I recently upgraded my home server from Ubuntu 10.04 to 12.04.1. It runs the linux-image-server kernel, x86_64 arch.

There's nothing particularly unusual running, I think -- a deluge daemon, apache2, iptables firewall with IP masquerading, DHCP server, bind DNS server which has zone files updated automatically with the hostnames DHCP clients identify themselves with, sshd, nfs server, a handful of other things. This machine is my router -- it sits between the internet and the local network.

Since the upgrade it has been failing intermittently. It'll be fine for a while after booting and then suddenly we'll lose our network connections on wifi. If I plug in a network cable I can't get an IP address from the DHCP server. If I set myself a static IP address I can continue to access the internet just fine. This makes it look like it's the DHCP server that fails (indeed, I run dhclient -v eth0 and nothing responds to the dhcpdiscover shoutouts), noticed when the clients try to renew their IP leases. But wired with a static IP I can still get to the internet, so iptables is still chugging along just fine.

So I try to log in to the machine over SSH, but it seems to hang. If I make ssh verbose I see that it does establish a connection to the server, then fails a little further down the line -- hard to see where exactly.

I note that if I try to grab a web page from its HTTP server I get the page I request but any extra requests made (for images, stylesheets, javascripts) are not served. I can however get these files if I request them directly, for instance from curl.

Is this suggesting that things are going downhill whenever something attempts to fork?

I dragged a monitor and keyboard to the server (it's usually headless) and had a look -- I see stack traces.

I switch to a fresh virtual terminal and attempt to log in. I get a stack trace (general protection fault) after entering my password. Here it is:

Jan  6 20:19:54 localhost kernel: [ 1475.178245] general protection fault: 0000 [#12] SMP 

Jan  6 20:19:54 localhost kernel: [ 1475.178292] CPU 1 

Jan  6 20:19:54 localhost kernel: [ 1475.178309] Modules linked in: btrfs zlib_deflate libcrc32c ufs qnx4 hfsplus hfs minix ntfs vfat msdos fat jfs xfs reiserfs ext2 nfsd nfs lockd fscache auth_rpcgss nfs_acl sunrpc dm_crypt ppdev ipt_REJECT ipt_LOG ipt_MASQUERADE xt_state iptable_mangle iptable_nat nf_nat nf_conntrack_ipv4 nf_conntrack nf_defrag_ipv4 iptable_filter ip_tables x_tables joydev sp5100_tco edac_core i2c_piix4 serio_raw k8temp edac_mce_amd snd_hda_codec_hdmi snd_hda_intel snd_hda_codec snd_hwdep snd_pcm snd_timer snd soundcore parport_pc snd_page_alloc mac_hid shpchp lp parport radeon 8139too ttm drm_kms_helper drm pata_atiixp i2c_algo_bit usbhid hid wmi r8169

Jan  6 20:19:54 localhost kernel: [ 1475.178911] 

Jan  6 20:19:54 localhost kernel: [ 1475.178927] Pid: 1305, comm: login Tainted: G    B D      3.2.0-35-generic #55-Ubuntu Gigabyte Technology Co., Ltd. GA-MA785GM-US2H/GA-MA785GM-US2H

Jan  6 20:19:54 localhost kernel: [ 1475.179028] RIP: 0010:[<ffffffff8116589a>]  [<ffffffff8116589a>] kmem_cache_alloc+0x5a/0x140

Jan  6 20:19:54 localhost kernel: [ 1475.179096] RSP: 0018:ffff88006b251d78  EFLAGS: 00010206

Jan  6 20:19:54 localhost kernel: [ 1475.179135] RAX: 0000000000000000 RBX: 00007f062bb91000 RCX: 000000000005b2ed

Jan  6 20:19:54 localhost kernel: [ 1475.179186] RDX: 000000000005b2ec RSI: 0000000000016da0 RDI: ffff88006d408a00

Jan  6 20:19:54 localhost kernel: [ 1475.179236] RBP: ffff88006b251dc8 R08: ffff88006fa96da0 R09: 0000000000000001

Jan  6 20:19:54 localhost kernel: [ 1475.179287] R10: 00000000000000d1 R11: ffff88006b23a8f0 R12: ffff88006d408a00

Jan  6 20:19:54 localhost kernel: [ 1475.179336] R13: 2665c4979a04b7b8 R14: ffffffff811447c5 R15: 00000000000080d0

Jan  6 20:19:54 localhost kernel: [ 1475.179387] FS:  00007f062bb81700(0000) GS:ffff88006fa80000(0000) knlGS:0000000000000000

