oom: add oom_kill_allocating_task sysctl

Adds a new sysctl, 'oom_kill_allocating_task', which will automatically kill the OOM-triggering task instead of scanning through the tasklist to find a memory-hogging target. This is helpful for systems with an insanely large number of tasks where scanning the tasklist significantly degrades performance. Cc: Andrea Arcangeli <andrea@suse.de> Acked-by: Christoph Lameter <clameter@sgi.com> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

oom: add oom_kill_allocating_task sysctl
Adds a new sysctl, 'oom_kill_allocating_task', which will automatically kill the OOM-triggering task instead of scanning through the tasklist to find a memory-hogging target. This is helpful for systems with an insanely large number of tasks where scanning the tasklist significantly degrades performance. Cc: Andrea Arcangeli <andrea@suse.de> Acked-by: Christoph Lameter <clameter@sgi.com> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
David Rientjes · Linus Torvalds
1 parent ff0ceb9deb
Showing 3 changed files with 39 additions and 5 deletions Side-by-side Diff
Documentation/sysctl/vm.txt
kernel/sysctl.c
mm/oom_kill.c
@@ -31,6 +31,7 @@
 - min_unmapped_ratio
 - min_slab_ratio
 - panic_on_oom
+- oom_kill_allocating_task
 - mmap_min_address
 - numa_zonelist_order
  
@@ -219,6 +220,27 @@
 The default value is 0.
 1 and 2 are for failover of clustering. Please select either
 according to your policy of failover.
+
+=============================================================
+
+oom_kill_allocating_task
+
+This enables or disables killing the OOM-triggering task in
+out-of-memory situations.
+
+If this is set to zero, the OOM killer will scan through the entire
+tasklist and select a task based on heuristics to kill.  This normally
+selects a rogue memory-hogging task that frees up a large amount of
+memory when killed.
+
+If this is set to non-zero, the OOM killer simply kills the task that
+triggered the out-of-memory condition.  This avoids the expensive
+tasklist scan.
+
+If panic_on_oom is selected, it takes precedence over whatever value
+is used in oom_kill_allocating_task.
+
+The default value is 0.
  
 ==============================================================
  
@@ -63,6 +63,7 @@
 extern int sysctl_overcommit_memory;
 extern int sysctl_overcommit_ratio;
 extern int sysctl_panic_on_oom;
+extern int sysctl_oom_kill_allocating_task;
 extern int max_threads;
 extern int core_uses_pid;
 extern int suid_dumpable;
@@ -777,6 +778,14 @@
 		.procname	= "panic_on_oom",
 		.data		= &sysctl_panic_on_oom,
 		.maxlen		= sizeof(sysctl_panic_on_oom),
+		.mode		= 0644,
+		.proc_handler	= &proc_dointvec,
+	},
+	{
+		.ctl_name	= CTL_UNNUMBERED,
+		.procname	= "oom_kill_allocating_task",
+		.data		= &sysctl_oom_kill_allocating_task,
+		.maxlen		= sizeof(sysctl_oom_kill_allocating_task),
 		.mode		= 0644,
 		.proc_handler	= &proc_dointvec,
 	},
@@ -27,6 +27,7 @@
 #include <linux/notifier.h>
  
 int sysctl_panic_on_oom;
+int sysctl_oom_kill_allocating_task;
 static DEFINE_MUTEX(zone_scan_mutex);
 /* #define DEBUG */
  
  
@@ -471,14 +472,16 @@
 				"No available memory (MPOL_BIND)");
 		break;
  
