Skip to contentSkip to conversation

kernel BUG at kernel/smpboot.c:125!

DISCUSSION Dave Jones 2 participants 2 messages Jan 18, 2013 – Jan 19, 2013

lore
Share & tools
lore

Dave Jones

author
#0 lore raw

Since 3.7, we've started seeing reports of this BUG_ON in kernel/smpboot.c being hit..

121 if (kthread_should_park()) {

122                         __set_current_state(TASK_RUNNING);
123                         preempt_enable();
124                         if (ht->park && td->status == HP_THREAD_ACTIVE) {
125                                 BUG_ON(td->cpu != smp_processor_id());
126                                 ht->park(td->cpu);
127                                 td->status = HP_THREAD_PARKED;
128                         }
129                         kthread_parkme();
130                         /* We might have been woken for stop */
131                         continue;

132 }

In all reports, it seems that it happens when coming back from suspend/hibernate.

Example trace below. More reports at https://bugzilla.redhat.com/show_bug.cgi?id=891088

Quoted 49 lines

kernel BUG at kernel/smpboot.c:125!
invalid opcode: 0000 [#1] SMP

Modules linked in: cpufreq_stats des_generic md4 nls_utf8 cifs dns_resolver

fscache fuse ipt_MASQUERADE nf_conntrack_netbios_ns nf_conntrack_broadcast ip6table_mangle ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 iptable_nat nf_nat_ipv4 nf_nat iptable_mangle nf_conntrack_ipv4 nf_defrag_ipv4 xt_conntrack nf_conntrack ebtable_filter ebtables ip6table_filter ip6_tables bnep snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_intel arc4 snd_hda_codec snd_hwdep snd_seq uvcvideo snd_seq_device iwldvm snd_pcm videobuf2_vmalloc iTCO_wdt videobuf2_memops iTCO_vendor_support videobuf2_core mac80211 videodev asus_nb_wmi asus_wmi media vhost_net sparse_keymap coretemp snd_page_alloc btusb tun iwlwifi macvtap macvlan bluetooth microcode snd_timer kvm_intel snd joydev serio_raw lpc_ich i2c_i801 cfg80211 soundcore kvm mfd_core mei rfkill uinput dm_crypt nouveau crc32c_intel i915 mxm_wmi i2c_algo_bit drm_kms_helper ghash_clmulni_intel ttm drm i2c_core wmi video
CPU 0
Pid: 22, comm: watchdog/3 Not tainted 3.7.1-1.fc19.x86_64 #1 ASUSTeK COMPUTER INC. UX32VD/UX32VD

RIP: 0010:[<ffffffff8108aae3>]  [<ffffffff8108aae3>]
smpboot_thread_fn+0x193/0x1a0
RSP: 0000:ffff88029fc71e78  EFLAGS: 00010206
RAX: ffffffff810eb4b0 RBX: ffff88029fc35c80 RCX: 0000000000000000
RDX: 0000000000000000 RSI: ffff88029fc35c80 RDI: 0000000000000003
RBP: ffff88029fc71eb8 R08: ffff88029fc70000 R09: 0000000000000001
R10: 0000000000000000 R11: 0000000000000001 R12: ffff8802a4400940
R13: ffffffff81c3b9a0 R14: ffff88029fc35c80 R15: 0000000000000000
FS:  0000000000000000(0000) GS:ffff8802aee00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f825e2b6000 CR3: 000000027a915000 CR4: 00000000001407f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process watchdog/3 (pid: 22, threadinfo ffff88029fc70000, task
ffff88029fc35c80)
Stack:
 ffff88029fc71e88 0000000000000000 ffff88029fc71eb8 0000000000000001
 ffff8802a00e5d18 ffff8802a4400940 ffffffff8108a950 0000000000000000
 ffff88029fc71f48 ffffffff81081d50 0000000000000001 0000000000000005
Call Trace:
 [<ffffffff8108a950>] ? lg_global_unlock+0x60/0x60
 [<ffffffff81081d50>] kthread+0xc0/0xd0
 [<ffffffff81010808>] ? perf_trace_xen_mmu_flush_tlb_all+0x88/0xc0
 [<ffffffff81081c90>] ? kthread_create_on_node+0x120/0x120
 [<ffffffff8163d92c>] ret_from_fork+0x7c/0xb0
 [<ffffffff81081c90>] ? kthread_create_on_node+0x120/0x120
Code: bf 77 00 0f a3 3a 19 d2 31 f6 85 d2 40 0f 95 c6 ff d0 4c 89 e7 e8 ce 4f

0f 00 48 83 c4 20 31 c0 5b 41 5c 41 5d 41 5e 5d c3 0f 0b <0f> 0b 66 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 55 48 c7
RIP [<ffffffff8108aae3>] smpboot_thread_fn+0x193/0x1a0
RSP <ffff88029fc71e78>

#1 lore raw
Quoted 7 lines · Dave Jones

Since 3.7, we've started seeing reports of this BUG_ON in kernel/smpboot.c being hit..

