[PATCH v2] notifier: Fix soft lockup for notifier_call_chain().

[Ding Tianhong <dingtianhong@huawei.com>]

The problem was occurs in my system that a lot of drviers register
its own handler to the notifiler call chain for netdev_chain, and
then create 4095 vlan dev for one nic, and add several ipv6 address
on each one of them, just like this:

for i in `seq 1 4095`; do ip link add link eth0 name eth0.$i type vlan id $i; done
for i in `seq 1 4095`; do ip -6 addr add 2001::$i dev eth0.$i; done
for i in `seq 1 4095`; do ip -6 addr add 2002::$i dev eth0.$i; done
for i in `seq 1 4095`; do ip -6 addr add 2003::$i dev eth0.$i; done

ifconfig eth0 up
ifconfig eth0 down

then it will halt several seconds, and occurs softlockup:

<0>[ 7620.364058]NMI watchdog: BUG: soft lockup - CPU#0 stuck for 23s! [ifconfig:19186]
<0>[ 7620.364592]Call trace:
<4>[ 7620.364599][<ffffffc000208f68>] dump_backtrace+0x0/0x220
<4>[ 7620.364603][<ffffffc0002091a8>] show_stack+0x20/0x28
<4>[ 7620.364607][<ffffffc000691fac>] dump_stack+0x90/0xb0
<4>[ 7620.364612][<ffffffc0002cacbc>] watchdog_timer_fn+0x41c/0x460
<4>[ 7620.364617][<ffffffc000289ec8>] __run_hrtimer+0x98/0x2d8
<4>[ 7620.364620][<ffffffc00028a3e0>] hrtimer_interrupt+0x110/0x288
<4>[ 7620.364624][<ffffffc00059c0c8>] arch_timer_handler_phys+0x38/0x48
<4>[ 7620.364628][<ffffffc000276d2c>] handle_percpu_devid_irq+0x9c/0x190
<4>[ 7620.364632][<ffffffc000271ef8>] generic_handle_irq+0x40/0x58
<4>[ 7620.364635][<ffffffc000272270>] __handle_domain_irq+0x68/0xc0
<4>[ 7620.364638][<ffffffc000200634>] gic_handle_irq+0xc4/0x1c8
<4>[ 7620.364641]Exception stack(0xffffffc0309b3640 to 0xffffffc0309b3770)
<4>[ 7620.364644]3640: 0000000000001000 0000000000000000 ffffffc0309b37c0 ffffffbfa1019cf8
<4>[ 7620.364647]3660: 0000000080000145 ffffffc0309b3958 0000000000000000 ffffffbfa1013008
<4>[ 7620.364651]3680: 00000000000007f0 ffffffbfa131b770 ffffffd08aaadc40 ffffffbfa1019cf8
<4>[ 7620.364654]36a0: ffffffbfa1019cc4 ffffffd089c2b000 ffffffd08eff8000 ffffffc0309b3958
<4>[ 7620.364656]36c0: ffffffbfa101c5c0 0000000000000000 0000000000000000 ffffffbfa101c66c
<4>[ 7620.364659]36e0: 7f7f7f7f7f7f7f7f 0000000000000030 ffffffffffffffff ffff000000000000
<4>[ 7620.364662]3700: 0000000000000000 0000000000000000 ffffffc000393d58 0000007f794d67b0
<4>[ 7620.364665]3720: 0000007fe62215d0 ffffffc0309b3830 ffffffc00021d8e0 ffffffbfa1049b68
<4>[ 7620.364668]3740: ffffffc000697578 ffffffc0006974b8 ffffffc0309b3958 0000000000000000
<4>[ 7620.364670]3760: ffffffbfa1013008 00000000000007f0
<4>[ 7620.364673][<ffffffc000203780>] el1_irq+0x80/0x100
<4>[ 7620.364692][<ffffffbfa1019ed4>] fib6_walk+0x3c/0x70 [ipv6]
<4>[ 7620.364710][<ffffffbfa1019f70>] fib6_clean_tree+0x68/0x90 [ipv6]
<4>[ 7620.364727][<ffffffbfa101a020>] __fib6_clean_all+0x88/0xc0 [ipv6]
<4>[ 7620.364746][<ffffffbfa101c760>] fib6_clean_all+0x28/0x30 [ipv6]
<4>[ 7620.364763][<ffffffbfa101933c>] rt6_ifdown+0x64/0x148 [ipv6]
<4>[ 7620.364781][<ffffffbfa100e6d8>] addrconf_ifdown+0x68/0x540 [ipv6]
<4>[ 7620.364798][<ffffffbfa1010f58>] addrconf_notify+0xd0/0x8b8 [ipv6]
<4>[ 7620.364801][<ffffffc00023f83c>] notifier_call_chain+0x5c/0xa0
<4>[ 7620.364804][<ffffffc00023f9e0>] raw_notifier_call_chain+0x20/0x28
<4>[ 7620.364809][<ffffffc0005cbab4>] call_netdevice_notifiers_info+0x4c/0x80
<4>[ 7620.364812][<ffffffc0005cbfc8>] dev_close_many+0xd0/0x138
<4>[ 7620.364821][<ffffffbfa33be6e8>] vlan_device_event+0x4a8/0x6a0 [8021q]
<4>[ 7620.364824][<ffffffc00023f83c>] notifier_call_chain+0x5c/0xa0
<4>[ 7620.364827][<ffffffc00023f9e0>] raw_notifier_call_chain+0x20/0x28
<4>[ 7620.364830][<ffffffc0005cbab4>] call_netdevice_notifiers_info+0x4c/0x80
<4>[ 7620.364833][<ffffffc0005d5148>] __dev_notify_flags+0xb8/0xe0
<4>[ 7620.364836][<ffffffc0005d5994>] dev_change_flags+0x54/0x68
<4>[ 7620.364840][<ffffffc00064a620>] devinet_ioctl+0x650/0x700
<4>[ 7620.364843][<ffffffc00064bea4>] inet_ioctl+0xa4/0xc8
<4>[ 7620.364847][<ffffffc0005b1094>] sock_do_ioctl+0x44/0x88
<4>[ 7620.364850][<ffffffc0005b1a3c>] sock_ioctl+0x23c/0x308
<4>[ 7620.364854][<ffffffc000393bc4>] do_vfs_ioctl+0x48c/0x620
<4>[ 7620.364857][<ffffffc000393dec>] SyS_ioctl+0x94/0xa8

=================================cut here========================================

It looks that the notifier_call_chain has to deal with too much handler, and will not
feed the watchdog until finish the work, and the notifier_call_chain may be called
in atomic context, so add touch_nmi_watchdog() in the loops to fix this problem,
and it will not panic again.

v2: add cond_resched() will break the atomic context, so feed the watchdog in
    the loops to fix this bug.

Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
---
 kernel/notifier.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/kernel/notifier.c b/kernel/notifier.c
index fd2c9ac..7eca3c1 100644
--- a/kernel/notifier.c
+++ b/kernel/notifier.c
@@ -5,6 +5,7 @@
 #include <linux/rcupdate.h>
 #include <linux/vmalloc.h>
 #include <linux/reboot.h>
+#include <linux/nmi.h>
 
 /*
  *	Notifier list for kernel code which wants to be called
@@ -92,6 +93,8 @@ static int notifier_call_chain(struct notifier_block **nl,
 #endif
 		ret = nb->notifier_call(nb, val, v);
 
+		touch_nmi_watchdog();
+
 		if (nr_calls)
 			(*nr_calls)++;
 
-- 
1.9.0