* [PATCH] mm: __set_page_dirty_nobuffers uses spin_lock_irqseve instead of spin_lock_irq
@ 2014-02-03 16:49 ` kosaki.motohiro
0 siblings, 0 replies; 10+ messages in thread
From: kosaki.motohiro @ 2014-02-03 16:49 UTC (permalink / raw
To: linux-kernel
Cc: linux-mm, akpm, KOSAKI Motohiro, Larry Woodman, Rik van Riel,
Johannes Weiner, stable
From: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
During aio stress test, we observed the following lockdep warning.
This mean AIO+numa_balancing is currently deadlockable.
The problem is, aio_migratepage disable interrupt, but __set_page_dirty_nobuffers
unintentionally enable it again.
Generally, all helper function should use spin_lock_irqsave()
instead of spin_lock_irq() because they don't know caller at all.
[ 599.843948] other info that might help us debug this:
[ 599.873748] Possible unsafe locking scenario:
[ 599.873748]
[ 599.900902] CPU0
[ 599.912701] ----
[ 599.924929] lock(&(&ctx->completion_lock)->rlock);
[ 599.950299] <Interrupt>
[ 599.962576] lock(&(&ctx->completion_lock)->rlock);
[ 599.985771]
[ 599.985771] *** DEADLOCK ***
[ 600.375623] [<ffffffff81678d3c>] dump_stack+0x19/0x1b
[ 600.398769] [<ffffffff816731aa>] print_usage_bug+0x1f7/0x208
[ 600.425092] [<ffffffff810df370>] ? print_shortest_lock_dependencies+0x1d0/0x1d0
[ 600.458981] [<ffffffff810e08dd>] mark_lock+0x21d/0x2a0
[ 600.482910] [<ffffffff810e0a19>] mark_held_locks+0xb9/0x140
[ 600.508956] [<ffffffff8168201c>] ? _raw_spin_unlock_irq+0x2c/0x50
[ 600.536825] [<ffffffff810e0ba5>] trace_hardirqs_on_caller+0x105/0x1d0
[ 600.566861] [<ffffffff810e0c7d>] trace_hardirqs_on+0xd/0x10
[ 600.593210] [<ffffffff8168201c>] _raw_spin_unlock_irq+0x2c/0x50
[ 600.620599] [<ffffffff8117f72c>] __set_page_dirty_nobuffers+0x8c/0xf0
[ 600.649992] [<ffffffff811d1094>] migrate_page_copy+0x434/0x540
[ 600.676635] [<ffffffff8123f5b1>] aio_migratepage+0xb1/0x140
[ 600.703126] [<ffffffff811d126d>] move_to_new_page+0x7d/0x230
[ 600.729022] [<ffffffff811d1b45>] migrate_pages+0x5e5/0x700
[ 600.754705] [<ffffffff811d0070>] ? buffer_migrate_lock_buffers+0xb0/0xb0
[ 600.785784] [<ffffffff811d29cc>] migrate_misplaced_page+0xbc/0xf0
[ 600.814029] [<ffffffff8119eb62>] do_numa_page+0x102/0x190
[ 600.839182] [<ffffffff8119ee31>] handle_pte_fault+0x241/0x970
[ 600.865875] [<ffffffff811a0345>] handle_mm_fault+0x265/0x370
[ 600.892071] [<ffffffff81686d82>] __do_page_fault+0x172/0x5a0
[ 600.918065] [<ffffffff81682cd8>] ? retint_swapgs+0x13/0x1b
[ 600.943493] [<ffffffff816871ca>] do_page_fault+0x1a/0x70
[ 600.968081] [<ffffffff81682ff8>] page_fault+0x28/0x30
Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Larry Woodman <lwoodman@redhat.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Johannes Weiner <jweiner@redhat.com>
Cc: stable@vger.kernel.org
---
mm/page-writeback.c | 5 +++--
1 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/mm/page-writeback.c b/mm/page-writeback.c
index 2d30e2c..7106cb1 100644
--- a/mm/page-writeback.c
+++ b/mm/page-writeback.c
@@ -2173,11 +2173,12 @@ int __set_page_dirty_nobuffers(struct page *page)
if (!TestSetPageDirty(page)) {
struct address_space *mapping = page_mapping(page);
struct address_space *mapping2;
+ unsigned long flags;
if (!mapping)
return 1;
- spin_lock_irq(&mapping->tree_lock);
+ spin_lock_irqsave(&mapping->tree_lock, flags);
mapping2 = page_mapping(page);
if (mapping2) { /* Race with truncate? */
BUG_ON(mapping2 != mapping);
@@ -2186,7 +2187,7 @@ int __set_page_dirty_nobuffers(struct page *page)
radix_tree_tag_set(&mapping->page_tree,
page_index(page), PAGECACHE_TAG_DIRTY);
}
- spin_unlock_irq(&mapping->tree_lock);
+ spin_unlock_irqrestore(&mapping->tree_lock, flags);
if (mapping->host) {
/* !PageAnon && !swapper_space */
__mark_inode_dirty(mapping->host, I_DIRTY_PAGES);
--
1.7.1
^ permalink raw reply related [flat|nested] 10+ messages in thread
* [PATCH] mm: __set_page_dirty_nobuffers uses spin_lock_irqseve instead of spin_lock_irq
@ 2014-02-03 16:49 ` kosaki.motohiro
0 siblings, 0 replies; 10+ messages in thread
From: kosaki.motohiro @ 2014-02-03 16:49 UTC (permalink / raw
To: linux-kernel
Cc: linux-mm, akpm, KOSAKI Motohiro, Larry Woodman, Rik van Riel,
Johannes Weiner, stable
From: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
During aio stress test, we observed the following lockdep warning.
This mean AIO+numa_balancing is currently deadlockable.
The problem is, aio_migratepage disable interrupt, but __set_page_dirty_nobuffers
unintentionally enable it again.
