lowlatency-flavour crashes and locks up alot

Bug #1275116 reported by Søren Holm
16
This bug affects 2 people
Affects Status Importance Assigned to Milestone
linux (Ubuntu)
Fix Released
High
Unassigned
linux-lowlatency (Ubuntu)
Fix Released
High
Unassigned

Bug Description

The 3.11 flavour that where upgraded to 3.13 today crashed 10 times this morning. I have the crashdums, but I figured that it might have been solved in 3.13. Unfortunatly it seems like it is not.

I will add crashdumps when I get crashes that do not just lockup the machine but currently I have only seem complete lockups with 3.13

So the essense is :

3.11-lowlatency crashed 10 time today most of the crashes reached the console so I could see the panic text.
3.13-lowlatency crash 1 time 10 minuttes after first boot. No VT swtich. During next reboot it also locked.

ProblemType: Bug
DistroRelease: Ubuntu 14.04
Package: linux-image-3.13.0-6-lowlatency (not installed)
ProcVersionSignature: Ubuntu 3.13.0-6.23-generic 3.13.0
Uname: Linux 3.13.0-6-generic i686
ApportVersion: 2.13.2-0ubuntu2
Architecture: i386
AudioDevicesInUse:
 USER PID ACCESS COMMAND
 /dev/snd/controlC0: sgh 2176 F.... pulseaudio
CurrentDesktop: KDE
Date: Fri Jan 31 23:38:10 2014
InstallationDate: Installed on 2013-12-24 (38 days ago)
InstallationMedia: Kubuntu 14.04 LTS "Trusty Tahr" - Alpha i386 (20131224)
MachineType: LENOVO 2356GCG
ProcFB: 0 inteldrmfb
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-3.13.0-6-generic root=UUID=6376d4d7-bb9e-4488-bb5f-798e6c41f3fc ro quiet splash vt.handoff=7
RelatedPackageVersions:
 linux-restricted-modules-3.13.0-6-generic N/A
 linux-backports-modules-3.13.0-6-generic N/A
 linux-firmware 1.123
SourcePackage: linux
UpgradeStatus: No upgrade log present (probably fresh install)
dmi.bios.date: 07/02/2012
dmi.bios.vendor: LENOVO
dmi.bios.version: G7ET31WW (1.13 )
dmi.board.asset.tag: Not Available
dmi.board.name: 2356GCG
dmi.board.vendor: LENOVO
dmi.board.version: Not Defined
dmi.chassis.asset.tag: No Asset Information
dmi.chassis.type: 10
dmi.chassis.vendor: LENOVO
dmi.chassis.version: Not Available
dmi.modalias: dmi:bvnLENOVO:bvrG7ET31WW(1.13):bd07/02/2012:svnLENOVO:pn2356GCG:pvrThinkPadT430s:rvnLENOVO:rn2356GCG:rvrNotDefined:cvnLENOVO:ct10:cvrNotAvailable:
dmi.product.name: 2356GCG
dmi.product.version: ThinkPad T430s
dmi.sys.vendor: LENOVO

Revision history for this message
Søren Holm (sgh) wrote :
Revision history for this message
Brad Figg (brad-figg) wrote : Status changed to Confirmed

This change was made by a bot.

Changed in linux (Ubuntu):
status: New → Confirmed
Revision history for this message
Søren Holm (sgh) wrote :
Download full text (31.1 KiB)

