Skip to contentSkip to conversation

mm/oom_kill: show oom eligibility when displaying the current memory state of all tasks

RFC PATCH Superseded Aaron Tomlin 4 participants 4 messages Jun 11, 2021 – Jun 12, 2021

Subsystems (2)

v1 of 3 Superseded Latest is v3 →

lore
Share & tools
lore
#0 lore raw

RFC PATCH 1/1 mm/oom_kill: show oom eligibility when displaying the current memory state of all tasks

At the present time, when showing potential OOM victims, we do not exclude tasks which already have MMF_OOM_SKIP set; it is possible that the last OOM killable victim was already OOM killed, yet the OOM reaper failed to reclaim memory and set MMF_OOM_SKIP.
This can be confusing/or perhaps even misleading, to the reader of the OOM report. Now, we already unconditionally display a task's oom_score_adj_min value that can be set to OOM_SCORE_ADJ_MIN which is indicative of an "unkillable" task i.e. is not eligible.

This patch provides a clear indication with regard to the OOM eligibility of each displayed task.

---
 mm/oom_kill.c | 31 +++++++++++++++++++++++++++----
 1 file changed, 27 insertions(+), 4 deletions(-)
mm/oom_kill.c +27 -4
index eefd3f5fde46..70781d681a6e 100644
@@ -160,6 +160,27 @@ static inline bool is_sysrq_oom(struct oom_control *oc)
 	return oc->order == -1;
 }
 
+/**
+ * is_task_eligible_oom - determine if and why a task cannot be OOM killed
+ * @tsk: task to check
+ *
+ * Needs to be called with task_lock().
+ */
+static const char * is_task_oom_eligible(struct task_struct *p)
+{
+	long adj;
+
+	adj = (long)p->signal->oom_score_adj;
+	if (adj == OOM_SCORE_ADJ_MIN)
+		return "no: oom score";
+	else if (test_bit(MMF_OOM_SKIP, &p->mm->flags)
+		return "no: oom reaped";
+	else if (in_vfork(p))
+		return "no: in vfork";
+	else
+		return "yes";
+}
+
 /* return true if the task is not adequate as candidate victim task. */
 static bool oom_unkillable_task(struct task_struct *p)
 {
@@ -401,12 +422,13 @@ static int dump_task(struct task_struct *p, void *arg)
 		return 0;
 	}
 
-	pr_info("[%7d] %5d %5d %8lu %8lu %8ld %8lu         %5hd %s\n",
+	pr_info("[%7d] %5d %5d %8lu %8lu %8ld %8lu         %5hd %-15s %s\n",
 		task->pid, from_kuid(&init_user_ns, task_uid(task)),
 		task->tgid, task->mm->total_vm, get_mm_rss(task->mm),
 		mm_pgtables_bytes(task->mm),
 		get_mm_counter(task->mm, MM_SWAPENTS),
-		task->signal->oom_score_adj, task->comm);
+		task->signal->oom_score_adj, is_task_oom_eligible(task),
+		task->comm);
 	task_unlock(task);
 
 	return 0;
@@ -420,12 +442,13 @@ static int dump_task(struct task_struct *p, void *arg)
  * memcg, not in the same cpuset, or bound to a disjoint set of mempolicy nodes
  * are not shown.
  * State information includes task's pid, uid, tgid, vm size, rss,
- * pgtables_bytes, swapents, oom_score_adj value, and name.
+ * pgtables_bytes, swapents, oom_score_adj value, oom eligible status
+ * and name.
  */
 static void dump_tasks(struct oom_control *oc)
 {
 	pr_info("Tasks state (memory values in pages):\n");
-	pr_info("[  pid  ]   uid  tgid total_vm      rss pgtables_bytes swapents oom_score_adj name\n");
+	pr_info("[  pid  ]   uid  tgid total_vm      rss pgtables_bytes swapents oom_score_adj oom eligible? name\n");
 
 	if (is_memcg_oom(oc))
 		mem_cgroup_scan_tasks(oc->memcg, dump_task, oc);
Signature · 1 line
2.26.3
#1 lore raw
Quoted 68 lines · Aaron Tomlin

At the present time, when showing potential OOM victims, we do not exclude tasks which already have MMF_OOM_SKIP set; it is possible that the last OOM killable victim was already OOM killed, yet the OOM reaper failed to reclaim memory and set MMF_OOM_SKIP.
This can be confusing/or perhaps even misleading, to the reader of the OOM report. Now, we already unconditionally display a task's oom_score_adj_min value that can be set to OOM_SCORE_ADJ_MIN which is indicative of an "unkillable" task i.e. is not eligible.