Generally, all helper function should use spin_lock_irqsave()
instead of spin_lock_irq() because they don't know caller at all.
[ 599.843948] other info that might help us debug this:
[ 599.873748] Possible unsafe locking scenario:
[ 599.873748]
[ 599.900902] CPU0
[ 599.912701] ----
[ 599.924929] lock(&(&ctx->completion_lock)->rlock);
[ 599.950299] <Interrupt>
[ 599.962576] lock(&(&ctx->completion_lock)->rlock);
[ 599.985771]
[ 599.985771] *** DEADLOCK ***
[ 600.375623] [<ffffffff81678d3c>] dump_stack+0x19/0x1b
[ 600.398769] [<ffffffff816731aa>] print_usage_bug+0x1f7/0x208
[ 600.425092] [<ffffffff810df370>] ? print_shortest_lock_dependencies+0x1d0/0x1d0
[ 600.458981] [<ffffffff810e08dd>] mark_lock+0x21d/0x2a0
[ 600.482910] [<ffffffff810e0a19>] mark_held_locks+0xb9/0x140
[ 600.508956] [<ffffffff8168201c>] ? _raw_spin_unlock_irq+0x2c/0x50
[ 600.536825] [<ffffffff810e0ba5>] trace_hardirqs_on_caller+0x105/0x1d0
[ 600.566861] [<ffffffff810e0c7d>] trace_hardirqs_on+0xd/0x10
[ 600.593210] [<ffffffff8168201c>] _raw_spin_unlock_irq+0x2c/0x50
[ 600.620599] [<ffffffff8117f72c>] __set_page_dirty_nobuffers+0x8c/0xf0
[ 600.649992] [<ffffffff811d1094>] migrate_page_copy+0x434/0x540
[ 600.676635] [<ffffffff8123f5b1>] aio_migratepage+0xb1/0x140
[ 600.703126] [<ffffffff811d126d>] move_to_new_page+0x7d/0x230
[ 600.729022] [<ffffffff811d1b45>] migrate_pages+0x5e5/0x700
[ 600.754705] [<ffffffff811d0070>] ? buffer_migrate_lock_buffers+0xb0/0xb0
[ 600.785784] [<ffffffff811d29cc>] migrate_misplaced_page+0xbc/0xf0
[ 600.814029] [<ffffffff8119eb62>] do_numa_page+0x102/0x190
[ 600.839182] [<ffffffff8119ee31>] handle_pte_fault+0x241/0x970
[ 600.865875] [<ffffffff811a0345>] handle_mm_fault+0x265/0x370
[ 600.892071] [<ffffffff81686d82>] __do_page_fault+0x172/0x5a0
[ 600.918065] [<ffffffff81682cd8>] ? retint_swapgs+0x13/0x1b
[ 600.943493] [<ffffffff816871ca>] do_page_fault+0x1a/0x70
[ 600.968081] [<ffffffff81682ff8>] page_fault+0x28/0x30
Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Larry Woodman <lwoodman@redhat.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Johannes Weiner <jweiner@redhat.com>
Cc: stable@vger.kernel.org
---
mm/page-writeback.c | 5 +++--
1 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/mm/page-writeback.c b/mm/page-writeback.c
index 2d30e2c..7106cb1 100644
--- a/mm/page-writeback.c
+++ b/mm/page-writeback.c
@@ -2173,11 +2173,12 @@ int __set_page_dirty_nobuffers(struct page *page)
if (!TestSetPageDirty(page)) {
struct address_space *mapping = page_mapping(page);
struct address_space *mapping2;
+ unsigned long flags;
if (!mapping)
return 1;
- spin_lock_irq(&mapping->tree_lock);
+ spin_lock_irqsave(&mapping->tree_lock, flags);
mapping2 = page_mapping(page);
if (mapping2) { /* Race with truncate? */
BUG_ON(mapping2 != mapping);
@@ -2186,7 +2187,7 @@ int __set_page_dirty_nobuffers(struct page *page)
radix_tree_tag_set(&mapping->page_tree,
page_index(page), PAGECACHE_TAG_DIRTY);
}
- spin_unlock_irq(&mapping->tree_lock);
+ spin_unlock_irqrestore(&mapping->tree_lock, flags);
if (mapping->host) {
/* !PageAnon && !swapper_space */
__mark_inode_dirty(mapping->host, I_DIRTY_PAGES);
--
1.7.1
--
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>
^ permalink raw reply related [flat|nested] 10+ messages in thread
* Re: [PATCH] mm: __set_page_dirty_nobuffers uses spin_lock_irqseve instead of spin_lock_irq
2014-02-03 16:49 ` kosaki.motohiro
@ 2014-02-03 21:11 ` David Rientjes
-1 siblings, 0 replies; 10+ messages in thread
From: David Rientjes @ 2014-02-03 21:11 UTC (permalink / raw
To: kosaki.motohiro
Cc: linux-kernel, linux-mm, akpm, KOSAKI Motohiro, Larry Woodman,
Rik van Riel, Johannes Weiner, stable
On Mon, 3 Feb 2014, kosaki.motohiro@gmail.com wrote:
> From: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
>
> During aio stress test, we observed the following lockdep warning.
> This mean AIO+numa_balancing is currently deadlockable.
>
> The problem is, aio_migratepage disable interrupt, but __set_page_dirty_nobuffers
> unintentionally enable it again.
>
> Generally, all helper function should use spin_lock_irqsave()
> instead of spin_lock_irq() because they don't know caller at all.