4,996,3231003672,-;------------[ cut here ]------------
2,997,3231003700,-;kernel BUG at /build/buildd/linux-3.13.0/kernel/seccomp.c:449!
4,998,3231003733,-;invalid opcode: 0000 [#1] PREEMPT SMP
4,999,3231003759,c;Modules linked in: configfs ctr ccm pci_stub vboxpci(OF) vboxnetadp(OF) vboxnetflt(OF) joydev vboxdrv(OF) intel_rapl x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm snd_hda_codec_hdmi crc32_pclmul aesni_intel snd_hda_codec_realtek parport_pc aes_i586 xts ppdev lrw gf128mul ablk_helper cryptd snd_hda_intel rfcomm snd_hda_codec bnep btusb bluetooth snd_hwdep uvcvideo videobuf2_vmalloc snd_pcm thinkpad_acpi cdc_mbim snd_page_alloc videobuf2_memops cdc_ncm nvram videobuf2_core videodev snd_seq_midi usbnet arc4 cdc_acm mii cdc_wdm snd_seq_midi_event iwldvm snd_rawmidi mac80211 snd_seq snd_seq_device psmouse iwlwifi binfmt_misc snd_timer serio_raw microcode lpc_ich snd cfg80211 soundcore mac_hid mei_me mei lp parport i915 i2c_algo_bit drm_kms_helper e1000e sdhci_pci ahci drm libahci
4,1000,3231004149,+; sdhci ptp pps_core wmi video
0,1001,3231004168,-;CPU: 1 PID: 13597 Comm: EMT-1 Tainted: GF O 3.13.0-6-lowlatency #23-Ubuntu
0,1002,3231004209,-;Hardware name: LENOVO 2356GCG/2356GCG, BIOS G7ET31WW (1.13 ) 07/02/2012
0,1003,3231004244,-;task: c2a20000 ti: d532a000 task.ti: cd378000
4,1004,3231004270,-;EIP: 0060:[<c10e8078>] EFLAGS: 00010293 CPU: 1
1,1005,3231004319,-;BUG: unable to handle kernel paging request at 946d8e89
1,1006,3231004362,-;IP: [<946d8e89>] 0x946d8e89
4,1007,3231004383,-;*pdpt = 00000000350c6001 *pde = 0000000173938067
4,1008,3231004412,-;Oops: 0011 [#2] PREEMPT SMP
4,1009,3231004433,c;Modules linked in: configfs ctr ccm pci_stub vboxpci(OF) vboxnetadp(OF) vboxnetflt(OF) joydev vboxdrv(OF) intel_rapl x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm snd_hda_codec_hdmi crc32_pclmul aesni_intel snd_hda_codec_realtek parport_pc aes_i586 xts ppdev lrw gf128mul ablk_helper cryptd snd_hda_intel rfcomm snd_hda_codec bnep btusb bluetooth snd_hwdep uvcvideo videobuf2_vmalloc snd_pcm thinkpad_acpi cdc_mbim snd_page_alloc videobuf2_memops cdc_ncm nvram videobuf2_core videodev snd_seq_midi usbnet arc4 cdc_acm mii cdc_wdm snd_seq_midi_event iwldvm snd_rawmidi mac80211 snd_seq snd_seq_device psmouse iwlwifi binfmt_misc snd_timer serio_raw microcode lpc_ich snd cfg80211 soundcore mac_hid mei_me mei lp parport i915 i2c_algo_bit drm_kms_helper e1000e sdhci_pci ahci drm libahci
4,1010,3231004817,+; sdhci ptp pps_core wmi video
0,1011,3231004834,-;CPU: 1 PID: 13597 Comm: EMT-1 Tainted: GF O 3.13.0-6-lowlatency #23-Ubuntu
0,1012,3231004874,-;Hardware name: LENOVO 2356GCG/2356GCG, BIOS G7ET31WW (1.13 ) 07/02/2012
0,1013,3231004909,-;task: c2a20000 ti: d532a000 task.ti: cd378000
4,1014,3231004934,-;EIP: 0060:[<946d8e89>] EFLAGS: 00010092 CPU: 1
4,1015,3231004961,-;EIP is at 0x946d8e89
4,1016,3231004976,-;EAX: 0000002e EBX: d532bf2c ECX: 00000006 EDX: 00000007
4,1017,3231005005,-;ESI: 00000001 EDI: d532bf68 EBP: f7b9f648 ESP: d532bddc
4,1018,3231005034,-; DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068
4,1019,3231005059,-;CR0: 80050033 CR2: 946d8e89 CR3: 2bd8e000 CR4: 001427f0
0,1020,3231005088,-;Stack:
...

Revision history for this message
Søren Holm (sgh) wrote :

Crash during web-site access using chromium.

description: updated
Revision history for this message
Søren Holm (sgh) wrote :

After the prebvious crash I powercycled the computer and got a stall here.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

Same problem here. Whole system locks sudently. I'm using same dist and kernel.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

I'm using arch amd64

Revision history for this message
Søren Holm (sgh) wrote :

I have tried setting up netconsole to cash what the kernel might think is happening but wireless is not supported.

https://wiki.ubuntu.com/Kernel/Netconsole

Jaime, maybe you have a wired setup and can try to set it up?

Revision history for this message
Søren Holm (sgh) wrote :

Jaime, are you running virtualbox or some other virtualization software?

Revision history for this message
Søren Holm (sgh) wrote :

I'm quite sure that this relates to running virtualbox - the system crashes within 5 minuttes. Not running virtualbox does not make the system unstable.