This patch provides a clear indication with regard to the OOM eligibility of each displayed task.

---
  mm/oom_kill.c | 31 +++++++++++++++++++++++++++----
  1 file changed, 27 insertions(+), 4 deletions(-)
mm/oom_kill.c +24 -2
index eefd3f5fde46..70781d681a6e 100644
@@ -160,6 +160,27 @@ static inline bool is_sysrq_oom(struct oom_control *oc)
  	return oc->order == -1;
  }
 
+/**
+ * is_task_eligible_oom - determine if and why a task cannot be OOM killed
+ * @tsk: task to check
+ *
+ * Needs to be called with task_lock().
+ */
+static const char * is_task_oom_eligible(struct task_struct *p)
+{
+	long adj;
+
+	adj = (long)p->signal->oom_score_adj;
+	if (adj == OOM_SCORE_ADJ_MIN)
+		return "no: oom score";
+	else if (test_bit(MMF_OOM_SKIP, &p->mm->flags)
+		return "no: oom reaped";
+	else if (in_vfork(p))
+		return "no: in vfork";
+	else
+		return "yes";
+}
+
  /* return true if the task is not adequate as candidate victim task. */
  static bool oom_unkillable_task(struct task_struct *p)
  {
@@ -401,12 +422,13 @@ static int dump_task(struct task_struct *p, void *arg)
  		return 0;
  	}
 
-	pr_info("[%7d] %5d %5d %8lu %8lu %8ld %8lu         %5hd %s\n",
+	pr_info("[%7d] %5d %5d %8lu %8lu %8ld %8lu         %5hd %-15s %s\n",
  		task->pid, from_kuid(&init_user_ns, task_uid(task)),
  		task->tgid, task->mm->total_vm, get_mm_rss(task->mm),
  		mm_pgtables_bytes(task->mm),
  		get_mm_counter(task->mm, MM_SWAPENTS),
-		task->signal->oom_score_adj, task->comm);
+		task->signal->oom_score_adj, is_task_oom_eligible(task),
+		task->comm);
  	task_unlock(task);
 
  	return 0;
@@ -420,12 +442,13 @@ static int dump_task(struct task_struct *p, void *arg)
   * memcg, not in the same cpuset, or bound to a disjoint set of mempolicy nodes
   * are not shown.

mm/oom_kill.c static int dump_task(struct task_struct *p, void *arg) · line 451

Quoted 10 lines
mm/oom_kill.c +3 -2
   * State information includes task's pid, uid, tgid, vm size, rss,
- * pgtables_bytes, swapents, oom_score_adj value, and name.
+ * pgtables_bytes, swapents, oom_score_adj value, oom eligible status
+ * and name.
   */
  static void dump_tasks(struct oom_control *oc)
  {
  	pr_info("Tasks state (memory values in pages):\n");
-	pr_info("[  pid  ]   uid  tgid total_vm      rss pgtables_bytes swapents oom_score_adj name\n");
+	pr_info("[  pid  ]   uid  tgid total_vm      rss pgtables_bytes swapents oom_score_adj oom eligible? name\n");

A minor nit:

"oom eligible?" has 13 characters. The field width is 15. Maybe you should pad 2 more spaces to make the proper alignment.

Cheers,
Longman

#2 lore raw
Quoted 3 lines · Waiman Long

A minor nit:

"oom eligible?" has 13 characters. The field width is 15. Maybe you should pad 2 more spaces to make the proper alignment.