-	case CONSTRAINT_CPUSET:
-		oom_kill_process(current, points,
-				"No available memory in cpuset");
-		break;
-
 	case CONSTRAINT_NONE:
 		if (sysctl_panic_on_oom)
 			panic("out of memory. panic_on_oom is selected\n");
+		/* Fall-through */
+	case CONSTRAINT_CPUSET:
+		if (sysctl_oom_kill_allocating_task) {
+			oom_kill_process(current, points,
+					"Out of memory (oom_kill_allocating_task)");
+			break;
+		}
 retry:
 		/*
 		 * Rambo mode: Shoot down a process and hope it solves whatever
...	...	@@ -31,6 +31,7 @@
31	31	- min_unmapped_ratio
32	32	- min_slab_ratio
33	33	- panic_on_oom
	34	+- oom_kill_allocating_task
34	35	- mmap_min_address
35	36	- numa_zonelist_order
36	37
...	...	@@ -219,6 +220,27 @@
219	220	The default value is 0.
220	221	1 and 2 are for failover of clustering. Please select either
221	222	according to your policy of failover.
	223	+
	224	+=============================================================
	225	+
	226	+oom_kill_allocating_task
	227	+
	228	+This enables or disables killing the OOM-triggering task in
	229	+out-of-memory situations.
	230	+
	231	+If this is set to zero, the OOM killer will scan through the entire
	232	+tasklist and select a task based on heuristics to kill. This normally
	233	+selects a rogue memory-hogging task that frees up a large amount of
	234	+memory when killed.
	235	+
	236	+If this is set to non-zero, the OOM killer simply kills the task that
	237	+triggered the out-of-memory condition. This avoids the expensive
	238	+tasklist scan.
	239	+
	240	+If panic_on_oom is selected, it takes precedence over whatever value
	241	+is used in oom_kill_allocating_task.
	242	+
	243	+The default value is 0.
222	244
223	245	==============================================================
224	246
...	...	@@ -63,6 +63,7 @@
63	63	extern int sysctl_overcommit_memory;
64	64	extern int sysctl_overcommit_ratio;
65	65	extern int sysctl_panic_on_oom;
	66	+extern int sysctl_oom_kill_allocating_task;
66	67	extern int max_threads;
67	68	extern int core_uses_pid;
68	69	extern int suid_dumpable;
...	...	@@ -777,6 +778,14 @@
777	778	.procname = "panic_on_oom",
778	779	.data = &sysctl_panic_on_oom,
779	780	.maxlen = sizeof(sysctl_panic_on_oom),
	781	+ .mode = 0644,
	782	+ .proc_handler = &proc_dointvec,
	783	+ },
	784	+ {
	785	+ .ctl_name = CTL_UNNUMBERED,
	786	+ .procname = "oom_kill_allocating_task",
	787	+ .data = &sysctl_oom_kill_allocating_task,
	788	+ .maxlen = sizeof(sysctl_oom_kill_allocating_task),
780	789	.mode = 0644,
781	790	.proc_handler = &proc_dointvec,
782	791	},
...	...	@@ -27,6 +27,7 @@
27	27	#include <linux/notifier.h>
28	28
29	29	int sysctl_panic_on_oom;
	30	+int sysctl_oom_kill_allocating_task;
30	31	static DEFINE_MUTEX(zone_scan_mutex);
31	32	/* #define DEBUG */
32	33
33	34
...	...	@@ -471,14 +472,16 @@
471	472	"No available memory (MPOL_BIND)");
472	473	break;
473	474
474		- case CONSTRAINT_CPUSET:
475		- oom_kill_process(current, points,
476		- "No available memory in cpuset");
477		- break;
478		-
479	475	case CONSTRAINT_NONE:
480	476	if (sysctl_panic_on_oom)
481	477	panic("out of memory. panic_on_oom is selected\n");
	478	+ /* Fall-through */
	479	+ case CONSTRAINT_CPUSET:
	480	+ if (sysctl_oom_kill_allocating_task) {
	481	+ oom_kill_process(current, points,
	482	+ "Out of memory (oom_kill_allocating_task)");
	483	+ break;
	484	+ }
482	485	retry:
483	486	/*
484	487	* Rambo mode: Shoot down a process and hope it solves whatever