Revision history for this message
Søren Holm (sgh) wrote :

Another crash

Revision history for this message
Jaime Pérez (jaime-91) wrote :

I'm running not virtualized ubuntu. I have wired conexion, but i don't know what is net console. I'm going to read the link you posted and I'll try to do something.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

However in my netbook I'm running lowlatency kernel and it seems to work fine. Perhaps is a driver issue or something.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

Ok its logging. I'm going to leave it for a while. Computer is blocked and still pasing info by netconsole. If log is too big i will compress it.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

Here it is. I thing that it was in a loop, because some info seems to reappear. If more information its needed I can do this again.

Revision history for this message
Søren Holm (sgh) wrote :

We seem some more skilled people than me on this. It seem like the error start at one thing, but then progresses and pull the IO subsystem down (the IO-error that I also get).

Revision history for this message
Søren Holm (sgh) wrote :
Download full text (3.6 KiB)

diff --git a/boot/config-3.13.0-6-generic b/boot/config-3.13.0-6-lowlatency
index 2684575..e86aff9 100644
--- a/boot/config-3.13.0-6-generic
+++ b/boot/config-3.13.0-6-lowlatency
@@ -1,6 +1,6 @@
 #
 # Automatically generated file; DO NOT EDIT.
-# Linux/i386 3.13.0-6-generic Kernel Configuration
+# Linux/i386 3.13.0-6-lowlatency Kernel Configuration
 #
 # CONFIG_64BIT is not set
 CONFIG_X86_32=y
@@ -64,7 +64,7 @@ CONFIG_KERNEL_GZIP=y
 # CONFIG_KERNEL_LZO is not set
 # CONFIG_KERNEL_LZ4 is not set
 CONFIG_DEFAULT_HOSTNAME="(none)"
-CONFIG_VERSION_SIGNATURE="Ubuntu 3.13.0-6.23-generic 3.13.0"
+CONFIG_VERSION_SIGNATURE="Ubuntu 3.13.0-6.23-lowlatency 3.13.0"
 CONFIG_SWAP=y
 CONFIG_SYSVIPC=y
 CONFIG_SYSVIPC_SYSCTL=y
@@ -86,7 +86,7 @@ CONFIG_GENERIC_IRQ_CHIP=y
 CONFIG_IRQ_DOMAIN=y
 # CONFIG_IRQ_DOMAIN_DEBUG is not set
 CONFIG_IRQ_FORCED_THREADING=y
-# CONFIG_IRQ_FORCED_THREADING_DEFAULT is not set
+CONFIG_IRQ_FORCED_THREADING_DEFAULT=y
 CONFIG_SPARSE_IRQ=y
 CONFIG_CLOCKSOURCE_WATCHDOG=y
 CONFIG_KTIME_SCALAR=y
@@ -121,14 +121,15 @@ CONFIG_TASK_IO_ACCOUNTING=y
 #
 # RCU Subsystem
 #
-CONFIG_TREE_RCU=y
-# CONFIG_PREEMPT_RCU is not set
+CONFIG_TREE_PREEMPT_RCU=y
+CONFIG_PREEMPT_RCU=y
 CONFIG_RCU_STALL_COMMON=y
 CONFIG_RCU_FANOUT=32
 CONFIG_RCU_FANOUT_LEAF=16
 # CONFIG_RCU_FANOUT_EXACT is not set
 CONFIG_RCU_FAST_NO_HZ=y
 # CONFIG_TREE_RCU_TRACE is not set
+# CONFIG_RCU_BOOST is not set
 # CONFIG_RCU_NOCB_CPU is not set
 # CONFIG_IKCONFIG is not set
 CONFIG_LOG_BUF_SHIFT=17
@@ -222,7 +223,6 @@ CONFIG_HAVE_OPROFILE=y
 CONFIG_OPROFILE_NMI_TIMER=y
 CONFIG_KPROBES=y
 CONFIG_JUMP_LABEL=y
-CONFIG_OPTPROBES=y
 CONFIG_KPROBES_ON_FTRACE=y
 CONFIG_UPROBES=y
 # CONFIG_HAVE_64BIT_ALIGNED_ACCESS is not set
@@ -342,11 +342,6 @@ CONFIG_PREEMPT_NOTIFIERS=y
 CONFIG_PADATA=y
 CONFIG_ASN1=y
 CONFIG_UNINLINE_SPIN_UNLOCK=y
-CONFIG_INLINE_SPIN_UNLOCK_IRQ=y
-CONFIG_INLINE_READ_UNLOCK=y
-CONFIG_INLINE_READ_UNLOCK_IRQ=y
-CONFIG_INLINE_WRITE_UNLOCK=y
-CONFIG_INLINE_WRITE_UNLOCK_IRQ=y
 CONFIG_MUTEX_SPIN_ON_OWNER=y
 CONFIG_FREEZER=y