>
> [ 599.843948] other info that might help us debug this:
> [ 599.873748] Possible unsafe locking scenario:
> [ 599.873748]
> [ 599.900902] CPU0
> [ 599.912701] ----
> [ 599.924929] lock(&(&ctx->completion_lock)->rlock);
> [ 599.950299] <Interrupt>
> [ 599.962576] lock(&(&ctx->completion_lock)->rlock);
> [ 599.985771]
> [ 599.985771] *** DEADLOCK ***
>
> [ 600.375623] [<ffffffff81678d3c>] dump_stack+0x19/0x1b
> [ 600.398769] [<ffffffff816731aa>] print_usage_bug+0x1f7/0x208
> [ 600.425092] [<ffffffff810df370>] ? print_shortest_lock_dependencies+0x1d0/0x1d0
> [ 600.458981] [<ffffffff810e08dd>] mark_lock+0x21d/0x2a0
> [ 600.482910] [<ffffffff810e0a19>] mark_held_locks+0xb9/0x140
> [ 600.508956] [<ffffffff8168201c>] ? _raw_spin_unlock_irq+0x2c/0x50
> [ 600.536825] [<ffffffff810e0ba5>] trace_hardirqs_on_caller+0x105/0x1d0
> [ 600.566861] [<ffffffff810e0c7d>] trace_hardirqs_on+0xd/0x10
> [ 600.593210] [<ffffffff8168201c>] _raw_spin_unlock_irq+0x2c/0x50
> [ 600.620599] [<ffffffff8117f72c>] __set_page_dirty_nobuffers+0x8c/0xf0
> [ 600.649992] [<ffffffff811d1094>] migrate_page_copy+0x434/0x540
> [ 600.676635] [<ffffffff8123f5b1>] aio_migratepage+0xb1/0x140
> [ 600.703126] [<ffffffff811d126d>] move_to_new_page+0x7d/0x230
> [ 600.729022] [<ffffffff811d1b45>] migrate_pages+0x5e5/0x700
> [ 600.754705] [<ffffffff811d0070>] ? buffer_migrate_lock_buffers+0xb0/0xb0
> [ 600.785784] [<ffffffff811d29cc>] migrate_misplaced_page+0xbc/0xf0
> [ 600.814029] [<ffffffff8119eb62>] do_numa_page+0x102/0x190
> [ 600.839182] [<ffffffff8119ee31>] handle_pte_fault+0x241/0x970
> [ 600.865875] [<ffffffff811a0345>] handle_mm_fault+0x265/0x370
> [ 600.892071] [<ffffffff81686d82>] __do_page_fault+0x172/0x5a0
> [ 600.918065] [<ffffffff81682cd8>] ? retint_swapgs+0x13/0x1b
> [ 600.943493] [<ffffffff816871ca>] do_page_fault+0x1a/0x70
> [ 600.968081] [<ffffffff81682ff8>] page_fault+0x28/0x30
>
> Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
> Cc: Larry Woodman <lwoodman@redhat.com>
> Cc: Rik van Riel <riel@redhat.com>
> Cc: Johannes Weiner <jweiner@redhat.com>
> Cc: stable@vger.kernel.org
> ---
> mm/page-writeback.c | 5 +++--
> 1 files changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/mm/page-writeback.c b/mm/page-writeback.c
> index 2d30e2c..7106cb1 100644
> --- a/mm/page-writeback.c
> +++ b/mm/page-writeback.c
> @@ -2173,11 +2173,12 @@ int __set_page_dirty_nobuffers(struct page *page)
> if (!TestSetPageDirty(page)) {
> struct address_space *mapping = page_mapping(page);
> struct address_space *mapping2;
> + unsigned long flags;
>
> if (!mapping)
> return 1;
>
> - spin_lock_irq(&mapping->tree_lock);
> + spin_lock_irqsave(&mapping->tree_lock, flags);
> mapping2 = page_mapping(page);
> if (mapping2) { /* Race with truncate? */
> BUG_ON(mapping2 != mapping);
> @@ -2186,7 +2187,7 @@ int __set_page_dirty_nobuffers(struct page *page)
> radix_tree_tag_set(&mapping->page_tree,
> page_index(page), PAGECACHE_TAG_DIRTY);
> }
> - spin_unlock_irq(&mapping->tree_lock);
> + spin_unlock_irqrestore(&mapping->tree_lock, flags);
> if (mapping->host) {
> /* !PageAnon && !swapper_space */
> __mark_inode_dirty(mapping->host, I_DIRTY_PAGES);
Indeed, good catch. Do we need the same treatment for
__set_page_dirty_buffers() that can be called by way of
clear_page_dirty_for_io()?
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH] mm: __set_page_dirty_nobuffers uses spin_lock_irqseve instead of spin_lock_irq
@ 2014-02-03 21:11 ` David Rientjes
0 siblings, 0 replies; 10+ messages in thread
From: David Rientjes @ 2014-02-03 21:11 UTC (permalink / raw
To: kosaki.motohiro
Cc: linux-kernel, linux-mm, akpm, KOSAKI Motohiro, Larry Woodman,
Rik van Riel, Johannes Weiner, stable
On Mon, 3 Feb 2014, kosaki.motohiro@gmail.com wrote:
> From: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
>
> During aio stress test, we observed the following lockdep warning.
> This mean AIO+numa_balancing is currently deadlockable.
>
> The problem is, aio_migratepage disable interrupt, but __set_page_dirty_nobuffers
> unintentionally enable it again.
>
> Generally, all helper function should use spin_lock_irqsave()
> instead of spin_lock_irq() because they don't know caller at all.