121 if (kthread_should_park()) {

122                         __set_current_state(TASK_RUNNING);
123                         preempt_enable();
124                         if (ht->park && td->status == HP_THREAD_ACTIVE) {

125 BUG_ON(td->cpu != smp_processor_id());

Unsafe to get the current cpu if kthread is not bound to it, would you please take a try for the following diff?

Hillf
---

kernel/kthread.c +1 -0
@@ -306,6 +306,7 @@ struct task_struct *kthread_create_on_cp
 		return p;
 	set_bit(KTHREAD_IS_PER_CPU, &to_kthread(p)->flags);
 	to_kthread(p)->cpu = cpu;
+	__kthread_bind(p, cpu);
 	/* Park the thread to get it out of TASK_UNINTERRUPTIBLE state */
 	kthread_park(p);
 	return p;

--

Quoted 62 lines

126 ht->park(td->cpu);
127 td->status = HP_THREAD_PARKED;

128                         }
129                         kthread_parkme();
130                         /* We might have been woken for stop */
131                         continue;

132 }

In all reports, it seems that it happens when coming back from suspend/hibernate.

Example trace below. More reports at https://bugzilla.redhat.com/show_bug.cgi?id=891088

Quoted 49 lines

kernel BUG at kernel/smpboot.c:125!
invalid opcode: 0000 [#1] SMP

Modules linked in: cpufreq_stats des_generic md4 nls_utf8 cifs dns_resolver

fscache fuse ipt_MASQUERADE nf_conntrack_netbios_ns nf_conntrack_broadcast ip6table_mangle ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 iptable_nat nf_nat_ipv4 nf_nat iptable_mangle nf_conntrack_ipv4 nf_defrag_ipv4 xt_conntrack nf_conntrack ebtable_filter ebtables ip6table_filter ip6_tables bnep snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_intel arc4 snd_hda_codec snd_hwdep snd_seq uvcvideo snd_seq_device iwldvm snd_pcm videobuf2_vmalloc iTCO_wdt videobuf2_memops iTCO_vendor_support videobuf2_core mac80211 videodev asus_nb_wmi asus_wmi media vhost_net sparse_keymap coretemp snd_page_alloc btusb tun iwlwifi macvtap macvlan bluetooth microcode snd_timer kvm_intel snd joydev serio_raw lpc_ich i2c_i801 cfg80211 soundcore kvm mfd_core mei rfkill uinput dm_crypt nouveau crc32c_intel i915 mxm_wmi i2c_algo_bit drm_kms_helper ghash_clmulni_intel ttm drm i2c_core wmi video
CPU 0
Pid: 22, comm: watchdog/3 Not tainted 3.7.1-1.fc19.x86_64 #1 ASUSTeK COMPUTER INC. UX32VD/UX32VD

RIP: 0010:[<ffffffff8108aae3>]  [<ffffffff8108aae3>]
smpboot_thread_fn+0x193/0x1a0
RSP: 0000:ffff88029fc71e78  EFLAGS: 00010206
RAX: ffffffff810eb4b0 RBX: ffff88029fc35c80 RCX: 0000000000000000
RDX: 0000000000000000 RSI: ffff88029fc35c80 RDI: 0000000000000003
RBP: ffff88029fc71eb8 R08: ffff88029fc70000 R09: 0000000000000001
R10: 0000000000000000 R11: 0000000000000001 R12: ffff8802a4400940
R13: ffffffff81c3b9a0 R14: ffff88029fc35c80 R15: 0000000000000000
FS:  0000000000000000(0000) GS:ffff8802aee00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f825e2b6000 CR3: 000000027a915000 CR4: 00000000001407f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process watchdog/3 (pid: 22, threadinfo ffff88029fc70000, task
ffff88029fc35c80)
Stack:
 ffff88029fc71e88 0000000000000000 ffff88029fc71eb8 0000000000000001
 ffff8802a00e5d18 ffff8802a4400940 ffffffff8108a950 0000000000000000
 ffff88029fc71f48 ffffffff81081d50 0000000000000001 0000000000000005
Call Trace:
 [<ffffffff8108a950>] ? lg_global_unlock+0x60/0x60
 [<ffffffff81081d50>] kthread+0xc0/0xd0
 [<ffffffff81010808>] ? perf_trace_xen_mmu_flush_tlb_all+0x88/0xc0
 [<ffffffff81081c90>] ? kthread_create_on_node+0x120/0x120
 [<ffffffff8163d92c>] ret_from_fork+0x7c/0xb0
 [<ffffffff81081c90>] ? kthread_create_on_node+0x120/0x120
Code: bf 77 00 0f a3 3a 19 d2 31 f6 85 d2 40 0f 95 c6 ff d0 4c 89 e7 e8 ce 4f

0f 00 48 83 c4 20 31 c0 5b 41 5c 41 5d 41 5e 5d c3 0f 0b <0f> 0b 66 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 55 48 c7
RIP [<ffffffff8108aae3>] smpboot_thread_fn+0x193/0x1a0
RSP <ffff88029fc71e78>