@@ -438,8 +433,9 @@ CONFIG_NR_CPUS=8
 CONFIG_SCHED_SMT=y
 CONFIG_SCHED_MC=y
 # CONFIG_PREEMPT_NONE is not set
-CONFIG_PREEMPT_VOLUNTARY=y
-# CONFIG_PREEMPT is not set
+# CONFIG_PREEMPT_VOLUNTARY is not set
+CONFIG_PREEMPT=y
+CONFIG_PREEMPT_COUNT=y
 CONFIG_X86_LOCAL_APIC=y
 CONFIG_X86_IO_APIC=y
 CONFIG_X86_REROUTE_FOR_BROKEN_BOOT_IRQS=y
@@ -534,10 +530,10 @@ CONFIG_EFI_STUB=y
 CONFIG_SECCOMP=y
 CONFIG_CC_STACKPROTECTOR=y
 # CONFIG_HZ_100 is not set
-CONFIG_HZ_250=y
+# CONFIG_HZ_250 is not set
 # CONFIG_HZ_300 is not set
-# CONFIG_HZ_1000 is not set
-CONFIG_HZ=250
+CONFIG_HZ_1000=y
+CONFIG_HZ=1000
 CONFIG_SCHED_HRTICK=y
 CONFIG_KEXEC=y
 CONFIG_CRASH_DUMP=y
@@ -4881,7 +4877,6 @@ CONFIG_DRM_NOUVEAU=m
 CONFIG_NOUVEAU_DEBUG=5
 CONFIG_NOUVEAU_DEBUG_DEFAULT=3
 CONFIG_DRM_NOUVEAU_BACKLIGHT=y
-CONFIG_DRM_I810=m
 CONFIG_DRM_I915=m
 CONFIG_DRM_I915_KMS=y
 CONFIG_DRM_I915_FBDEV=y
@@ -7191,6 +7186,7 @@ CONFIG_PANIC_ON_OOPS_VALUE=0
 CONFIG_SCHED_DEBUG=y
 CONFIG_SCHEDSTATS=y
 CONFIG_TIMER_STATS=y
+# CONFIG_DEBUG_PREEMPT is not set

 #
 # Lock Debugging (spinlocks, mutexes, etc...)
@@ -7217,9 +7213,11 @@ CONFIG_DEBUG_BUGVERBOSE=y
 #
 # RCU Debugging
 #
+# CONFIG_PROV...

Read more...

Revision history for this message
Søren Holm (sgh) wrote :

The diff does not reveal any huge differences between lowlatency and generic, but looking at it maybe others can gen an ideas as to wht might be causing this all-over-the-place panic.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

It seems that CPU gets blocked, so perhaps is something of theese:
+CONFIG_IRQ_FORCED_THREADING_DEFAULT=y
-CONFIG_INLINE_SPIN_UNLOCK_IRQ=y
-CONFIG_INLINE_READ_UNLOCK=y
-CONFIG_INLINE_READ_UNLOCK_IRQ=y
-CONFIG_INLINE_WRITE_UNLOCK=y
-CONFIG_INLINE_WRITE_UNLOCK_IRQ=y

Revision history for this message
Søren Holm (sgh) wrote :

Apparently CONFIG_IRQ_FORCED_THREADING_DEFAULT is not an upstream config flag.

The resulting param force_irqthreads is not set no true unless "threadirqs" are on the kernel commandline.

In the problematic kernel it is set to true by default - something that will never happen on a vanilla kernel.

It can be turned off using "nothreadirqs" on the commandline - which is exactly what I'm going to try now.

Revision history for this message
Søren Holm (sgh) wrote :

Well - nothreadirq has the effect that my virtualbox crashes. Someway around the time where I would expect the whole system to crash.

Jaime - I thing adding "nothreadirqs" would be something that you should try also.

1 comments hidden view all 198 comments
Revision history for this message
Jaime Pérez (jaime-91) wrote : Re: [Bug 1275116] Re: lowlatency-flavour crashes and locks up alot

ok i will try tomorrow

Revision history for this message
Jaime Pérez (jaime-91) wrote :

Ok, I will try tomorrow

1 comments hidden view all 198 comments
Revision history for this message
Søren Holm (sgh) wrote :

I can add that my system locked up in the end but the characteristics of that crash was for sure different. So no real cure for me. Looking forward to hearing what it does to your side of things.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

How can I add the "nothreadsirqs" setting? I open grub console and type it, but I got an error. So I put it in /etc/grub/defaults and execute update-grub. It's that ok?

