KVM: VMX: Properly return error to userspace on vmentry failure
[linux-2.6.git] / mm / oom_kill.c
index a6879f2..709aedf 100644 (file)
@@ -17,6 +17,8 @@
 
 #include <linux/oom.h>
 #include <linux/mm.h>
+#include <linux/err.h>
+#include <linux/gfp.h>
 #include <linux/sched.h>
 #include <linux/swap.h>
 #include <linux/timex.h>
 #include <linux/cpuset.h>
 #include <linux/module.h>
 #include <linux/notifier.h>
+#include <linux/memcontrol.h>
+#include <linux/security.h>
 
 int sysctl_panic_on_oom;
+int sysctl_oom_kill_allocating_task;
+int sysctl_oom_dump_tasks;
+static DEFINE_SPINLOCK(zone_scan_lock);
 /* #define DEBUG */
 
+/*
+ * Is all threads of the target process nodes overlap ours?
+ */
+static int has_intersects_mems_allowed(struct task_struct *tsk)
+{
+       struct task_struct *t;
+
+       t = tsk;
+       do {
+               if (cpuset_mems_allowed_intersects(current, t))
+                       return 1;
+               t = next_thread(t);
+       } while (t != tsk);
+
+       return 0;
+}
+
 /**
  * badness - calculate a numeric value for how bad this task has been
  * @p: task struct of which task we should calculate
@@ -49,9 +73,16 @@ int sysctl_panic_on_oom;
 
 unsigned long badness(struct task_struct *p, unsigned long uptime)
 {
-       unsigned long points, cpu_time, run_time, s;
+       unsigned long points, cpu_time, run_time;
        struct mm_struct *mm;
        struct task_struct *child;
+       int oom_adj = p->signal->oom_adj;
+       struct task_cputime task_time;
+       unsigned long utime;
+       unsigned long stime;
+
+       if (oom_adj == OOM_DISABLE)
+               return 0;
 
        task_lock(p);
        mm = p->mm;
@@ -61,12 +92,6 @@ unsigned long badness(struct task_struct *p, unsigned long uptime)
        }
 
        /*
-        * swapoff can easily use up all memory, so kill those first.
-        */
-       if (p->flags & PF_SWAPOFF)
-               return ULONG_MAX;
-
-       /*
         * The memory size of the process is the basis for the badness.
         */
        points = mm->total_vm;
@@ -77,6 +102,12 @@ unsigned long badness(struct task_struct *p, unsigned long uptime)
        task_unlock(p);
 
        /*
+        * swapoff can easily use up all memory, so kill those first.
+        */
+       if (p->flags & PF_OOM_ORIGIN)
+               return ULONG_MAX;
+
+       /*
         * Processes which fork a lot of child processes are likely
         * a good choice. We add half the vmsize of the children if they
         * have an own mm. This prevents forking servers to flood the
@@ -96,20 +127,21 @@ unsigned long badness(struct task_struct *p, unsigned long uptime)
          * of seconds. There is no particular reason for this other than
          * that it turned out to work very well in practice.
         */
-       cpu_time = (cputime_to_jiffies(p->utime) + cputime_to_jiffies(p->stime))
-               >> (SHIFT_HZ + 3);
+       thread_group_cputime(p, &task_time);
+       utime = cputime_to_jiffies(task_time.utime);
+       stime = cputime_to_jiffies(task_time.stime);
+       cpu_time = (utime + stime) >> (SHIFT_HZ + 3);
+
 
        if (uptime >= p->start_time.tv_sec)
                run_time = (uptime - p->start_time.tv_sec) >> 10;
        else
                run_time = 0;
 
-       s = int_sqrt(cpu_time);
-       if (s)
-               points /= s;
-       s = int_sqrt(int_sqrt(run_time));
-       if (s)
-               points /= s;
+       if (cpu_time)
+               points /= int_sqrt(cpu_time);
+       if (run_time)
+               points /= int_sqrt(int_sqrt(run_time));
 