>
> [ 599.843948] other info that might help us debug this:
> [ 599.873748] Possible unsafe locking scenario:
> [ 599.873748]
> [ 599.900902] CPU0
> [ 599.912701] ----
> [ 599.924929] lock(&(&ctx->completion_lock)->rlock);
> [ 599.950299] <Interrupt>
> [ 599.962576] lock(&(&ctx->completion_lock)->rlock);
> [ 599.985771]
> [ 599.985771] *** DEADLOCK ***
>
> [ 600.375623] [<ffffffff81678d3c>] dump_stack+0x19/0x1b
> [ 600.398769] [<ffffffff816731aa>] print_usage_bug+0x1f7/0x208
> [ 600.425092] [<ffffffff810df370>] ? print_shortest_lock_dependencies+0x1d0/0x1d0
> [ 600.458981] [<ffffffff810e08dd>] mark_lock+0x21d/0x2a0
> [ 600.482910] [<ffffffff810e0a19>] mark_held_locks+0xb9/0x140
> [ 600.508956] [<ffffffff8168201c>] ? _raw_spin_unlock_irq+0x2c/0x50
> [ 600.536825] [<ffffffff810e0ba5>] trace_hardirqs_on_caller+0x105/0x1d0
> [ 600.566861] [<ffffffff810e0c7d>] trace_hardirqs_on+0xd/0x10
> [ 600.593210] [<ffffffff8168201c>] _raw_spin_unlock_irq+0x2c/0x50
> [ 600.620599] [<ffffffff8117f72c>] __set_page_dirty_nobuffers+0x8c/0xf0
> [ 600.649992] [<ffffffff811d1094>] migrate_page_copy+0x434/0x540
> [ 600.676635] [<ffffffff8123f5b1>] aio_migratepage+0xb1/0x140
> [ 600.703126] [<ffffffff811d126d>] move_to_new_page+0x7d/0x230
> [ 600.729022] [<ffffffff811d1b45>] migrate_pages+0x5e5/0x700
> [ 600.754705] [<ffffffff811d0070>] ? buffer_migrate_lock_buffers+0xb0/0xb0
> [ 600.785784] [<ffffffff811d29cc>] migrate_misplaced_page+0xbc/0xf0
> [ 600.814029] [<ffffffff8119eb62>] do_numa_page+0x102/0x190
> [ 600.839182] [<ffffffff8119ee31>] handle_pte_fault+0x241/0x970
> [ 600.865875] [<ffffffff811a0345>] handle_mm_fault+0x265/0x370
> [ 600.892071] [<ffffffff81686d82>] __do_page_fault+0x172/0x5a0
> [ 600.918065] [<ffffffff81682cd8>] ? retint_swapgs+0x13/0x1b
> [ 600.943493] [<ffffffff816871ca>] do_page_fault+0x1a/0x70
> [ 600.968081] [<ffffffff81682ff8>] page_fault+0x28/0x30
>
> Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
> Cc: Larry Woodman <lwoodman@redhat.com>
> Cc: Rik van Riel <riel@redhat.com>
> Cc: Johannes Weiner <jweiner@redhat.com>
> Cc: stable@vger.kernel.org
> ---
> mm/page-writeback.c | 5 +++--
> 1 files changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/mm/page-writeback.c b/mm/page-writeback.c
> index 2d30e2c..7106cb1 100644
> --- a/mm/page-writeback.c
> +++ b/mm/page-writeback.c
> @@ -2173,11 +2173,12 @@ int __set_page_dirty_nobuffers(struct page *page)
> if (!TestSetPageDirty(page)) {
> struct address_space *mapping = page_mapping(page);
> struct address_space *mapping2;
> + unsigned long flags;
>
> if (!mapping)
> return 1;
>
> - spin_lock_irq(&mapping->tree_lock);
> + spin_lock_irqsave(&mapping->tree_lock, flags);
> mapping2 = page_mapping(page);
> if (mapping2) { /* Race with truncate? */
> BUG_ON(mapping2 != mapping);
> @@ -2186,7 +2187,7 @@ int __set_page_dirty_nobuffers(struct page *page)
> radix_tree_tag_set(&mapping->page_tree,
> page_index(page), PAGECACHE_TAG_DIRTY);
> }
> - spin_unlock_irq(&mapping->tree_lock);
> + spin_unlock_irqrestore(&mapping->tree_lock, flags);
> if (mapping->host) {
> /* !PageAnon && !swapper_space */
> __mark_inode_dirty(mapping->host, I_DIRTY_PAGES);
Indeed, good catch. Do we need the same treatment for
__set_page_dirty_buffers() that can be called by way of
clear_page_dirty_for_io()?
--
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>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH] mm: __set_page_dirty_nobuffers uses spin_lock_irqseve instead of spin_lock_irq
2014-02-03 21:11 ` David Rientjes
@ 2014-02-04 17:10 ` KOSAKI Motohiro
-1 siblings, 0 replies; 10+ messages in thread
From: KOSAKI Motohiro @ 2014-02-04 17:10 UTC (permalink / raw
To: David Rientjes
Cc: LKML, linux-mm@kvack.org, Andrew Morton, Larry Woodman,
Rik van Riel, Johannes Weiner, stable
> Indeed, good catch. Do we need the same treatment for
> __set_page_dirty_buffers() that can be called by way of
> clear_page_dirty_for_io()?
Indeed. I posted a patch fixed __set_page_dirty() too. plz see
Subject: [PATCH] __set_page_dirty uses spin_lock_irqsave instead of
spin_lock_irq
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH] mm: __set_page_dirty_nobuffers uses spin_lock_irqseve instead of spin_lock_irq
@ 2014-02-04 17:10 ` KOSAKI Motohiro
0 siblings, 0 replies; 10+ messages in thread
From: KOSAKI Motohiro @ 2014-02-04 17:10 UTC (permalink / raw
To: David Rientjes
Cc: LKML, linux-mm@kvack.org, Andrew Morton, Larry Woodman,
Rik van Riel, Johannes Weiner, stable
> Indeed, good catch. Do we need the same treatment for
> __set_page_dirty_buffers() that can be called by way of
> clear_page_dirty_for_io()?
Indeed. I posted a patch fixed __set_page_dirty() too. plz see
Subject: [PATCH] __set_page_dirty uses spin_lock_irqsave instead of
spin_lock_irq
--
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>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH] mm: __set_page_dirty_nobuffers uses spin_lock_irqseve instead of spin_lock_irq
2014-02-03 16:49 ` kosaki.motohiro
@ 2014-02-06 6:35 ` Yasuaki Ishimatsu
-1 siblings, 0 replies; 10+ messages in thread
From: Yasuaki Ishimatsu @ 2014-02-06 6:35 UTC (permalink / raw
To: kosaki.motohiro
Cc: linux-kernel, linux-mm, akpm, KOSAKI Motohiro, Larry Woodman,
Rik van Riel, Johannes Weiner, stable
(2014/02/04 1:49), kosaki.motohiro@gmail.com wrote:
> From: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
>
> During aio stress test, we observed the following lockdep warning.
