sched/fair: Care divide error in update_task_scan_period()

PATCH In review Yasuaki Ishimatsu 3 participants 5 messages Oct 22, 2014 – Oct 28, 2014 Scheduler

lore

Share & tools

lore

Yasuaki Ishimatsu

author

2014-10-22 07:05 UTC #0 lore raw

PATCH v3 1/1 sched/fair: Care divide error in update_task_scan_period()

While offling node by hot removing memory, the following divide error occurs:

divide error: 0000 [#1] SMP
[...]

  Call Trace:
   [...] handle_mm_fault
   [...] ? try_to_wake_up
   [...] ? wake_up_state
   [...] __do_page_fault
   [...] ? do_futex
   [...] ? put_prev_entity
   [...] ? __switch_to
   [...] do_page_fault
   [...] page_fault
  [...]
  RIP  [<ffffffff810a7081>] task_numa_fault
   RSP <ffff88084eb2bcb0>

The issue occurs as follows:
1. When page fault occurs and page is allocated from node 1, task_struct->numa_faults_buffer_memory[] of node 1 is incremented and p->numa_faults_locality[] is also incremented as follows:

o numa_faults_buffer_memory[] o numa_faults_locality[]
NR_NUMA_HINT_FAULT_TYPES
| 0 | 1 |
---------------------------------- ----------------------
node 0 | 0 | 0 | remote | 0 |
node 1 | 0 | 1 | locale | 1 |
---------------------------------- ----------------------

2. node 1 is offlined by hot removing memory.

3. When page fault occurs, fault_types[] is calculated by using

     p->numa_faults_buffer_memory[] of all online nodes in
     task_numa_placement(). But node 1 was offline by step 2. So
     the fault_types[] is calculated by using only
     p->numa_faults_buffer_memory[] of node 0. So both of fault_types[]
     are set to 0.

4. The values(0) of fault_types[] pass to update_task_scan_period().

5. numa_faults_locality[1] is set to 1. So the following division is

     calculated.

        static void update_task_scan_period(struct task_struct *p,
                                unsigned long shared, unsigned long private){
        ...
                ratio = DIV_ROUND_UP(private * NUMA_PERIOD_SLOTS, (private + shared));
        }

6. But both of private and shared are set to 0. So divide error occurs here.

The divide error is rare case because the trigger is node offline. This patch always increments denominator for avoiding divide error.

Signed-off-by Yasuaki Ishimatsu <[email protected]>

---
v2:
- Simply increment a denominator

 kernel/sched/fair.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

kernel/sched/fair.c +1 -1

index 0b069bf..f3b492d 100644
@@ -1520,7 +1520,7 @@ static void update_task_scan_period(struct task_struct *p,
 		 * scanning faster if shared accesses dominate as it may
 		 * simply bounce migrations uselessly
 		 */
-		ratio = DIV_ROUND_UP(private * NUMA_PERIOD_SLOTS, (private + shared));
+		ratio = DIV_ROUND_UP(private * NUMA_PERIOD_SLOTS, (private + shared + 1));
 		diff = (diff * ratio) / NUMA_PERIOD_SLOTS;
 	}

Signature · 1 line

1.8.3.1

Wanpeng Li

Reviewed-by on v3 1/1

2014-10-22 07:09 UTC #1 lore raw

10/22/14, 3:04 PM, Yasuaki Ishimatsu:

Quoted 53 lines

While offling node by hot removing memory, the following divide error occurs:
divide error: 0000 [#1] SMP
[...]
  Call Trace:
   [...] handle_mm_fault
   [...] ? try_to_wake_up
   [...] ? wake_up_state
   [...] __do_page_fault
   [...] ? do_futex
   [...] ? put_prev_entity
   [...] ? __switch_to
   [...] do_page_fault
   [...] page_fault
  [...]
  RIP  [<ffffffff810a7081>] task_numa_fault
   RSP <ffff88084eb2bcb0>
The issue occurs as follows:
1. When page fault occurs and page is allocated from node 1, task_struct->numa_faults_buffer_memory[] of node 1 is incremented and p->numa_faults_locality[] is also incremented as follows:
o numa_faults_buffer_memory[] o numa_faults_locality[]
NR_NUMA_HINT_FAULT_TYPES
| 0 | 1 |
---------------------------------- ----------------------
node 0 | 0 | 0 | remote | 0 |
node 1 | 0 | 1 | locale | 1 |
---------------------------------- ----------------------
2. node 1 is offlined by hot removing memory.
3. When page fault occurs, fault_types[] is calculated by using
     p->numa_faults_buffer_memory[] of all online nodes in
     task_numa_placement(). But node 1 was offline by step 2. So
     the fault_types[] is calculated by using only
     p->numa_faults_buffer_memory[] of node 0. So both of fault_types[]
     are set to 0.
4. The values(0) of fault_types[] pass to update_task_scan_period().
5. numa_faults_locality[1] is set to 1. So the following division is
     calculated.

        static void update_task_scan_period(struct task_struct *p,
                                unsigned long shared, unsigned long private){
        ...
                ratio = DIV_ROUND_UP(private * NUMA_PERIOD_SLOTS, (private + shared));
        }

Quoted 7 lines

6. But both of private and shared are set to 0. So divide error occurs here.
The divide error is rare case because the trigger is node offline. This patch always increments denominator for avoiding divide error.
Signed-off-by Yasuaki Ishimatsu <[email protected]>

Reviewed-by Wanpeng Li <[email protected]>

Quoted 20 lines

---
v2:
- Simply increment a denominator

 kernel/sched/fair.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

kernel/sched/fair.c +1 -1

index 0b069bf..f3b492d 100644
@@ -1520,7 +1520,7 @@ static void update_task_scan_period(struct task_struct *p,
 		 * scanning faster if shared accesses dominate as it may
 		 * simply bounce migrations uselessly
 		 */
-		ratio = DIV_ROUND_UP(private * NUMA_PERIOD_SLOTS, (private + shared));
+		ratio = DIV_ROUND_UP(private * NUMA_PERIOD_SLOTS, (private + shared + 1));
 		diff = (diff * ratio) / NUMA_PERIOD_SLOTS;
 	}

Yasuaki Ishimatsu

author

2014-10-22 07:43 UTC #2 lore raw

Quoted 57 lines · Wanpeng Li

10/22/14, 3:04 PM, Yasuaki Ishimatsu:
Quoted 56 lines
While offling node by hot removing memory, the following divide error occurs:
   divide error: 0000 [#1] SMP
   [...]
   Call Trace:
    [...] handle_mm_fault
    [...] ? try_to_wake_up
    [...] ? wake_up_state
    [...] __do_page_fault
    [...] ? do_futex
    [...] ? put_prev_entity
    [...] ? __switch_to
    [...] do_page_fault
    [...] page_fault
   [...]
   RIP  [<ffffffff810a7081>] task_numa_fault
    RSP <ffff88084eb2bcb0>
The issue occurs as follows:
1. When page fault occurs and page is allocated from node 1, task_struct->numa_faults_buffer_memory[] of node 1 is incremented and p->numa_faults_locality[] is also incremented as follows:
o numa_faults_buffer_memory[] o numa_faults_locality[]
NR_NUMA_HINT_FAULT_TYPES
| 0 | 1 |
---------------------------------- ----------------------
node 0 | 0 | 0 | remote | 0 |
node 1 | 0 | 1 | locale | 1 |
---------------------------------- ----------------------
2. node 1 is offlined by hot removing memory.
3. When page fault occurs, fault_types[] is calculated by using
      p->numa_faults_buffer_memory[] of all online nodes in
      task_numa_placement(). But node 1 was offline by step 2. So
      the fault_types[] is calculated by using only
      p->numa_faults_buffer_memory[] of node 0. So both of fault_types[]
      are set to 0.
4. The values(0) of fault_types[] pass to update_task_scan_period().
   5. numa_faults_locality[1] is set to 1. So the following division is
      calculated.