Revision history for this message
Søren Holm (sgh) wrote :

Problably.... but I just press and hold "shift" to get the bootmenu. This I press 'e' to edit the current entry and add "nothreadirqs" to the linux-line.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

I think I have done it in the wrong way. I will check when I return home.

 Søren Holm <email address hidden> wrote:

Problably.... but I just press and hold "shift" to get the bootmenu.
This I press 'e' to edit the current entry and add "nothreadirqs" to the
linux-line.

--
You received this bug notification because you are subscribed to the bug
report.
https://bugs.launchpad.net/bugs/1275116

Title:
  lowlatency-flavour crashes and locks up alot

Status in “linux” package in Ubuntu:
  Confirmed

Bug description:
  The 3.11 flavour that where upgraded to 3.13 today crashed 10 times
  this morning. I have the crashdums, but I figured that it might have
  been solved in 3.13. Unfortunatly it seems like it is not.

  I will add crashdumps when I get crashes that do not just lockup the
  machine but currently I have only seem complete lockups with 3.13

  So the essense is :

  3.11-lowlatency crashed 10 time today most of the crashes reached the console so I could see the panic text.
  3.13-lowlatency crash 1 time 10 minuttes after first boot. No VT swtich. During next reboot it also locked.

  ProblemType: Bug
  DistroRelease: Ubuntu 14.04
  Package: linux-image-3.13.0-6-lowlatency (not installed)
  ProcVersionSignature: Ubuntu 3.13.0-6.23-generic 3.13.0
  Uname: Linux 3.13.0-6-generic i686
  ApportVersion: 2.13.2-0ubuntu2
  Architecture: i386
  AudioDevicesInUse:
   USER PID ACCESS COMMAND
   /dev/snd/controlC0: sgh 2176 F.... pulseaudio
  CurrentDesktop: KDE
  Date: Fri Jan 31 23:38:10 2014
  InstallationDate: Installed on 2013-12-24 (38 days ago)
  InstallationMedia: Kubuntu 14.04 LTS "Trusty Tahr" - Alpha i386 (20131224)
  MachineType: LENOVO 2356GCG
  ProcFB: 0 inteldrmfb
  ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-3.13.0-6-generic root=UUID=6376d4d7-bb9e-4488-bb5f-798e6c41f3fc ro quiet splash vt.handoff=7
  RelatedPackageVersions:
   linux-restricted-modules-3.13.0-6-generic N/A
   linux-backports-modules-3.13.0-6-generic N/A
   linux-firmware 1.123
  SourcePackage: linux
  UpgradeStatus: No upgrade log present (probably fresh install)
  dmi.bios.date: 07/02/2012
  dmi.bios.vendor: LENOVO
  dmi.bios.version: G7ET31WW (1.13 )
  dmi.board.asset.tag: Not Available
  dmi.board.name: 2356GCG
  dmi.board.vendor: LENOVO
  dmi.board.version: Not Defined
  dmi.chassis.asset.tag: No Asset Information
  dmi.chassis.type: 10
  dmi.chassis.vendor: LENOVO
  dmi.chassis.version: Not Available
  dmi.modalias: dmi:bvnLENOVO:bvrG7ET31WW(1.13):bd07/02/2012:svnLENOVO:pn2356GCG:pvrThinkPadT430s:rvnLENOVO:rn2356GCG:rvrNotDefined:cvnLENOVO:ct10:cvrNotAvailable:
  dmi.product.name: 2356GCG
  dmi.product.version: ThinkPad T430s
  dmi.sys.vendor: LENOVO

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1275116/+subscriptions

Revision history for this message
Jaime Pérez (jaime-91) wrote :

Ok, it seems to be working. However I will be checking if it breaks.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

This is working fine. You are using ubuntu through virtualbox? Perhaps if you use only one processor the problem goes off.

Revision history for this message
Søren Holm (sgh) wrote :

No I'm running Ubuntu on the host system. I run Wintendo in virtualbox.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

Ah ok! isn't there any change in config related to virtualization?

 Søren Holm <email address hidden> wrote:

No I'm running Ubuntu on the host system. I run Wintendo in virtualbox.

--
You received this bug notification because you are subscribed to the bug
report.
https://bugs.launchpad.net/bugs/1275116

Title:
  lowlatency-flavour crashes and locks up alot

Status in “linux” package in Ubuntu:
  Confirmed

Bug description:
  The 3.11 flavour that where upgraded to 3.13 today crashed 10 times
  this morning. I have the crashdums, but I figured that it might have
  been solved in 3.13. Unfortunatly it seems like it is not.

