ---
>  mm/compaction.c     |  8 +-------
>  mm/internal.h       |  2 ++
>  mm/page_alloc.c     | 22 +++++++++++++---------
>  mm/page_isolation.c |  4 +---
>  4 files changed, 17 insertions(+), 19 deletions(-)
>
> diff --git a/mm/compaction.c b/mm/compaction.c
> index 6043ef8..e15d350 100644
> --- a/mm/compaction.c
> +++ b/mm/compaction.c
> @@ -75,14 +75,8 @@ static void map_pages(struct list_head *list)
>
>  		order = page_private(page);
>  		nr_pages = 1 << order;
> -		set_page_private(page, 0);
> -		set_page_refcounted(page);
>
> -		arch_alloc_page(page, order);
> -		kernel_map_pages(page, nr_pages, 1);
> -		kasan_alloc_pages(page, order);
> -
> -		set_page_owner(page, order, __GFP_MOVABLE);
> +		post_alloc_hook(page, order, __GFP_MOVABLE);
>  		if (order)
>  			split_page(page, order);
>
> diff --git a/mm/internal.h b/mm/internal.h
> index b6ead95..420bbe3 100644
> --- a/mm/internal.h
> +++ b/mm/internal.h
> @@ -153,6 +153,8 @@ extern int __isolate_free_page(struct page *page, unsigned int order);
>  extern void __free_pages_bootmem(struct page *page, unsigned long pfn,
>  					unsigned int order);
>  extern void prep_compound_page(struct page *page, unsigned int order);
> +extern void post_alloc_hook(struct page *page, unsigned int order,
> +					gfp_t gfp_flags);
>  extern int user_min_free_kbytes;
>
>  #if defined CONFIG_COMPACTION || defined CONFIG_CMA
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 616ada9..baa5999 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -1722,6 +1722,18 @@ static bool check_new_pages(struct page *page, unsigned int order)
>  	return false;
>  }
>
> +void post_alloc_hook(struct page *page, unsigned int order, gfp_t gfp_flags)
> +{
> +	set_page_private(page, 0);
> +	set_page_refcounted(page);
> +
> +	arch_alloc_page(page, order);
> +	kernel_map_pages(page, 1 << order, 1);
> +	kernel_poison_pages(page, 1 << order, 1);
> +	kasan_alloc_pages(page, order);
> +	set_page_owner(page, order, gfp_flags);
> +}
> +
>  static void prep_new_page(struct page *page, unsigned int order, gfp_t gfp_flags,
>  							unsigned int alloc_flags)
>  {
> @@ -1734,13 +1746,7 @@ static void prep_new_page(struct page *page, unsigned int order, gfp_t gfp_flags
>  			poisoned &= page_is_poisoned(p);
>  	}
>
> -	set_page_private(page, 0);
> -	set_page_refcounted(page);
> -
> -	arch_alloc_page(page, order);
> -	kernel_map_pages(page, 1 << order, 1);
> -	kernel_poison_pages(page, 1 << order, 1);
> -	kasan_alloc_pages(page, order);
> +	post_alloc_hook(page, order, gfp_flags);
>
>  	if (!free_pages_prezeroed(poisoned) && (gfp_flags & __GFP_ZERO))
>  		for (i = 0; i < (1 << order); i++)
> @@ -1749,8 +1755,6 @@ static void prep_new_page(struct page *page, unsigned int order, gfp_t gfp_flags
>  	if (order && (gfp_flags & __GFP_COMP))
>  		prep_compound_page(page, order);
>
> -	set_page_owner(page, order, gfp_flags);
> -
>  	/*
>  	 * page is set pfmemalloc when ALLOC_NO_WATERMARKS was necessary to
>  	 * allocate the page. The expectation is that the caller is taking
> diff --git a/mm/page_isolation.c b/mm/page_isolation.c
> index 927f5ee..4639163 100644
> --- a/mm/page_isolation.c
> +++ b/mm/page_isolation.c
> @@ -128,9 +128,7 @@ static void unset_migratetype_isolate(struct page *page, unsigned migratetype)
>  out:
>  	spin_unlock_irqrestore(&zone->lock, flags);
>  	if (isolated_page) {
> -		kernel_map_pages(page, (1 << order), 1);
> -		set_page_refcounted(page);
> -		set_page_owner(page, order, __GFP_MOVABLE);
> +		post_alloc_hook(page, order, __GFP_MOVABLE);
>  		__free_pages(isolated_page, order);
>  	}
>  }
>

Re: [PATCH v2 7/7] mm/page_alloc: introduce post allocation processing on page allocator

[Joonsoo Kim]

On Mon, Jun 06, 2016 at 05:21:45PM +0200, Vlastimil Babka wrote:
> On 05/26/2016 04:37 AM, js1304@gmail.com wrote:
> >From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
> >
> >This patch is motivated from Hugh and Vlastimil's concern [1].
> >
> >There are two ways to get freepage from the allocator. One is using
> >normal memory allocation API and the other is __isolate_free_page() which
> >is internally used for compaction and pageblock isolation. Later usage is
> >rather tricky since it doesn't do whole post allocation processing
> >done by normal API.
> >
> >One problematic thing I already know is that poisoned page would not be
> >checked if it is allocated by __isolate_free_page(). Perhaps, there would
> >be more.
> >
> >We could add more debug logic for allocated page in the future and this
> >separation would cause more problem. I'd like to fix this situation
> >at this time. Solution is simple. This patch commonize some logic
> >for newly allocated page and uses it on all sites. This will solve
> >the problem.
> >
> >[1] http://marc.info/?i=alpine.LSU.2.11.1604270029350.7066%40eggly.anvils%3E
> >
> >Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
> 
> Yes that's much better. Hopefully introducing a function call into
> prep_new_page() (or can compiler still inline it there?) doesn't
> impact the fast paths though.

Looks like it is already inlined in my build environment but I will
add inline attribute.

> 
> Acked-by: Vlastimil Babka <vbabka@suse.cz>

Thanks.

Re: [PATCH v2 1/7] mm/compaction: split freepages without holding the zone lock

[Vlastimil Babka]

On 05/26/2016 04:37 AM, js1304@gmail.com wrote:
> From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>
> We don't need to split freepages with holding the zone lock. It will cause
> more contention on zone lock so not desirable.
>
> Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>

So it wasn't possible to at least move this code from compaction.c to 
page_alloc.c? Or better, reuse prep_new_page() with some forged gfp/alloc_flags? 
As we discussed in v1...

> ---
>  include/linux/mm.h |  1 -
>  mm/compaction.c    | 42 ++++++++++++++++++++++++++++++------------
>  mm/page_alloc.c    | 27 ---------------------------
>  3 files changed, 30 insertions(+), 40 deletions(-)
>
> diff --git a/include/linux/mm.h b/include/linux/mm.h
> index a00ec81..1a1782c 100644
> --- a/include/linux/mm.h
> +++ b/include/linux/mm.h
> @@ -537,7 +537,6 @@ void __put_page(struct page *page);
>  void put_pages_list(struct list_head *pages);
>
>  void split_page(struct page *page, unsigned int order);
> -int split_free_page(struct page *page);
>
>  /*
>   * Compound pages have a destructor function.  Provide a
> diff --git a/mm/compaction.c b/mm/compaction.c
> index 1427366..8e013eb 100644
> --- a/mm/compaction.c
> +++ b/mm/compaction.c
> @@ -65,13 +65,31 @@ static unsigned long release_freepages(struct list_head *freelist)
>
>  static void map_pages(struct list_head *list)
>  {
> -	struct page *page;
> +	unsigned int i, order, nr_pages;
> +	struct page *page, *next;
> +	LIST_HEAD(tmp_list);
> +
> +	list_for_each_entry_safe(page, next, list, lru) {
> +		list_del(&page->lru);
> +
> +		order = page_private(page);
> +		nr_pages = 1 << order;
> +		set_page_private(page, 0);
> +		set_page_refcounted(page);
> +
> +		arch_alloc_page(page, order);
> +		kernel_map_pages(page, nr_pages, 1);
> +		kasan_alloc_pages(page, order);
> +		if (order)
> +			split_page(page, order);
>
> -	list_for_each_entry(page, list, lru) {
> -		arch_alloc_page(page, 0);
> -		kernel_map_pages(page, 1, 1);
> -		kasan_alloc_pages(page, 0);
> +		for (i = 0; i < nr_pages; i++) {
> +			list_add(&page->lru, &tmp_list);
> +			page++;
> +		}
>  	}
> +
> +	list_splice(&tmp_list, list);
>  }
>
>  static inline bool migrate_async_suitable(int migratetype)
> @@ -368,12 +386,13 @@ static unsigned long isolate_freepages_block(struct compact_control *cc,
>  	unsigned long flags = 0;
>  	bool locked = false;
>  	unsigned long blockpfn = *start_pfn;
> +	unsigned int order;
>
>  	cursor = pfn_to_page(blockpfn);
>
>  	/* Isolate free pages. */
>  	for (; blockpfn < end_pfn; blockpfn++, cursor++) {
> -		int isolated, i;
> +		int isolated;
>  		struct page *page = cursor;
>
>  		/*
> @@ -439,13 +458,12 @@ static unsigned long isolate_freepages_block(struct compact_control *cc,
>  				goto isolate_fail;
>  		}
>
> -		/* Found a free page, break it into order-0 pages */
> -		isolated = split_free_page(page);
> +		/* Found a free page, will break it into order-0 pages */
> +		order = page_order(page);
> +		isolated = __isolate_free_page(page, page_order(page));
> +		set_page_private(page, order);
>  		total_isolated += isolated;
> -		for (i = 0; i < isolated; i++) {
> -			list_add(&page->lru, freelist);
> -			page++;
> -		}
> +		list_add_tail(&page->lru, freelist);
>
>  		/* If a page was split, advance to the end of it */
>  		if (isolated) {
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index d27e8b9..5134f46 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -2525,33 +2525,6 @@ int __isolate_free_page(struct page *page, unsigned int order)
>  }
>
>  /*
> - * Similar to split_page except the page is already free. As this is only
> - * being used for migration, the migratetype of the block also changes.
> - * As this is called with interrupts disabled, the caller is responsible
> - * for calling arch_alloc_page() and kernel_map_page() after interrupts
> - * are enabled.
> - *
> - * Note: this is probably too low level an operation for use in drivers.
> - * Please consult with lkml before using this in your driver.
> - */
> -int split_free_page(struct page *page)
> -{
> -	unsigned int order;
> -	int nr_pages;
> -
> -	order = page_order(page);
> -
> -	nr_pages = __isolate_free_page(page, order);
> -	if (!nr_pages)
> -		return 0;
> -
> -	/* Split into individual pages */
> -	set_page_refcounted(page);
> -	split_page(page, order);
> -	return nr_pages;
> -}
> -
> -/*
>   * Update NUMA hit/miss statistics
>   *
>   * Must be called with interrupts disabled.
>

Re: [PATCH v2 1/7] mm/compaction: split freepages without holding the zone lock

[Joonsoo Kim]

2016-06-03 19:10 GMT+09:00 Vlastimil Babka <vbabka@suse.cz>:
> On 05/26/2016 04:37 AM, js1304@gmail.com wrote:
>>
>> From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>>
>> We don't need to split freepages with holding the zone lock. It will cause
>> more contention on zone lock so not desirable.
>>
>> Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>
>
> So it wasn't possible to at least move this code from compaction.c to
> page_alloc.c? Or better, reuse prep_new_page() with some forged
> gfp/alloc_flags? As we discussed in v1...

Sorry for not mentioning that I did it as a separate patch,
Please see below link which is the last one within this patchset.

Link: http://lkml.kernel.org/r/1464230275-25791-7-git-send-email-iamjoonsoo.kim@lge.com

Thanks.

Re: [PATCH v2 1/7] mm/compaction: split freepages without holding the zone lock

[Vlastimil Babka]

On 06/03/2016 02:45 PM, Joonsoo Kim wrote:
> 2016-06-03 19:10 GMT+09:00 Vlastimil Babka <vbabka@suse.cz>:
>> On 05/26/2016 04:37 AM, js1304@gmail.com wrote:
>>>
>>> From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>>>
>>> We don't need to split freepages with holding the zone lock. It will cause
>>> more contention on zone lock so not desirable.
>>>
>>> Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>>
>>
>> So it wasn't possible to at least move this code from compaction.c to
>> page_alloc.c? Or better, reuse prep_new_page() with some forged
>> gfp/alloc_flags? As we discussed in v1...
>
> Sorry for not mentioning that I did it as a separate patch,
> Please see below link which is the last one within this patchset.
>
> Link: http://lkml.kernel.org/r/1464230275-25791-7-git-send-email-iamjoonsoo.kim@lge.com

Ah I see. In that case,

Acked-by: Vlastimil Babka <vbabka@suse.cz>


> Thanks.
>

Re: [PATCH v2 1/7] mm/compaction: split freepages without holding the zone lock

[Sasha Levin]

On 05/25/2016 10:37 PM, js1304@gmail.com wrote:
> From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
> 
> We don't need to split freepages with holding the zone lock. It will cause
> more contention on zone lock so not desirable.
> 
> Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>

Hey Joonsoo,

I'm seeing the following corruption/crash which seems to be related to
this patch:

[ 3777.807224] ------------[ cut here ]------------

[ 3777.807834] WARNING: CPU: 5 PID: 3270 at lib/list_debug.c:62 __list_del_entry+0x14e/0x280

[ 3777.808562] list_del corruption. next->prev should be ffffea0004a76120, but was ffffea0004a72120

[ 3777.809498] Modules linked in:

[ 3777.809923] CPU: 5 PID: 3270 Comm: khugepaged Tainted: G        W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101

[ 3777.811014]  1ffff100f9315d7b 000000000bb7299a ffff8807c98aec60 ffffffffa0035b2b

[ 3777.811816]  ffffffff00000005 fffffbfff5630bf4 0000000041b58ab3 ffffffffaaaf18e0

[ 3777.812662]  ffffffffa00359bc ffffffff9e54d4a0 ffffffffa8b2ade0 ffff8807c98aece0

[ 3777.813493] Call Trace:

[ 3777.813796] dump_stack (lib/dump_stack.c:53)
[ 3777.814310] ? arch_local_irq_restore (./arch/x86/include/asm/paravirt.h:134)
[ 3777.814947] ? is_module_text_address (kernel/module.c:4185)
[ 3777.815571] ? __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3777.816174] ? vprintk_default (kernel/printk/printk.c:1886)
[ 3777.816761] ? __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3777.817381] __warn (kernel/panic.c:518)
[ 3777.817867] warn_slowpath_fmt (kernel/panic.c:526)
[ 3777.818428] ? __warn (kernel/panic.c:526)
[ 3777.819001] ? __schedule (kernel/sched/core.c:2858 kernel/sched/core.c:3345)
[ 3777.819541] __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3777.820116] ? __list_add (lib/list_debug.c:45)
[ 3777.820721] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
[ 3777.821347] list_del (lib/list_debug.c:78)
[ 3777.821829] __isolate_free_page (mm/page_alloc.c:2514)
[ 3777.822400] ? __zone_watermark_ok (mm/page_alloc.c:2493)
[ 3777.823007] isolate_freepages_block (mm/compaction.c:498)
[ 3777.823629] ? compact_unlock_should_abort (mm/compaction.c:417)
[ 3777.824312] compaction_alloc (mm/compaction.c:1112 mm/compaction.c:1156)
[ 3777.824871] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3777.825512] ? __page_cache_release (mm/swap.c:73)
[ 3777.826127] migrate_pages (mm/migrate.c:1079 mm/migrate.c:1325)
[ 3777.826712] ? __reset_isolation_suitable (mm/compaction.c:1175)
[ 3777.827398] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3777.828109] ? buffer_migrate_page (mm/migrate.c:1301)
[ 3777.828727] compact_zone (mm/compaction.c:1555)
[ 3777.829290] ? compaction_restarting (mm/compaction.c:1476)
[ 3777.829969] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3777.830607] compact_zone_order (mm/compaction.c:1653)
[ 3777.831204] ? kick_process (kernel/sched/core.c:2692)
[ 3777.831774] ? compact_zone (mm/compaction.c:1637)
[ 3777.832336] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3777.832934] try_to_compact_pages (mm/compaction.c:1717)
[ 3777.833550] ? compaction_zonelist_suitable (mm/compaction.c:1679)
[ 3777.834265] __alloc_pages_direct_compact (mm/page_alloc.c:3180)
[ 3777.834922] ? get_page_from_freelist (mm/page_alloc.c:3172)
[ 3777.835549] __alloc_pages_slowpath (mm/page_alloc.c:3741)
[ 3777.836210] ? kvm_clock_read (./arch/x86/include/asm/preempt.h:84 arch/x86/kernel/kvmclock.c:92)
[ 3777.836744] ? __alloc_pages_direct_compact (mm/page_alloc.c:3546)
[ 3777.837429] ? get_page_from_freelist (mm/page_alloc.c:2950)
[ 3777.838072] ? release_pages (mm/swap.c:731)
[ 3777.838610] ? __isolate_free_page (mm/page_alloc.c:2883)
[ 3777.839209] ? ___might_sleep (kernel/sched/core.c:7540 (discriminator 1))
[ 3777.839826] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3777.840427] __alloc_pages_nodemask (mm/page_alloc.c:3841)
[ 3777.841071] ? rwsem_wake (kernel/locking/rwsem-xadd.c:580)
[ 3777.841608] ? __alloc_pages_slowpath (mm/page_alloc.c:3757)
[ 3777.842253] ? call_rwsem_wake (arch/x86/lib/rwsem.S:129)
[ 3777.842839] ? up_write (kernel/locking/rwsem.c:112)
[ 3777.843350] ? pmdp_huge_clear_flush (mm/pgtable-generic.c:131)
[ 3777.844125] khugepaged_alloc_page (mm/khugepaged.c:752)
[ 3777.844719] collapse_huge_page (mm/khugepaged.c:948)
[ 3777.845332] ? khugepaged_scan_shmem (mm/khugepaged.c:922)
[ 3777.846020] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3777.846608] ? remove_wait_queue (kernel/sched/wait.c:292)
[ 3777.847181] khugepaged (mm/khugepaged.c:1724 mm/khugepaged.c:1799 mm/khugepaged.c:1848)
[ 3777.847704] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3777.848297] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3777.848950] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3777.849555] ? default_wake_function (kernel/sched/core.c:3544)
[ 3777.850161] ? __wake_up_common (kernel/sched/wait.c:73)
[ 3777.850724] ? __kthread_parkme (kernel/kthread.c:168)
[ 3777.851306] kthread (kernel/kthread.c:209)
[ 3777.851819] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3777.852448] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3777.853045] ret_from_fork (arch/x86/entry/entry_64.S:390)
[ 3777.853605] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3777.854173] ---[ end trace 8cbbecee22435cc2 ]---

[ 3777.854602] ------------[ cut here ]------------

[ 3777.855101] WARNING: CPU: 5 PID: 3270 at lib/list_debug.c:62 __list_del_entry+0x14e/0x280

[ 3777.855863] list_del corruption. next->prev should be ffffea0004a761e0, but was ffffea0004a72020

[ 3777.856701] Modules linked in:

[ 3777.857041] CPU: 5 PID: 3270 Comm: khugepaged Tainted: G        W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101

[ 3777.858153]  1ffff100f9315d7b 000000000bb7299a ffff8807c98aec60 ffffffffa0035b2b

[ 3777.858963]  ffffffff00000005 fffffbfff5630bf4 0000000041b58ab3 ffffffffaaaf18e0

[ 3777.859724]  ffffffffa00359bc ffffffff9e54d4a0 ffffffffa8b2ade0 ffff8807c98aece0

[ 3777.860494] Call Trace:

[ 3777.860752] dump_stack (lib/dump_stack.c:53)
[ 3777.861266] ? arch_local_irq_restore (./arch/x86/include/asm/paravirt.h:134)
[ 3777.861878] ? is_module_text_address (kernel/module.c:4185)
[ 3777.862487] ? __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3777.863059] ? vprintk_default (kernel/printk/printk.c:1886)
[ 3777.863643] ? __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3777.864230] __warn (kernel/panic.c:518)
[ 3777.864720] warn_slowpath_fmt (kernel/panic.c:526)
[ 3777.865314] ? __warn (kernel/panic.c:526)
[ 3777.865808] ? __schedule (kernel/sched/core.c:2858 kernel/sched/core.c:3345)
[ 3777.866343] __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3777.866911] ? __list_add (lib/list_debug.c:45)
[ 3777.867427] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
[ 3777.868045] list_del (lib/list_debug.c:78)
[ 3777.868517] __isolate_free_page (mm/page_alloc.c:2514)
[ 3777.869091] ? __zone_watermark_ok (mm/page_alloc.c:2493)
[ 3777.869676] isolate_freepages_block (mm/compaction.c:498)
[ 3777.870350] ? compact_unlock_should_abort (mm/compaction.c:417)
[ 3777.871014] compaction_alloc (mm/compaction.c:1112 mm/compaction.c:1156)
[ 3777.871561] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3777.872216] ? __page_cache_release (mm/swap.c:73)
[ 3777.872846] migrate_pages (mm/migrate.c:1079 mm/migrate.c:1325)
[ 3777.873416] ? __reset_isolation_suitable (mm/compaction.c:1175)
[ 3777.874163] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3777.874811] ? buffer_migrate_page (mm/migrate.c:1301)
[ 3777.875421] compact_zone (mm/compaction.c:1555)
[ 3777.875980] ? compaction_restarting (mm/compaction.c:1476)
[ 3777.876600] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3777.877256] compact_zone_order (mm/compaction.c:1653)
[ 3777.877821] ? kick_process (kernel/sched/core.c:2692)
[ 3777.878412] ? compact_zone (mm/compaction.c:1637)
[ 3777.879053] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3777.879657] try_to_compact_pages (mm/compaction.c:1717)
[ 3777.880252] ? compaction_zonelist_suitable (mm/compaction.c:1679)
[ 3777.880957] __alloc_pages_direct_compact (mm/page_alloc.c:3180)
[ 3777.881624] ? get_page_from_freelist (mm/page_alloc.c:3172)
[ 3777.882300] __alloc_pages_slowpath (mm/page_alloc.c:3741)
[ 3777.882998] ? kvm_clock_read (./arch/x86/include/asm/preempt.h:84 arch/x86/kernel/kvmclock.c:92)
[ 3777.883588] ? __alloc_pages_direct_compact (mm/page_alloc.c:3546)
[ 3777.884319] ? get_page_from_freelist (mm/page_alloc.c:2950)
[ 3777.884988] ? release_pages (mm/swap.c:731)
[ 3777.885539] ? __isolate_free_page (mm/page_alloc.c:2883)
[ 3777.886163] ? ___might_sleep (kernel/sched/core.c:7540 (discriminator 1))
[ 3777.886756] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3777.887305] __alloc_pages_nodemask (mm/page_alloc.c:3841)
[ 3777.887914] ? rwsem_wake (kernel/locking/rwsem-xadd.c:580)
[ 3777.888432] ? __alloc_pages_slowpath (mm/page_alloc.c:3757)
[ 3777.889092] ? call_rwsem_wake (arch/x86/lib/rwsem.S:129)
[ 3777.889642] ? up_write (kernel/locking/rwsem.c:112)
[ 3777.890192] ? pmdp_huge_clear_flush (mm/pgtable-generic.c:131)
[ 3777.890836] khugepaged_alloc_page (mm/khugepaged.c:752)
[ 3777.891436] collapse_huge_page (mm/khugepaged.c:948)
[ 3777.892023] ? khugepaged_scan_shmem (mm/khugepaged.c:922)
[ 3777.892630] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3777.893181] ? remove_wait_queue (kernel/sched/wait.c:292)
[ 3777.893769] khugepaged (mm/khugepaged.c:1724 mm/khugepaged.c:1799 mm/khugepaged.c:1848)
[ 3777.894318] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3777.894912] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3777.895542] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3777.896181] ? default_wake_function (kernel/sched/core.c:3544)
[ 3777.896851] ? __wake_up_common (kernel/sched/wait.c:73)
[ 3777.897425] ? __kthread_parkme (kernel/kthread.c:168)
[ 3777.898017] kthread (kernel/kthread.c:209)
[ 3777.898501] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3777.899086] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3777.899682] ret_from_fork (arch/x86/entry/entry_64.S:390)
[ 3777.900241] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3777.900887] ---[ end trace 8cbbecee22435cc3 ]---

[ 3777.901335] ------------[ cut here ]------------

[ 3777.901782] WARNING: CPU: 5 PID: 3270 at lib/list_debug.c:62 __list_del_entry+0x14e/0x280

[ 3777.902545] list_del corruption. next->prev should be ffffea0004a76220, but was ffffea0005a4f420

[ 3777.903383] Modules linked in:

[ 3777.903706] CPU: 5 PID: 3270 Comm: khugepaged Tainted: G        W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101

[ 3777.904734]  1ffff100f9315d7b 000000000bb7299a ffff8807c98aec60 ffffffffa0035b2b

[ 3777.905497]  ffffffff00000005 fffffbfff5630bf4 0000000041b58ab3 ffffffffaaaf18e0

[ 3777.906250]  ffffffffa00359bc ffffffff9e54d4a0 ffffffffa8b2ade0 ffff8807c98aece0

[ 3777.907002] Call Trace:

[ 3777.907258] dump_stack (lib/dump_stack.c:53)
[ 3777.907822] ? arch_local_irq_restore (./arch/x86/include/asm/paravirt.h:134)
[ 3777.908425] ? is_module_text_address (kernel/module.c:4185)
[ 3777.909040] ? __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3777.909633] ? vprintk_default (kernel/printk/printk.c:1886)
[ 3777.910233] ? __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3777.910813] __warn (kernel/panic.c:518)
[ 3777.911278] warn_slowpath_fmt (kernel/panic.c:526)
[ 3777.911852] ? __warn (kernel/panic.c:526)
[ 3777.912343] ? __schedule (kernel/sched/core.c:2858 kernel/sched/core.c:3345)
[ 3777.912883] __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3777.913443] ? __list_add (lib/list_debug.c:45)
[ 3777.913974] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
[ 3777.914607] list_del (lib/list_debug.c:78)
[ 3777.915064] __isolate_free_page (mm/page_alloc.c:2514)
[ 3777.915647] ? __zone_watermark_ok (mm/page_alloc.c:2493)
[ 3777.916281] isolate_freepages_block (mm/compaction.c:498)
[ 3777.916930] ? compact_unlock_should_abort (mm/compaction.c:417)
[ 3777.917646] compaction_alloc (mm/compaction.c:1112 mm/compaction.c:1156)
[ 3777.918216] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3777.918862] ? __page_cache_release (mm/swap.c:73)
[ 3777.919495] migrate_pages (mm/migrate.c:1079 mm/migrate.c:1325)
[ 3777.920087] ? __reset_isolation_suitable (mm/compaction.c:1175)
[ 3777.920733] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3777.921345] ? buffer_migrate_page (mm/migrate.c:1301)
[ 3777.921971] compact_zone (mm/compaction.c:1555)
[ 3777.922543] ? compaction_restarting (mm/compaction.c:1476)
[ 3777.923189] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3777.923814] compact_zone_order (mm/compaction.c:1653)
[ 3777.924397] ? kick_process (kernel/sched/core.c:2692)
[ 3777.924941] ? compact_zone (mm/compaction.c:1637)
[ 3777.925502] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3777.926102] try_to_compact_pages (mm/compaction.c:1717)
[ 3777.926708] ? compaction_zonelist_suitable (mm/compaction.c:1679)
[ 3777.927403] __alloc_pages_direct_compact (mm/page_alloc.c:3180)
[ 3777.928074] ? get_page_from_freelist (mm/page_alloc.c:3172)
[ 3777.928745] __alloc_pages_slowpath (mm/page_alloc.c:3741)
[ 3777.929384] ? kvm_clock_read (./arch/x86/include/asm/preempt.h:84 arch/x86/kernel/kvmclock.c:92)
[ 3777.929968] ? __alloc_pages_direct_compact (mm/page_alloc.c:3546)
[ 3777.930644] ? get_page_from_freelist (mm/page_alloc.c:2950)
[ 3777.931350] ? release_pages (mm/swap.c:731)
[ 3777.931963] ? __isolate_free_page (mm/page_alloc.c:2883)
[ 3777.932566] ? ___might_sleep (kernel/sched/core.c:7540 (discriminator 1))
[ 3777.933157] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3777.936090] __alloc_pages_nodemask (mm/page_alloc.c:3841)
[ 3777.936721] ? rwsem_wake (kernel/locking/rwsem-xadd.c:580)
[ 3777.937281] ? __alloc_pages_slowpath (mm/page_alloc.c:3757)
[ 3777.937945] ? call_rwsem_wake (arch/x86/lib/rwsem.S:129)
[ 3777.938489] ? up_write (kernel/locking/rwsem.c:112)
[ 3777.938991] ? pmdp_huge_clear_flush (mm/pgtable-generic.c:131)
[ 3777.939604] khugepaged_alloc_page (mm/khugepaged.c:752)
[ 3777.940235] collapse_huge_page (mm/khugepaged.c:948)
[ 3777.940812] ? khugepaged_scan_shmem (mm/khugepaged.c:922)
[ 3777.941435] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3777.941974] ? remove_wait_queue (kernel/sched/wait.c:292)
[ 3777.942552] khugepaged (mm/khugepaged.c:1724 mm/khugepaged.c:1799 mm/khugepaged.c:1848)
[ 3777.943067] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3777.943667] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3777.944262] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3777.944846] ? default_wake_function (kernel/sched/core.c:3544)
[ 3777.945452] ? __wake_up_common (kernel/sched/wait.c:73)
[ 3777.946076] ? __kthread_parkme (kernel/kthread.c:168)
[ 3777.946647] kthread (kernel/kthread.c:209)
[ 3777.947188] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3777.947788] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3777.948366] ret_from_fork (arch/x86/entry/entry_64.S:390)
[ 3777.948892] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3777.949477] ---[ end trace 8cbbecee22435cc4 ]---

[ 3777.949918] ------------[ cut here ]------------

[ 3777.950377] WARNING: CPU: 5 PID: 3270 at lib/list_debug.c:62 __list_del_entry+0x14e/0x280

[ 3777.951149] list_del corruption. next->prev should be ffffea0004a76320, but was ffffea0004a76020

[ 3777.951976] Modules linked in:

[ 3777.952314] CPU: 5 PID: 3270 Comm: khugepaged Tainted: G        W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101

[ 3777.953347]  1ffff100f9315d7b 000000000bb7299a ffff8807c98aec60 ffffffffa0035b2b

[ 3777.954104]  ffffffff00000005 fffffbfff5630bf4 0000000041b58ab3 ffffffffaaaf18e0

