On Mon, Jan 24, 2011 at 6:01 PM, CAI Qian <caiqian@xxxxxxxxxx> wrote: > This is not always reproducible so far but I thought it would send it out anyway in case someone could spot the problem. Running some memory allocation using ksm with swapping workloads (like oom02 in LTP) caused hung tasks in 2.6.38-rc2 kernel. > > INFO: task ksmd:278 blocked for more than 120 seconds. > "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. > ksmd      ÂD ffff88045ddab9e0   0  278   Â2 0x00000000 > Âffff88045ddadc20 0000000000000046 0000000000000000 0000000100000000 > Â0000000000014d40 ffff88045ddab480 ffff88045ddab9e0 ffff88045ddadfd8 > Âffff88045ddab9e8 0000000000014d40 ffff88045ddac010 0000000000014d40 > Call Trace: > Â[<ffffffff814a290d>] schedule_timeout+0x20d/0x2e0 > Â[<ffffffff8104d08d>] ? task_rq_lock+0x5d/0xa0 > Â[<ffffffff81058463>] ? try_to_wake_up+0xc3/0x420 > Â[<ffffffff814a253d>] wait_for_common+0x11d/0x190 > Â[<ffffffff810587c0>] ? default_wake_function+0x0/0x20 > Â[<ffffffff814a268d>] wait_for_completion+0x1d/0x20 > Â[<ffffffff81079530>] flush_work+0x30/0x40 > Â[<ffffffff81078230>] ? wq_barrier_func+0x0/0x20 > Â[<ffffffff81079603>] schedule_on_each_cpu+0xc3/0x110 > Â[<ffffffff81100ac5>] lru_add_drain_all+0x15/0x20 > Â[<ffffffff81137e58>] ksm_scan_thread+0x8e8/0xcb0 > Â[<ffffffff8107fd50>] ? autoremove_wake_function+0x0/0x40 > Â[<ffffffff81137570>] ? ksm_scan_thread+0x0/0xcb0 > Â[<ffffffff8107f6c6>] kthread+0x96/0xa0 > Â[<ffffffff8100cdc4>] kernel_thread_helper+0x4/0x10 > Â[<ffffffff8107f630>] ? kthread+0x0/0xa0 > Â[<ffffffff8100cdc0>] ? kernel_thread_helper+0x0/0x10 Oooh, that is unfair, I put that lru_add_drain_all() into ksmd just for you, and now you accuse it of hanging :) More seriously, it looks like the draining never gets a chance to run on one (or more) of the cpus, presumably something else is keeping that cpu unnaturally busy. I think this is either another manifestation of the same problem as in your "kswapd hung tasks" posting, or an example of kswapd itself too busy, as reported by David: see his patch to linux-mm today, http://marc.info/?l=linux-mm&m=129582353117092&w=2 Or perhaps those two cases are themselves related, though kswapd appears to be too busy in one case, and prevented from getting busy in the other. Hugh > INFO: task jbd2/dm-0-8:1179 blocked for more than 120 seconds. > "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. > jbd2/dm-0-8   D ffff88045d9a3120   0 Â1179   Â2 0x00000000 > Âffff88045cd97c20 0000000000000046 ffff88045d4d4fc0 ffff88045de984f8 > Â0000000000014d40 ffff88045d9a2bc0 ffff88045d9a3120 ffff88045cd97fd8 > Âffff88045d9a3128 0000000000014d40 ffff88045cd96010 0000000000014d40 > Call Trace: > Â[<ffffffff81175fb0>] ? sync_buffer+0x0/0x50 > Â[<ffffffff814a2330>] io_schedule+0x70/0xc0 > Â[<ffffffff81175ff0>] sync_buffer+0x40/0x50 > Â[<ffffffff814a2bef>] __wait_on_bit+0x5f/0x90 > Â[<ffffffff81175fb0>] ? sync_buffer+0x0/0x50 > Â[<ffffffff814a2c98>] out_of_line_wait_on_bit+0x78/0x90 > Â[<ffffffff8107fd90>] ? wake_bit_function+0x0/0x50 > Â[<ffffffff81175fae>] __wait_on_buffer+0x2e/0x30 > Â[<ffffffffa0082fe8>] jbd2_journal_commit_transaction+0x908/0x13f0 [jbd2] > Â[<ffffffff8106e2a3>] ? try_to_del_timer_sync+0x83/0xe0 > Â[<ffffffffa00882c8>] kjournald2+0xb8/0x220 [jbd2] > Â[<ffffffff8107fd50>] ? autoremove_wake_function+0x0/0x40 > Â[<ffffffffa0088210>] ? kjournald2+0x0/0x220 [jbd2] > Â[<ffffffff8107f6c6>] kthread+0x96/0xa0 > Â[<ffffffff8100cdc4>] kernel_thread_helper+0x4/0x10 > Â[<ffffffff8107f630>] ? kthread+0x0/0xa0 > Â[<ffffffff8100cdc0>] ? kernel_thread_helper+0x0/0x10 > INFO: task irqbalance:3321 blocked for more than 120 seconds. > "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. > irqbalance   ÂD ffff880c5d995020   0 Â3321   Â1 0x00000080 > Âffff880c5d4f7b08 0000000000000082 ffff880c5d4f7ab8 ffffffffffffffff > Â0000000000014d40 ffff880c5d994ac0 ffff880c5d995020 ffff880c5d4f7fd8 > Âffff880c5d995028 0000000000014d40 ffff880c5d4f6010 0000000000014d40 > Call Trace: > Â[<ffffffff810f5260>] ? sync_page+0x0/0x50 > Â[<ffffffff814a2330>] io_schedule+0x70/0xc0 > Â[<ffffffff810f52a0>] sync_page+0x40/0x50 > Â[<ffffffff814a2bef>] __wait_on_bit+0x5f/0x90 > Â[<ffffffff810f5463>] wait_on_page_bit+0x73/0x80 > Â[<ffffffff8107fd90>] ? wake_bit_function+0x0/0x50 > Â[<ffffffff810f550a>] __lock_page_or_retry+0x3a/0x60 > Â[<ffffffff810f6537>] filemap_fault+0x2d7/0x4c0 > Â[<ffffffff81119694>] __do_fault+0x54/0x570 > Â[<ffffffff81141357>] ? mem_cgroup_uncharge_swap+0x27/0x80 > Â[<ffffffff81119ca7>] handle_pte_fault+0xf7/0xb20 > Â[<ffffffff81115727>] ? __pte_alloc+0xd7/0xf0 > Â[<ffffffff8111a826>] handle_mm_fault+0x156/0x250 > Â[<ffffffff814a7da3>] do_page_fault+0x143/0x4b0 > Â[<ffffffff810848d4>] ? hrtimer_nanosleep+0xc4/0x180 > Â[<ffffffff810836f0>] ? hrtimer_wakeup+0x0/0x30 > Â[<ffffffff81084704>] ? hrtimer_start_range_ns+0x14/0x20 > Â[<ffffffff814a4b15>] page_fault+0x25/0x30 > INFO: task rpcbind:3335 blocked for more than 120 seconds. > "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. > rpcbind     D ffff880c5dfc99a0   0 Â3335   Â1 0x00000080 > Âffff880c7f4ab738 0000000000000082 000000000000001e 0000004000000000 > Â0000000000014d40 ffff880c5dfc9440 ffff880c5dfc99a0 ffff880c7f4abfd8 > Âffff880c5dfc99a8 0000000000014d40 ffff880c7f4aa010 0000000000014d40 > Call Trace: > Â[<ffffffff810f5260>] ? sync_page+0x0/0x50 > Â[<ffffffff814a2330>] io_schedule+0x70/0xc0 > Â[<ffffffff810f52a0>] sync_page+0x40/0x50 > Â[<ffffffff814a2bef>] __wait_on_bit+0x5f/0x90 > Â[<ffffffff8112824f>] ? read_swap_cache_async+0x4f/0x140 > Â[<ffffffff810f5463>] wait_on_page_bit+0x73/0x80 > Â[<ffffffff8107fd90>] ? wake_bit_function+0x0/0x50 > Â[<ffffffff810f550a>] __lock_page_or_retry+0x3a/0x60 > Â[<ffffffff8111a673>] handle_pte_fault+0xac3/0xb20 > Â[<ffffffff8111a826>] handle_mm_fault+0x156/0x250 > Â[<ffffffff814a7da3>] do_page_fault+0x143/0x4b0 > Â[<ffffffff81083af1>] ? lock_hrtimer_base+0x31/0x60 > Â[<ffffffff8108474d>] ? hrtimer_try_to_cancel+0x3d/0xd0 > Â[<ffffffff81084802>] ? hrtimer_cancel+0x22/0x30 > Â[<ffffffff814a4b15>] page_fault+0x25/0x30 > Â[<ffffffff8115be47>] ? do_sys_poll+0x357/0x540 > Â[<ffffffff8115be2f>] ? do_sys_poll+0x33f/0x540 > Â[<ffffffff8115b020>] ? __pollwait+0x0/0xf0 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8111a826>] ? handle_mm_fault+0x156/0x250 > Â[<ffffffff814a7e38>] ? do_page_fault+0x1d8/0x4b0 > Â[<ffffffff8116562f>] ? vfsmount_lock_global_unlock_online+0x4f/0x60 > Â[<ffffffff8116604c>] ? mntput_no_expire+0x19c/0x1c0 > Â[<ffffffff81013219>] ? read_tsc+0x9/0x20 > Â[<ffffffff810899f3>] ? ktime_get_ts+0xb3/0xf0 > Â[<ffffffff8115aecd>] ? poll_select_set_timeout+0x8d/0xa0 > Â[<ffffffff8115c22c>] sys_poll+0x7c/0x110 > Â[<ffffffff8100bf82>] system_call_fastpath+0x16/0x1b > INFO: task hald:6503 blocked for more than 120 seconds. > "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. > hald      ÂD ffff88105d8f3a20   0 Â6503   Â1 0x00000080 > Âffff88105d30d738 0000000000000086 000000005d30d698 0000004000000000 > Â0000000000014d40 ffff88105d8f34c0 ffff88105d8f3a20 ffff88105d30dfd8 > Âffff88105d8f3a28 0000000000014d40 ffff88105d30c010 0000000000014d40 > Call Trace: > Â[<ffffffff810f5260>] ? sync_page+0x0/0x50 > Â[<ffffffff814a2330>] io_schedule+0x70/0xc0 > Â[<ffffffff810f52a0>] sync_page+0x40/0x50 > Â[<ffffffff814a2bef>] __wait_on_bit+0x5f/0x90 > Â[<ffffffff8112824f>] ? read_swap_cache_async+0x4f/0x140 > Â[<ffffffff810f5463>] wait_on_page_bit+0x73/0x80 > Â[<ffffffff8107fd90>] ? wake_bit_function+0x0/0x50 > Â[<ffffffff810f550a>] __lock_page_or_retry+0x3a/0x60 > Â[<ffffffff8111a673>] handle_pte_fault+0xac3/0xb20 > Â[<ffffffff8111a826>] handle_mm_fault+0x156/0x250 > Â[<ffffffff814a7da3>] do_page_fault+0x143/0x4b0 > Â[<ffffffff81083af1>] ? lock_hrtimer_base+0x31/0x60 > Â[<ffffffff8108474d>] ? hrtimer_try_to_cancel+0x3d/0xd0 > Â[<ffffffff81084802>] ? hrtimer_cancel+0x22/0x30 > Â[<ffffffff814a4b15>] page_fault+0x25/0x30 > Â[<ffffffff8115be47>] ? do_sys_poll+0x357/0x540 > Â[<ffffffff8115be2f>] ? do_sys_poll+0x33f/0x540 > Â[<ffffffff8115b020>] ? __pollwait+0x0/0xf0 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff81013219>] ? read_tsc+0x9/0x20 > Â[<ffffffff810899f3>] ? ktime_get_ts+0xb3/0xf0 > Â[<ffffffff8115aecd>] ? poll_select_set_timeout+0x8d/0xa0 > Â[<ffffffff8115c22c>] sys_poll+0x7c/0x110 > Â[<ffffffff8100bf82>] system_call_fastpath+0x16/0x1b > INFO: task ntpd:6604 blocked for more than 120 seconds. > "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. > ntpd      ÂD ffff8808563085e0   0 Â6604   Â1 0x00000080 > Âffff88085d9f7958 0000000000000082 000000005df1eb40 00000040ffffffff > Â0000000000014d40 ffff880856308080 ffff8808563085e0 ffff88085d9f7fd8 > Âffff8808563085e8 0000000000014d40 ffff88085d9f6010 0000000000014d40 > Call Trace: > Â[<ffffffff810f5260>] ? sync_page+0x0/0x50 > Â[<ffffffff814a2330>] io_schedule+0x70/0xc0 > Â[<ffffffff810f52a0>] sync_page+0x40/0x50 > Â[<ffffffff814a2bef>] __wait_on_bit+0x5f/0x90 > Â[<ffffffff8112824f>] ? read_swap_cache_async+0x4f/0x140 > Â[<ffffffff810f5463>] wait_on_page_bit+0x73/0x80 > Â[<ffffffff8107fd90>] ? wake_bit_function+0x0/0x50 > Â[<ffffffff810f550a>] __lock_page_or_retry+0x3a/0x60 > Â[<ffffffff8111a673>] handle_pte_fault+0xac3/0xb20 > Â[<ffffffff8111a826>] handle_mm_fault+0x156/0x250 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff814a7da3>] do_page_fault+0x143/0x4b0 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff814a4b15>] page_fault+0x25/0x30 > Â[<ffffffff8122405d>] ? copy_user_generic_string+0x2d/0x40 > Â[<ffffffff8115bad8>] ? set_fd_set+0x48/0x60 > Â[<ffffffff8115c515>] core_sys_select+0x1f5/0x2f0 > Â[<ffffffff81079025>] ? queue_work_on+0x25/0x30 > Â[<ffffffff8107906f>] ? queue_work+0x1f/0x30 > Â[<ffffffff8107986d>] ? queue_delayed_work+0x2d/0x40 > Â[<ffffffff8107989b>] ? schedule_delayed_work+0x1b/0x20 > Â[<ffffffff8108a51b>] ? do_adjtimex+0x1ab/0x670 > Â[<ffffffff81013219>] ? read_tsc+0x9/0x20 > Â[<ffffffff810899f3>] ? ktime_get_ts+0xb3/0xf0 > Â[<ffffffff8115c867>] sys_select+0x47/0x110 > Â[<ffffffff8100bf82>] system_call_fastpath+0x16/0x1b > INFO: task master:6680 blocked for more than 120 seconds. > "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. > master     ÂD ffff88085cef2660   0 Â6680   Â1 0x00000080 > Âffff88085d0abb08 0000000000000086 ffff88085d0abab8 ffffffffffffffff > Â0000000000014d40 ffff88085cef2100 ffff88085cef2660 ffff88085d0abfd8 > Âffff88085cef2668 0000000000014d40 ffff88085d0aa010 0000000000014d40 > Call Trace: > Â[<ffffffff810f5260>] ? sync_page+0x0/0x50 > Â[<ffffffff814a2330>] io_schedule+0x70/0xc0 > Â[<ffffffff810f52a0>] sync_page+0x40/0x50 > Â[<ffffffff814a2bef>] __wait_on_bit+0x5f/0x90 > Â[<ffffffff810f5463>] wait_on_page_bit+0x73/0x80 > Â[<ffffffff8107fd90>] ? wake_bit_function+0x0/0x50 > Â[<ffffffff810f550a>] __lock_page_or_retry+0x3a/0x60 > Â[<ffffffff810f6537>] filemap_fault+0x2d7/0x4c0 > Â[<ffffffff81119694>] __do_fault+0x54/0x570 > Â[<ffffffff81141357>] ? mem_cgroup_uncharge_swap+0x27/0x80 > Â[<ffffffff81119ca7>] handle_pte_fault+0xf7/0xb20 > Â[<ffffffff81115727>] ? __pte_alloc+0xd7/0xf0 > Â[<ffffffff8111a826>] handle_mm_fault+0x156/0x250 > Â[<ffffffff814a7da3>] do_page_fault+0x143/0x4b0 > Â[<ffffffff81185100>] ? sys_epoll_wait+0xa0/0x450 > Â[<ffffffff810587c0>] ? default_wake_function+0x0/0x20 > Â[<ffffffff814a4b15>] page_fault+0x25/0x30 > INFO: task abrtd:6689 blocked for more than 120 seconds. > "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. > abrtd      D ffff88045de2c660   0 Â6689   Â1 0x00000080 > Âffff88045e1f7738 0000000000000086 000000005e1f7698 0000004000000000 > Â0000000000014d40 ffff88045de2c100 ffff88045de2c660 ffff88045e1f7fd8 > Âffff88045de2c668 0000000000014d40 ffff88045e1f6010 0000000000014d40 > Call Trace: > Â[<ffffffff810f5260>] ? sync_page+0x0/0x50 > Â[<ffffffff814a2330>] io_schedule+0x70/0xc0 > Â[<ffffffff810f52a0>] sync_page+0x40/0x50 > Â[<ffffffff814a2bef>] __wait_on_bit+0x5f/0x90 > Â[<ffffffff8112824f>] ? read_swap_cache_async+0x4f/0x140 > Â[<ffffffff810f5463>] wait_on_page_bit+0x73/0x80 > Â[<ffffffff8107fd90>] ? wake_bit_function+0x0/0x50 > Â[<ffffffff810f550a>] __lock_page_or_retry+0x3a/0x60 > Â[<ffffffff8111a673>] handle_pte_fault+0xac3/0xb20 > Â[<ffffffff8111a826>] handle_mm_fault+0x156/0x250 > Â[<ffffffff814a7da3>] do_page_fault+0x143/0x4b0 > Â[<ffffffff81083af1>] ? lock_hrtimer_base+0x31/0x60 > Â[<ffffffff8108474d>] ? hrtimer_try_to_cancel+0x3d/0xd0 > Â[<ffffffff81084802>] ? hrtimer_cancel+0x22/0x30 > Â[<ffffffff814a4b15>] page_fault+0x25/0x30 > Â[<ffffffff8115be47>] ? do_sys_poll+0x357/0x540 > Â[<ffffffff8115be2f>] ? do_sys_poll+0x33f/0x540 > Â[<ffffffff8115b020>] ? __pollwait+0x0/0xf0 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff8115b110>] ? pollwake+0x0/0x60 > Â[<ffffffff811d68f1>] ? inode_has_perm+0x51/0xa0 > Â[<ffffffff8116608d>] ? mntput+0x1d/0x30 > Â[<ffffffff811539e2>] ? path_put+0x22/0x30 > Â[<ffffffff811543f7>] ? finish_open+0x117/0x1f0 > Â[<ffffffff81157566>] ? do_path_lookup+0x76/0x130 > Â[<ffffffff811d6e2b>] ? dentry_has_perm+0x5b/0x80 > Â[<ffffffff8114db58>] ? cp_new_stat+0xf8/0x110 > Â[<ffffffff81013219>] ? read_tsc+0x9/0x20 > Â[<ffffffff810899f3>] ? ktime_get_ts+0xb3/0xf0 > Â[<ffffffff8115aecd>] ? poll_select_set_timeout+0x8d/0xa0 > Â[<ffffffff8115c22c>] sys_poll+0x7c/0x110 > Â[<ffffffff8100bf82>] system_call_fastpath+0x16/0x1b > INFO: task qmgr:6691 blocked for more than 120 seconds. > "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. > qmgr      ÂD ffff88105f9eb9a0   0 Â6691  6680 0x00000084 > Âffff88105a74db08 0000000000000086 ffff88105a74dab8 ffffffff00000000 > Â0000000000014d40 ffff88105f9eb440 ffff88105f9eb9a0 ffff88105a74dfd8 > Âffff88105f9eb9a8 0000000000014d40 ffff88105a74c010 0000000000014d40 > Call Trace: > Â[<ffffffff810f5260>] ? sync_page+0x0/0x50 > Â[<ffffffff814a2330>] io_schedule+0x70/0xc0 > Â[<ffffffff810f52a0>] sync_page+0x40/0x50 > Â[<ffffffff814a2bef>] __wait_on_bit+0x5f/0x90 > Â[<ffffffff810f5463>] wait_on_page_bit+0x73/0x80 > Â[<ffffffff8107fd90>] ? wake_bit_function+0x0/0x50 > Â[<ffffffff810f550a>] __lock_page_or_retry+0x3a/0x60 > Â[<ffffffff810f6537>] filemap_fault+0x2d7/0x4c0 > Â[<ffffffff81119694>] __do_fault+0x54/0x570 > Â[<ffffffff81141357>] ? mem_cgroup_uncharge_swap+0x27/0x80 > Â[<ffffffff81119ca7>] handle_pte_fault+0xf7/0xb20 > Â[<ffffffff81115727>] ? __pte_alloc+0xd7/0xf0 > Â[<ffffffff8111a826>] handle_mm_fault+0x156/0x250 > Â[<ffffffff814a7da3>] do_page_fault+0x143/0x4b0 > Â[<ffffffff81185100>] ? sys_epoll_wait+0xa0/0x450 > Â[<ffffffff810587c0>] ? default_wake_function+0x0/0x20 > Â[<ffffffff814a4b15>] page_fault+0x25/0x30 > INFO: task oom02:7543 blocked for more than 120 seconds. > "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. > oom02      D ffff88085db07ae0   0 Â7543  7388 0x00000080 > Âffff880855de1518 0000000000000082 0000000000011210 ffff880600000000 > Â0000000000014d40 ffff88085db07580 ffff88085db07ae0 ffff880855de1fd8 > Âffff88085db07ae8 0000000000014d40 ffff880855de0010 0000000000014d40 > Call Trace: > Â[<ffffffff814a2330>] io_schedule+0x70/0xc0 > Â[<ffffffff8120463a>] get_request_wait+0xca/0x1a0 > Â[<ffffffff8107fd50>] ? autoremove_wake_function+0x0/0x40 > Â[<ffffffff811fd197>] ? elv_merge+0x1d7/0x210 > Â[<ffffffff8120477b>] __make_request+0x6b/0x4d0 > Â[<ffffffff8120259f>] generic_make_request+0x21f/0x5b0 > Â[<ffffffff810f7885>] ? mempool_alloc_slab+0x15/0x20 > Â[<ffffffff810f7a23>] ? mempool_alloc+0x63/0x150 > Â[<ffffffff812029b6>] submit_bio+0x86/0x110 > Â[<ffffffff810ffb96>] ? test_set_page_writeback+0x106/0x190 > Â[<ffffffff81127f63>] swap_writepage+0x83/0xd0 > Â[<ffffffff811048fe>] pageout+0x12e/0x310 > Â[<ffffffff81104efa>] shrink_page_list+0x41a/0x5a0 > Â[<ffffffff81105686>] shrink_inactive_list+0x166/0x480 > Â[<ffffffff81106023>] shrink_zone+0x363/0x4d0 > Â[<ffffffff8110470e>] ? shrink_slab+0x14e/0x180 > Â[<ffffffff8110675f>] do_try_to_free_pages+0xaf/0x4a0 > Â[<ffffffff81106dc2>] try_to_free_pages+0x92/0x130 > Â[<ffffffff810fbc21>] ? get_page_from_freelist+0x3c1/0x810 > Â[<ffffffff810fc46b>] __alloc_pages_nodemask+0x3fb/0x760 > Â[<ffffffff81135313>] alloc_pages_vma+0x93/0x150 > Â[<ffffffff8111a2e2>] handle_pte_fault+0x732/0xb20 > Â[<ffffffff81115727>] ? __pte_alloc+0xd7/0xf0 > Â[<ffffffff8111a826>] handle_mm_fault+0x156/0x250 > Â[<ffffffff814a7da3>] do_page_fault+0x143/0x4b0 > Â[<ffffffff814a1cfe>] ? schedule+0x44e/0xa10 > Â[<ffffffff814a4b15>] page_fault+0x25/0x30 -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxxx For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom policy in Canada: sign http://dissolvethecrtc.ca/ Don't email: <a href