  I will add crashdumps when I get crashes that do not just lockup the
  machine but currently I have only seem complete lockups with 3.13

  So the essense is :

  3.11-lowlatency crashed 10 time today most of the crashes reached the console so I could see the panic text.
  3.13-lowlatency crash 1 time 10 minuttes after first boot. No VT swtich. During next reboot it also locked.

  ProblemType: Bug
  DistroRelease: Ubuntu 14.04
  Package: linux-image-3.13.0-6-lowlatency (not installed)
  ProcVersionSignature: Ubuntu 3.13.0-6.23-generic 3.13.0
  Uname: Linux 3.13.0-6-generic i686
  ApportVersion: 2.13.2-0ubuntu2
  Architecture: i386
  AudioDevicesInUse:
   USER PID ACCESS COMMAND
   /dev/snd/controlC0: sgh 2176 F.... pulseaudio
  CurrentDesktop: KDE
  Date: Fri Jan 31 23:38:10 2014
  InstallationDate: Installed on 2013-12-24 (38 days ago)
  InstallationMedia: Kubuntu 14.04 LTS "Trusty Tahr" - Alpha i386 (20131224)
  MachineType: LENOVO 2356GCG
  ProcFB: 0 inteldrmfb
  ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-3.13.0-6-generic root=UUID=6376d4d7-bb9e-4488-bb5f-798e6c41f3fc ro quiet splash vt.handoff=7
  RelatedPackageVersions:
   linux-restricted-modules-3.13.0-6-generic N/A
   linux-backports-modules-3.13.0-6-generic N/A
   linux-firmware 1.123
  SourcePackage: linux
  UpgradeStatus: No upgrade log present (probably fresh install)
  dmi.bios.date: 07/02/2012
  dmi.bios.vendor: LENOVO
  dmi.bios.version: G7ET31WW (1.13 )
  dmi.board.asset.tag: Not Available
  dmi.board.name: 2356GCG
  dmi.board.vendor: LENOVO
  dmi.board.version: Not Defined
  dmi.chassis.asset.tag: No Asset Information
  dmi.chassis.type: 10
  dmi.chassis.vendor: LENOVO
  dmi.chassis.version: Not Available
  dmi.modalias: dmi:bvnLENOVO:bvrG7ET31WW(1.13):bd07/02/2012:svnLENOVO:pn2356GCG:pvrThinkPadT430s:rvnLENOVO:rn2356GCG:rvrNotDefined:cvnLENOVO:ct10:cvrNotAvailable:
  dmi.product.name: 2356GCG
  dmi.product.version: ThinkPad T430s
  dmi.sys.vendor: LENOVO

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1275116/+subscriptions

Changed in linux (Ubuntu):
importance: Undecided → High
tags: added: kernel-da-key linux-lowlatency
Revision history for this message
Joseph Salisbury (jsalisbury) wrote :

I built an amd64 lowlatency test kernel with PREEMPT disabled. This kernel can be downloaded from:

http://kernel.ubuntu.com/~jsalisbury/lp1275116/

Can you test this kernel and see if it also exhibits this bug? Please note, this is a diagnostic kernel and not a fix.

Thanks in advance!

tags: added: kernel-key
Revision history for this message
Joseph Salisbury (jsalisbury) wrote :

@Soren, I see you have an i686 machine, so I'll build a 32 bit kernel as well.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

I will test it. Should I put the kernel option off?

Revision history for this message
Jaime Pérez (jaime-91) wrote :

Failed to process /etc/kernel/postinst.d at /var/lib/dpkg/info/linux-image-3.13.0-6-lowlatency.postinst line 1025.
dpkg: error processing package linux-image-3.13.0-6-lowlatency (--install):
 subprocess installed post-installation script returned error exit status 2
Setting up linux-tools-3.13.0-6-lowlatency (3.13.0-6.23~lp1275116PREEMPTDISABLED) ...
Errors were encountered while processing:
 linux-image-3.13.0-6-lowlatency

Revision history for this message
Jaime Pérez (jaime-91) wrote :

Ok sorry I put my foot on it. Ignore last message.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

It seems to work all right. But I thought it was due to someone of theese options:
+CONFIG_IRQ_FORCED_THREADING_DEFAULT=y
-CONFIG_INLINE_SPIN_UNLOCK_IRQ=y
-CONFIG_INLINE_READ_UNLOCK=y
-CONFIG_INLINE_READ_UNLOCK_IRQ=y
-CONFIG_INLINE_WRITE_UNLOCK=y
-CONFIG_INLINE_WRITE_UNLOCK_IRQ=y

It seems I was wrong.