        /*
         * Niced processes are most likely less important, so double
@@ -122,8 +154,8 @@ unsigned long badness(struct task_struct *p, unsigned long uptime)
         * Superuser processes are usually more important, so we make it
         * less likely that we kill those.
         */
-       if (cap_t(p->cap_effective) & CAP_TO_MASK(CAP_SYS_ADMIN) ||
-                               p->uid == 0 || p->euid == 0)
+       if (has_capability_noaudit(p, CAP_SYS_ADMIN) ||
+           has_capability_noaudit(p, CAP_SYS_RESOURCE))
                points /= 4;
 
        /*
@@ -132,7 +164,7 @@ unsigned long badness(struct task_struct *p, unsigned long uptime)
         * tend to only have this flag set on applications they think
         * of as important.
         */
-       if (cap_t(p->cap_effective) & CAP_TO_MASK(CAP_SYS_RAWIO))
+       if (has_capability_noaudit(p, CAP_SYS_RAWIO))
                points /= 4;
 
        /*
@@ -140,54 +172,71 @@ unsigned long badness(struct task_struct *p, unsigned long uptime)
         * because p may have allocated or otherwise mapped memory on
         * this node before. However it will be less likely.
         */
-       if (!cpuset_excl_nodes_overlap(p))
+       if (!has_intersects_mems_allowed(p))
                points /= 8;
 
        /*
-        * Adjust the score by oomkilladj.
+        * Adjust the score by oom_adj.
         */
-       if (p->oomkilladj) {
-               if (p->oomkilladj > 0)
-                       points <<= p->oomkilladj;
-               else
-                       points >>= -(p->oomkilladj);
+       if (oom_adj) {
+               if (oom_adj > 0) {
+                       if (!points)
+                               points = 1;
+                       points <<= oom_adj;
+               } else
+                       points >>= -(oom_adj);
        }
 
 #ifdef DEBUG
-       printk(KERN_DEBUG "OOMkill: task %d (%s) got %d points\n",
+       printk(KERN_DEBUG "OOMkill: task %d (%s) got %lu points\n",
        p->pid, p->comm, points);
 #endif
        return points;
 }
 
 /*
- * Types of limitations to the nodes from which allocations may occur
- */
-#define CONSTRAINT_NONE 1
-#define CONSTRAINT_MEMORY_POLICY 2
-#define CONSTRAINT_CPUSET 3
-
-/*
  * Determine the type of allocation constraint.
  */
-static inline int constrained_alloc(struct zonelist *zonelist, gfp_t gfp_mask)
-{
 #ifdef CONFIG_NUMA
-       struct zone **z;
-       nodemask_t nodes = node_online_map;
+static enum oom_constraint constrained_alloc(struct zonelist *zonelist,
+                                   gfp_t gfp_mask, nodemask_t *nodemask)
+{
+       struct zone *zone;
+       struct zoneref *z;
+       enum zone_type high_zoneidx = gfp_zone(gfp_mask);
 
-       for (z = zonelist->zones; *z; z++)
-               if (cpuset_zone_allowed(*z, gfp_mask))
-                       node_clear(zone_to_nid(*z), nodes);
-               else
-                       return CONSTRAINT_CPUSET;
+       /*
+        * Reach here only when __GFP_NOFAIL is used. So, we should avoid
+        * to kill current.We have to random task kill in this case.
+        * Hopefully, CONSTRAINT_THISNODE...but no way to handle it, now.
+        */
+       if (gfp_mask & __GFP_THISNODE)
+               return CONSTRAINT_NONE;
 
-       if (!nodes_empty(nodes))
+       /*
+        * The nodemask here is a nodemask passed to alloc_pages(). Now,
+        * cpuset doesn't use this nodemask for its hardwall/softwall/hierarchy
+        * feature. mempolicy is an only user of nodemask here.
+        * check mempolicy's nodemask contains all N_HIGH_MEMORY
+        */
+       if (nodemask && !nodes_subset(node_states[N_HIGH_MEMORY], *nodemask))
                return CONSTRAINT_MEMORY_POLICY;
-#endif
 
+       /* Check this allocation failure is caused by cpuset's wall function */
+       for_each_zone_zonelist_nodemask(zone, z, zonelist,
+                       high_zoneidx, nodemask)
+               if (!cpuset_zone_allowed_softwall(zone, gfp_mask))
+                       return CONSTRAINT_CPUSET;
+
+       return CONSTRAINT_NONE;
+}
+#else
+static enum oom_constraint constrained_alloc(struct zonelist *zonelist,
+                               gfp_t gfp_mask, nodemask_t *nodemask)
+{
        return CONSTRAINT_NONE;
 }
+#endif
 