Jan  6 20:19:54 localhost kernel: [ 1475.179445] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033

Jan  6 20:19:54 localhost kernel: [ 1475.179486] CR2: 00007f9b4d79da00 CR3: 0000000059a34000 CR4: 00000000000006e0

Jan  6 20:19:54 localhost kernel: [ 1475.179536] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000

Jan  6 20:19:54 localhost kernel: [ 1475.179586] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400

Jan  6 20:19:54 localhost kernel: [ 1475.179637] Process login (pid: 1305, threadinfo ffff88006b250000, task ffff880036058000)

Jan  6 20:19:54 localhost kernel: [ 1475.179695] Stack:

Jan  6 20:19:54 localhost kernel: [ 1475.179711]  ffff880036058000 0000000000000041 0000000000000001 ffffffff81188cec

Jan  6 20:19:54 localhost kernel: [ 1475.179777]  0000000000000282 00007f062bb91000 ffff88006822ce00 0000000000000001

Jan  6 20:19:54 localhost kernel: [ 1475.179841]  0000000000001000 0000000000000000 ffff88006b251e88 ffffffff811447c5

Jan  6 20:19:54 localhost kernel: [ 1475.179905] Call Trace:

Jan  6 20:19:54 localhost kernel: [ 1475.179928]  [<ffffffff81188cec>] ? path_openat+0xfc/0x3f0

Jan  6 20:19:54 localhost kernel: [ 1475.179971]  [<ffffffff811447c5>] mmap_region+0x2a5/0x4f0

Jan  6 20:19:54 localhost kernel: [ 1475.180012]  [<ffffffff81144d58>] do_mmap_pgoff+0x348/0x360

Jan  6 20:19:54 localhost kernel: [ 1475.180054]  [<ffffffff81144e36>] sys_mmap_pgoff+0xc6/0x230

Jan  6 20:19:54 localhost kernel: [ 1475.180098]  [<ffffffff81018b12>] sys_mmap+0x22/0x30

Jan  6 20:19:54 localhost kernel: [ 1475.180136]  [<ffffffff816655c2>] system_call_fastpath+0x16/0x1b

Jan  6 20:19:54 localhost kernel: [ 1475.180180] Code: 00 4d 8b 04 24 65 4c 03 04 25 50 da 00 00 49 8b 50 08 4d 8b 28 4d 85 ed 0f 84 d8 00 00 00 49 63 44 24 20 49 8b 34 24 48 8d 4a 01 <49> 8b 5c 05 00 4c 89 e8 65 48 0f c7 0e 0f 94 c0 84 c0 74 c2 4d 

Jan  6 20:19:54 localhost kernel: [ 1475.180503] RIP  [<ffffffff8116589a>] kmem_cache_alloc+0x5a/0x140

Jan  6 20:19:54 localhost kernel: [ 1475.180552]  RSP <ffff88006b251d78>

Jan  6 20:19:54 localhost kernel: [ 1475.180603] ---[ end trace 766ef1ef52f774b9 ]---

If I watch long enough, I see more general protection faults. I've seen them for login, apache2, deluge-web, head, powerbtn.sh so far.

I have to hard reset the machine to get it back to a working state (I even get a general protection fault for powerbtn.sh when I hit the power button), but it's not long before it gets like this again.

I haven't yet figured out how to reproduce this on demand -- it seems to happen randomly.

In case it's useful, I looked through kern.log and found the first such fault. There's a ton of them all in a row starting with zsh, then deluged, apache2, cron, head, console-kit-dae, irqbalance, nmbd... Here's the zsh one and the bad page state error which comes right after:

Jan  6 20:13:35 localhost kernel: [ 1096.184250] general protection fault: 0000 [#1] SMP 

Jan  6 20:13:35 localhost kernel: [ 1096.186339] CPU 1 

Jan  6 20:13:35 localhost kernel: [ 1096.186355] Modules linked in: btrfs zlib_deflate libcrc32c ufs qnx4 hfsplus hfs minix ntfs vfat msdos fat jfs xfs reiserfs ext2 nfsd nfs lockd fscache auth_rpcgss nfs_acl sunrpc dm_crypt ppdev ipt_REJECT ipt_LOG ipt_MASQUERADE xt_state iptable_mangle iptable_nat nf_nat nf_conntrack_ipv4 nf_conntrack nf_defrag_ipv4 iptable_filter ip_tables x_tables joydev sp5100_tco edac_core i2c_piix4 serio_raw k8temp edac_mce_amd snd_hda_codec_hdmi snd_hda_intel snd_hda_codec snd_hwdep snd_pcm snd_timer snd soundcore parport_pc snd_page_alloc mac_hid shpchp lp parport radeon 8139too ttm drm_kms_helper drm pata_atiixp i2c_algo_bit usbhid hid wmi r8169