         static void update_task_scan_period(struct task_struct *p,
                                 unsigned long shared, unsigned long private){
         ...
                 ratio = DIV_ROUND_UP(private * NUMA_PERIOD_SLOTS, (private + shared));
         }

   6. But both of private and shared are set to 0. So divide error
      occurs here.

Quoted 6 lines

Quoted 4 lines
The divide error is rare case because the trigger is node offline. This patch always increments denominator for avoiding divide error.
Signed-off-by Yasuaki Ishimatsu <[email protected]>
Reviewed-by Wanpeng Li <[email protected]>

Thank you for your review.

Thanks,
Yasuaki Ishimatsu

Quoted 22 lines

Quoted 21 lines

---
v2:
- Simply increment a denominator

  kernel/sched/fair.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)

kernel/sched/fair.c +1 -1

index 0b069bf..f3b492d 100644
@@ -1520,7 +1520,7 @@ static void update_task_scan_period(struct task_struct *p,
  		 * scanning faster if shared accesses dominate as it may
  		 * simply bounce migrations uselessly
  		 */
-		ratio = DIV_ROUND_UP(private * NUMA_PERIOD_SLOTS, (private + shared));
+		ratio = DIV_ROUND_UP(private * NUMA_PERIOD_SLOTS, (private + shared + 1));
  		diff = (diff * ratio) / NUMA_PERIOD_SLOTS;
  	}

Yasuaki Ishimatsu

author

2014-10-27 05:46 UTC #3 lore raw

In reply to Yasuaki Ishimatsu

Could you review my patch?

Thanks,
Yasuaki Ishimatsu

Quoted 80 lines · Yasuaki Ishimatsu

While offling node by hot removing memory, the following divide error occurs:
   divide error: 0000 [#1] SMP
   [...]
   Call Trace:
    [...] handle_mm_fault
    [...] ? try_to_wake_up
    [...] ? wake_up_state
    [...] __do_page_fault
    [...] ? do_futex
    [...] ? put_prev_entity
    [...] ? __switch_to
    [...] do_page_fault
    [...] page_fault
   [...]
   RIP  [<ffffffff810a7081>] task_numa_fault
    RSP <ffff88084eb2bcb0>
The issue occurs as follows:
1. When page fault occurs and page is allocated from node 1, task_struct->numa_faults_buffer_memory[] of node 1 is incremented and p->numa_faults_locality[] is also incremented as follows:
o numa_faults_buffer_memory[] o numa_faults_locality[]
NR_NUMA_HINT_FAULT_TYPES
| 0 | 1 |
---------------------------------- ----------------------
node 0 | 0 | 0 | remote | 0 |
node 1 | 0 | 1 | locale | 1 |
---------------------------------- ----------------------
2. node 1 is offlined by hot removing memory.
3. When page fault occurs, fault_types[] is calculated by using
      p->numa_faults_buffer_memory[] of all online nodes in
      task_numa_placement(). But node 1 was offline by step 2. So
      the fault_types[] is calculated by using only
      p->numa_faults_buffer_memory[] of node 0. So both of fault_types[]
      are set to 0.
4. The values(0) of fault_types[] pass to update_task_scan_period().
   5. numa_faults_locality[1] is set to 1. So the following division is
      calculated.

         static void update_task_scan_period(struct task_struct *p,
                                 unsigned long shared, unsigned long private){
         ...
                 ratio = DIV_ROUND_UP(private * NUMA_PERIOD_SLOTS, (private + shared));
         }