 /*
  * Simple selection loop. We chose the process with the highest
@@ -195,15 +244,16 @@ static inline int constrained_alloc(struct zonelist *zonelist, gfp_t gfp_mask)
  *
  * (not docbooked, we don't want this one cluttering up the manual)
  */
-static struct task_struct *select_bad_process(unsigned long *ppoints)
+static struct task_struct *select_bad_process(unsigned long *ppoints,
+                                               struct mem_cgroup *mem)
 {
-       struct task_struct *g, *p;
+       struct task_struct *p;
        struct task_struct *chosen = NULL;
        struct timespec uptime;
        *ppoints = 0;
 
        do_posix_clock_monotonic_gettime(&uptime);
-       do_each_thread(g, p) {
+       for_each_process(p) {
                unsigned long points;
 
                /*
@@ -213,7 +263,9 @@ static struct task_struct *select_bad_process(unsigned long *ppoints)
                if (!p->mm)
                        continue;
                /* skip the init task */
-               if (is_init(p))
+               if (is_global_init(p))
+                       continue;
+               if (mem && !task_in_mem_cgroup(p, mem))
                        continue;
 
                /*
@@ -246,7 +298,7 @@ static struct task_struct *select_bad_process(unsigned long *ppoints)
                        *ppoints = ULONG_MAX;
                }
 
-               if (p->oomkilladj == OOM_DISABLE)
+               if (p->signal->oom_adj == OOM_DISABLE)
                        continue;
 
                points = badness(p, uptime.tv_sec);
@@ -254,39 +306,112 @@ static struct task_struct *select_bad_process(unsigned long *ppoints)
                        chosen = p;
                        *ppoints = points;
                }
-       } while_each_thread(g, p);
+       }
 
        return chosen;
 }
 
 /**
+ * dump_tasks - dump current memory state of all system tasks
+ * @mem: target memory controller
+ *
+ * Dumps the current memory state of all system tasks, excluding kernel threads.
+ * State information includes task's pid, uid, tgid, vm size, rss, cpu, oom_adj
+ * score, and name.
+ *
+ * If the actual is non-NULL, only tasks that are a member of the mem_cgroup are
+ * shown.
+ *
+ * Call with tasklist_lock read-locked.
+ */
+static void dump_tasks(const struct mem_cgroup *mem)
+{
+       struct task_struct *g, *p;
+
+       printk(KERN_INFO "[ pid ]   uid  tgid total_vm      rss cpu oom_adj "
+              "name\n");
+       do_each_thread(g, p) {
+               struct mm_struct *mm;
+
+               if (mem && !task_in_mem_cgroup(p, mem))
+                       continue;
+               if (!thread_group_leader(p))
+                       continue;
+
+               task_lock(p);
+               mm = p->mm;
+               if (!mm) {
+                       /*
+                        * total_vm and rss sizes do not exist for tasks with no
+                        * mm so there's no need to report them; they can't be
+                        * oom killed anyway.
+                        */
+                       task_unlock(p);
+                       continue;
+               }
+               printk(KERN_INFO "[%5d] %5d %5d %8lu %8lu %3d     %3d %s\n",
+                      p->pid, __task_cred(p)->uid, p->tgid, mm->total_vm,
+                      get_mm_rss(mm), (int)task_cpu(p), p->signal->oom_adj,
+                      p->comm);
+               task_unlock(p);
+       } while_each_thread(g, p);
+}
+
+static void dump_header(struct task_struct *p, gfp_t gfp_mask, int order,
+                                                       struct mem_cgroup *mem)
+{
+       pr_warning("%s invoked oom-killer: gfp_mask=0x%x, order=%d, "
+               "oom_adj=%d\n",
+               current->comm, gfp_mask, order, current->signal->oom_adj);
+       task_lock(current);
+       cpuset_print_task_mems_allowed(current);
+       task_unlock(current);
+       dump_stack();
+       mem_cgroup_print_oom_info(mem, p);
+       show_mem();
+       if (sysctl_oom_dump_tasks)
+               dump_tasks(mem);
+}
+
+#define K(x) ((x) << (PAGE_SHIFT-10))
+
+/*
  * Send SIGKILL to the selected  process irrespective of  CAP_SYS_RAW_IO
  * flag though it's unlikely that  we select a process with CAP_SYS_RAW_IO
  * set.
  */
 static void __oom_kill_task(struct task_struct *p, int verbose)
 {
-       if (is_init(p)) {
+       if (is_global_init(p)) {
                WARN_ON(1);
                printk(KERN_WARNING "tried to kill init!\n");
                return;
        }
 