Jan  6 20:13:35 localhost kernel: [ 1096.188008] 

Jan  6 20:13:35 localhost kernel: [ 1096.188008] Pid: 2564, comm: zsh Not tainted 3.2.0-35-generic #55-Ubuntu Gigabyte Technology Co., Ltd. GA-MA785GM-US2H/GA-MA785GM-US2H

Jan  6 20:13:35 localhost kernel: [ 1096.188008] RIP: 0010:[<ffffffff8116589a>]  [<ffffffff8116589a>] kmem_cache_alloc+0x5a/0x140

Jan  6 20:13:35 localhost kernel: [ 1096.188008] RSP: 0018:ffff880059877d78  EFLAGS: 00010206

Jan  6 20:13:35 localhost kernel: [ 1096.188008] RAX: 0000000000000000 RBX: 00007f202c59d000 RCX: 000000000005b2ed

Jan  6 20:13:35 localhost kernel: [ 1096.188008] RDX: 000000000005b2ec RSI: 0000000000016da0 RDI: ffff88006d408a00

Jan  6 20:13:35 localhost kernel: [ 1096.188008] RBP: ffff880059877dc8 R08: ffff88006fa96da0 R09: 0000000000000001

Jan  6 20:13:35 localhost kernel: [ 1096.188008] R10: 0000000000100073 R11: ffff880059dbb2c0 R12: ffff88006d408a00

Jan  6 20:13:35 localhost kernel: [ 1096.188008] R13: 2665c4979a04b7b8 R14: ffffffff811447c5 R15: 00000000000080d0

Jan  6 20:13:35 localhost kernel: [ 1096.188008] FS:  00007f202c5ac700(0000) GS:ffff88006fa80000(0000) knlGS:0000000000000000

Jan  6 20:13:35 localhost kernel: [ 1096.188008] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033

Jan  6 20:13:35 localhost kernel: [ 1096.188008] CR2: 00000000025991f0 CR3: 0000000059dbc000 CR4: 00000000000006e0

Jan  6 20:13:35 localhost kernel: [ 1096.188008] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000

Jan  6 20:13:35 localhost kernel: [ 1096.188008] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400

Jan  6 20:13:35 localhost kernel: [ 1096.188008] Process zsh (pid: 2564, threadinfo ffff880059876000, task ffff88006b6b5c00)

Jan  6 20:13:35 localhost kernel: [ 1096.188008] Stack:

Jan  6 20:13:35 localhost kernel: [ 1096.188008]  0000000000000001 0000000000001000 0000000000000001 ffffffff8129e2e0

Jan  6 20:13:35 localhost kernel: [ 1096.188008]  0000000000000001 00007f202c59d000 ffff88006822f480 0000000000000001

Jan  6 20:13:35 localhost kernel: [ 1096.188008]  0000000000001000 0000000000000000 ffff880059877e88 ffffffff811447c5

Jan  6 20:13:35 localhost kernel: [ 1096.188008] Call Trace:

Jan  6 20:13:35 localhost kernel: [ 1096.188008]  [<ffffffff8129e2e0>] ? cap_vm_enough_memory+0x50/0x60

Jan  6 20:13:35 localhost kernel: [ 1096.188008]  [<ffffffff811447c5>] mmap_region+0x2a5/0x4f0

Jan  6 20:13:35 localhost kernel: [ 1096.188008]  [<ffffffff81144d58>] do_mmap_pgoff+0x348/0x360

Jan  6 20:13:35 localhost kernel: [ 1096.188008]  [<ffffffff81144eb1>] sys_mmap_pgoff+0x141/0x230

Jan  6 20:13:35 localhost kernel: [ 1096.188008]  [<ffffffff81018b12>] sys_mmap+0x22/0x30

Jan  6 20:13:35 localhost kernel: [ 1096.188008]  [<ffffffff816655c2>] system_call_fastpath+0x16/0x1b

Jan  6 20:13:35 localhost kernel: [ 1096.188008] Code: 00 4d 8b 04 24 65 4c 03 04 25 50 da 00 00 49 8b 50 08 4d 8b 28 4d 85 ed 0f 84 d8 00 00 00 49 63 44 24 20 49 8b 34 24 48 8d 4a 01 <49> 8b 5c 05 00 4c 89 e8 65 48 0f c7 0e 0f 94 c0 84 c0 74 c2 4d 