   6. But both of private and shared are set to 0. So divide error
      occurs here.
The divide error is rare case because the trigger is node offline. This patch always increments denominator for avoiding divide error.
Signed-off-by Yasuaki Ishimatsu <[email protected]>
---
v2:
- Simply increment a denominator
  kernel/sched/fair.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
kernel/sched/fair.c +1 -1
index 0b069bf..f3b492d 100644
@@ -1520,7 +1520,7 @@ static void update_task_scan_period(struct task_struct *p,
  		 * scanning faster if shared accesses dominate as it may
  		 * simply bounce migrations uselessly
  		 */
-		ratio = DIV_ROUND_UP(private * NUMA_PERIOD_SLOTS, (private + shared));
+		ratio = DIV_ROUND_UP(private * NUMA_PERIOD_SLOTS, (private + shared + 1));
  		diff = (diff * ratio) / NUMA_PERIOD_SLOTS;
  	}
 

tip-bot for Yasuaki Ishimatsu

merge bot

2014-10-28 11:03 UTC #4 lore raw

In reply to Yasuaki Ishimatsu

sched/fair: Care divide error in update_task_scan_period()

Commit-ID:  2847c90e1b3ae95379af24894fc4f98e7f2fd705
Gitweb:     http://git.kernel.org/tip/2847c90e1b3ae95379af24894fc4f98e7f2fd705
Author:     Yasuaki Ishimatsu <[email protected]>
AuthorDate: Wed, 22 Oct 2014 16:04:35 +0900
Committer:  Ingo Molnar <[email protected]>
CommitDate: Tue, 28 Oct 2014 10:46:03 +0100

sched/fair: Care divide error in update_task_scan_period()

While offling node by hot removing memory, the following divide error occurs:

divide error: 0000 [#1] SMP
[...]

  Call Trace:
   [...] handle_mm_fault
   [...] ? try_to_wake_up
   [...] ? wake_up_state
   [...] __do_page_fault
   [...] ? do_futex
   [...] ? put_prev_entity
   [...] ? __switch_to
   [...] do_page_fault
   [...] page_fault
  [...]
  RIP  [<ffffffff810a7081>] task_numa_fault
   RSP <ffff88084eb2bcb0>

2. node 1 is offlined by hot removing memory.

3. When page fault occurs, fault_types[] is calculated by using

     p->numa_faults_buffer_memory[] of all online nodes in
     task_numa_placement(). But node 1 was offline by step 2. So
     the fault_types[] is calculated by using only
     p->numa_faults_buffer_memory[] of node 0. So both of fault_types[]
     are set to 0.

4. The values(0) of fault_types[] pass to update_task_scan_period().

5. numa_faults_locality[1] is set to 1. So the following division is

     calculated.

        static void update_task_scan_period(struct task_struct *p,
                                unsigned long shared, unsigned long private){
        ...
                ratio = DIV_ROUND_UP(private * NUMA_PERIOD_SLOTS, (private + shared));
        }

6. But both of private and shared are set to 0. So divide error occurs here.

The divide error is rare case because the trigger is node offline. This patch always increments denominator for avoiding divide error.

Signed-off-by Yasuaki Ishimatsu <[email protected]>

Signed-off-by Peter Zijlstra (Intel) <[email protected]>

Cc Linus Torvalds <[email protected]>

Link http://lkml.kernel.org/r/[email protected] open here

Signed-off-by Ingo Molnar <[email protected]>

---
 kernel/sched/fair.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

kernel/sched/fair.c +1 -1

index fbc0b82..e9abd4e 100644
@@ -1530,7 +1530,7 @@ static void update_task_scan_period(struct task_struct *p,
 		 * scanning faster if shared accesses dominate as it may
 		 * simply bounce migrations uselessly
 		 */
-		ratio = DIV_ROUND_UP(private * NUMA_PERIOD_SLOTS, (private + shared));
+		ratio = DIV_ROUND_UP(private * NUMA_PERIOD_SLOTS, (private + shared + 1));
 		diff = (diff * ratio) / NUMA_PERIOD_SLOTS;
 	}