> This mean AIO+numa_balancing is currently deadlockable.
>
> The problem is, aio_migratepage disable interrupt, but __set_page_dirty_nobuffers
> unintentionally enable it again.
>
> Generally, all helper function should use spin_lock_irqsave()
> instead of spin_lock_irq() because they don't know caller at all.
>
> [ 599.843948] other info that might help us debug this:
> [ 599.873748] Possible unsafe locking scenario:
> [ 599.873748]
> [ 599.900902] CPU0
> [ 599.912701] ----
> [ 599.924929] lock(&(&ctx->completion_lock)->rlock);
> [ 599.950299] <Interrupt>
> [ 599.962576] lock(&(&ctx->completion_lock)->rlock);
> [ 599.985771]
> [ 599.985771] *** DEADLOCK ***
>
> [ 600.375623] [<ffffffff81678d3c>] dump_stack+0x19/0x1b
> [ 600.398769] [<ffffffff816731aa>] print_usage_bug+0x1f7/0x208
> [ 600.425092] [<ffffffff810df370>] ? print_shortest_lock_dependencies+0x1d0/0x1d0
> [ 600.458981] [<ffffffff810e08dd>] mark_lock+0x21d/0x2a0
> [ 600.482910] [<ffffffff810e0a19>] mark_held_locks+0xb9/0x140
> [ 600.508956] [<ffffffff8168201c>] ? _raw_spin_unlock_irq+0x2c/0x50
> [ 600.536825] [<ffffffff810e0ba5>] trace_hardirqs_on_caller+0x105/0x1d0
> [ 600.566861] [<ffffffff810e0c7d>] trace_hardirqs_on+0xd/0x10
> [ 600.593210] [<ffffffff8168201c>] _raw_spin_unlock_irq+0x2c/0x50
> [ 600.620599] [<ffffffff8117f72c>] __set_page_dirty_nobuffers+0x8c/0xf0
> [ 600.649992] [<ffffffff811d1094>] migrate_page_copy+0x434/0x540
> [ 600.676635] [<ffffffff8123f5b1>] aio_migratepage+0xb1/0x140
> [ 600.703126] [<ffffffff811d126d>] move_to_new_page+0x7d/0x230
> [ 600.729022] [<ffffffff811d1b45>] migrate_pages+0x5e5/0x700
> [ 600.754705] [<ffffffff811d0070>] ? buffer_migrate_lock_buffers+0xb0/0xb0
> [ 600.785784] [<ffffffff811d29cc>] migrate_misplaced_page+0xbc/0xf0
> [ 600.814029] [<ffffffff8119eb62>] do_numa_page+0x102/0x190
> [ 600.839182] [<ffffffff8119ee31>] handle_pte_fault+0x241/0x970
> [ 600.865875] [<ffffffff811a0345>] handle_mm_fault+0x265/0x370
> [ 600.892071] [<ffffffff81686d82>] __do_page_fault+0x172/0x5a0
> [ 600.918065] [<ffffffff81682cd8>] ? retint_swapgs+0x13/0x1b
> [ 600.943493] [<ffffffff816871ca>] do_page_fault+0x1a/0x70
> [ 600.968081] [<ffffffff81682ff8>] page_fault+0x28/0x30
>
> Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
> Cc: Larry Woodman <lwoodman@redhat.com>
> Cc: Rik van Riel <riel@redhat.com>
> Cc: Johannes Weiner <jweiner@redhat.com>
> Cc: stable@vger.kernel.org
> ---
Tested-by: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Thank you for posting the patch.
The same issue occurred on my box. And I confirmed that the issue
disappeared by the patch.
Thanks,
Yasuaki Ishimatsu
> mm/page-writeback.c | 5 +++--
> 1 files changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/mm/page-writeback.c b/mm/page-writeback.c
> index 2d30e2c..7106cb1 100644
> --- a/mm/page-writeback.c
> +++ b/mm/page-writeback.c
> @@ -2173,11 +2173,12 @@ int __set_page_dirty_nobuffers(struct page *page)
> if (!TestSetPageDirty(page)) {
> struct address_space *mapping = page_mapping(page);
> struct address_space *mapping2;
> + unsigned long flags;
>
> if (!mapping)
> return 1;
>
> - spin_lock_irq(&mapping->tree_lock);
> + spin_lock_irqsave(&mapping->tree_lock, flags);
> mapping2 = page_mapping(page);
> if (mapping2) { /* Race with truncate? */
> BUG_ON(mapping2 != mapping);
> @@ -2186,7 +2187,7 @@ int __set_page_dirty_nobuffers(struct page *page)
> radix_tree_tag_set(&mapping->page_tree,
> page_index(page), PAGECACHE_TAG_DIRTY);
> }
> - spin_unlock_irq(&mapping->tree_lock);
> + spin_unlock_irqrestore(&mapping->tree_lock, flags);
> if (mapping->host) {
> /* !PageAnon && !swapper_space */
> __mark_inode_dirty(mapping->host, I_DIRTY_PAGES);
>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH] mm: __set_page_dirty_nobuffers uses spin_lock_irqseve instead of spin_lock_irq
@ 2014-02-06 6:35 ` Yasuaki Ishimatsu
0 siblings, 0 replies; 10+ messages in thread
From: Yasuaki Ishimatsu @ 2014-02-06 6:35 UTC (permalink / raw
To: kosaki.motohiro
Cc: linux-kernel, linux-mm, akpm, KOSAKI Motohiro, Larry Woodman,
Rik van Riel, Johannes Weiner, stable
(2014/02/04 1:49), kosaki.motohiro@gmail.com wrote:
> From: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
>
> During aio stress test, we observed the following lockdep warning.