Maybe single character (e.g. 'S' for OOM_SCORE_ADJ_MIN, 'R' for MMF_OOM_SKIP, 'V' for in_vfork(), none for eligible) is better because dump_task() might print many hundreds of lines?

#3 lore raw
In reply to Aaron Tomlin
Quoted 26 lines · Aaron Tomlin

At the present time, when showing potential OOM victims, we do not exclude tasks which already have MMF_OOM_SKIP set; it is possible that the last OOM killable victim was already OOM killed, yet the OOM reaper failed to reclaim memory and set MMF_OOM_SKIP.
This can be confusing/or perhaps even misleading, to the reader of the OOM report. Now, we already unconditionally display a task's oom_score_adj_min value that can be set to OOM_SCORE_ADJ_MIN which is indicative of an "unkillable" task i.e. is not eligible.

This patch provides a clear indication with regard to the OOM eligibility of each displayed task.

---
 mm/oom_kill.c | 31 +++++++++++++++++++++++++++----
 1 file changed, 27 insertions(+), 4 deletions(-)
mm/oom_kill.c +1 -0
index eefd3f5fde46..70781d681a6e 100644
@@ -160,6 +160,27 @@ static inline bool is_sysrq_oom(struct oom_control *oc)
 	return oc->order == -1;
 }
 
+/**

mm/oom_kill.c static inline bool is_sysrq_oom(struct oom_control *oc) · line 173

Quoted 10 lines
mm/oom_kill.c +10 -0
+ * is_task_eligible_oom - determine if and why a task cannot be OOM killed
+ * @tsk: task to check
+ *
+ * Needs to be called with task_lock().
+ */
+static const char * is_task_oom_eligible(struct task_struct *p)
+{
+	long adj;
+
+	adj = (long)p->signal->oom_score_adj;

accoring to the origin type, adj this place maybe use *short* instead.

Quoted 45 lines
mm/oom_kill.c +16 -4
+	if (adj == OOM_SCORE_ADJ_MIN)
+		return "no: oom score";
+	else if (test_bit(MMF_OOM_SKIP, &p->mm->flags)
+		return "no: oom reaped";
+	else if (in_vfork(p))
+		return "no: in vfork";
+	else
+		return "yes";
+}
+
 /* return true if the task is not adequate as candidate victim task. */
 static bool oom_unkillable_task(struct task_struct *p)
 {
@@ -401,12 +422,13 @@ static int dump_task(struct task_struct *p, void *arg)
 		return 0;
 	}
 
-	pr_info("[%7d] %5d %5d %8lu %8lu %8ld %8lu         %5hd %s\n",
+	pr_info("[%7d] %5d %5d %8lu %8lu %8ld %8lu         %5hd %-15s %s\n",
 		task->pid, from_kuid(&init_user_ns, task_uid(task)),
 		task->tgid, task->mm->total_vm, get_mm_rss(task->mm),
 		mm_pgtables_bytes(task->mm),
 		get_mm_counter(task->mm, MM_SWAPENTS),
-		task->signal->oom_score_adj, task->comm);
+		task->signal->oom_score_adj, is_task_oom_eligible(task),
+		task->comm);
 	task_unlock(task);
 
 	return 0;
@@ -420,12 +442,13 @@ static int dump_task(struct task_struct *p, void *arg)
  * memcg, not in the same cpuset, or bound to a disjoint set of mempolicy nodes
  * are not shown.
  * State information includes task's pid, uid, tgid, vm size, rss,
- * pgtables_bytes, swapents, oom_score_adj value, and name.
+ * pgtables_bytes, swapents, oom_score_adj value, oom eligible status
+ * and name.
  */
 static void dump_tasks(struct oom_control *oc)
 {
 	pr_info("Tasks state (memory values in pages):\n");
-	pr_info("[  pid  ]   uid  tgid total_vm      rss pgtables_bytes swapents oom_score_adj name\n");
+	pr_info("[  pid  ]   uid  tgid total_vm      rss pgtables_bytes swapents oom_score_adj oom eligible? name\n");
 
 	if (is_memcg_oom(oc))
 		mem_cgroup_scan_tasks(oc->memcg, dump_task, oc);