memcg: ensure list is empty at rmdir
Daisuke Nishimura [Sat, 16 Jan 2010 01:01:30 +0000 (17:01 -0800)]
Current mem_cgroup_force_empty() only ensures mem->res.usage == 0 on
success.  But this doesn't guarantee memcg's LRU is really empty, because
there are some cases in which !PageCgrupUsed pages exist on memcg's LRU.

For example:
- Pages can be uncharged by its owner process while they are on LRU.
- race between mem_cgroup_add_lru_list() and __mem_cgroup_uncharge_common().

So there can be a case in which the usage is zero but some of the LRUs are not empty.

OTOH, mem_cgroup_del_lru_list(), which can be called asynchronously with
rmdir, accesses the mem_cgroup, so this access can cause a problem if it
races with rmdir because the mem_cgroup might have been freed by rmdir.

Actually, I saw a bug which seems to be caused by this race.

[1530745.949906] BUG: unable to handle kernel NULL pointer dereference at 0000000000000230
[1530745.950651] IP: [<ffffffff810fbc11>] mem_cgroup_del_lru_list+0x30/0x80
[1530745.950651] PGD 3863de067 PUD 3862c7067 PMD 0
[1530745.950651] Oops: 0002 [#1] SMP
[1530745.950651] last sysfs file: /sys/devices/system/cpu/cpu7/cache/index1/shared_cpu_map
[1530745.950651] CPU 3
[1530745.950651] Modules linked in: configs ipt_REJECT xt_tcpudp iptable_filter ip_tables x_tables bridge stp nfsd nfs_acl auth_rpcgss exportfs autofs4 hidp rfcomm l2cap crc16 bluetooth lockd sunrpc ib_iser rdma_cm ib_cm iw_cm ib_sa ib_mad ib_core ib_addr iscsi_tcp bnx2i cnic uio ipv6 cxgb3i cxgb3 mdio libiscsi_tcp libiscsi scsi_transport_iscsi dm_mirror dm_multipath scsi_dh video output sbs sbshc battery ac lp kvm_intel kvm sg ide_cd_mod cdrom serio_raw tpm_tis tpm tpm_bios acpi_memhotplug button parport_pc parport rtc_cmos rtc_core rtc_lib e1000 i2c_i801 i2c_core pcspkr dm_region_hash dm_log dm_mod ata_piix libata shpchp megaraid_mbox sd_mod scsi_mod megaraid_mm ext3 jbd uhci_hcd ohci_hcd ehci_hcd [last unloaded: freq_table]
[1530745.950651] Pid: 19653, comm: shmem_test_02 Tainted: G   M       2.6.32-mm1-00701-g2b04386 #3 Express5800/140Rd-4 [N8100-1065]
[1530745.950651] RIP: 0010:[<ffffffff810fbc11>]  [<ffffffff810fbc11>] mem_cgroup_del_lru_list+0x30/0x80
[1530745.950651] RSP: 0018:ffff8803863ddcb8  EFLAGS: 00010002
[1530745.950651] RAX: 00000000000001e0 RBX: ffff8803abc02238 RCX: 00000000000001e0
[1530745.950651] RDX: 0000000000000000 RSI: ffff88038611a000 RDI: ffff8803abc02238
[1530745.950651] RBP: ffff8803863ddcc8 R08: 0000000000000002 R09: ffff8803a04c8643
[1530745.950651] R10: 0000000000000000 R11: ffffffff810c7333 R12: 0000000000000000
[1530745.950651] R13: ffff880000017f00 R14: 0000000000000092 R15: ffff8800179d0310
[1530745.950651] FS:  0000000000000000(0000) GS:ffff880017800000(0000) knlGS:0000000000000000
[1530745.950651] CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[1530745.950651] CR2: 0000000000000230 CR3: 0000000379d87000 CR4: 00000000000006e0
[1530745.950651] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[1530745.950651] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[1530745.950651] Process shmem_test_02 (pid: 19653, threadinfo ffff8803863dc000, task ffff88038612a8a0)
[1530745.950651] Stack:
[1530745.950651]  ffffea00040c2fe8 0000000000000000 ffff8803863ddd98 ffffffff810c739a
[1530745.950651] <0> 00000000863ddd18 000000000000000c 0000000000000000 0000000000000000
[1530745.950651] <0> 0000000000000002 0000000000000000 ffff8803863ddd68 0000000000000046
[1530745.950651] Call Trace:
[1530745.950651]  [<ffffffff810c739a>] release_pages+0x142/0x1e7
[1530745.950651]  [<ffffffff810c778f>] ? pagevec_move_tail+0x6e/0x112
[1530745.950651]  [<ffffffff810c781e>] pagevec_move_tail+0xfd/0x112
[1530745.950651]  [<ffffffff810c78a9>] lru_add_drain+0x76/0x94
[1530745.950651]  [<ffffffff810dba0c>] exit_mmap+0x6e/0x145
[1530745.950651]  [<ffffffff8103f52d>] mmput+0x5e/0xcf
[1530745.950651]  [<ffffffff81043ea8>] exit_mm+0x11c/0x129
[1530745.950651]  [<ffffffff8108fb29>] ? audit_free+0x196/0x1c9
[1530745.950651]  [<ffffffff81045353>] do_exit+0x1f5/0x6b7
[1530745.950651]  [<ffffffff8106133f>] ? up_read+0x2b/0x2f
[1530745.950651]  [<ffffffff8137d187>] ? lockdep_sys_exit_thunk+0x35/0x67
[1530745.950651]  [<ffffffff81045898>] do_group_exit+0x83/0xb0
[1530745.950651]  [<ffffffff810458dc>] sys_exit_group+0x17/0x1b
[1530745.950651]  [<ffffffff81002c1b>] system_call_fastpath+0x16/0x1b
[1530745.950651] Code: 54 53 0f 1f 44 00 00 83 3d cc 29 7c 00 00 41 89 f4 75 63 eb 4e 48 83 7b 08 00 75 04 0f 0b eb fe 48 89 df e8 18 f3 ff ff 44 89 e2 <48> ff 4c d0 50 48 8b 05 2b 2d 7c 00 48 39 43 08 74 39 48 8b 4b
[1530745.950651] RIP  [<ffffffff810fbc11>] mem_cgroup_del_lru_list+0x30/0x80
[1530745.950651]  RSP <ffff8803863ddcb8>
[1530745.950651] CR2: 0000000000000230
[1530745.950651] ---[ end trace c3419c1bb8acc34f ]---
[1530745.950651] Fixing recursive fault but reboot is needed!