Revision history for this message
Joseph Salisbury (jsalisbury) wrote :

@Soren, there is an i386 kernel available at that link now as well.

Revision history for this message
Søren Holm (sgh) wrote :

I can't see it - am I looking the wrong place?

Revision history for this message
Jaime Pérez (jaime-91) wrote :

I thought preemtible kernel wasn't the problem. I don't know, but the fix worked for me, so thanks!

tags: added: patch
Jaime Pérez (jaime-91)
affects: linux → virtualbox
Jaime Pérez (jaime-91)
no longer affects: virtualbox
Jaime Pérez (jaime-91)
affects: linux (Ubuntu) → linux-lowlatency (Ubuntu)
Changed in linux (Ubuntu):
status: New → Confirmed
Jaime Pérez (jaime-91)
affects: linux-meta (Ubuntu) → linux (Ubuntu)
Brad Figg (brad-figg)
Changed in linux (Ubuntu):
status: New → Confirmed
118 comments hidden view all 198 comments
Revision history for this message
Jaime Pérez (jaime-91) wrote :

so the change that produced the error is there, no?

Revision history for this message
Søren Holm (sgh) wrote :

I don't know what changes has gone into the trusty 3.11 or 3.12 kernel. but if I should bisect based in "threadirqs" I would make 3.12 as good and 3.13 as bad based on my comment from yesterday :

<quote>
v3.12-trusty/ seems to work for me. So to me it seem to have broken between 3.12 and 3.13. For sure it did break since where saucy where around christmas, because I ran that with it's lowlatency until then.
</quote>

Revision history for this message
Jaime Pérez (jaime-91) wrote :

3.13.0-generic stucks with threadirqs

Revision history for this message
Jaime Pérez (jaime-91) wrote :

But 3.11.0-17-generic seems to work fine with threadirqs

Revision history for this message
Søren Holm (sgh) wrote :

Well then it's clear. Do we have a kernel bug on this?

Revision history for this message
Jaime Pérez (jaime-91) wrote :

yes, I filled one. the link is somewhere upwards

Revision history for this message
Joseph Salisbury (jsalisbury) wrote :

There is some other bug reports that indicate this may be fixed in the mainline kernel. Can you see if that is the case by testing 3.14-rc4:

http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.14-rc4-trusty/

If the bug still exists there, then the other bug report may not be related.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

Testing

Revision history for this message
Søren Holm (sgh) wrote :

If 3.14-rc4-lowlatency is as good to set with "threadirqs" as the generic kernel I can say that 3.14 works as a charm.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

It seems to be working for me, but it can fail it any moment...

Revision history for this message
Jaime Pérez (jaime-91) wrote :

The only bad point is that I don't have HDMI audio

Revision history for this message
Joseph Salisbury (jsalisbury) wrote :

That is good news that the bug appears to be fixed in 3.14-rc4. If it is stable and the bug is in fact gone after some time, we can perform a reverse bisect to identify the commit that fixes this in 3.14-rc4.

@jamie, it should be ok that HDMI is not working with this kernel. We are only interested in testing it to identify the commit that fixes this original bug. Once we find the fix, we will backport or cherry pick it into Trusty.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

nd how can we find the change?

Revision history for this message
Jaime Pérez (jaime-91) wrote :
Revision history for this message
Jaime Pérez (jaime-91) wrote :

A backport to saucy is needed too.

Revision history for this message
Joseph Salisbury (jsalisbury) wrote :

The following commit has landed in upstream v3.14-rc5:

commit a1227f3c1030e96ebc51d677d2f636268845c5fb
Author: Stanislaw Gruszka <email address hidden>
Date: Wed Feb 19 10:29:01 2014 +0100

    usb: ehci: fix deadlock when threadirqs option is used

It has also been cc'd to stable, so it will make it's way into the stable releases through the usual stable update process.