[ 3777.954896]  ffffffffa00359bc ffffffff9e54d4a0 ffffffffa8b2ade0 ffff8807c98aece0

[ 3777.955693] Call Trace:

[ 3777.955955] dump_stack (lib/dump_stack.c:53)
[ 3777.956493] ? arch_local_irq_restore (./arch/x86/include/asm/paravirt.h:134)
[ 3777.957179] ? is_module_text_address (kernel/module.c:4185)
[ 3777.957781] ? __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3777.958373] ? vprintk_default (kernel/printk/printk.c:1886)
[ 3777.958999] ? __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3777.959583] __warn (kernel/panic.c:518)
[ 3777.960103] warn_slowpath_fmt (kernel/panic.c:526)
[ 3777.960688] ? __warn (kernel/panic.c:526)
[ 3777.961251] ? __schedule (kernel/sched/core.c:2858 kernel/sched/core.c:3345)
[ 3777.961831] __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3777.962457] ? __list_add (lib/list_debug.c:45)
[ 3777.963032] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
[ 3777.963642] list_del (lib/list_debug.c:78)
[ 3777.964125] __isolate_free_page (mm/page_alloc.c:2514)
[ 3777.964716] ? __zone_watermark_ok (mm/page_alloc.c:2493)
[ 3777.965316] isolate_freepages_block (mm/compaction.c:498)
[ 3777.965944] ? compact_unlock_should_abort (mm/compaction.c:417)
[ 3777.966605] compaction_alloc (mm/compaction.c:1112 mm/compaction.c:1156)
[ 3777.967167] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3777.967804] ? __page_cache_release (mm/swap.c:73)
[ 3777.968409] migrate_pages (mm/migrate.c:1079 mm/migrate.c:1325)
[ 3777.969006] ? __reset_isolation_suitable (mm/compaction.c:1175)
[ 3777.969700] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3777.970366] ? buffer_migrate_page (mm/migrate.c:1301)
[ 3777.970996] compact_zone (mm/compaction.c:1555)
[ 3777.971603] ? compaction_restarting (mm/compaction.c:1476)
[ 3777.972204] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3777.972833] compact_zone_order (mm/compaction.c:1653)
[ 3777.973420] ? kick_process (kernel/sched/core.c:2692)
[ 3777.974074] ? compact_zone (mm/compaction.c:1637)
[ 3777.974679] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3777.975288] try_to_compact_pages (mm/compaction.c:1717)
[ 3777.975873] ? compaction_zonelist_suitable (mm/compaction.c:1679)
[ 3777.976559] __alloc_pages_direct_compact (mm/page_alloc.c:3180)
[ 3777.977241] ? get_page_from_freelist (mm/page_alloc.c:3172)
[ 3777.977915] __alloc_pages_slowpath (mm/page_alloc.c:3741)
[ 3777.978537] ? kvm_clock_read (./arch/x86/include/asm/preempt.h:84 arch/x86/kernel/kvmclock.c:92)
[ 3777.979084] ? __alloc_pages_direct_compact (mm/page_alloc.c:3546)
[ 3777.979739] ? get_page_from_freelist (mm/page_alloc.c:2950)
[ 3777.980373] ? release_pages (mm/swap.c:731)
[ 3777.980964] ? __isolate_free_page (mm/page_alloc.c:2883)
[ 3777.981566] ? ___might_sleep (kernel/sched/core.c:7540 (discriminator 1))
[ 3777.982129] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3777.982684] __alloc_pages_nodemask (mm/page_alloc.c:3841)
[ 3777.983306] ? rwsem_wake (kernel/locking/rwsem-xadd.c:580)
[ 3777.983850] ? __alloc_pages_slowpath (mm/page_alloc.c:3757)
[ 3777.984482] ? call_rwsem_wake (arch/x86/lib/rwsem.S:129)
[ 3777.985070] ? up_write (kernel/locking/rwsem.c:112)
[ 3777.985559] ? pmdp_huge_clear_flush (mm/pgtable-generic.c:131)
[ 3777.986162] khugepaged_alloc_page (mm/khugepaged.c:752)
[ 3777.986809] collapse_huge_page (mm/khugepaged.c:948)
[ 3777.987404] ? khugepaged_scan_shmem (mm/khugepaged.c:922)
[ 3777.988057] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3777.988595] ? remove_wait_queue (kernel/sched/wait.c:292)
[ 3777.989164] khugepaged (mm/khugepaged.c:1724 mm/khugepaged.c:1799 mm/khugepaged.c:1848)
[ 3777.989690] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3777.990285] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3777.990916] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3777.991514] ? default_wake_function (kernel/sched/core.c:3544)
[ 3777.992122] ? __wake_up_common (kernel/sched/wait.c:73)
[ 3777.992722] ? __kthread_parkme (kernel/kthread.c:168)
[ 3777.993291] kthread (kernel/kthread.c:209)
[ 3777.993810] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3777.994433] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3777.995060] ret_from_fork (arch/x86/entry/entry_64.S:390)
[ 3777.995587] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3777.996176] ---[ end trace 8cbbecee22435cc5 ]---

[ 3777.997195] ------------[ cut here ]------------

[ 3777.997838] WARNING: CPU: 16 PID: 3730 at lib/list_debug.c:29 __list_add+0xc6/0x240

[ 3777.998568] list_add corruption. next->prev should be prev (ffff8801b7fd8068), but was ffffea0004f56420. (next=ffffea0004f48e20).

[ 3777.999804] Modules linked in:

[ 3778.000141] CPU: 16 PID: 3730 Comm: kswapd0 Tainted: G        W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101

[ 3778.001223]  1ffff10066e36de3 00000000be0b1ffd ffff8803371b6fa0 ffffffffa0035b2b

[ 3778.001958]  ffffffff00000010 fffffbfff5630bf4 0000000041b58ab3 ffffffffaaaf18e0

[ 3778.002726]  ffffffffa00359bc ffffffff9e54d4a0 ffffffffa8b2ab20 ffff8803371b7020

[ 3778.003547] Call Trace:

[ 3778.003876] dump_stack (lib/dump_stack.c:53)
[ 3778.004400] ? arch_local_irq_restore (./arch/x86/include/asm/paravirt.h:134)
[ 3778.005025] ? is_module_text_address (kernel/module.c:4185)
[ 3778.005652] ? __list_add (lib/list_debug.c:30 (discriminator 3))
[ 3778.006148] ? vprintk_default (kernel/printk/printk.c:1886)
[ 3778.006714] ? __list_add (lib/list_debug.c:30 (discriminator 3))
[ 3778.007257] __warn (kernel/panic.c:518)
[ 3778.007752] ? kvm_sched_clock_read (arch/x86/kernel/kvmclock.c:104)
[ 3778.008381] warn_slowpath_fmt (kernel/panic.c:526)
[ 3778.008958] ? __warn (kernel/panic.c:526)
[ 3778.009479] ? qstat_write (kernel/locking/qspinlock.c:411)
[ 3778.010044] ? __radix_tree_lookup (lib/radix-tree.c:673)
[ 3778.010621] __list_add (lib/list_debug.c:30 (discriminator 3))
[ 3778.011109] ? __this_cpu_preempt_check (lib/list_debug.c:25)
[ 3778.011703] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
[ 3778.012290] free_pcppages_bulk (mm/page_alloc.c:878 mm/page_alloc.c:1135)
[ 3778.012905] ? __free_page_frag (mm/page_alloc.c:1084)
[ 3778.013468] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.014078] free_hot_cold_page (mm/page_alloc.c:2442 (discriminator 1))
[ 3778.014627] ? drain_all_pages (mm/page_alloc.c:2403)
[ 3778.015186] ? uncharge_list (mm/memcontrol.c:5536)
[ 3778.015757] ? __this_cpu_preempt_check (lib/list_debug.c:25)
[ 3778.016359] free_hot_cold_page_list (mm/page_alloc.c:2456 (discriminator 3))
[ 3778.016950] shrink_page_list (include/linux/compiler.h:222 include/linux/list.h:189 include/linux/list.h:296 mm/vmscan.c:1238)
[ 3778.017505] ? putback_lru_page (mm/vmscan.c:889)
[ 3778.018098] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.018736] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.019343] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
[ 3778.019964] ? __mod_zone_page_state (mm/vmstat.c:249)
[ 3778.020551] shrink_inactive_list (include/linux/spinlock.h:332 mm/vmscan.c:1629)
[ 3778.021130] ? putback_inactive_pages (mm/vmscan.c:1573)
[ 3778.021728] ? zone_reclaimable_pages (mm/vmscan.c:208)
[ 3778.022336] ? _find_next_bit (lib/find_bit.c:54)
[ 3778.022875] ? zone_reclaimable (include/linux/vmstat.h:166 mm/vmscan.c:212)
[ 3778.023432] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.024068] ? get_scan_count (mm/vmscan.c:1954)
[ 3778.024656] ? blk_start_plug (block/blk-core.c:3170 (discriminator 1))
[ 3778.025206] ? blk_lld_busy (block/blk-core.c:3170)
[ 3778.025724] shrink_zone_memcg (mm/vmscan.c:1932 mm/vmscan.c:2224)
[ 3778.026299] ? shrink_active_list (mm/vmscan.c:2181)
[ 3778.026887] ? css_next_descendant_pre (kernel/cgroup.c:4028)
[ 3778.027507] ? mem_cgroup_iter (mm/memcontrol.c:889)
[ 3778.028101] ? preempt_count_add (include/linux/ftrace.h:724 kernel/sched/core.c:3074 kernel/sched/core.c:3099)
[ 3778.028661] shrink_zone (mm/vmscan.c:2407)
[ 3778.029174] ? mem_cgroup_split_huge_fixup (mm/memcontrol.c:2607)
[ 3778.029849] ? shrink_zone_memcg (mm/vmscan.c:2374)
[ 3778.030457] ? zone_watermark_ok_safe (mm/page_alloc.c:2839)
[ 3778.031062] kswapd (./arch/x86/include/asm/bitops.h:113 mm/vmscan.c:3090 mm/vmscan.c:3239 mm/vmscan.c:3427)
[ 3778.031539] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.032270] ? mem_cgroup_shrink_node_zone (mm/vmscan.c:3351)
[ 3778.033010] ? __schedule (kernel/sched/core.c:2858 kernel/sched/core.c:3345)
[ 3778.035456] ? remove_wait_queue (kernel/sched/wait.c:292)
[ 3778.036004] ? default_wake_function (kernel/sched/core.c:3544)
[ 3778.037227] ? __kthread_parkme (kernel/kthread.c:168)
[ 3778.037792] kthread (kernel/kthread.c:209)
[ 3778.038280] ? mem_cgroup_shrink_node_zone (mm/vmscan.c:3351)
[ 3778.038921] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.039506] ret_from_fork (arch/x86/entry/entry_64.S:390)
[ 3778.040082] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.040655] ---[ end trace 8cbbecee22435cc6 ]---

[ 3778.041122] ------------[ cut here ]------------

[ 3778.041560] WARNING: CPU: 16 PID: 3730 at lib/list_debug.c:33 __list_add+0x11f/0x240

[ 3778.042272] list_add corruption. prev->next should be next (ffff8801b7fd8000), but was ffffea00050f0a20. (prev=ffffea00050f09a0).

[ 3778.043351] Modules linked in:

[ 3778.043666] CPU: 16 PID: 3730 Comm: kswapd0 Tainted: G        W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101

[ 3778.044643]  1ffff10066e36de3 00000000be0b1ffd ffff8803371b6fa0 ffffffffa0035b2b

[ 3778.045400]  ffffffff00000010 fffffbfff5630bf4 0000000041b58ab3 ffffffffaaaf18e0

[ 3778.046136]  ffffffffa00359bc ffffffff9e54d4a0 ffffffffa8b2abe0 ffff8803371b7020

[ 3778.046865] Call Trace:

[ 3778.047112] dump_stack (lib/dump_stack.c:53)
[ 3778.047627] ? arch_local_irq_restore (./arch/x86/include/asm/paravirt.h:134)
[ 3778.048251] ? is_module_text_address (kernel/module.c:4185)
[ 3778.048827] ? __list_add (lib/list_debug.c:34 (discriminator 3))
[ 3778.049325] ? vprintk_default (kernel/printk/printk.c:1886)
[ 3778.049851] ? __list_add (lib/list_debug.c:34 (discriminator 3))
[ 3778.050346] __warn (kernel/panic.c:518)
[ 3778.050879] ? kvm_sched_clock_read (arch/x86/kernel/kvmclock.c:104)
[ 3778.051429] warn_slowpath_fmt (kernel/panic.c:526)
[ 3778.051954] ? __warn (kernel/panic.c:526)
[ 3778.052479] ? qstat_write (kernel/locking/qspinlock.c:411)
[ 3778.053006] ? __radix_tree_lookup (lib/radix-tree.c:673)
[ 3778.053580] __list_add (lib/list_debug.c:34 (discriminator 3))
[ 3778.054154] ? __this_cpu_preempt_check (lib/list_debug.c:25)
[ 3778.054736] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
[ 3778.055375] ? __mod_zone_page_state (mm/vmstat.c:249)
[ 3778.055964] free_pcppages_bulk (include/linux/list.h:77 mm/page_alloc.c:870 mm/page_alloc.c:1135)
[ 3778.056548] ? __free_page_frag (mm/page_alloc.c:1084)
[ 3778.057210] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.057847] free_hot_cold_page (mm/page_alloc.c:2442 (discriminator 1))
[ 3778.058410] ? drain_all_pages (mm/page_alloc.c:2403)
[ 3778.059016] ? uncharge_list (mm/memcontrol.c:5536)
[ 3778.059627] ? __this_cpu_preempt_check (lib/list_debug.c:25)
[ 3778.060270] free_hot_cold_page_list (mm/page_alloc.c:2456 (discriminator 3))
[ 3778.060867] shrink_page_list (include/linux/compiler.h:222 include/linux/list.h:189 include/linux/list.h:296 mm/vmscan.c:1238)
[ 3778.061425] ? putback_lru_page (mm/vmscan.c:889)
[ 3778.061988] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.062631] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.063254] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
[ 3778.063873] ? __mod_zone_page_state (mm/vmstat.c:249)
[ 3778.064472] shrink_inactive_list (include/linux/spinlock.h:332 mm/vmscan.c:1629)
[ 3778.065054] ? putback_inactive_pages (mm/vmscan.c:1573)
[ 3778.065703] ? zone_reclaimable_pages (mm/vmscan.c:208)
[ 3778.066355] ? _find_next_bit (lib/find_bit.c:54)
[ 3778.066914] ? zone_reclaimable (include/linux/vmstat.h:166 mm/vmscan.c:212)
[ 3778.067478] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.068118] ? get_scan_count (mm/vmscan.c:1954)
[ 3778.068675] ? blk_start_plug (block/blk-core.c:3170 (discriminator 1))
[ 3778.069211] ? blk_lld_busy (block/blk-core.c:3170)
[ 3778.069807] shrink_zone_memcg (mm/vmscan.c:1932 mm/vmscan.c:2224)
[ 3778.070336] ? shrink_active_list (mm/vmscan.c:2181)
[ 3778.070891] ? css_next_descendant_pre (kernel/cgroup.c:4028)
[ 3778.071508] ? mem_cgroup_iter (mm/memcontrol.c:889)
[ 3778.072052] ? preempt_count_add (include/linux/ftrace.h:724 kernel/sched/core.c:3074 kernel/sched/core.c:3099)
[ 3778.072607] shrink_zone (mm/vmscan.c:2407)
[ 3778.073103] ? mem_cgroup_split_huge_fixup (mm/memcontrol.c:2607)
[ 3778.073756] ? shrink_zone_memcg (mm/vmscan.c:2374)
[ 3778.074310] ? zone_watermark_ok_safe (mm/page_alloc.c:2839)
[ 3778.074912] kswapd (./arch/x86/include/asm/bitops.h:113 mm/vmscan.c:3090 mm/vmscan.c:3239 mm/vmscan.c:3427)
[ 3778.075381] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.075951] ? mem_cgroup_shrink_node_zone (mm/vmscan.c:3351)
[ 3778.076681] ? __schedule (kernel/sched/core.c:2858 kernel/sched/core.c:3345)
[ 3778.077200] ? remove_wait_queue (kernel/sched/wait.c:292)
[ 3778.077745] ? default_wake_function (kernel/sched/core.c:3544)
[ 3778.078319] ? __kthread_parkme (kernel/kthread.c:168)
[ 3778.078868] kthread (kernel/kthread.c:209)
[ 3778.079339] ? mem_cgroup_shrink_node_zone (mm/vmscan.c:3351)
[ 3778.079972] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.080530] ret_from_fork (arch/x86/entry/entry_64.S:390)
[ 3778.081022] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.081614] ---[ end trace 8cbbecee22435cc7 ]---

[ 3778.082029] ------------[ cut here ]------------

[ 3778.082480] WARNING: CPU: 16 PID: 3730 at lib/list_debug.c:29 __list_add+0xc6/0x240

[ 3778.083285] list_add corruption. next->prev should be prev (ffff8801b7fd80d0), but was ffffea00050f07a0. (next=ffffea00050f0820).

[ 3778.084308] Modules linked in:

[ 3778.084611] CPU: 16 PID: 3730 Comm: kswapd0 Tainted: G        W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101

[ 3778.085601]  1ffff10066e36de3 00000000be0b1ffd ffff8803371b6fa0 ffffffffa0035b2b

[ 3778.086387]  ffffffff00000010 fffffbfff5630bf4 0000000041b58ab3 ffffffffaaaf18e0

[ 3778.087130]  ffffffffa00359bc ffffffff9e54d4a0 ffffffffa8b2ab20 ffff8803371b7020

[ 3778.087909] Call Trace:

[ 3778.088153] dump_stack (lib/dump_stack.c:53)
[ 3778.088644] ? arch_local_irq_restore (./arch/x86/include/asm/paravirt.h:134)
[ 3778.089278] ? is_module_text_address (kernel/module.c:4185)
[ 3778.090006] ? __list_add (lib/list_debug.c:30 (discriminator 3))
[ 3778.090511] ? vprintk_default (kernel/printk/printk.c:1886)
[ 3778.091079] ? __list_add (lib/list_debug.c:30 (discriminator 3))
[ 3778.091618] __warn (kernel/panic.c:518)
[ 3778.092094] ? kvm_sched_clock_read (arch/x86/kernel/kvmclock.c:104)
[ 3778.092680] warn_slowpath_fmt (kernel/panic.c:526)
[ 3778.093219] ? __warn (kernel/panic.c:526)
[ 3778.093718] ? qstat_write (kernel/locking/qspinlock.c:411)
[ 3778.094303] ? __radix_tree_lookup (lib/radix-tree.c:673)
[ 3778.094912] __list_add (lib/list_debug.c:30 (discriminator 3))
[ 3778.095396] ? __this_cpu_preempt_check (lib/list_debug.c:25)
[ 3778.096015] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
[ 3778.096627] free_pcppages_bulk (mm/page_alloc.c:878 mm/page_alloc.c:1135)
[ 3778.097182] ? __free_page_frag (mm/page_alloc.c:1084)
[ 3778.097764] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.098404] free_hot_cold_page (mm/page_alloc.c:2442 (discriminator 1))
[ 3778.098964] ? drain_all_pages (mm/page_alloc.c:2403)
[ 3778.099545] ? uncharge_list (mm/memcontrol.c:5536)
[ 3778.100084] ? __this_cpu_preempt_check (lib/list_debug.c:25)
[ 3778.100728] free_hot_cold_page_list (mm/page_alloc.c:2456 (discriminator 3))
[ 3778.101349] shrink_page_list (include/linux/compiler.h:222 include/linux/list.h:189 include/linux/list.h:296 mm/vmscan.c:1238)
[ 3778.101976] ? putback_lru_page (mm/vmscan.c:889)
[ 3778.102560] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.103184] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.103817] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
[ 3778.104424] ? __mod_zone_page_state (mm/vmstat.c:249)
[ 3778.105013] shrink_inactive_list (include/linux/spinlock.h:332 mm/vmscan.c:1629)
[ 3778.105585] ? putback_inactive_pages (mm/vmscan.c:1573)
[ 3778.106190] ? zone_reclaimable_pages (mm/vmscan.c:208)
[ 3778.106835] ? _find_next_bit (lib/find_bit.c:54)
[ 3778.107370] ? zone_reclaimable (include/linux/vmstat.h:166 mm/vmscan.c:212)
[ 3778.107976] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.108605] ? get_scan_count (mm/vmscan.c:1954)
[ 3778.109158] ? blk_start_plug (block/blk-core.c:3170 (discriminator 1))
[ 3778.109715] ? blk_lld_busy (block/blk-core.c:3170)
[ 3778.110221] shrink_zone_memcg (mm/vmscan.c:1932 mm/vmscan.c:2224)
[ 3778.110813] ? shrink_active_list (mm/vmscan.c:2181)
[ 3778.111384] ? css_next_descendant_pre (kernel/cgroup.c:4028)
[ 3778.111993] ? mem_cgroup_iter (mm/memcontrol.c:889)
[ 3778.112576] ? preempt_count_add (include/linux/ftrace.h:724 kernel/sched/core.c:3074 kernel/sched/core.c:3099)
[ 3778.113139] shrink_zone (mm/vmscan.c:2407)
[ 3778.113704] ? mem_cgroup_split_huge_fixup (mm/memcontrol.c:2607)
[ 3778.114336] ? shrink_zone_memcg (mm/vmscan.c:2374)
[ 3778.114904] ? zone_watermark_ok_safe (mm/page_alloc.c:2839)
[ 3778.115509] kswapd (./arch/x86/include/asm/bitops.h:113 mm/vmscan.c:3090 mm/vmscan.c:3239 mm/vmscan.c:3427)
[ 3778.115996] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.116572] ? mem_cgroup_shrink_node_zone (mm/vmscan.c:3351)
[ 3778.117234] ? __schedule (kernel/sched/core.c:2858 kernel/sched/core.c:3345)
[ 3778.117761] ? remove_wait_queue (kernel/sched/wait.c:292)
[ 3778.118312] ? default_wake_function (kernel/sched/core.c:3544)
[ 3778.118892] ? __kthread_parkme (kernel/kthread.c:168)
[ 3778.119450] kthread (kernel/kthread.c:209)
[ 3778.119937] ? mem_cgroup_shrink_node_zone (mm/vmscan.c:3351)
[ 3778.120624] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.121186] ret_from_fork (arch/x86/entry/entry_64.S:390)
[ 3778.121689] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.122249] ---[ end trace 8cbbecee22435cc8 ]---

[ 3778.122683] ------------[ cut here ]------------

[ 3778.123128] WARNING: CPU: 16 PID: 3730 at lib/list_debug.c:29 __list_add+0xc6/0x240

[ 3778.124008] list_add corruption. next->prev should be prev (ffff8801b7fd8138), but was ffffea0005116820. (next=ffffea00050f0020).

[ 3778.125058] Modules linked in:

[ 3778.125367] CPU: 16 PID: 3730 Comm: kswapd0 Tainted: G        W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101

[ 3778.126483]  1ffff10066e36de3 00000000be0b1ffd ffff8803371b6fa0 ffffffffa0035b2b

[ 3778.127213]  ffffffff00000010 fffffbfff5630bf4 0000000041b58ab3 ffffffffaaaf18e0

[ 3778.127952]  ffffffffa00359bc ffffffff9e54d4a0 ffffffffa8b2ab20 ffff8803371b7020

[ 3778.128693] Call Trace:

[ 3778.128941] dump_stack (lib/dump_stack.c:53)
[ 3778.129443] ? arch_local_irq_restore (./arch/x86/include/asm/paravirt.h:134)
[ 3778.130070] ? is_module_text_address (kernel/module.c:4185)
[ 3778.130670] ? __list_add (lib/list_debug.c:30 (discriminator 3))
[ 3778.131443] ? vprintk_default (kernel/printk/printk.c:1886)
[ 3778.132001] ? __list_add (lib/list_debug.c:30 (discriminator 3))
[ 3778.132506] __warn (kernel/panic.c:518)
[ 3778.133010] ? kvm_sched_clock_read (arch/x86/kernel/kvmclock.c:104)
[ 3778.134072] warn_slowpath_fmt (kernel/panic.c:526)
[ 3778.134684] ? __warn (kernel/panic.c:526)
[ 3778.135174] ? qstat_write (kernel/locking/qspinlock.c:411)
[ 3778.135698] ? __radix_tree_lookup (lib/radix-tree.c:673)
[ 3778.136319] __list_add (lib/list_debug.c:30 (discriminator 3))
[ 3778.136823] ? __this_cpu_preempt_check (lib/list_debug.c:25)
[ 3778.137421] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
[ 3778.138058] free_pcppages_bulk (mm/page_alloc.c:878 mm/page_alloc.c:1135)
[ 3778.138636] ? __free_page_frag (mm/page_alloc.c:1084)
[ 3778.139207] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.139817] free_hot_cold_page (mm/page_alloc.c:2442 (discriminator 1))
[ 3778.140379] ? drain_all_pages (mm/page_alloc.c:2403)
[ 3778.140929] ? uncharge_list (mm/memcontrol.c:5536)
[ 3778.141516] ? __this_cpu_preempt_check (lib/list_debug.c:25)
[ 3778.142116] free_hot_cold_page_list (mm/page_alloc.c:2456 (discriminator 3))
[ 3778.142693] shrink_page_list (include/linux/compiler.h:222 include/linux/list.h:189 include/linux/list.h:296 mm/vmscan.c:1238)
[ 3778.143295] ? putback_lru_page (mm/vmscan.c:889)
[ 3778.143850] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.144570] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.145153] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
[ 3778.145732] ? __mod_zone_page_state (mm/vmstat.c:249)
[ 3778.146301] shrink_inactive_list (include/linux/spinlock.h:332 mm/vmscan.c:1629)
[ 3778.146860] ? putback_inactive_pages (mm/vmscan.c:1573)
[ 3778.147428] ? zone_reclaimable_pages (mm/vmscan.c:208)
[ 3778.148055] ? _find_next_bit (lib/find_bit.c:54)
[ 3778.148613] ? zone_reclaimable (include/linux/vmstat.h:166 mm/vmscan.c:212)
[ 3778.149209] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.149817] ? get_scan_count (mm/vmscan.c:1954)
[ 3778.150349] ? blk_start_plug (block/blk-core.c:3170 (discriminator 1))
[ 3778.150895] ? blk_lld_busy (block/blk-core.c:3170)
[ 3778.151427] shrink_zone_memcg (mm/vmscan.c:1932 mm/vmscan.c:2224)
[ 3778.152028] ? shrink_active_list (mm/vmscan.c:2181)
[ 3778.152638] ? css_next_descendant_pre (kernel/cgroup.c:4028)
[ 3778.153269] ? mem_cgroup_iter (mm/memcontrol.c:889)
[ 3778.153897] ? preempt_count_add (include/linux/ftrace.h:724 kernel/sched/core.c:3074 kernel/sched/core.c:3099)
[ 3778.154489] shrink_zone (mm/vmscan.c:2407)
[ 3778.155015] ? mem_cgroup_split_huge_fixup (mm/memcontrol.c:2607)
[ 3778.155683] ? shrink_zone_memcg (mm/vmscan.c:2374)
[ 3778.156360] ? zone_watermark_ok_safe (mm/page_alloc.c:2839)
[ 3778.157029] kswapd (./arch/x86/include/asm/bitops.h:113 mm/vmscan.c:3090 mm/vmscan.c:3239 mm/vmscan.c:3427)
[ 3778.157527] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.158217] ? mem_cgroup_shrink_node_zone (mm/vmscan.c:3351)
[ 3778.158972] ? __schedule (kernel/sched/core.c:2858 kernel/sched/core.c:3345)
[ 3778.159508] ? remove_wait_queue (kernel/sched/wait.c:292)
[ 3778.160112] ? default_wake_function (kernel/sched/core.c:3544)
[ 3778.160732] ? __kthread_parkme (kernel/kthread.c:168)
[ 3778.161287] kthread (kernel/kthread.c:209)
[ 3778.161811] ? mem_cgroup_shrink_node_zone (mm/vmscan.c:3351)
[ 3778.162513] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.163082] ret_from_fork (arch/x86/entry/entry_64.S:390)
[ 3778.163590] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.164229] ---[ end trace 8cbbecee22435cc9 ]---

[ 3778.170922] ------------[ cut here ]------------

[ 3778.171447] WARNING: CPU: 16 PID: 3730 at lib/list_debug.c:29 __list_add+0xc6/0x240

[ 3778.171740] ------------[ cut here ]------------

[ 3778.171760] WARNING: CPU: 5 PID: 3270 at lib/list_debug.c:62 __list_del_entry+0x14e/0x280

[ 3778.171767] list_del corruption. next->prev should be ffffea0005116820, but was ffffea0005292220

[ 3778.171772] Modules linked in:

[ 3778.171786] CPU: 5 PID: 3270 Comm: khugepaged Tainted: G        W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101

[ 3778.171799]  1ffff100f9315da5 000000000bb7299a ffff8807c98aedb0 ffffffffa0035b2b

[ 3778.171808]  ffffffff00000005 fffffbfff5630bf4 0000000041b58ab3 ffffffffaaaf18e0

[ 3778.171818]  ffffffffa00359bc ffffffff9e54d4a0 ffffffffa8b2ade0 ffff8807c98aee30

[ 3778.171820] Call Trace:

[ 3778.171837] dump_stack (lib/dump_stack.c:53)
[ 3778.171850] ? arch_local_irq_restore (./arch/x86/include/asm/paravirt.h:134)
[ 3778.171863] ? is_module_text_address (kernel/module.c:4185)
[ 3778.171874] ? __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3778.171893] ? vprintk_default (kernel/printk/printk.c:1886)
[ 3778.171905] ? __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3778.171917] __warn (kernel/panic.c:518)
[ 3778.171928] ? kick_process (kernel/sched/core.c:2692)
[ 3778.171939] warn_slowpath_fmt (kernel/panic.c:526)
[ 3778.171948] ? __warn (kernel/panic.c:526)
[ 3778.171964] ? kernel_poison_pages (mm/page_poison.c:169)
[ 3778.171975] __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3778.171985] ? post_alloc_hook (mm/page_alloc.c:1740 (discriminator 3) include/linux/page_owner.h:27 (discriminator 3) mm/page_alloc.c:1741 (discriminator 3))
[ 3778.171995] ? __list_add (lib/list_debug.c:45)
[ 3778.172011] ? __this_cpu_preempt_check (lib/list_debug.c:25)
[ 3778.172022] list_del (lib/list_debug.c:78)
[ 3778.172032] map_pages (mm/compaction.c:75)
[ 3778.172041] ? compaction_free (mm/compaction.c:67)
[ 3778.172051] compaction_alloc (mm/compaction.c:1136 mm/compaction.c:1156)
[ 3778.172061] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3778.172073] ? __page_cache_release (mm/swap.c:73)
[ 3778.172087] migrate_pages (mm/migrate.c:1079 mm/migrate.c:1325)
[ 3778.172097] ? __reset_isolation_suitable (mm/compaction.c:1175)
[ 3778.172107] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3778.172119] ? buffer_migrate_page (mm/migrate.c:1301)
[ 3778.172128] compact_zone (mm/compaction.c:1555)
[ 3778.172139] ? compaction_restarting (mm/compaction.c:1476)
[ 3778.172149] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.172159] compact_zone_order (mm/compaction.c:1653)
[ 3778.172170] ? kick_process (kernel/sched/core.c:2692)
[ 3778.172179] ? compact_zone (mm/compaction.c:1637)
[ 3778.172191] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3778.172202] try_to_compact_pages (mm/compaction.c:1717)
[ 3778.172213] ? compaction_zonelist_suitable (mm/compaction.c:1679)
[ 3778.172224] __alloc_pages_direct_compact (mm/page_alloc.c:3180)
[ 3778.172239] ? get_page_from_freelist (mm/page_alloc.c:3172)
[ 3778.172250] __alloc_pages_slowpath (mm/page_alloc.c:3741)
[ 3778.172265] ? kvm_clock_read (./arch/x86/include/asm/preempt.h:84 arch/x86/kernel/kvmclock.c:92)
[ 3778.172276] ? __alloc_pages_direct_compact (mm/page_alloc.c:3546)
[ 3778.172287] ? get_page_from_freelist (mm/page_alloc.c:2950)
[ 3778.172298] ? release_pages (mm/swap.c:731)
[ 3778.172308] ? __isolate_free_page (mm/page_alloc.c:2883)
[ 3778.172319] ? ___might_sleep (kernel/sched/core.c:7540 (discriminator 1))
[ 3778.172331] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3778.172342] __alloc_pages_nodemask (mm/page_alloc.c:3841)
[ 3778.172352] ? rwsem_wake (kernel/locking/rwsem-xadd.c:580)
[ 3778.172362] ? __alloc_pages_slowpath (mm/page_alloc.c:3757)
[ 3778.172372] ? call_rwsem_wake (arch/x86/lib/rwsem.S:129)
[ 3778.172381] ? up_write (kernel/locking/rwsem.c:112)
[ 3778.172393] ? pmdp_huge_clear_flush (mm/pgtable-generic.c:131)
[ 3778.172403] khugepaged_alloc_page (mm/khugepaged.c:752)
[ 3778.172413] collapse_huge_page (mm/khugepaged.c:948)
[ 3778.172423] ? khugepaged_scan_shmem (mm/khugepaged.c:922)
[ 3778.172434] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3778.172445] ? remove_wait_queue (kernel/sched/wait.c:292)
[ 3778.172454] khugepaged (mm/khugepaged.c:1724 mm/khugepaged.c:1799 mm/khugepaged.c:1848)
[ 3778.172464] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.172475] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3778.172486] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3778.172496] ? default_wake_function (kernel/sched/core.c:3544)
[ 3778.172507] ? __wake_up_common (kernel/sched/wait.c:73)
[ 3778.172516] ? __kthread_parkme (kernel/kthread.c:168)
[ 3778.172525] kthread (kernel/kthread.c:209)
[ 3778.172538] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3778.172548] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.172559] ret_from_fork (arch/x86/entry/entry_64.S:390)
[ 3778.172568] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.172605] ---[ end trace 8cbbecee22435cca ]---

[ 3778.172614] ------------[ cut here ]------------

[ 3778.172628] WARNING: CPU: 5 PID: 3270 at lib/list_debug.c:62 __list_del_entry+0x14e/0x280

[ 3778.172634] list_del corruption. next->prev should be ffffea00050f07a0, but was ffffea0005293120

[ 3778.172637] Modules linked in:

[ 3778.172648] CPU: 5 PID: 3270 Comm: khugepaged Tainted: G        W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101

[ 3778.172658]  1ffff100f9315da5 000000000bb7299a ffff8807c98aedb0 ffffffffa0035b2b

[ 3778.172668]  ffffffff00000005 fffffbfff5630bf4 0000000041b58ab3 ffffffffaaaf18e0

[ 3778.172677]  ffffffffa00359bc ffffffff9e54d4a0 ffffffffa8b2ade0 ffff8807c98aee30

[ 3778.172679] Call Trace:

[ 3778.172698] dump_stack (lib/dump_stack.c:53)
[ 3778.172711] ? arch_local_irq_restore (./arch/x86/include/asm/paravirt.h:134)
[ 3778.172722] ? is_module_text_address (kernel/module.c:4185)
[ 3778.172733] ? __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3778.172744] ? vprintk_default (kernel/printk/printk.c:1886)
[ 3778.172755] ? __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3778.172765] __warn (kernel/panic.c:518)
[ 3778.172776] ? kick_process (kernel/sched/core.c:2692)
[ 3778.172786] warn_slowpath_fmt (kernel/panic.c:526)
[ 3778.172796] ? __warn (kernel/panic.c:526)
[ 3778.172808] ? kernel_poison_pages (mm/page_poison.c:169)
[ 3778.172819] __list_del_entry (lib/list_debug.c:60 (discriminator 1))
[ 3778.172829] ? post_alloc_hook (mm/page_alloc.c:1740 (discriminator 3) include/linux/page_owner.h:27 (discriminator 3) mm/page_alloc.c:1741 (discriminator 3))
[ 3778.172839] ? __list_add (lib/list_debug.c:45)
[ 3778.172850] ? __this_cpu_preempt_check (lib/list_debug.c:25)
[ 3778.172860] list_del (lib/list_debug.c:78)
[ 3778.172869] map_pages (mm/compaction.c:75)
[ 3778.172878] ? compaction_free (mm/compaction.c:67)
[ 3778.172887] compaction_alloc (mm/compaction.c:1136 mm/compaction.c:1156)
[ 3778.172898] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3778.172909] ? __page_cache_release (mm/swap.c:73)
[ 3778.172920] migrate_pages (mm/migrate.c:1079 mm/migrate.c:1325)
[ 3778.172930] ? __reset_isolation_suitable (mm/compaction.c:1175)
[ 3778.172940] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3778.172952] ? buffer_migrate_page (mm/migrate.c:1301)
[ 3778.172961] compact_zone (mm/compaction.c:1555)
[ 3778.172972] ? compaction_restarting (mm/compaction.c:1476)
[ 3778.172982] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.172992] compact_zone_order (mm/compaction.c:1653)
[ 3778.173007] ? kick_process (kernel/sched/core.c:2692)
[ 3778.173020] ? compact_zone (mm/compaction.c:1637)
[ 3778.173033] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3778.173043] try_to_compact_pages (mm/compaction.c:1717)
[ 3778.173054] ? compaction_zonelist_suitable (mm/compaction.c:1679)
[ 3778.173066] __alloc_pages_direct_compact (mm/page_alloc.c:3180)
[ 3778.173076] ? get_page_from_freelist (mm/page_alloc.c:3172)
[ 3778.173087] __alloc_pages_slowpath (mm/page_alloc.c:3741)
[ 3778.173098] ? kvm_clock_read (./arch/x86/include/asm/preempt.h:84 arch/x86/kernel/kvmclock.c:92)
[ 3778.173109] ? __alloc_pages_direct_compact (mm/page_alloc.c:3546)
[ 3778.173120] ? get_page_from_freelist (mm/page_alloc.c:2950)
[ 3778.173131] ? release_pages (mm/swap.c:731)
[ 3778.173141] ? __isolate_free_page (mm/page_alloc.c:2883)
[ 3778.173152] ? ___might_sleep (kernel/sched/core.c:7540 (discriminator 1))
[ 3778.173163] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3778.173216] __alloc_pages_nodemask (mm/page_alloc.c:3841)
[ 3778.173226] ? rwsem_wake (kernel/locking/rwsem-xadd.c:580)
[ 3778.173237] ? __alloc_pages_slowpath (mm/page_alloc.c:3757)
[ 3778.173247] ? call_rwsem_wake (arch/x86/lib/rwsem.S:129)
[ 3778.173256] ? up_write (kernel/locking/rwsem.c:112)
[ 3778.173267] ? pmdp_huge_clear_flush (mm/pgtable-generic.c:131)
[ 3778.173277] khugepaged_alloc_page (mm/khugepaged.c:752)
[ 3778.173287] collapse_huge_page (mm/khugepaged.c:948)
[ 3778.173298] ? khugepaged_scan_shmem (mm/khugepaged.c:922)
[ 3778.173309] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3778.173320] ? remove_wait_queue (kernel/sched/wait.c:292)
[ 3778.173330] khugepaged (mm/khugepaged.c:1724 mm/khugepaged.c:1799 mm/khugepaged.c:1848)
[ 3778.173339] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.173350] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3778.173362] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3778.173372] ? default_wake_function (kernel/sched/core.c:3544)
[ 3778.173382] ? __wake_up_common (kernel/sched/wait.c:73)
[ 3778.173391] ? __kthread_parkme (kernel/kthread.c:168)
[ 3778.173400] kthread (kernel/kthread.c:209)
[ 3778.173410] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3778.173419] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.173429] ret_from_fork (arch/x86/entry/entry_64.S:390)
[ 3778.173443] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.173472] ---[ end trace 8cbbecee22435ccb ]---

[ 3778.173506] ================================================================================

[ 3778.173536] UBSAN: Undefined behaviour in mm/compaction.c:76:16

[ 3778.173551] shift exponent 88341344 is too large for 32-bit type 'int'

[ 3778.173562] CPU: 5 PID: 3270 Comm: khugepaged Tainted: G        W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101

[ 3778.173572]  1ffff100f9315daa 000000000bb7299a ffff8807c98aedd8 ffffffffa0035b2b

[ 3778.173582]  ffffffff00000005 fffffbfff5630bf4 0000000041b58ab3 ffffffffaaaf18e0

[ 3778.173591]  ffffffffa00359bc ffff8807c98aee00 ffff8807c98aeda0 000000000bb7299a

[ 3778.173593] Call Trace:

[ 3778.173606] dump_stack (lib/dump_stack.c:53)
[ 3778.173618] ? arch_local_irq_restore (./arch/x86/include/asm/paravirt.h:134)
[ 3778.173650] ubsan_epilogue (lib/ubsan.c:165)
[ 3778.173662] __ubsan_handle_shift_out_of_bounds (lib/ubsan.c:421)
[ 3778.173672] ? warn_slowpath_fmt (kernel/panic.c:526)
[ 3778.173684] ? __ubsan_handle_load_invalid_value (lib/ubsan.c:388)
[ 3778.173696] ? kernel_poison_pages (mm/page_poison.c:169)
[ 3778.173707] ? kasan_unpoison_shadow (mm/kasan/kasan.c:59)
[ 3778.173718] ? kasan_alloc_pages (mm/kasan/kasan.c:344)
[ 3778.173727] ? post_alloc_hook (mm/page_alloc.c:1740 (discriminator 3) include/linux/page_owner.h:27 (discriminator 3) mm/page_alloc.c:1741 (discriminator 3))
[ 3778.173738] ? __list_add (lib/list_debug.c:45)
[ 3778.173749] ? __this_cpu_preempt_check (lib/list_debug.c:25)
[ 3778.173758] map_pages (mm/compaction.c:76 (discriminator 1))
[ 3778.173766] ? map_pages (mm/compaction.c:76 (discriminator 1))
[ 3778.173775] ? compaction_free (mm/compaction.c:67)
[ 3778.173785] compaction_alloc (mm/compaction.c:1136 mm/compaction.c:1156)
[ 3778.173795] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3778.173805] ? __page_cache_release (mm/swap.c:73)
[ 3778.173817] migrate_pages (mm/migrate.c:1079 mm/migrate.c:1325)
[ 3778.173827] ? __reset_isolation_suitable (mm/compaction.c:1175)
[ 3778.173836] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3778.173848] ? buffer_migrate_page (mm/migrate.c:1301)
[ 3778.173857] compact_zone (mm/compaction.c:1555)
[ 3778.173867] ? compaction_restarting (mm/compaction.c:1476)
[ 3778.173877] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.173887] compact_zone_order (mm/compaction.c:1653)
[ 3778.173901] ? kick_process (kernel/sched/core.c:2692)
[ 3778.173911] ? compact_zone (mm/compaction.c:1637)
[ 3778.173923] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3778.173933] try_to_compact_pages (mm/compaction.c:1717)
[ 3778.173944] ? compaction_zonelist_suitable (mm/compaction.c:1679)
[ 3778.173955] __alloc_pages_direct_compact (mm/page_alloc.c:3180)
[ 3778.173966] ? get_page_from_freelist (mm/page_alloc.c:3172)
[ 3778.173977] __alloc_pages_slowpath (mm/page_alloc.c:3741)
[ 3778.173987] ? kvm_clock_read (./arch/x86/include/asm/preempt.h:84 arch/x86/kernel/kvmclock.c:92)
[ 3778.173998] ? __alloc_pages_direct_compact (mm/page_alloc.c:3546)
[ 3778.174012] ? get_page_from_freelist (mm/page_alloc.c:2950)
[ 3778.174023] ? release_pages (mm/swap.c:731)
[ 3778.174033] ? __isolate_free_page (mm/page_alloc.c:2883)
[ 3778.174044] ? ___might_sleep (kernel/sched/core.c:7540 (discriminator 1))
[ 3778.174055] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3778.174066] __alloc_pages_nodemask (mm/page_alloc.c:3841)
[ 3778.174075] ? rwsem_wake (kernel/locking/rwsem-xadd.c:580)
[ 3778.174086] ? __alloc_pages_slowpath (mm/page_alloc.c:3757)
[ 3778.174095] ? call_rwsem_wake (arch/x86/lib/rwsem.S:129)
[ 3778.174104] ? up_write (kernel/locking/rwsem.c:112)
[ 3778.174115] ? pmdp_huge_clear_flush (mm/pgtable-generic.c:131)
[ 3778.174125] khugepaged_alloc_page (mm/khugepaged.c:752)
[ 3778.174135] collapse_huge_page (mm/khugepaged.c:948)
[ 3778.174145] ? khugepaged_scan_shmem (mm/khugepaged.c:922)
[ 3778.174157] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3778.174167] ? remove_wait_queue (kernel/sched/wait.c:292)
[ 3778.174177] khugepaged (mm/khugepaged.c:1724 mm/khugepaged.c:1799 mm/khugepaged.c:1848)
[ 3778.174186] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.174197] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3778.174208] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3778.174218] ? default_wake_function (kernel/sched/core.c:3544)
[ 3778.174228] ? __wake_up_common (kernel/sched/wait.c:73)
[ 3778.174237] ? __kthread_parkme (kernel/kthread.c:168)
[ 3778.174246] kthread (kernel/kthread.c:209)
[ 3778.174255] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3778.174264] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.174274] ret_from_fork (arch/x86/entry/entry_64.S:390)
[ 3778.174284] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.174287] ================================================================================

[ 3778.174354] ================================================================================

[ 3778.174362] UBSAN: Undefined behaviour in mm/page_alloc.c:1738:27

[ 3778.174369] shift exponent 88341344 is too large for 32-bit type 'int'