+       task_lock(p);
        if (!p->mm) {
                WARN_ON(1);
-               printk(KERN_WARNING "tried to kill an mm-less task!\n");
+               printk(KERN_WARNING "tried to kill an mm-less task %d (%s)!\n",
+                       task_pid_nr(p), p->comm);
+               task_unlock(p);
                return;
        }
 
        if (verbose)
-               printk(KERN_ERR "Killed process %d (%s)\n", p->pid, p->comm);
+               printk(KERN_ERR "Killed process %d (%s) "
+                      "vsz:%lukB, anon-rss:%lukB, file-rss:%lukB\n",
+                      task_pid_nr(p), p->comm,
+                      K(p->mm->total_vm),
+                      K(get_mm_counter(p->mm, MM_ANONPAGES)),
+                      K(get_mm_counter(p->mm, MM_FILEPAGES)));
+       task_unlock(p);
 
        /*
         * We give our sacrificial lamb high priority and access to
         * all the memory it needs. That way it should be able to
         * exit() and clear out its resources quickly...
         */
-       p->time_slice = HZ;
+       p->rt.time_slice = HZ;
        set_tsk_thread_flag(p, TIF_MEMDIE);
 
        force_sig(SIGKILL, p);
@@ -294,11 +419,6 @@ static void __oom_kill_task(struct task_struct *p, int verbose)
 
 static int oom_kill_task(struct task_struct *p)
 {
-       struct mm_struct *mm;
-       struct task_struct *g, *q;
-
-       mm = p->mm;
-
        /* WARNING: mm may not be dereferenced since we did not obtain its
         * value from get_task_mm(p).  This is OK since all we need to do is
         * compare mm to q->mm below.
@@ -307,37 +427,22 @@ static int oom_kill_task(struct task_struct *p)
         * change to NULL at any time since we do not hold task_lock(p).
         * However, this is of no concern to us.
         */
-
-       if (mm == NULL)
+       if (!p->mm || p->signal->oom_adj == OOM_DISABLE)
                return 1;
 
-       /*
-        * Don't kill the process if any threads are set to OOM_DISABLE
-        */
-       do_each_thread(g, q) {
-               if (q->mm == mm && p->oomkilladj == OOM_DISABLE)
-                       return 1;
-       } while_each_thread(g, q);
-
        __oom_kill_task(p, 1);
 
-       /*
-        * kill all processes that share the ->mm (i.e. all threads),
-        * but are in a different thread group.
-        */
-       do_each_thread(g, q) {
-               if (q->mm == mm && q->tgid != p->tgid)
-                       __oom_kill_task(q, 1);
-       } while_each_thread(g, q);
-
        return 0;
 }
 
-static int oom_kill_process(struct task_struct *p, unsigned long points,
-               const char *message)
+static int oom_kill_process(struct task_struct *p, gfp_t gfp_mask, int order,
+                           unsigned long points, struct mem_cgroup *mem,
+                           const char *message)
 {
        struct task_struct *c;
-       struct list_head *tsk;
+
+       if (printk_ratelimit())
+               dump_header(p, gfp_mask, order, mem);
 
        /*
         * If the task is already exiting, don't alarm the sysadmin or kill
@@ -349,19 +454,42 @@ static int oom_kill_process(struct task_struct *p, unsigned long points,
        }
 
        printk(KERN_ERR "%s: kill process %d (%s) score %li or a child\n",
-                                       message, p->pid, p->comm, points);
+                                       message, task_pid_nr(p), p->comm, points);
 