> This mean AIO+numa_balancing is currently deadlockable.
>
> The problem is, aio_migratepage disable interrupt, but __set_page_dirty_nobuffers
> unintentionally enable it again.
>
> Generally, all helper function should use spin_lock_irqsave()
> instead of spin_lock_irq() because they don't know caller at all.
>
> [ 599.843948] other info that might help us debug this:
> [ 599.873748] Possible unsafe locking scenario:
> [ 599.873748]
> [ 599.900902] CPU0
> [ 599.912701] ----
> [ 599.924929] lock(&(&ctx->completion_lock)->rlock);
> [ 599.950299] <Interrupt>
> [ 599.962576] lock(&(&ctx->completion_lock)->rlock);
> [ 599.985771]
> [ 599.985771] *** DEADLOCK ***
>
> [ 600.375623] [<ffffffff81678d3c>] dump_stack+0x19/0x1b
> [ 600.398769] [<ffffffff816731aa>] print_usage_bug+0x1f7/0x208
> [ 600.425092] [<ffffffff810df370>] ? print_shortest_lock_dependencies+0x1d0/0x1d0
> [ 600.458981] [<ffffffff810e08dd>] mark_lock+0x21d/0x2a0
> [ 600.482910] [<ffffffff810e0a19>] mark_held_locks+0xb9/0x140
> [ 600.508956] [<ffffffff8168201c>] ? _raw_spin_unlock_irq+0x2c/0x50
> [ 600.536825] [<ffffffff810e0ba5>] trace_hardirqs_on_caller+0x105/0x1d0
> [ 600.566861] [<ffffffff810e0c7d>] trace_hardirqs_on+0xd/0x10
> [ 600.593210] [<ffffffff8168201c>] _raw_spin_unlock_irq+0x2c/0x50
> [ 600.620599] [<ffffffff8117f72c>] __set_page_dirty_nobuffers+0x8c/0xf0
> [ 600.649992] [<ffffffff811d1094>] migrate_page_copy+0x434/0x540
> [ 600.676635] [<ffffffff8123f5b1>] aio_migratepage+0xb1/0x140
> [ 600.703126] [<ffffffff811d126d>] move_to_new_page+0x7d/0x230
> [ 600.729022] [<ffffffff811d1b45>] migrate_pages+0x5e5/0x700
> [ 600.754705] [<ffffffff811d0070>] ? buffer_migrate_lock_buffers+0xb0/0xb0
> [ 600.785784] [<ffffffff811d29cc>] migrate_misplaced_page+0xbc/0xf0
> [ 600.814029] [<ffffffff8119eb62>] do_numa_page+0x102/0x190
> [ 600.839182] [<ffffffff8119ee31>] handle_pte_fault+0x241/0x970
> [ 600.865875] [<ffffffff811a0345>] handle_mm_fault+0x265/0x370
> [ 600.892071] [<ffffffff81686d82>] __do_page_fault+0x172/0x5a0
> [ 600.918065] [<ffffffff81682cd8>] ? retint_swapgs+0x13/0x1b
> [ 600.943493] [<ffffffff816871ca>] do_page_fault+0x1a/0x70
> [ 600.968081] [<ffffffff81682ff8>] page_fault+0x28/0x30
>
> Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
> Cc: Larry Woodman <lwoodman@redhat.com>
> Cc: Rik van Riel <riel@redhat.com>
> Cc: Johannes Weiner <jweiner@redhat.com>
> Cc: stable@vger.kernel.org
> ---
Tested-by: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Thank you for posting the patch.
The same issue occurred on my box. And I confirmed that the issue
disappeared by the patch.
Thanks,
Yasuaki Ishimatsu
> mm/page-writeback.c | 5 +++--
> 1 files changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/mm/page-writeback.c b/mm/page-writeback.c
> index 2d30e2c..7106cb1 100644
> --- a/mm/page-writeback.c
> +++ b/mm/page-writeback.c
> @@ -2173,11 +2173,12 @@ int __set_page_dirty_nobuffers(struct page *page)
> if (!TestSetPageDirty(page)) {
> struct address_space *mapping = page_mapping(page);
> struct address_space *mapping2;
> + unsigned long flags;
>
> if (!mapping)
> return 1;
>
> - spin_lock_irq(&mapping->tree_lock);
> + spin_lock_irqsave(&mapping->tree_lock, flags);
> mapping2 = page_mapping(page);
> if (mapping2) { /* Race with truncate? */
> BUG_ON(mapping2 != mapping);
> @@ -2186,7 +2187,7 @@ int __set_page_dirty_nobuffers(struct page *page)
> radix_tree_tag_set(&mapping->page_tree,
> page_index(page), PAGECACHE_TAG_DIRTY);
> }
> - spin_unlock_irq(&mapping->tree_lock);
> + spin_unlock_irqrestore(&mapping->tree_lock, flags);
> if (mapping->host) {
> /* !PageAnon && !swapper_space */
> __mark_inode_dirty(mapping->host, I_DIRTY_PAGES);
>
--
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>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH] mm: __set_page_dirty_nobuffers uses spin_lock_irqseve instead of spin_lock_irq
2014-02-03 16:49 ` kosaki.motohiro
@ 2014-02-06 8:04 ` Tang Chen
-1 siblings, 0 replies; 10+ messages in thread
From: Tang Chen @ 2014-02-06 8:04 UTC (permalink / raw
To: kosaki.motohiro
Cc: linux-kernel, linux-mm, akpm, KOSAKI Motohiro, Larry Woodman,
Rik van Riel, Johannes Weiner, stable
Hi,
Tested-by: Tang Chen <tangchen@cn.fujitsu.com>
Have tested this patch, and the problem is fixed.
Thanks.