Jan  6 20:13:35 localhost kernel: [ 1096.188008] RIP  [<ffffffff8116589a>] kmem_cache_alloc+0x5a/0x140

Jan  6 20:13:35 localhost kernel: [ 1096.188008]  RSP <ffff880059877d78>

Jan  6 20:13:35 localhost kernel: [ 1096.274513] ---[ end trace 766ef1ef52f774ae ]---

Jan  6 20:13:37 localhost kernel: [ 1097.836149] BUG: Bad page state in process swapper/0  pfn:59a33

Jan  6 20:13:37 localhost kernel: [ 1097.838885] page:ffffea0001668cc0 count:0 mapcount:-1 mapping:          (null) index:0xffff880059a33160

Jan  6 20:13:37 localhost kernel: [ 1097.841673] page flags: 0x100000000000000()

Jan  6 20:13:37 localhost kernel: [ 1097.844440] Modules linked in: btrfs zlib_deflate libcrc32c ufs qnx4 hfsplus hfs minix ntfs vfat msdos fat jfs xfs reiserfs ext2 nfsd nfs lockd fscache auth_rpcgss nfs_acl sunrpc dm_crypt ppdev ipt_REJECT ipt_LOG ipt_MASQUERADE xt_state iptable_mangle iptable_nat nf_nat nf_conntrack_ipv4 nf_conntrack nf_defrag_ipv4 iptable_filter ip_tables x_tables joydev sp5100_tco edac_core i2c_piix4 serio_raw k8temp edac_mce_amd snd_hda_codec_hdmi snd_hda_intel snd_hda_codec snd_hwdep snd_pcm snd_timer snd soundcore parport_pc snd_page_alloc mac_hid shpchp lp parport radeon 8139too ttm drm_kms_helper drm pata_atiixp i2c_algo_bit usbhid hid wmi r8169

Jan  6 20:13:37 localhost kernel: [ 1097.856881] Pid: 0, comm: swapper/0 Tainted: G      D      3.2.0-35-generic #55-Ubuntu

Jan  6 20:13:37 localhost kernel: [ 1097.860020] Call Trace:

Jan  6 20:13:37 localhost kernel: [ 1097.863063]  <IRQ>  [<ffffffff8111fe8f>] bad_page.part.61+0x9f/0xf0

Jan  6 20:13:37 localhost kernel: [ 1097.866119]  [<ffffffff8111fef8>] bad_page+0x18/0x30

Jan  6 20:13:37 localhost kernel: [ 1097.869158]  [<ffffffff8112098e>] free_pages_prepare+0x10e/0x120

Jan  6 20:13:37 localhost kernel: [ 1097.872178]  [<ffffffff81120af9>] free_hot_cold_page+0x49/0x1a0

Jan  6 20:13:37 localhost kernel: [ 1097.875183]  [<ffffffff81120c7d>] __free_pages+0x2d/0x40

Jan  6 20:13:37 localhost kernel: [ 1097.878163]  [<ffffffff8159a8fb>] tcp_v4_destroy_sock+0x25b/0x2c0

Jan  6 20:13:37 localhost kernel: [ 1097.881105]  [<ffffffff81582695>] inet_csk_destroy_sock+0x55/0x140

Jan  6 20:13:37 localhost kernel: [ 1097.883970]  [<ffffffff815849b0>] tcp_done+0x50/0x90

Jan  6 20:13:37 localhost kernel: [ 1097.886853]  [<ffffffff81591d92>] tcp_rcv_state_process+0x422/0x5f0

Jan  6 20:13:37 localhost kernel: [ 1097.889724]  [<ffffffff8159a597>] tcp_v4_do_rcv+0xc7/0x1d0

Jan  6 20:13:37 localhost kernel: [ 1097.892513]  [<ffffffff8159c1f1>] tcp_v4_rcv+0x581/0x820

Jan  6 20:13:37 localhost kernel: [ 1097.895301]  [<ffffffff81577b60>] ? ip_rcv_finish+0x370/0x370

Jan  6 20:13:37 localhost kernel: [ 1097.898110]  [<ffffffff81577b60>] ? ip_rcv_finish+0x370/0x370

Jan  6 20:13:37 localhost kernel: [ 1097.900915]  [<ffffffff81577c3d>] ip_local_deliver_finish+0xdd/0x280