        /* Try to kill a child first */
-       list_for_each(tsk, &p->children) {
-               c = list_entry(tsk, struct task_struct, sibling);
+       list_for_each_entry(c, &p->children, sibling) {
                if (c->mm == p->mm)
                        continue;
+               if (mem && !task_in_mem_cgroup(c, mem))
+                       continue;
                if (!oom_kill_task(c))
                        return 0;
        }
        return oom_kill_task(p);
 }
 
+#ifdef CONFIG_CGROUP_MEM_RES_CTLR
+void mem_cgroup_out_of_memory(struct mem_cgroup *mem, gfp_t gfp_mask)
+{
+       unsigned long points = 0;
+       struct task_struct *p;
+
+       if (sysctl_panic_on_oom == 2)
+               panic("out of memory(memcg). panic_on_oom is selected.\n");
+       read_lock(&tasklist_lock);
+retry:
+       p = select_bad_process(&points, mem);
+       if (!p || PTR_ERR(p) == -1UL)
+               goto out;
+
+       if (oom_kill_process(p, gfp_mask, 0, points, mem,
+                               "Memory cgroup out of memory"))
+               goto retry;
+out:
+       read_unlock(&tasklist_lock);
+}
+#endif
+
 static BLOCKING_NOTIFIER_HEAD(oom_notify_list);
 
 int register_oom_notifier(struct notifier_block *nb)
@@ -376,80 +504,170 @@ int unregister_oom_notifier(struct notifier_block *nb)
 }
 EXPORT_SYMBOL_GPL(unregister_oom_notifier);
 
+/*
+ * Try to acquire the OOM killer lock for the zones in zonelist.  Returns zero
+ * if a parallel OOM killing is already taking place that includes a zone in
+ * the zonelist.  Otherwise, locks all zones in the zonelist and returns 1.
+ */
+int try_set_zone_oom(struct zonelist *zonelist, gfp_t gfp_mask)
+{
+       struct zoneref *z;
+       struct zone *zone;
+       int ret = 1;
+
+       spin_lock(&zone_scan_lock);
+       for_each_zone_zonelist(zone, z, zonelist, gfp_zone(gfp_mask)) {
+               if (zone_is_oom_locked(zone)) {
+                       ret = 0;
+                       goto out;
+               }
+       }
+
+       for_each_zone_zonelist(zone, z, zonelist, gfp_zone(gfp_mask)) {
+               /*
+                * Lock each zone in the zonelist under zone_scan_lock so a
+                * parallel invocation of try_set_zone_oom() doesn't succeed
+                * when it shouldn't.
+                */
+               zone_set_flag(zone, ZONE_OOM_LOCKED);
+       }
+
+out:
+       spin_unlock(&zone_scan_lock);
+       return ret;
+}
+
+/*
+ * Clears the ZONE_OOM_LOCKED flag for all zones in the zonelist so that failed
+ * allocation attempts with zonelists containing them may now recall the OOM
+ * killer, if necessary.
+ */
+void clear_zonelist_oom(struct zonelist *zonelist, gfp_t gfp_mask)
+{
+       struct zoneref *z;
+       struct zone *zone;
+
+       spin_lock(&zone_scan_lock);
+       for_each_zone_zonelist(zone, z, zonelist, gfp_zone(gfp_mask)) {
+               zone_clear_flag(zone, ZONE_OOM_LOCKED);
+       }
+       spin_unlock(&zone_scan_lock);
+}
+
+/*
+ * Must be called with tasklist_lock held for read.
+ */
+static void __out_of_memory(gfp_t gfp_mask, int order)
+{
+       struct task_struct *p;
+       unsigned long points;
+
+       if (sysctl_oom_kill_allocating_task)
+               if (!oom_kill_process(current, gfp_mask, order, 0, NULL,
+                               "Out of memory (oom_kill_allocating_task)"))
+                       return;
+retry:
+       /*
+        * Rambo mode: Shoot down a process and hope it solves whatever
+        * issues we may have.
+        */
+       p = select_bad_process(&points, NULL);
+
+       if (PTR_ERR(p) == -1UL)
+               return;
+
+       /* Found nothing?!?! Either we hang forever, or we panic. */
+       if (!p) {
+               read_unlock(&tasklist_lock);
+               dump_header(NULL, gfp_mask, order, NULL);
+               panic("Out of memory and no killable processes...\n");
+       }
+
+       if (oom_kill_process(p, gfp_mask, order, points, NULL,
+                            "Out of memory"))
+               goto retry;
+}
+
+/*
+ * pagefault handler calls into here because it is out of memory but
+ * doesn't know exactly how or why.
+ */
+void pagefault_out_of_memory(void)
+{
+       unsigned long freed = 0;
+
+       blocking_notifier_call_chain(&oom_notify_list, 0, &freed);
+       if (freed > 0)
+               /* Got some memory back in the last second. */
+               return;
+
+       if (sysctl_panic_on_oom)
+               panic("out of memory from page fault. panic_on_oom is selected.\n");
+
+       read_lock(&tasklist_lock);
+       __out_of_memory(0, 0); /* unknown gfp_mask and order */
+       read_unlock(&tasklist_lock);
+
+       /*
+        * Give "p" a good chance of killing itself before we
+        * retry to allocate memory.
+        */
+       if (!test_thread_flag(TIF_MEMDIE))
+               schedule_timeout_uninterruptible(1);
+}
+
 /**
  * out_of_memory - kill the "best" process when we run out of memory
+ * @zonelist: zonelist pointer
+ * @gfp_mask: memory allocation flags
+ * @order: amount of memory being requested as a power of 2
  *
  * If we run out of memory, we have the choice between either
  * killing a random task (bad), letting the system crash (worse)
  * OR try to be smart about which process to kill. Note that we
  * don't have to be perfect here, we just have to be good.
  */
-void out_of_memory(struct zonelist *zonelist, gfp_t gfp_mask, int order)
+void out_of_memory(struct zonelist *zonelist, gfp_t gfp_mask,
+               int order, nodemask_t *nodemask)
 {
-       struct task_struct *p;
-       unsigned long points = 0;
        unsigned long freed = 0;
+       enum oom_constraint constraint;
 