On 02/04/2014 12:49 AM, kosaki.motohiro@gmail.com wrote:
> From: KOSAKI Motohiro<kosaki.motohiro@jp.fujitsu.com>
>
> During aio stress test, we observed the following lockdep warning.
> This mean AIO+numa_balancing is currently deadlockable.
>
> The problem is, aio_migratepage disable interrupt, but __set_page_dirty_nobuffers
> unintentionally enable it again.
>
> Generally, all helper function should use spin_lock_irqsave()
> instead of spin_lock_irq() because they don't know caller at all.
>
> [ 599.843948] other info that might help us debug this:
> [ 599.873748] Possible unsafe locking scenario:
> [ 599.873748]
> [ 599.900902] CPU0
> [ 599.912701] ----
> [ 599.924929] lock(&(&ctx->completion_lock)->rlock);
> [ 599.950299]<Interrupt>
> [ 599.962576] lock(&(&ctx->completion_lock)->rlock);
> [ 599.985771]
> [ 599.985771] *** DEADLOCK ***
>
> [ 600.375623] [<ffffffff81678d3c>] dump_stack+0x19/0x1b
> [ 600.398769] [<ffffffff816731aa>] print_usage_bug+0x1f7/0x208
> [ 600.425092] [<ffffffff810df370>] ? print_shortest_lock_dependencies+0x1d0/0x1d0
> [ 600.458981] [<ffffffff810e08dd>] mark_lock+0x21d/0x2a0
> [ 600.482910] [<ffffffff810e0a19>] mark_held_locks+0xb9/0x140
> [ 600.508956] [<ffffffff8168201c>] ? _raw_spin_unlock_irq+0x2c/0x50
> [ 600.536825] [<ffffffff810e0ba5>] trace_hardirqs_on_caller+0x105/0x1d0
> [ 600.566861] [<ffffffff810e0c7d>] trace_hardirqs_on+0xd/0x10
> [ 600.593210] [<ffffffff8168201c>] _raw_spin_unlock_irq+0x2c/0x50
> [ 600.620599] [<ffffffff8117f72c>] __set_page_dirty_nobuffers+0x8c/0xf0
> [ 600.649992] [<ffffffff811d1094>] migrate_page_copy+0x434/0x540
> [ 600.676635] [<ffffffff8123f5b1>] aio_migratepage+0xb1/0x140
> [ 600.703126] [<ffffffff811d126d>] move_to_new_page+0x7d/0x230
> [ 600.729022] [<ffffffff811d1b45>] migrate_pages+0x5e5/0x700
> [ 600.754705] [<ffffffff811d0070>] ? buffer_migrate_lock_buffers+0xb0/0xb0
> [ 600.785784] [<ffffffff811d29cc>] migrate_misplaced_page+0xbc/0xf0
> [ 600.814029] [<ffffffff8119eb62>] do_numa_page+0x102/0x190
> [ 600.839182] [<ffffffff8119ee31>] handle_pte_fault+0x241/0x970
> [ 600.865875] [<ffffffff811a0345>] handle_mm_fault+0x265/0x370
> [ 600.892071] [<ffffffff81686d82>] __do_page_fault+0x172/0x5a0
> [ 600.918065] [<ffffffff81682cd8>] ? retint_swapgs+0x13/0x1b
> [ 600.943493] [<ffffffff816871ca>] do_page_fault+0x1a/0x70
> [ 600.968081] [<ffffffff81682ff8>] page_fault+0x28/0x30
>
> Signed-off-by: KOSAKI Motohiro<kosaki.motohiro@jp.fujitsu.com>
> Cc: Larry Woodman<lwoodman@redhat.com>
> Cc: Rik van Riel<riel@redhat.com>
> Cc: Johannes Weiner<jweiner@redhat.com>
> Cc: stable@vger.kernel.org
> ---
> mm/page-writeback.c | 5 +++--
> 1 files changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/mm/page-writeback.c b/mm/page-writeback.c
> index 2d30e2c..7106cb1 100644
> --- a/mm/page-writeback.c
> +++ b/mm/page-writeback.c
> @@ -2173,11 +2173,12 @@ int __set_page_dirty_nobuffers(struct page *page)
> if (!TestSetPageDirty(page)) {
> struct address_space *mapping = page_mapping(page);
> struct address_space *mapping2;
> + unsigned long flags;
>
> if (!mapping)
> return 1;
>
> - spin_lock_irq(&mapping->tree_lock);
> + spin_lock_irqsave(&mapping->tree_lock, flags);
> mapping2 = page_mapping(page);
> if (mapping2) { /* Race with truncate? */
> BUG_ON(mapping2 != mapping);
> @@ -2186,7 +2187,7 @@ int __set_page_dirty_nobuffers(struct page *page)
> radix_tree_tag_set(&mapping->page_tree,
> page_index(page), PAGECACHE_TAG_DIRTY);
> }
> - spin_unlock_irq(&mapping->tree_lock);
> + spin_unlock_irqrestore(&mapping->tree_lock, flags);
> if (mapping->host) {
> /* !PageAnon&& !swapper_space */
> __mark_inode_dirty(mapping->host, I_DIRTY_PAGES);
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH] mm: __set_page_dirty_nobuffers uses spin_lock_irqseve instead of spin_lock_irq
@ 2014-02-06 8:04 ` Tang Chen
0 siblings, 0 replies; 10+ messages in thread
From: Tang Chen @ 2014-02-06 8:04 UTC (permalink / raw
To: kosaki.motohiro
Cc: linux-kernel, linux-mm, akpm, KOSAKI Motohiro, Larry Woodman,
Rik van Riel, Johannes Weiner, stable
Hi,
Tested-by: Tang Chen <tangchen@cn.fujitsu.com>
Have tested this patch, and the problem is fixed.
Thanks.