[ 3778.174380] CPU: 5 PID: 3270 Comm: khugepaged Tainted: G        W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101

[ 3778.174391]  1ffff100f9315d96 000000000bb7299a ffff8807c98aed38 ffffffffa0035b2b

[ 3778.174400]  ffffffff00000005 fffffbfff5630bf4 0000000041b58ab3 ffffffffaaaf18e0

[ 3778.174410]  ffffffffa00359bc ffff8807c98aed60 ffff8807c98aed00 000000000bb7299a

[ 3778.174412] Call Trace:

[ 3778.174426] dump_stack (lib/dump_stack.c:53)
[ 3778.174438] ? arch_local_irq_restore (./arch/x86/include/asm/paravirt.h:134)
[ 3778.174448] ubsan_epilogue (lib/ubsan.c:165)
[ 3778.174459] __ubsan_handle_shift_out_of_bounds (lib/ubsan.c:421)
[ 3778.174471] ? __ubsan_handle_load_invalid_value (lib/ubsan.c:388)
[ 3778.174481] ? _raw_spin_unlock_irqrestore (kernel/locking/spinlock.c:192)
[ 3778.174493] ? __ubsan_handle_shift_out_of_bounds (lib/ubsan.c:388 (discriminator 1))
[ 3778.174503] ? warn_slowpath_fmt (kernel/panic.c:526)
[ 3778.174520] ? __ubsan_handle_load_invalid_value (lib/ubsan.c:388)
[ 3778.174534] ? kernel_poison_pages (mm/page_poison.c:169)
[ 3778.174544] post_alloc_hook (mm/page_alloc.c:1739 (discriminator 1))
[ 3778.174553] ? post_alloc_hook (mm/page_alloc.c:1739 (discriminator 1))
[ 3778.174563] ? clear_zone_contiguous (mm/page_alloc.c:1733)
[ 3778.174572] map_pages (mm/compaction.c:78 (discriminator 1))
[ 3778.174582] ? compaction_free (mm/compaction.c:67)
[ 3778.174592] compaction_alloc (mm/compaction.c:1136 mm/compaction.c:1156)
[ 3778.174602] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3778.174613] ? __page_cache_release (mm/swap.c:73)
[ 3778.174629] migrate_pages (mm/migrate.c:1079 mm/migrate.c:1325)
[ 3778.174640] ? __reset_isolation_suitable (mm/compaction.c:1175)
[ 3778.174650] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3778.174661] ? buffer_migrate_page (mm/migrate.c:1301)
[ 3778.174671] compact_zone (mm/compaction.c:1555)
[ 3778.174681] ? compaction_restarting (mm/compaction.c:1476)
[ 3778.174691] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.174701] compact_zone_order (mm/compaction.c:1653)
[ 3778.174712] ? kick_process (kernel/sched/core.c:2692)
[ 3778.174721] ? compact_zone (mm/compaction.c:1637)
[ 3778.174733] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3778.174743] try_to_compact_pages (mm/compaction.c:1717)
[ 3778.174754] ? compaction_zonelist_suitable (mm/compaction.c:1679)
[ 3778.174765] __alloc_pages_direct_compact (mm/page_alloc.c:3180)
[ 3778.174775] ? get_page_from_freelist (mm/page_alloc.c:3172)
[ 3778.174786] __alloc_pages_slowpath (mm/page_alloc.c:3741)
[ 3778.174796] ? kvm_clock_read (./arch/x86/include/asm/preempt.h:84 arch/x86/kernel/kvmclock.c:92)
[ 3778.174807] ? __alloc_pages_direct_compact (mm/page_alloc.c:3546)
[ 3778.174818] ? get_page_from_freelist (mm/page_alloc.c:2950)
[ 3778.174829] ? release_pages (mm/swap.c:731)
[ 3778.174839] ? __isolate_free_page (mm/page_alloc.c:2883)
[ 3778.174850] ? ___might_sleep (kernel/sched/core.c:7540 (discriminator 1))
[ 3778.174861] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3778.174871] __alloc_pages_nodemask (mm/page_alloc.c:3841)
[ 3778.174881] ? rwsem_wake (kernel/locking/rwsem-xadd.c:580)
[ 3778.174891] ? __alloc_pages_slowpath (mm/page_alloc.c:3757)
[ 3778.174901] ? call_rwsem_wake (arch/x86/lib/rwsem.S:129)
[ 3778.174909] ? up_write (kernel/locking/rwsem.c:112)
[ 3778.174920] ? pmdp_huge_clear_flush (mm/pgtable-generic.c:131)
[ 3778.174930] khugepaged_alloc_page (mm/khugepaged.c:752)
[ 3778.174940] collapse_huge_page (mm/khugepaged.c:948)
[ 3778.174950] ? khugepaged_scan_shmem (mm/khugepaged.c:922)
[ 3778.174962] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3778.174972] ? remove_wait_queue (kernel/sched/wait.c:292)
[ 3778.174982] khugepaged (mm/khugepaged.c:1724 mm/khugepaged.c:1799 mm/khugepaged.c:1848)
[ 3778.174991] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.175006] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3778.175019] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3778.175029] ? default_wake_function (kernel/sched/core.c:3544)
[ 3778.175040] ? __wake_up_common (kernel/sched/wait.c:73)
[ 3778.175049] ? __kthread_parkme (kernel/kthread.c:168)
[ 3778.175058] kthread (kernel/kthread.c:209)
[ 3778.175067] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3778.175077] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.175087] ret_from_fork (arch/x86/entry/entry_64.S:390)
[ 3778.175096] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.175100] ================================================================================

[ 3778.175130] ================================================================================

[ 3778.175142] UBSAN: Undefined behaviour in mm/page_alloc.c:1739:30

[ 3778.175149] shift exponent 88341344 is too large for 32-bit type 'int'

[ 3778.175160] CPU: 5 PID: 3270 Comm: khugepaged Tainted: G        W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101

[ 3778.175171]  1ffff100f9315d96 000000000bb7299a ffff8807c98aed38 ffffffffa0035b2b

[ 3778.175180]  ffffffff00000005 fffffbfff5630bf4 0000000041b58ab3 ffffffffaaaf18e0

[ 3778.175190]  ffffffffa00359bc ffff8807c98aed60 ffff8807c98aed00 000000000bb7299a

[ 3778.175192] Call Trace:

[ 3778.175205] dump_stack (lib/dump_stack.c:53)
[ 3778.175217] ? arch_local_irq_restore (./arch/x86/include/asm/paravirt.h:134)
[ 3778.175227] ubsan_epilogue (lib/ubsan.c:165)
[ 3778.175239] __ubsan_handle_shift_out_of_bounds (lib/ubsan.c:421)
[ 3778.175250] ? __ubsan_handle_load_invalid_value (lib/ubsan.c:388)
[ 3778.175261] ? _raw_spin_unlock_irqrestore (kernel/locking/spinlock.c:192)
[ 3778.175273] ? __ubsan_handle_shift_out_of_bounds (lib/ubsan.c:388 (discriminator 1))
[ 3778.175283] ? warn_slowpath_fmt (kernel/panic.c:526)
[ 3778.175295] ? __ubsan_handle_load_invalid_value (lib/ubsan.c:388)
[ 3778.175307] ? kernel_poison_pages (mm/page_poison.c:169)
[ 3778.175317] post_alloc_hook (mm/page_alloc.c:1739 (discriminator 3))
[ 3778.175326] ? post_alloc_hook (mm/page_alloc.c:1739 (discriminator 3))
[ 3778.175336] ? clear_zone_contiguous (mm/page_alloc.c:1733)
[ 3778.175345] map_pages (mm/compaction.c:78 (discriminator 1))
[ 3778.175354] ? compaction_free (mm/compaction.c:67)
[ 3778.175364] compaction_alloc (mm/compaction.c:1136 mm/compaction.c:1156)
[ 3778.175374] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3778.175385] ? __page_cache_release (mm/swap.c:73)
[ 3778.175397] migrate_pages (mm/migrate.c:1079 mm/migrate.c:1325)
[ 3778.175407] ? __reset_isolation_suitable (mm/compaction.c:1175)
[ 3778.175417] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3778.175428] ? buffer_migrate_page (mm/migrate.c:1301)
[ 3778.175438] compact_zone (mm/compaction.c:1555)
[ 3778.175448] ? compaction_restarting (mm/compaction.c:1476)
[ 3778.175458] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.175467] compact_zone_order (mm/compaction.c:1653)
[ 3778.175478] ? kick_process (kernel/sched/core.c:2692)
[ 3778.175487] ? compact_zone (mm/compaction.c:1637)
[ 3778.175499] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3778.175509] try_to_compact_pages (mm/compaction.c:1717)
[ 3778.175520] ? compaction_zonelist_suitable (mm/compaction.c:1679)
[ 3778.175531] __alloc_pages_direct_compact (mm/page_alloc.c:3180)
[ 3778.175542] ? get_page_from_freelist (mm/page_alloc.c:3172)
[ 3778.175552] __alloc_pages_slowpath (mm/page_alloc.c:3741)
[ 3778.175562] ? kvm_clock_read (./arch/x86/include/asm/preempt.h:84 arch/x86/kernel/kvmclock.c:92)
[ 3778.175573] ? __alloc_pages_direct_compact (mm/page_alloc.c:3546)
[ 3778.175584] ? get_page_from_freelist (mm/page_alloc.c:2950)
[ 3778.175595] ? release_pages (mm/swap.c:731)
[ 3778.175605] ? __isolate_free_page (mm/page_alloc.c:2883)
[ 3778.175616] ? ___might_sleep (kernel/sched/core.c:7540 (discriminator 1))
[ 3778.175627] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3778.175637] __alloc_pages_nodemask (mm/page_alloc.c:3841)
[ 3778.175647] ? rwsem_wake (kernel/locking/rwsem-xadd.c:580)
[ 3778.175657] ? __alloc_pages_slowpath (mm/page_alloc.c:3757)
[ 3778.175666] ? call_rwsem_wake (arch/x86/lib/rwsem.S:129)
[ 3778.175675] ? up_write (kernel/locking/rwsem.c:112)
[ 3778.175686] ? pmdp_huge_clear_flush (mm/pgtable-generic.c:131)
[ 3778.175696] khugepaged_alloc_page (mm/khugepaged.c:752)
[ 3778.175705] collapse_huge_page (mm/khugepaged.c:948)
[ 3778.175719] ? khugepaged_scan_shmem (mm/khugepaged.c:922)
[ 3778.175731] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3778.175742] ? remove_wait_queue (kernel/sched/wait.c:292)
[ 3778.175752] khugepaged (mm/khugepaged.c:1724 mm/khugepaged.c:1799 mm/khugepaged.c:1848)
[ 3778.175761] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.175772] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3778.175784] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3778.175794] ? default_wake_function (kernel/sched/core.c:3544)
[ 3778.175809] ? __wake_up_common (kernel/sched/wait.c:73)
[ 3778.175819] ? __kthread_parkme (kernel/kthread.c:168)
[ 3778.175827] kthread (kernel/kthread.c:209)
[ 3778.175837] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3778.175847] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.175857] ret_from_fork (arch/x86/entry/entry_64.S:390)
[ 3778.175866] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.175870] ================================================================================

[ 3778.175918] kasan: CONFIG_KASAN_INLINE enabled

[ 3778.175937] kasan: GPF could be caused by NULL-ptr deref or user memory access

[ 3778.175938] general protection fault: 0000 [#1] PREEMPT SMP KASAN

[ 3778.175942] Modules linked in:

[ 3778.175952] CPU: 5 PID: 3270 Comm: khugepaged Tainted: G        W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101

[ 3778.175959] task: ffff8807c9892000 ti: ffff8807c98a8000 task.ti: ffff8807c98a8000

[ 3778.175980] RIP: memset_orig (arch/x86/lib/memset_64.S:88)
[ 3778.175985] RSP: 0000:ffff8807c98aeeb8  EFLAGS: 00010216

[ 3778.175991] RAX: 0000000000000000 RBX: fffcdf0dbfebfe00 RCX: 00000007ffffffff

[ 3778.175997] RDX: 0000020000000000 RSI: 0000000000000000 RDI: fffcdd0dbfebfe00

[ 3778.176006] RBP: ffff8807c98aeed0 R08: 0000000000000001 R09: 0000000000000000

[ 3778.176012] R10: fffcdd0dbfebfe00 R11: 00000000e27286ff R12: 0000100000000000

[ 3778.176020] R13: 0000000000000008 R14: 000000000543fb60 R15: 0000000000000001

[ 3778.176028] FS:  0000000000000000(0000) GS:ffff880951c00000(0000) knlGS:0000000000000000

[ 3778.176034] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033

[ 3778.176039] CR2: 00007f2eb1c17576 CR3: 000000002ae23000 CR4: 00000000000006a0

[ 3778.176050] Stack:

[ 3778.176061]  ffffffff9e7a02a5 ffff8801b7fd7fe0 1ffff100f9315ddf ffff8807c98aeee0

[ 3778.176071]  ffffffff9e7a05e8 ffff8807c98aef80 ffffffff9e67716a 000000000543fb60

[ 3778.176080]  0000000041b58ab3 ffffffffaaaf18e0 ffffffff9e676f60 ffffea0005294f20

[ 3778.176082] Call Trace:

[ 3778.176094] ? kasan_unpoison_shadow (mm/kasan/kasan.c:59)
[ 3778.176105] kasan_alloc_pages (mm/kasan/kasan.c:344)
[ 3778.176115] post_alloc_hook (mm/page_alloc.c:1740 (discriminator 3) include/linux/page_owner.h:27 (discriminator 3) mm/page_alloc.c:1741 (discriminator 3))
[ 3778.176125] ? clear_zone_contiguous (mm/page_alloc.c:1733)
[ 3778.176135] map_pages (mm/compaction.c:78 (discriminator 1))
[ 3778.176144] ? compaction_free (mm/compaction.c:67)
[ 3778.176154] compaction_alloc (mm/compaction.c:1136 mm/compaction.c:1156)
[ 3778.176165] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3778.176176] ? __page_cache_release (mm/swap.c:73)
[ 3778.176188] migrate_pages (mm/migrate.c:1079 mm/migrate.c:1325)
[ 3778.176199] ? __reset_isolation_suitable (mm/compaction.c:1175)
[ 3778.176209] ? isolate_freepages_block (mm/compaction.c:1146)
[ 3778.176221] ? buffer_migrate_page (mm/migrate.c:1301)
[ 3778.176230] compact_zone (mm/compaction.c:1555)
[ 3778.176241] ? compaction_restarting (mm/compaction.c:1476)
[ 3778.176251] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.176261] compact_zone_order (mm/compaction.c:1653)
[ 3778.176272] ? kick_process (kernel/sched/core.c:2692)
[ 3778.176281] ? compact_zone (mm/compaction.c:1637)
[ 3778.176294] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3778.176308] try_to_compact_pages (mm/compaction.c:1717)
[ 3778.176320] ? compaction_zonelist_suitable (mm/compaction.c:1679)
[ 3778.176331] __alloc_pages_direct_compact (mm/page_alloc.c:3180)
[ 3778.176342] ? get_page_from_freelist (mm/page_alloc.c:3172)
[ 3778.176353] __alloc_pages_slowpath (mm/page_alloc.c:3741)
[ 3778.176363] ? kvm_clock_read (./arch/x86/include/asm/preempt.h:84 arch/x86/kernel/kvmclock.c:92)
[ 3778.176375] ? __alloc_pages_direct_compact (mm/page_alloc.c:3546)
[ 3778.176386] ? get_page_from_freelist (mm/page_alloc.c:2950)
[ 3778.176397] ? release_pages (mm/swap.c:731)
[ 3778.176408] ? __isolate_free_page (mm/page_alloc.c:2883)
[ 3778.176419] ? ___might_sleep (kernel/sched/core.c:7540 (discriminator 1))
[ 3778.176431] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3778.176441] __alloc_pages_nodemask (mm/page_alloc.c:3841)
[ 3778.176451] ? rwsem_wake (kernel/locking/rwsem-xadd.c:580)
[ 3778.176462] ? __alloc_pages_slowpath (mm/page_alloc.c:3757)
[ 3778.176471] ? call_rwsem_wake (arch/x86/lib/rwsem.S:129)
[ 3778.176480] ? up_write (kernel/locking/rwsem.c:112)
[ 3778.176496] ? pmdp_huge_clear_flush (mm/pgtable-generic.c:131)
[ 3778.176506] khugepaged_alloc_page (mm/khugepaged.c:752)
[ 3778.176516] collapse_huge_page (mm/khugepaged.c:948)
[ 3778.176527] ? khugepaged_scan_shmem (mm/khugepaged.c:922)
[ 3778.176538] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
[ 3778.176549] ? remove_wait_queue (kernel/sched/wait.c:292)
[ 3778.176559] khugepaged (mm/khugepaged.c:1724 mm/khugepaged.c:1799 mm/khugepaged.c:1848)
[ 3778.176569] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.176580] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3778.176591] ? io_schedule_timeout (kernel/sched/core.c:3266)
[ 3778.176601] ? default_wake_function (kernel/sched/core.c:3544)
[ 3778.176612] ? __wake_up_common (kernel/sched/wait.c:73)
[ 3778.176621] ? __kthread_parkme (kernel/kthread.c:168)
[ 3778.176630] kthread (kernel/kthread.c:209)
[ 3778.176640] ? collapse_huge_page (mm/khugepaged.c:1840)
[ 3778.176649] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.176660] ret_from_fork (arch/x86/entry/entry_64.S:390)
[ 3778.176669] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.176769] Code: b8 01 01 01 01 01 01 01 01 48 0f af c1 41 89 f9 41 83 e1 07 75 70 48 89 d1 48 c1 e9 06 74 39 66 0f 1f 84 00 00 00 00 00 48 ff c9 <48> 89 07 48 89 47 08 48 89 47 10 48 89 47 18 48 89 47 20 48 89

All code
========
   0:	b8 01 01 01 01       	mov    $0x1010101,%eax
   5:	01 01                	add    %eax,(%rcx)
   7:	01 01                	add    %eax,(%rcx)
   9:	48 0f af c1          	imul   %rcx,%rax
   d:	41 89 f9             	mov    %edi,%r9d
  10:	41 83 e1 07          	and    $0x7,%r9d
  14:	75 70                	jne    0x86
  16:	48 89 d1             	mov    %rdx,%rcx
  19:	48 c1 e9 06          	shr    $0x6,%rcx
  1d:	74 39                	je     0x58
  1f:	66 0f 1f 84 00 00 00 	nopw   0x0(%rax,%rax,1)
  26:	00 00
  28:	48 ff c9             	dec    %rcx
  2b:*	48 89 07             	mov    %rax,(%rdi)		<-- trapping instruction
  2e:	48 89 47 08          	mov    %rax,0x8(%rdi)
  32:	48 89 47 10          	mov    %rax,0x10(%rdi)
  36:	48 89 47 18          	mov    %rax,0x18(%rdi)
  3a:	48 89 47 20          	mov    %rax,0x20(%rdi)
  3e:	48 89 00             	mov    %rax,(%rax)

Code starting with the faulting instruction
===========================================
   0:	48 89 07             	mov    %rax,(%rdi)
   3:	48 89 47 08          	mov    %rax,0x8(%rdi)
   7:	48 89 47 10          	mov    %rax,0x10(%rdi)
   b:	48 89 47 18          	mov    %rax,0x18(%rdi)
   f:	48 89 47 20          	mov    %rax,0x20(%rdi)
  13:	48 89 00             	mov    %rax,(%rax)
[ 3778.176779] RIP memset_orig (arch/x86/lib/memset_64.S:88)
[ 3778.176782]  RSP <ffff8807c98aeeb8>

[ 3778.177119] ---[ end trace 8cbbecee22435ccc ]---

[ 3778.177125] Kernel panic - not syncing: Fatal exception

[ 3778.443313] list_add corruption. next->prev should be prev (ffff8801b7fd8208), but was ffffea0004e45ba0. (next=ffffea0004e46020).

[ 3778.444993] Modules linked in:

[ 3778.445301] CPU: 16 PID: 3730 Comm: kswapd0 Tainted: G      D W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101

[ 3778.446284]  1ffff10066e36de3 00000000be0b1ffd ffff8803371b6fa0 ffffffffa0035b2b

[ 3778.447026]  ffffffff00000010 fffffbfff5630bf4 0000000041b58ab3 ffffffffaaaf18e0

[ 3778.447753]  ffffffffa00359bc ffffffff9e54d4a0 ffffffffa8b2ab20 ffff8803371b7020

[ 3778.448495] Call Trace:

[ 3778.448751] dump_stack (lib/dump_stack.c:53)
[ 3778.449235] ? arch_local_irq_restore (./arch/x86/include/asm/paravirt.h:134)
[ 3778.449828] ? is_module_text_address (kernel/module.c:4185)
[ 3778.450417] ? __list_add (lib/list_debug.c:30 (discriminator 3))
[ 3778.450911] ? vprintk_default (kernel/printk/printk.c:1886)
[ 3778.451459] ? __list_add (lib/list_debug.c:30 (discriminator 3))
[ 3778.451968] __warn (kernel/panic.c:518)
[ 3778.452424] warn_slowpath_fmt (kernel/panic.c:526)
[ 3778.452954] ? __warn (kernel/panic.c:526)
[ 3778.453443] ? __radix_tree_lookup (lib/radix-tree.c:673)
[ 3778.454018] __list_add (lib/list_debug.c:30 (discriminator 3))
[ 3778.454498] ? __this_cpu_preempt_check (lib/list_debug.c:25)
[ 3778.455084] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
[ 3778.455677] free_pcppages_bulk (mm/page_alloc.c:878 mm/page_alloc.c:1135)
[ 3778.456236] ? __free_page_frag (mm/page_alloc.c:1084)
[ 3778.456775] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.457395] free_hot_cold_page (mm/page_alloc.c:2442 (discriminator 1))
[ 3778.457940] ? drain_all_pages (mm/page_alloc.c:2403)
[ 3778.458480] ? uncharge_list (mm/memcontrol.c:5536)
[ 3778.459007] ? __this_cpu_preempt_check (lib/list_debug.c:25)
[ 3778.459587] free_hot_cold_page_list (mm/page_alloc.c:2456 (discriminator 3))
[ 3778.460152] shrink_page_list (include/linux/compiler.h:222 include/linux/list.h:189 include/linux/list.h:296 mm/vmscan.c:1238)
[ 3778.460695] ? putback_lru_page (mm/vmscan.c:889)
[ 3778.461235] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.461874] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.462474] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
[ 3778.463069] ? __mod_zone_page_state (mm/vmstat.c:249)
[ 3778.463648] shrink_inactive_list (include/linux/spinlock.h:332 mm/vmscan.c:1629)
[ 3778.464217] ? putback_inactive_pages (mm/vmscan.c:1573)
[ 3778.464829] ? zone_reclaimable_pages (mm/vmscan.c:208)
[ 3778.465434] ? _find_next_bit (lib/find_bit.c:54)
[ 3778.466010] ? zone_reclaimable (include/linux/vmstat.h:166 mm/vmscan.c:212)
[ 3778.466570] ? get_scan_count (mm/vmscan.c:1954)
[ 3778.467127] ? blk_start_plug (block/blk-core.c:3170 (discriminator 1))
[ 3778.467658] ? blk_lld_busy (block/blk-core.c:3170)
[ 3778.468165] shrink_zone_memcg (mm/vmscan.c:1932 mm/vmscan.c:2224)
[ 3778.468700] ? shrink_active_list (mm/vmscan.c:2181)
[ 3778.469267] ? css_next_descendant_pre (kernel/cgroup.c:4028)
[ 3778.469959] ? mem_cgroup_iter (mm/memcontrol.c:889)
[ 3778.470506] ? preempt_count_add (include/linux/ftrace.h:724 kernel/sched/core.c:3074 kernel/sched/core.c:3099)
[ 3778.471043] shrink_zone (mm/vmscan.c:2407)
[ 3778.471536] ? mem_cgroup_split_huge_fixup (mm/memcontrol.c:2607)
[ 3778.472192] ? shrink_zone_memcg (mm/vmscan.c:2374)
[ 3778.472747] ? zone_watermark_ok_safe (mm/page_alloc.c:2839)
[ 3778.473340] kswapd (./arch/x86/include/asm/bitops.h:113 mm/vmscan.c:3090 mm/vmscan.c:3239 mm/vmscan.c:3427)
[ 3778.473825] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.474407] ? mem_cgroup_shrink_node_zone (mm/vmscan.c:3351)
[ 3778.475182] ? __schedule (kernel/sched/core.c:2858 kernel/sched/core.c:3345)
[ 3778.475703] ? remove_wait_queue (kernel/sched/wait.c:292)
[ 3778.476267] ? default_wake_function (kernel/sched/core.c:3544)
[ 3778.476901] ? __kthread_parkme (kernel/kthread.c:168)
[ 3778.477489] kthread (kernel/kthread.c:209)
[ 3778.478009] ? mem_cgroup_shrink_node_zone (mm/vmscan.c:3351)
[ 3778.478703] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.479376] ret_from_fork (arch/x86/entry/entry_64.S:390)
[ 3778.479951] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.480508] ---[ end trace 8cbbecee22435ccd ]---

[ 3778.480937] kasan: CONFIG_KASAN_INLINE enabled

[ 3778.481333] kasan: GPF could be caused by NULL-ptr deref or user memory access[ 3778.482013] general protection fault: 0000 [#2] PREEMPT SMP KASAN

[ 3778.482568] Modules linked in:

[ 3778.482877] CPU: 16 PID: 3730 Comm: kswapd0 Tainted: G      D W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101

[ 3778.483856] task: ffff880337162000 ti: ffff8803371b0000 task.ti: ffff8803371b0000

[ 3778.484527] RIP: __list_add (lib/list_debug.c:30 (discriminator 3))
[ 3778.485307] RSP: 0000:ffff8803371b70d8  EFLAGS: 00010002

[ 3778.485782] RAX: dffffc0000000000 RBX: ffffea0005294020 RCX: ffffea000529409c

[ 3778.486410] RDX: 1bd5a00000000040 RSI: dead000000000200 RDI: ffffea0005294020

[ 3778.487054] RBP: ffff8803371b7160 R08: dead000000000200 R09: ffff8801b7fd7000

[ 3778.487716] R10: 0000000000000010 R11: ffff880a8a24ecef R12: dead000000000200

[ 3778.488378] R13: ffff8801b7fd8000 R14: 1ffff10066e36e1b R15: ffff8801b7fd8008

[ 3778.489091] FS:  0000000000000000(0000) GS:ffff8801b1b00000(0000) knlGS:0000000000000000

[ 3778.489816] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033

[ 3778.490342] CR2: 00007f2eb1c17576 CR3: 000000002ae23000 CR4: 00000000000006a0

[ 3778.491003] Stack:

[ 3778.491204]  0000000041b58ab3 ffffffffaaaf1975 ffffffffa00adae0 000000000002fcf2

[ 3778.491943]  0000000000000014 ffff8801b7fd7000 ffff8803371b7120 ffffffffa00adadc

[ 3778.492684]  0000000000000000 ffff8803371b7160 ffffffff9e6ccc36 00000000be0b1ffd

[ 3778.493418] Call Trace:

[ 3778.493674] ? __this_cpu_preempt_check (lib/list_debug.c:25)
[ 3778.494270] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
[ 3778.494853] ? __mod_zone_page_state (mm/vmstat.c:249)
[ 3778.495431] free_pcppages_bulk (include/linux/list.h:77 mm/page_alloc.c:870 mm/page_alloc.c:1135)
[ 3778.496024] ? __free_page_frag (mm/page_alloc.c:1084)
[ 3778.496571] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.497222] free_hot_cold_page (mm/page_alloc.c:2442 (discriminator 1))
[ 3778.497793] ? drain_all_pages (mm/page_alloc.c:2403)
[ 3778.498370] ? uncharge_list (mm/memcontrol.c:5536)
[ 3778.498913] ? __this_cpu_preempt_check (lib/list_debug.c:25)
[ 3778.499526] free_hot_cold_page_list (mm/page_alloc.c:2456 (discriminator 3))
[ 3778.500130] shrink_page_list (include/linux/compiler.h:222 include/linux/list.h:189 include/linux/list.h:296 mm/vmscan.c:1238)
[ 3778.500704] ? putback_lru_page (mm/vmscan.c:889)
[ 3778.501271] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.501898] ? check_preemption_disabled (lib/smp_processor_id.c:52)
[ 3778.502546] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
[ 3778.503135] ? __mod_zone_page_state (mm/vmstat.c:249)
[ 3778.503737] shrink_inactive_list (include/linux/spinlock.h:332 mm/vmscan.c:1629)
[ 3778.504310] ? putback_inactive_pages (mm/vmscan.c:1573)
[ 3778.504917] ? zone_reclaimable_pages (mm/vmscan.c:208)
[ 3778.505515] ? _find_next_bit (lib/find_bit.c:54)
[ 3778.506044] ? zone_reclaimable (include/linux/vmstat.h:166 mm/vmscan.c:212)
[ 3778.506599] ? get_scan_count (mm/vmscan.c:1954)
[ 3778.507156] ? blk_start_plug (block/blk-core.c:3170 (discriminator 1))
[ 3778.507688] ? blk_lld_busy (block/blk-core.c:3170)
[ 3778.508212] shrink_zone_memcg (mm/vmscan.c:1932 mm/vmscan.c:2224)
[ 3778.508767] ? shrink_active_list (mm/vmscan.c:2181)
[ 3778.509342] ? css_next_descendant_pre (kernel/cgroup.c:4028)
[ 3778.509935] ? mem_cgroup_iter (mm/memcontrol.c:889)
[ 3778.510493] ? preempt_count_add (include/linux/ftrace.h:724 kernel/sched/core.c:3074 kernel/sched/core.c:3099)
[ 3778.511044] shrink_zone (mm/vmscan.c:2407)
[ 3778.511540] ? mem_cgroup_split_huge_fixup (mm/memcontrol.c:2607)
[ 3778.512185] ? shrink_zone_memcg (mm/vmscan.c:2374)
[ 3778.512760] ? zone_watermark_ok_safe (mm/page_alloc.c:2839)
[ 3778.513374] kswapd (./arch/x86/include/asm/bitops.h:113 mm/vmscan.c:3090 mm/vmscan.c:3239 mm/vmscan.c:3427)
[ 3778.513911] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
[ 3778.514491] ? mem_cgroup_shrink_node_zone (mm/vmscan.c:3351)
[ 3778.515141] ? __schedule (kernel/sched/core.c:2858 kernel/sched/core.c:3345)
[ 3778.515665] ? remove_wait_queue (kernel/sched/wait.c:292)
[ 3778.516212] ? default_wake_function (kernel/sched/core.c:3544)
[ 3778.516795] ? __kthread_parkme (kernel/kthread.c:168)
[ 3778.517345] kthread (kernel/kthread.c:209)
[ 3778.517816] ? mem_cgroup_shrink_node_zone (mm/vmscan.c:3351)
[ 3778.518438] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.518983] ret_from_fork (arch/x86/entry/entry_64.S:390)
[ 3778.519496] ? kthread_worker_fn (kernel/kthread.c:178)
[ 3778.520040] Code: b2 a8 e8 ea 1b 2d fe 4d 85 e4 75 0e 31 f6 48 c7 c7 40 34 71 ae e8 c3 89 08 00 48 b8 00 00 00 00 00 fc ff df 4c 89 e2 48 c1 ea 03 <80> 3c 02 00 74 08 4c 89 e7 e8 88 3d 6f fe 4d 8b 04 24 4d 39 e8

All code
========
   0:	b2 a8                	mov    $0xa8,%dl
   2:	e8 ea 1b 2d fe       	callq  0xfffffffffe2d1bf1
   7:	4d 85 e4             	test   %r12,%r12
   a:	75 0e                	jne    0x1a
   c:	31 f6                	xor    %esi,%esi
   e:	48 c7 c7 40 34 71 ae 	mov    $0xffffffffae713440,%rdi
  15:	e8 c3 89 08 00       	callq  0x889dd
  1a:	48 b8 00 00 00 00 00 	movabs $0xdffffc0000000000,%rax
  21:	fc ff df
  24:	4c 89 e2             	mov    %r12,%rdx
  27:	48 c1 ea 03          	shr    $0x3,%rdx
  2b:*	80 3c 02 00          	cmpb   $0x0,(%rdx,%rax,1)		<-- trapping instruction
  2f:	74 08                	je     0x39
  31:	4c 89 e7             	mov    %r12,%rdi
  34:	e8 88 3d 6f fe       	callq  0xfffffffffe6f3dc1
  39:	4d 8b 04 24          	mov    (%r12),%r8
  3d:	4d 39 e8             	cmp    %r13,%r8
	...