        blocking_notifier_call_chain(&oom_notify_list, 0, &freed);
        if (freed > 0)
                /* Got some memory back in the last second. */
                return;
 
-       if (printk_ratelimit()) {
-               printk(KERN_WARNING "%s invoked oom-killer: "
-                       "gfp_mask=0x%x, order=%d, oomkilladj=%d\n",
-                       current->comm, gfp_mask, order, current->oomkilladj);
-               dump_stack();
-               show_mem();
+       if (sysctl_panic_on_oom == 2) {
+               dump_header(NULL, gfp_mask, order, NULL);
+               panic("out of memory. Compulsory panic_on_oom is selected.\n");
        }
 
-       cpuset_lock();
-       read_lock(&tasklist_lock);
-
        /*
         * Check if there were limitations on the allocation (only relevant for
         * NUMA) that may require different handling.
         */
-       switch (constrained_alloc(zonelist, gfp_mask)) {
+       constraint = constrained_alloc(zonelist, gfp_mask, nodemask);
+       read_lock(&tasklist_lock);
+
+       switch (constraint) {
        case CONSTRAINT_MEMORY_POLICY:
-               oom_kill_process(current, points,
+               oom_kill_process(current, gfp_mask, order, 0, NULL,
                                "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)
+               if (sysctl_panic_on_oom) {
+                       dump_header(NULL, gfp_mask, order, NULL);
                        panic("out of memory. panic_on_oom is selected\n");
-retry:
-               /*
-                * Rambo mode: Shoot down a process and hope it solves whatever
-                * issues we may have.
-                */
-               p = select_bad_process(&points);
-
-               if (PTR_ERR(p) == -1UL)
-                       goto out;
-
-               /* Found nothing?!?! Either we hang forever, or we panic. */
-               if (!p) {
-                       read_unlock(&tasklist_lock);
-                       cpuset_unlock();
-                       panic("Out of memory and no killable processes...\n");
                }
-
-               if (oom_kill_process(p, points, "Out of memory"))
-                       goto retry;
-
+               /* Fall-through */
+       case CONSTRAINT_CPUSET:
+               __out_of_memory(gfp_mask, order);
                break;
        }
 
-out:
        read_unlock(&tasklist_lock);
-       cpuset_unlock();
 
        /*
         * Give "p" a good chance of killing itself before we