On 02/04/2014 12:49 AM, kosaki.motohiro@gmail.com wrote:
> From: KOSAKI Motohiro<kosaki.motohiro@jp.fujitsu.com>
>
> During aio stress test, we observed the following lockdep warning.
> This mean AIO+numa_balancing is currently deadlockable.
>
> The problem is, aio_migratepage disable interrupt, but __set_page_dirty_nobuffers
> unintentionally enable it again.
>
> Generally, all helper function should use spin_lock_irqsave()
> instead of spin_lock_irq() because they don't know caller at all.
>
> [ 599.843948] other info that might help us debug this:
> [ 599.873748] Possible unsafe locking scenario:
> [ 599.873748]
> [ 599.900902] CPU0
> [ 599.912701] ----
> [ 599.924929] lock(&(&ctx->completion_lock)->rlock);
> [ 599.950299]<Interrupt>
> [ 599.962576] lock(&(&ctx->completion_lock)->rlock);
> [ 599.985771]
> [ 599.985771] *** DEADLOCK ***
>
> [ 600.375623] [<ffffffff81678d3c>] dump_stack+0x19/0x1b
> [ 600.398769] [<ffffffff816731aa>] print_usage_bug+0x1f7/0x208
> [ 600.425092] [<ffffffff810df370>] ? print_shortest_lock_dependencies+0x1d0/0x1d0
> [ 600.458981] [<ffffffff810e08dd>] mark_lock+0x21d/0x2a0
> [ 600.482910] [<ffffffff810e0a19>] mark_held_locks+0xb9/0x140
> [ 600.508956] [<ffffffff8168201c>] ? _raw_spin_unlock_irq+0x2c/0x50
> [ 600.536825] [<ffffffff810e0ba5>] trace_hardirqs_on_caller+0x105/0x1d0
> [ 600.566861] [<ffffffff810e0c7d>] trace_hardirqs_on+0xd/0x10
> [ 600.593210] [<ffffffff8168201c>] _raw_spin_unlock_irq+0x2c/0x50
> [ 600.620599] [<ffffffff8117f72c>] __set_page_dirty_nobuffers+0x8c/0xf0
> [ 600.649992] [<ffffffff811d1094>] migrate_page_copy+0x434/0x540
> [ 600.676635] [<ffffffff8123f5b1>] aio_migratepage+0xb1/0x140
> [ 600.703126] [<ffffffff811d126d>] move_to_new_page+0x7d/0x230
> [ 600.729022] [<ffffffff811d1b45>] migrate_pages+0x5e5/0x700
> [ 600.754705] [<ffffffff811d0070>] ? buffer_migrate_lock_buffers+0xb0/0xb0
> [ 600.785784] [<ffffffff811d29cc>] migrate_misplaced_page+0xbc/0xf0
> [ 600.814029] [<ffffffff8119eb62>] do_numa_page+0x102/0x190
> [ 600.839182] [<ffffffff8119ee31>] handle_pte_fault+0x241/0x970
> [ 600.865875] [<ffffffff811a0345>] handle_mm_fault+0x265/0x370
> [ 600.892071] [<ffffffff81686d82>] __do_page_fault+0x172/0x5a0
> [ 600.918065] [<ffffffff81682cd8>] ? retint_swapgs+0x13/0x1b
> [ 600.943493] [<ffffffff816871ca>] do_page_fault+0x1a/0x70
> [ 600.968081] [<ffffffff81682ff8>] page_fault+0x28/0x30
>
> Signed-off-by: KOSAKI Motohiro<kosaki.motohiro@jp.fujitsu.com>
> Cc: Larry Woodman<lwoodman@redhat.com>
> Cc: Rik van Riel<riel@redhat.com>
> Cc: Johannes Weiner<jweiner@redhat.com>
> Cc: stable@vger.kernel.org
> ---
> mm/page-writeback.c | 5 +++--
> 1 files changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/mm/page-writeback.c b/mm/page-writeback.c
> index 2d30e2c..7106cb1 100644
> --- a/mm/page-writeback.c
> +++ b/mm/page-writeback.c
> @@ -2173,11 +2173,12 @@ int __set_page_dirty_nobuffers(struct page *page)
> if (!TestSetPageDirty(page)) {
> struct address_space *mapping = page_mapping(page);
> struct address_space *mapping2;
> + unsigned long flags;
>
> if (!mapping)
> return 1;
>
> - spin_lock_irq(&mapping->tree_lock);
> + spin_lock_irqsave(&mapping->tree_lock, flags);
> mapping2 = page_mapping(page);
> if (mapping2) { /* Race with truncate? */
> BUG_ON(mapping2 != mapping);
> @@ -2186,7 +2187,7 @@ int __set_page_dirty_nobuffers(struct page *page)
> radix_tree_tag_set(&mapping->page_tree,
> page_index(page), PAGECACHE_TAG_DIRTY);
> }
> - spin_unlock_irq(&mapping->tree_lock);
> + spin_unlock_irqrestore(&mapping->tree_lock, flags);
> if (mapping->host) {
> /* !PageAnon&& !swapper_space */
> __mark_inode_dirty(mapping->host, I_DIRTY_PAGES);
--
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>
^ permalink raw reply [flat|nested] 10+ messages in thread
end of thread, other threads:[~2014-02-06 8:01 UTC | newest]
Thread overview: 10+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-02-03 16:49 [PATCH] mm: __set_page_dirty_nobuffers uses spin_lock_irqseve instead of spin_lock_irq kosaki.motohiro
2014-02-03 16:49 ` kosaki.motohiro
2014-02-03 21:11 ` David Rientjes
2014-02-03 21:11 ` David Rientjes
2014-02-04 17:10 ` KOSAKI Motohiro
2014-02-04 17:10 ` KOSAKI Motohiro
2014-02-06 6:35 ` Yasuaki Ishimatsu
2014-02-06 6:35 ` Yasuaki Ishimatsu
2014-02-06 8:04 ` Tang Chen
2014-02-06 8:04 ` Tang Chen
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.