The problem here is pages on LRU may contain pointer to stale memcg.  To
make res->usage to be 0, all pages on memcg must be uncharged or moved to
another(parent) memcg.  Moved page_cgroup have already removed from
original LRU, but uncharged page_cgroup contains pointer to memcg withou
PCG_USED bit.  (This asynchronous LRU work is for improving performance.)
If PCG_USED bit is not set, page_cgroup will never be added to memcg's
LRU.  So, about pages not on LRU, they never access stale pointer.  Then,
what we have to take care of is page_cgroup _on_ LRU list.  This patch
fixes this problem by making mem_cgroup_force_empty() visit all LRUs
before exiting its loop and guarantee there are no pages on its LRU.

Signed-off-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Balbir Singh <balbir@linux.vnet.ibm.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/memcontrol.c

index 488b644..954032b 100644 (file)
@@ -2586,7 +2586,7 @@ static int mem_cgroup_force_empty(struct mem_cgroup *mem, bool free_all)
        if (free_all)
                goto try_to_free;
 move_account:
-       while (mem->res.usage > 0) {
+       do {
                ret = -EBUSY;
                if (cgroup_task_count(cgrp) || !list_empty(&cgrp->children))
                        goto out;
@@ -2614,8 +2614,8 @@ move_account:
                if (ret == -ENOMEM)
                        goto try_to_free;
                cond_resched();
-       }
-       ret = 0;
+       /* "ret" should also be checked to ensure all lists are empty. */
+       } while (mem->res.usage > 0 || ret);
 out:
        css_put(&mem->css);
        return ret;
@@ -2648,10 +2648,7 @@ try_to_free:
        }
        lru_add_drain();
        /* try move_account...there may be some *locked* pages. */
-       if (mem->res.usage)
-               goto move_account;
-       ret = 0;
-       goto out;
+       goto move_account;
 }
 
 int mem_cgroup_force_empty_write(struct cgroup *cont, unsigned int event)