Jan  6 20:13:37 localhost kernel: [ 1097.903716]  [<ffffffff81577fa8>] ip_local_deliver+0x88/0x90

Jan  6 20:13:37 localhost kernel: [ 1097.906502]  [<ffffffff815778fd>] ip_rcv_finish+0x10d/0x370

Jan  6 20:13:37 localhost kernel: [ 1097.909279]  [<ffffffff815781e5>] ip_rcv+0x235/0x300

Jan  6 20:13:37 localhost kernel: [ 1097.912067]  [<ffffffff81613dc7>] ? packet_rcv_spkt+0x47/0x190

Jan  6 20:13:37 localhost kernel: [ 1097.914831]  [<ffffffff81543446>] __netif_receive_skb+0x4d6/0x550

Jan  6 20:13:37 localhost kernel: [ 1097.917624]  [<ffffffff81544230>] netif_receive_skb+0x80/0x90

Jan  6 20:13:37 localhost kernel: [ 1097.920415]  [<ffffffff81536474>] ? __netdev_alloc_skb+0x24/0x50

Jan  6 20:13:37 localhost kernel: [ 1097.923124]  [<ffffffffa00d6e90>] rtl8139_rx+0x150/0x2b0 [8139too]

Jan  6 20:13:37 localhost kernel: [ 1097.925754]  [<ffffffffa00d704a>] rtl8139_poll+0x5a/0xd0 [8139too]

Jan  6 20:13:37 localhost kernel: [ 1097.928274]  [<ffffffff81544bd4>] net_rx_action+0x134/0x290

Jan  6 20:13:37 localhost kernel: [ 1097.930698]  [<ffffffff8103df8b>] ? native_safe_halt+0xb/0x10

Jan  6 20:13:37 localhost kernel: [ 1097.933115]  [<ffffffff8106f6e8>] __do_softirq+0xa8/0x210

Jan  6 20:13:37 localhost kernel: [ 1097.935495]  [<ffffffff810967f5>] ? do_timer+0x25/0x30

Jan  6 20:13:37 localhost kernel: [ 1097.937836]  [<ffffffff81035dc2>] ? ack_apic_level+0x72/0x190

Jan  6 20:13:37 localhost kernel: [ 1097.940163]  [<ffffffff8166782c>] call_softirq+0x1c/0x30

Jan  6 20:13:37 localhost kernel: [ 1097.942464]  [<ffffffff81016305>] do_softirq+0x65/0xa0

Jan  6 20:13:37 localhost kernel: [ 1097.944778]  [<ffffffff8106face>] irq_exit+0x8e/0xb0

Jan  6 20:13:37 localhost kernel: [ 1097.947068]  [<ffffffff816680e3>] do_IRQ+0x63/0xe0

Jan  6 20:13:37 localhost kernel: [ 1097.949327]  [<ffffffff8165d46e>] common_interrupt+0x6e/0x6e

Jan  6 20:13:37 localhost kernel: [ 1097.951597]  <EOI>  [<ffffffff8103df8b>] ? native_safe_halt+0xb/0x10

Jan  6 20:13:37 localhost kernel: [ 1097.953891]  [<ffffffff810900a8>] ? hrtimer_start+0x18/0x20

Jan  6 20:13:37 localhost kernel: [ 1097.956171]  [<ffffffff8101c983>] default_idle+0x53/0x1d0

Jan  6 20:13:37 localhost kernel: [ 1097.958426]  [<ffffffff8101cb5d>] amd_e400_idle+0x5d/0x120

Jan  6 20:13:37 localhost kernel: [ 1097.960704]  [<ffffffff81013236>] cpu_idle+0xd6/0x120

Jan  6 20:13:37 localhost kernel: [ 1097.962970]  [<ffffffff816235ee>] rest_init+0x72/0x74

Jan  6 20:13:37 localhost kernel: [ 1097.965195]  [<ffffffff81cfbc03>] start_kernel+0x3b0/0x3bd

Jan  6 20:13:37 localhost kernel: [ 1097.967421]  [<ffffffff81cfb388>] x86_64_start_reservations+0x132/0x136

Jan  6 20:13:37 localhost kernel: [ 1097.969660]  [<ffffffff81cfb140>] ? early_idt_handlers+0x140/0x140

Jan  6 20:13:37 localhost kernel: [ 1097.971888]  [<ffffffff81cfb459>] x86_64_start_kernel+0xcd/0xdc

What's happening here? What can I do?