Changed in linux (Ubuntu):
importance: Undecided → High
Revision history for this message
Jaime Pérez (jaime-91) wrote :

Great!=)

Revision history for this message
Jaime Pérez (jaime-91) wrote :

Thanks for your help Joseph

Revision history for this message
Jaime Pérez (jaime-91) wrote :

The solution of 3.14-rc5

Revision history for this message
Joseph Salisbury (jsalisbury) wrote :

I built a Trusty test kernel that has a cherry pick of commit a1227f3c1030e96ebc51d677d2f636268845c5fb. This test kernel can be downloaded from:

http://kernel.ubuntu.com/~jsalisbury/lp1275116/

Can you test this kernel and confirm it resolves this bug? I built an i386 and amd_64 version of the kernel. You will need to install both the linux-image and linux-image-extra packages.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

I suppose that we have to test them with "threadirqs"

Revision history for this message
Joseph Salisbury (jsalisbury) wrote :

Yes, that would be great. I'll also build a lowlatency specific kernel and post it shortly.

Revision history for this message
Joseph Salisbury (jsalisbury) wrote :

There are also lowlatency kernels available for testing at the link:
http://kernel.ubuntu.com/~jsalisbury/lp1275116/

Revision history for this message
Jaime Pérez (jaime-91) wrote :

It seems to work OK.

Changed in linux (Ubuntu):
status: Confirmed → Fix Committed
Changed in linux-lowlatency (Ubuntu):
status: Confirmed → Fix Committed
Revision history for this message
Jaime Pérez (jaime-91) wrote :

My system blocked again :( This time screen got black.

Revision history for this message
Joseph Salisbury (jsalisbury) wrote :

Per comments #167 and #168, it sounds like this bug is resolved in v3.14-rc4. It also sounds like commit a1227f3c is not what fixes this. It is probably best for us to perform a reverse bisect to identify the real commit that fixes this in 3.14. The first step is to identify the first 3.14 kernel version that fixes the bug. We know 3.14-rc4 has the fix, so can you test 3.14-rc3:

http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.14-rc3-trusty/

If v3.14-rc3 also has the fix, then we would want to test 3.14-rc2, etc until we find the first kernel that exhibits the bug again.

Changed in linux-lowlatency (Ubuntu):
status: Fix Committed → Confirmed
Changed in linux (Ubuntu):
status: Fix Committed → Confirmed
Revision history for this message
Jaime Pérez (jaime-91) wrote :

The thing is that perhaps 3.14-rc4 has the error too, because with the patch my pc was switched on for all day and got stucked at night. I will try 3.14-rc3

Revision history for this message
Jaime Pérez (jaime-91) wrote :

Im testing rc5 again, only to see if it has the error and I didn't wait enogh

Revision history for this message
Jaime Pérez (jaime-91) wrote :

I have looked for changes related with deadlocks. In 3.14-rc2 there is no-one, so we can avoid testing that one.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

netconsole3.2.log
[ 31.830432] usb 2-3.3: USB disconnect, device number 3
[ 71.972061] INFO: rcu_sched detected stalls on CPUs/tasks: { 1} (detected by 2, t=15002 jiffies)

3.14-rc1
Paul E. McKenney (54):
      rcu: Break call_rcu() deadlock involving scheduler and perf

comment 3
1,1046,3231011669,-;BUG: unable to handle kernel NULL pointer dereference at (null)

netconsole.log
[ 11.924118] BUG: unable to handle kernel NULL pointer dereference at 0000000000000010
this bug appear before lockup

Revision history for this message
Jaime Pérez (jaime-91) wrote :

Now Im using rc5 and it seems to work fine.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

I had a lockup with rc4, so the change must be between rc4 and rc5, but it seems is not the one of ehci.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

Does http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.14-rc5-trusty/CHANGES show all changes, or only those produced by ubuntu team?

Revision history for this message
Søren Holm (sgh) wrote :

 3.13.0-16-lowlatency has no issues for me as far as I can see. At least I virtualbox running all day - no crashes.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

but you have issues with previous version?

Revision history for this message
Søren Holm (sgh) wrote :

I've been busy, so I did not try the last versions that Joseph posted.

Revision history for this message
Jaime Pérez (jaime-91) wrote :

I have upgraded to 14.04 again. I'll download rc5, rc4 etc

Revision history for this message
Jaime Pérez (jaime-91) wrote :

Well, first I will test 3.13.0-16-lowlatency

Revision history for this message
Jaime Pérez (jaime-91) wrote :

3.13.0-16-lowlatency seems to be working fine

Revision history for this message
Jaime Pérez (jaime-91) wrote :

Still works, so it seems fix has been released.

Jaime Pérez (jaime-91)
Changed in linux-lowlatency (Ubuntu):
status: Confirmed → Fix Released
Changed in linux (Ubuntu):
status: Confirmed → Fix Released
Displaying first 40 and last 40 comments. View all 198 comments or add a comment.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.