Code starting with the faulting instruction
===========================================
   0:	80 3c 02 00          	cmpb   $0x0,(%rdx,%rax,1)
   4:	74 08                	je     0xe
   6:	4c 89 e7             	mov    %r12,%rdi
   9:	e8 88 3d 6f fe       	callq  0xfffffffffe6f3d96
   e:	4d 8b 04 24          	mov    (%r12),%r8
  12:	4d 39 e8             	cmp    %r13,%r8
	...
[ 3778.522780] RIP __list_add (lib/list_debug.c:30 (discriminator 3))
[ 3778.523311]  RSP <ffff8803371b70d8>

[ 3778.523654] ---[ end trace 8cbbecee22435cce ]---

[ 3779.304858] Shutting down cpus with NMI

[ 3779.306134] Kernel Offset: 0x1d000000 from 0xffffffff81000000 (relocation range: 0xffffffff80000000-0xffffffffbfffffff)

[ 3779.307309] Rebooting in 1 seconds..

Re: [PATCH v2 1/7] mm/compaction: split freepages without holding the zone lock

[Joonsoo Kim]

On Mon, Jun 13, 2016 at 04:31:15PM -0400, Sasha Levin wrote:
> On 05/25/2016 10:37 PM, js1304@gmail.com wrote:
> > From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
> > 
> > We don't need to split freepages with holding the zone lock. It will cause
> > more contention on zone lock so not desirable.
> > 
> > Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
> 
> Hey Joonsoo,

Hello, Sasha.
> 
> I'm seeing the following corruption/crash which seems to be related to
> this patch:

Could you tell me why you think that following corruption is related
to this patch? list_del() in __isolate_free_page() is unchanged part.

Before this patch, we did it by split_free_page() ->
__isolate_free_page() -> list_del(). With this patch, we do it by
calling __isolate_free_page() directly.

Thanks.


> [ 3777.807224] ------------[ cut here ]------------
> 
> [ 3777.807834] WARNING: CPU: 5 PID: 3270 at lib/list_debug.c:62 __list_del_entry+0x14e/0x280
> 
> [ 3777.808562] list_del corruption. next->prev should be ffffea0004a76120, but was ffffea0004a72120
> 
> [ 3777.809498] Modules linked in:
> 
> [ 3777.809923] CPU: 5 PID: 3270 Comm: khugepaged Tainted: G        W       4.7.0-rc2-next-20160609-sasha-00024-g30ecaf6 #3101
> 
> [ 3777.811014]  1ffff100f9315d7b 000000000bb7299a ffff8807c98aec60 ffffffffa0035b2b
> 
> [ 3777.811816]  ffffffff00000005 fffffbfff5630bf4 0000000041b58ab3 ffffffffaaaf18e0
> 
> [ 3777.812662]  ffffffffa00359bc ffffffff9e54d4a0 ffffffffa8b2ade0 ffff8807c98aece0
> 
> [ 3777.813493] Call Trace:
> 
> [ 3777.813796] dump_stack (lib/dump_stack.c:53)
> [ 3777.814310] ? arch_local_irq_restore (./arch/x86/include/asm/paravirt.h:134)
> [ 3777.814947] ? is_module_text_address (kernel/module.c:4185)
> [ 3777.815571] ? __list_del_entry (lib/list_debug.c:60 (discriminator 1))
> [ 3777.816174] ? vprintk_default (kernel/printk/printk.c:1886)
> [ 3777.816761] ? __list_del_entry (lib/list_debug.c:60 (discriminator 1))
> [ 3777.817381] __warn (kernel/panic.c:518)
> [ 3777.817867] warn_slowpath_fmt (kernel/panic.c:526)
> [ 3777.818428] ? __warn (kernel/panic.c:526)
> [ 3777.819001] ? __schedule (kernel/sched/core.c:2858 kernel/sched/core.c:3345)
> [ 3777.819541] __list_del_entry (lib/list_debug.c:60 (discriminator 1))
> [ 3777.820116] ? __list_add (lib/list_debug.c:45)
> [ 3777.820721] ? __this_cpu_preempt_check (lib/smp_processor_id.c:63)
> [ 3777.821347] list_del (lib/list_debug.c:78)
> [ 3777.821829] __isolate_free_page (mm/page_alloc.c:2514)
> [ 3777.822400] ? __zone_watermark_ok (mm/page_alloc.c:2493)
> [ 3777.823007] isolate_freepages_block (mm/compaction.c:498)
> [ 3777.823629] ? compact_unlock_should_abort (mm/compaction.c:417)
> [ 3777.824312] compaction_alloc (mm/compaction.c:1112 mm/compaction.c:1156)
> [ 3777.824871] ? isolate_freepages_block (mm/compaction.c:1146)
> [ 3777.825512] ? __page_cache_release (mm/swap.c:73)
> [ 3777.826127] migrate_pages (mm/migrate.c:1079 mm/migrate.c:1325)
> [ 3777.826712] ? __reset_isolation_suitable (mm/compaction.c:1175)
> [ 3777.827398] ? isolate_freepages_block (mm/compaction.c:1146)
> [ 3777.828109] ? buffer_migrate_page (mm/migrate.c:1301)
> [ 3777.828727] compact_zone (mm/compaction.c:1555)
> [ 3777.829290] ? compaction_restarting (mm/compaction.c:1476)
> [ 3777.829969] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
> [ 3777.830607] compact_zone_order (mm/compaction.c:1653)
> [ 3777.831204] ? kick_process (kernel/sched/core.c:2692)
> [ 3777.831774] ? compact_zone (mm/compaction.c:1637)
> [ 3777.832336] ? io_schedule_timeout (kernel/sched/core.c:3266)
> [ 3777.832934] try_to_compact_pages (mm/compaction.c:1717)
> [ 3777.833550] ? compaction_zonelist_suitable (mm/compaction.c:1679)
> [ 3777.834265] __alloc_pages_direct_compact (mm/page_alloc.c:3180)
> [ 3777.834922] ? get_page_from_freelist (mm/page_alloc.c:3172)
> [ 3777.835549] __alloc_pages_slowpath (mm/page_alloc.c:3741)
> [ 3777.836210] ? kvm_clock_read (./arch/x86/include/asm/preempt.h:84 arch/x86/kernel/kvmclock.c:92)
> [ 3777.836744] ? __alloc_pages_direct_compact (mm/page_alloc.c:3546)
> [ 3777.837429] ? get_page_from_freelist (mm/page_alloc.c:2950)
> [ 3777.838072] ? release_pages (mm/swap.c:731)
> [ 3777.838610] ? __isolate_free_page (mm/page_alloc.c:2883)
> [ 3777.839209] ? ___might_sleep (kernel/sched/core.c:7540 (discriminator 1))
> [ 3777.839826] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
> [ 3777.840427] __alloc_pages_nodemask (mm/page_alloc.c:3841)
> [ 3777.841071] ? rwsem_wake (kernel/locking/rwsem-xadd.c:580)
> [ 3777.841608] ? __alloc_pages_slowpath (mm/page_alloc.c:3757)
> [ 3777.842253] ? call_rwsem_wake (arch/x86/lib/rwsem.S:129)
> [ 3777.842839] ? up_write (kernel/locking/rwsem.c:112)
> [ 3777.843350] ? pmdp_huge_clear_flush (mm/pgtable-generic.c:131)
> [ 3777.844125] khugepaged_alloc_page (mm/khugepaged.c:752)
> [ 3777.844719] collapse_huge_page (mm/khugepaged.c:948)
> [ 3777.845332] ? khugepaged_scan_shmem (mm/khugepaged.c:922)
> [ 3777.846020] ? __might_sleep (kernel/sched/core.c:7532 (discriminator 14))
> [ 3777.846608] ? remove_wait_queue (kernel/sched/wait.c:292)
> [ 3777.847181] khugepaged (mm/khugepaged.c:1724 mm/khugepaged.c:1799 mm/khugepaged.c:1848)
> [ 3777.847704] ? _raw_spin_unlock_irq (./arch/x86/include/asm/preempt.h:92 include/linux/spinlock_api_smp.h:171 kernel/locking/spinlock.c:199)
> [ 3777.848297] ? collapse_huge_page (mm/khugepaged.c:1840)
> [ 3777.848950] ? io_schedule_timeout (kernel/sched/core.c:3266)
> [ 3777.849555] ? default_wake_function (kernel/sched/core.c:3544)
> [ 3777.850161] ? __wake_up_common (kernel/sched/wait.c:73)
> [ 3777.850724] ? __kthread_parkme (kernel/kthread.c:168)
> [ 3777.851306] kthread (kernel/kthread.c:209)
> [ 3777.851819] ? collapse_huge_page (mm/khugepaged.c:1840)
> [ 3777.852448] ? kthread_worker_fn (kernel/kthread.c:178)
> [ 3777.853045] ret_from_fork (arch/x86/entry/entry_64.S:390)
> [ 3777.853605] ? kthread_worker_fn (kernel/kthread.c:178)
> [ 3777.854173] ---[ end trace 8cbbecee22435cc2 ]---
> 

Re: [PATCH v2 1/7] mm/compaction: split freepages without holding the zone lock

[Sasha Levin]

On 06/14/2016 01:52 AM, Joonsoo Kim wrote:
> On Mon, Jun 13, 2016 at 04:31:15PM -0400, Sasha Levin wrote:
>> > On 05/25/2016 10:37 PM, js1304@gmail.com wrote:
>>> > > From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>>> > > 
>>> > > We don't need to split freepages with holding the zone lock. It will cause
>>> > > more contention on zone lock so not desirable.
>>> > > 
>>> > > Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>> > 
>> > Hey Joonsoo,
> Hello, Sasha.
>> > 
>> > I'm seeing the following corruption/crash which seems to be related to
>> > this patch:
> Could you tell me why you think that following corruption is related
> to this patch? list_del() in __isolate_free_page() is unchanged part.
> 
> Before this patch, we did it by split_free_page() ->
> __isolate_free_page() -> list_del(). With this patch, we do it by
> calling __isolate_free_page() directly.

I haven't bisected it, but it's the first time I see this issue and this
commit seems to have done related changes that might cause this.

I can go ahead with bisection if you don't think it's related.


Thanks,
Sasha

Re: [PATCH v2 1/7] mm/compaction: split freepages without holding the zone lock

[Joonsoo Kim]

On Tue, Jun 14, 2016 at 03:10:21PM -0400, Sasha Levin wrote:
> On 06/14/2016 01:52 AM, Joonsoo Kim wrote:
> > On Mon, Jun 13, 2016 at 04:31:15PM -0400, Sasha Levin wrote:
> >> > On 05/25/2016 10:37 PM, js1304@gmail.com wrote:
> >>> > > From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
> >>> > > 
> >>> > > We don't need to split freepages with holding the zone lock. It will cause
> >>> > > more contention on zone lock so not desirable.
> >>> > > 
> >>> > > Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
> >> > 
> >> > Hey Joonsoo,
> > Hello, Sasha.
> >> > 
> >> > I'm seeing the following corruption/crash which seems to be related to
> >> > this patch:
> > Could you tell me why you think that following corruption is related
> > to this patch? list_del() in __isolate_free_page() is unchanged part.
> > 
> > Before this patch, we did it by split_free_page() ->
> > __isolate_free_page() -> list_del(). With this patch, we do it by
> > calling __isolate_free_page() directly.
> 
> I haven't bisected it, but it's the first time I see this issue and this
> commit seems to have done related changes that might cause this.
> 
> I can go ahead with bisection if you don't think it's related.

Hmm... I can't find a bug in this patch for now. There are more candidates
on this area hat changed by me and it would be very helpful if you can
do bisection.

Thanks.

Re: [PATCH v2 1/7] mm/compaction: split freepages without holding the zone lock

[Joonsoo Kim]

On Wed, Jun 15, 2016 at 11:27:31AM +0900, Joonsoo Kim wrote:
> On Tue, Jun 14, 2016 at 03:10:21PM -0400, Sasha Levin wrote:
> > On 06/14/2016 01:52 AM, Joonsoo Kim wrote:
> > > On Mon, Jun 13, 2016 at 04:31:15PM -0400, Sasha Levin wrote:
> > >> > On 05/25/2016 10:37 PM, js1304@gmail.com wrote:
> > >>> > > From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
> > >>> > > 
> > >>> > > We don't need to split freepages with holding the zone lock. It will cause
> > >>> > > more contention on zone lock so not desirable.
> > >>> > > 
> > >>> > > Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
> > >> > 
> > >> > Hey Joonsoo,
> > > Hello, Sasha.
> > >> > 
> > >> > I'm seeing the following corruption/crash which seems to be related to
> > >> > this patch:
> > > Could you tell me why you think that following corruption is related
> > > to this patch? list_del() in __isolate_free_page() is unchanged part.
> > > 
> > > Before this patch, we did it by split_free_page() ->
> > > __isolate_free_page() -> list_del(). With this patch, we do it by
> > > calling __isolate_free_page() directly.
> > 
> > I haven't bisected it, but it's the first time I see this issue and this
> > commit seems to have done related changes that might cause this.
> > 
> > I can go ahead with bisection if you don't think it's related.
> 
> Hmm... I can't find a bug in this patch for now. There are more candidates
> on this area hat changed by me and it would be very helpful if you can
> do bisection.

Hello, Sasha.

You are right! Minchan found the bug in this patch! I will send
updated patch soon.

http://marc.info/?i=20160616100932.GS17127%40bbox

Thanks.