Commit 7f4d454d authored by Daisuke Nishimura's avatar Daisuke Nishimura Committed by Linus Torvalds
Browse files

memcg: avoid deadlock caused by race between oom and cpuset_attach



mpol_rebind_mm(), which can be called from cpuset_attach(), does
down_write(mm->mmap_sem).  This means down_write(mm->mmap_sem) can be
called under cgroup_mutex.

OTOH, page fault path does down_read(mm->mmap_sem) and calls
mem_cgroup_try_charge_xxx(), which may eventually calls
mem_cgroup_out_of_memory().  And mem_cgroup_out_of_memory() calls
cgroup_lock().  This means cgroup_lock() can be called under
down_read(mm->mmap_sem).

If those two paths race, deadlock can happen.

This patch avoid this deadlock by:
  - remove cgroup_lock() from mem_cgroup_out_of_memory().
  - define new mutex (memcg_tasklist) and serialize mem_cgroup_move_task()
    (->attach handler of memory cgroup) and mem_cgroup_out_of_memory.
Signed-off-by: default avatarDaisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Reviewed-by: default avatarKAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Acked-by: default avatarBalbir Singh <balbir@linux.vnet.ibm.com>
Signed-off-by: default avatarAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
parent a5e924f5
...@@ -51,6 +51,7 @@ static int really_do_swap_account __initdata = 1; /* for remember boot option*/ ...@@ -51,6 +51,7 @@ static int really_do_swap_account __initdata = 1; /* for remember boot option*/
#define do_swap_account (0) #define do_swap_account (0)
#endif #endif
static DEFINE_MUTEX(memcg_tasklist); /* can be hold under cgroup_mutex */
/* /*
* Statistics for memory cgroup. * Statistics for memory cgroup.
...@@ -827,7 +828,9 @@ static int __mem_cgroup_try_charge(struct mm_struct *mm, ...@@ -827,7 +828,9 @@ static int __mem_cgroup_try_charge(struct mm_struct *mm,
if (!nr_retries--) { if (!nr_retries--) {
if (oom) { if (oom) {
mutex_lock(&memcg_tasklist);
mem_cgroup_out_of_memory(mem_over_limit, gfp_mask); mem_cgroup_out_of_memory(mem_over_limit, gfp_mask);
mutex_unlock(&memcg_tasklist);
mem_over_limit->last_oom_jiffies = jiffies; mem_over_limit->last_oom_jiffies = jiffies;
} }
goto nomem; goto nomem;
...@@ -2211,10 +2214,12 @@ static void mem_cgroup_move_task(struct cgroup_subsys *ss, ...@@ -2211,10 +2214,12 @@ static void mem_cgroup_move_task(struct cgroup_subsys *ss,
struct cgroup *old_cont, struct cgroup *old_cont,
struct task_struct *p) struct task_struct *p)
{ {
mutex_lock(&memcg_tasklist);
/* /*
* FIXME: It's better to move charges of this process from old * FIXME: It's better to move charges of this process from old
* memcg to new memcg. But it's just on TODO-List now. * memcg to new memcg. But it's just on TODO-List now.
*/ */
mutex_unlock(&memcg_tasklist);
} }
struct cgroup_subsys mem_cgroup_subsys = { struct cgroup_subsys mem_cgroup_subsys = {
......
...@@ -429,7 +429,6 @@ void mem_cgroup_out_of_memory(struct mem_cgroup *mem, gfp_t gfp_mask) ...@@ -429,7 +429,6 @@ void mem_cgroup_out_of_memory(struct mem_cgroup *mem, gfp_t gfp_mask)
unsigned long points = 0; unsigned long points = 0;
struct task_struct *p; struct task_struct *p;
cgroup_lock();
read_lock(&tasklist_lock); read_lock(&tasklist_lock);
retry: retry:
p = select_bad_process(&points, mem); p = select_bad_process(&points, mem);
...@@ -444,7 +443,6 @@ retry: ...@@ -444,7 +443,6 @@ retry:
goto retry; goto retry;
out: out:
read_unlock(&tasklist_lock); read_unlock(&tasklist_lock);
cgroup_unlock();
} }
#endif #endif
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment