[Dell Inspiron 3451] WARNING: CPU: 1 PID: 1297 at /build/linux-lts-utopic-vy2yyy/linux-lts-utopic-3.16.0/drivers/gpu/drm/i915/intel_display.c:1743 vlv_wait_port_ready+0x126/0x170 [i915]()

[Andrew]

After upgrading from stock Ubuntu 14.04 to 16.04 my laptop consistently freezes up and is unresponsive to keyboard or mouse. The only way is to do a hard shutdown. I'm usually in a browser when this happens, and often watching a video or two. It seems to happen more often in Google Chrome than Firefox, but it happens with both. It has happened at least once with no videos loaded. I haven’t observed it freeze / hang when I’m not in a browser, but that's mainly what i use my laptop for.

This issue is correlated to documented Intel CPU errata noted in:
https://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/pentium-n3520-j2850-celeron-n2920-n2820-n2815-n2806-j1850-j1750-spec-update.pdf

Call trace:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1575467/+attachment/4648682/+files/kern.log.2.gz

WORKAROUND: Use out-of-tree patch from https://bugzilla.kernel.org/show_bug.cgi?id=109051 . Patch file:
https://bugzilla.kernel.org/attachment.cgi?id=247621

---
ApportVersion: 2.20.1-0ubuntu2
Architecture: amd64
AudioDevicesInUse:
 USER PID ACCESS COMMAND
 /dev/snd/controlC0: andrew 1949 F.... pulseaudio
DistroRelease: Ubuntu 16.04
HibernationDevice: RESUME=UUID=e8aaee7c-d956-4519-94f0-cf7aad745020
InstallationDate: Installed on 2015-08-08 (263 days ago)
InstallationMedia: Ubuntu 14.04.2 LTS "Trusty Tahr" - Release amd64 (20150218.1)
MachineType: Dell Inc. Inspiron 3451
Package: linux (not installed)
ProcEnviron:
 LANGUAGE=en_US
 TERM=xterm-256color
 PATH=(custom, no user)
 LANG=en_US.UTF-8
 SHELL=/bin/bash
ProcFB: 0 inteldrmfb
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-4.4.0-21-generic.efi.signed root=UUID=d01f07be-c906-4024-8eef-a58ba3935298 ro quiet splash vt.handoff=7
ProcVersionSignature: Ubuntu 4.4.0-21.37-generic 4.4.6
PulseList:
 Error: command ['pacmd', 'list'] failed with exit code 1: Home directory not accessible: Permission denied
 No PulseAudio daemon running, or not running as session daemon.
RelatedPackageVersions:
 linux-restricted-modules-4.4.0-21-generic N/A
 linux-backports-modules-4.4.0-21-generic N/A
 linux-firmware 1.157
Tags: xenial
Uname: Linux 4.4.0-21-generic x86_64
UpgradeStatus: Upgraded to xenial on 2016-04-22 (4 days ago)
UserGroups:

_MarkForUpload: True
dmi.bios.date: 12/05/2014
dmi.bios.vendor: Dell Inc.
dmi.bios.version: A00
dmi.board.name: 0H4MK6
dmi.board.vendor: Dell Inc.
dmi.board.version: A00
dmi.chassis.type: 8
dmi.chassis.vendor: Dell Inc.
dmi.chassis.version: Not Specified
dmi.modalias: dmi:bvnDellInc.:bvrA00:bd12/05/2014:svnDellInc.:pnInspiron3451:pvrA00:rvnDellInc.:rn0H4MK6:rvrA00:cvnDellInc.:ct8:cvrNotSpecified:
dmi.product.name: Inspiron 3451
dmi.product.version: A00
dmi.sys.vendor: Dell Inc.

[Sebastien Bacher (seb128)]

seems locks are more likely to be a kernel/driver issue than a desktop ui one

[Brad Figg (brad-figg)]

This bug is missing log files that will aid in diagnosing the problem. From a terminal window please run:

apport-collect 1575467

and then change the status of the bug to 'Confirmed'.

If, due to the nature of the issue you have encountered, you are unable to run this command, please add a comment stating that fact and change the bug status to 'Confirmed'.

This change has been made by an automated script, maintained by the Ubuntu Kernel Team.

[[e-mail] (k-pant)]

I have the same problem !!
Lenovo G50-30 with Intel Pentium N3540 CPU

[Andrew (keen101)]

I did wonder if it was a driver issue related to my processor, so that or a kernel issue makes sense. Not sure how to confirm that though. In the additional drivers thingy it says it is using a backport driver for something that i assume is processor related. My understanding was that backport drivers are unstable and should not be used unless absolutely necessary. There was also a processor microcode firmware proprietary driver that was not being used, so i tried it to see if that would fix the problem, it did not.

I will try to do the apport log thing that the automatic script suggested. Hopefully you guy's can figure this out as this is a pretty serious issue. Let me know if you need any more logs. It sounds like the other guy with the same processor as my laptop does also has the same problem.

[Andrew (keen101)]

apport information

[Andrew (keen101)]

apport information

[Andrew (keen101)]

apport information

[Andrew (keen101)]

apport information

[Andrew (keen101)]

apport information

[Andrew (keen101)]

apport information

[Andrew (keen101)]

apport information

[Andrew (keen101)]

apport information

[Andrew (keen101)]

apport information

[Andrew (keen101)]

apport information

[Andrew (keen101)]

apport information

[Andrew (keen101)]

apport information

[Andrew (keen101)]

apport information

[Andrew (keen101)]

Both of these bugs sound like the same problem i'm having, but these two bugs are expired now, but the thing we all have in common is Intel Pentium processor N3540 and often (but not always) freezes because we play videos in Chrome browser. So, is this a kernel problem or a video driver problem or what? I didn't have problems in 14.04, but i do in 16.04.

https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1530623

https://bugs.launchpad.net/ubuntu/+source/xorg/+bug/1516108

[Javier Antonio Nisa Avila (javiernisa)]

The same bug reported in this bug https://bugs.launchpad.net/ubuntu/+source/xorg/+bug/1575878

Intel N2940

Canonical should give a solution; it is a bug that has many users

[Tim Ryan (tim-ryan)]

I have the same issue. The work around with "GRUB_CMDLINE_LINUX_DEFAULT="quiet splash intel_idle.max_cstate=1" seems to have solved the issue for now on my Acer E511 with Intel Pentium N3540 CHIP. And the temp has not increased very much after recent hard use.

[Tim Ryan (tim-ryan)]

Oh sorry forgot to mention I am using Ubuntu 16.04. 64 bit.

[Andrew (keen101)]

Thanks Tim, good to see other people have the same problem as usually that means this is a bug that will get fixed quickly. And good to see that there is a potential workaround to potentially solve the problem. But isn't the "fix released" status meant for developers to mark? Meaning that code has actually been changed in an upstream package and future update? If your not a developer then i'm not sure the "fix released" status is appropriate as it probably means no one will no longer look at this bug because they think someone else already fixed it.

In the past i may have ventured to go fiddling around with grub or some other file (as i did in 14.04 with my non-working bluetooth), but for something like this i'm only going to wait for an official fix. And to me this is a pretty serious issue. If my understanding of how launchpad works, then just ignore my comments.

I am also using 16.04 64bit. (not sure if many people still use 32bit or if 32bit users are unaffected).

[Tim Ryan (tim-ryan)]

Oh sorry I clicked some thing wrong... I never said a fix was released lol... how can I remove that part?

[Andrew (keen101)]

It's up at the top. No worries. I think i changed it back. I do hope someone is able to help us figure out what specific package this affects as getting it to the right person should help speed up the bug fixing process. I'm still not sure if this is a memory issue, or a kernel issue, or a video driver issue, or a grub issue.

[Andrew (keen101)]

Tim, i found this similar bug which also posts the same solution you did as a way to bypass the problem.

https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1509723

If it is indeed the same problem then it seems like it is a serious kernel bug. The weird part is that there is no blinking caps lock led to indicate a kernel panic.

[Gavin Conaghty (gavsue)]

I have the same problem. Ubuntu 16.04 on HP pavilion g series RT9350

[Andrew (keen101)]

Thanks Christopher,

I just upgraded my BIOS from version A00 to A05. So far everything seems to work great. I never knew a BIOS bug could affect an operating system. But these new EFI BIOSs are very different from the old BIOSs i grew up with so i'm not really sure how they work. Apparently more integration (which in this case was not a good thing).

But anyway, so far it seems i can use my browser with multiple tabs and multiple videos running no problem. No freezes. As of this time i will go ahead and mark this bug as invalid as you requested. I assume if i have any more problems i can come back and change the status later.

Again thanks for the tip to update my BIOS.
-Andrew

[Andrew (keen101)]

Nevermind, it seems that i jumped the gun. It happened again. If i had to guess i'd say that the freezes are more likely to occur when the computer has been on for awhile and is in moderate use. Others have hinted at processor heat problems, often when one freeze occurs and the computer is immediately rebooted the freezes occur with more frequency (perhaps because the processor is already hot?).

sudo dmidecode -s bios-version && sudo dmidecode -s bios-release-date
[sudo] password for andrew:
A05
09/30/2015

[Andrew (keen101)]

i want to say that the updated bios may have helped. Hard to say at this point. If it did it may not be as severe as it was, but perhaps did not eliminate the problem entirely. Even having a buggy bios i find it hard to believe that i would not have had problems in Ubuntu 14.04 but did after upgrading. Perhaps there is a bug that is made worse by a buggy BIOS. Just my thoughts. I will try to keep monitoring the situation and try to add anything new if i notice anything.

[Terry Coles (lp-hadrian-way)]

I appear to have the same problem with my Dell Optiplex desktop machine and my Dell XPS 13 Developer Edition. The XPS is brand new and has been updated to the latest BIOS version. The Optiplex has a somewhat aged BIOS, but I've never seen any problems until now.

Both machines have been updated by doing a clean install of 16.04; Kubuntu on this machine and Ubuntu on the XPS.

One thing that I've discovered is that the machine can be pinged when it's down, but otherwise the symptoms are identical to those reported above. Entering coredumpctl returns 'No coredumps found'.

The problem has been observed when I am typing, moving the mouse, reading text or simply doing nothing.

[Andrew (keen101)]

Just updating to say that it has happened several times today while having multiple tabs open in Chrome. Not sure if that is related or not. And that the processor fan seems to be running loud around the time of the freeze ups.

I wanted to specifically say that the BIOS update did not help after all. I would like to request that the importance status be changed to be higher than "low" as this seems to affect multiple users on multiple systems and is pretty critical to workflow. For some people it might even be considered unusable.

[Tim Ryan (tim-ryan)]

Thank you Andrew for the link. I am looking every where for other's that might have the same problem or a fix more solid than messing with cstates . Yes this is a big wide spread bug. I read about many with the same Intel Pentium N354O that I have.

[penalvch (penalvch)]

Andrew, in order to allow additional upstream developers to examine the issue, at your earliest convenience, could you please test the latest upstream kernel available from http://kernel.ubuntu.com/~kernel-ppa/mainline/?C=N;O=D ? Please keep in mind the following:
1) The one to test is at the very top line at the top of the page (not the daily folder).
2) The release names are irrelevant.
3) The folder time stamps aren't indicative of when the kernel actually was released upstream.
4) Install instructions are available at https://wiki.ubuntu.com/Kernel/MainlineBuilds .

If testing on your main install would be inconvenient, one may:
1) Install Ubuntu to a different partition and then test this there.
2) Backup, or clone the primary install.

If the latest kernel did not allow you to test to the issue (ex. you couldn't boot into the OS) please make a comment in your report about this, and continue to test the next most recent kernel version until you can test to the issue. Once you've tested the upstream kernel, please comment on which kernel version specifically you tested. If this issue is fixed in the mainline kernel, please add the following tags by clicking on the yellow circle with a black pencil icon, next to the word Tags, located at the bottom of the report description:
kernel-fixed-upstream
kernel-fixed-upstream-X.Y-rcZ

Where X, and Y are the first two numbers of the kernel version, and Z is the release candidate number if it exists.

If the mainline kernel does not fix the issue, please add the following tags:
kernel-bug-exists-upstream
kernel-bug-exists-upstream-X.Y-rcZ

Please note, an error to install the kernel does not fit the criteria of kernel-bug-exists-upstream.

Also, you don't need to apport-collect further unless specifically requested to do so.

Once testing of the latest upstream kernel is complete, please mark this report Status Confirmed. Please let us know your results.

Thank you for your understanding.

[Andrew (keen101)]

ok, i will attempt to test the latest upstream kernel as per instructions. During deb depackaging there were two errors, but it seemed to complete anyway. Here is the text from my terminal just in case:

andrew@andrew-Inspiron-3451:~/Downloads/upstream$ ls
linux-headers-4.6.0-040600rc5_4.6.0-040600rc5.201604242031_all.deb
linux-headers-4.6.0-040600rc5-generic_4.6.0-040600rc5.201604242031_amd64.deb
linux-image-4.6.0-040600rc5-generic_4.6.0-040600rc5.201604242031_amd64.deb
andrew@andrew-Inspiron-3451:~/Downloads/upstream$ sudo dpkg -i *.deb
[sudo] password for andrew:
Selecting previously unselected package linux-headers-4.6.0-040600rc5.
(Reading database ... 238152 files and directories currently installed.)
Preparing to unpack linux-headers-4.6.0-040600rc5_4.6.0-040600rc5.201604242031_all.deb ...
Unpacking linux-headers-4.6.0-040600rc5 (4.6.0-040600rc5.201604242031) ...
Selecting previously unselected package linux-headers-4.6.0-040600rc5-generic.
Preparing to unpack linux-headers-4.6.0-040600rc5-generic_4.6.0-040600rc5.201604242031_amd64.deb ...
Unpacking linux-headers-4.6.0-040600rc5-generic (4.6.0-040600rc5.201604242031) ...
Selecting previously unselected package linux-image-4.6.0-040600rc5-generic.
Preparing to unpack linux-image-4.6.0-040600rc5-generic_4.6.0-040600rc5.201604242031_amd64.deb ...
Examining /etc/kernel/preinst.d/
run-parts: executing /etc/kernel/preinst.d/intel-microcode 4.6.0-040600rc5-generic /boot/vmlinuz-4.6.0-040600rc5-generic
Done.
Unpacking linux-image-4.6.0-040600rc5-generic (4.6.0-040600rc5.201604242031) ...
Setting up linux-headers-4.6.0-040600rc5 (4.6.0-040600rc5.201604242031) ...
Setting up linux-headers-4.6.0-040600rc5-generic (4.6.0-040600rc5.201604242031) ...
Examining /etc/kernel/header_postinst.d.
run-parts: executing /etc/kernel/header_postinst.d/dkms 4.6.0-040600rc5-generic /boot/vmlinuz-4.6.0-040600rc5-generic
Error! The dkms.conf for this module includes a BUILD_EXCLUSIVE directive which
does not match this kernel/arch. This indicates that it should not be built.

Setting up linux-image-4.6.0-040600rc5-generic (4.6.0-040600rc5.201604242031) ...
Running depmod.
update-initramfs: deferring update (hook will be called later)
Examining /etc/kernel/postinst.d.
run-parts: executing /etc/kernel/postinst.d/apt-auto-removal 4.6.0-040600rc5-generic /boot/vmlinuz-4.6.0-040600rc5-generic
run-parts: executing /etc/kernel/postinst.d/dkms 4.6.0-040600rc5-generic /boot/vmlinuz-4.6.0-040600rc5-generic
Error! The dkms.conf for this module includes a BUILD_EXCLUSIVE directive which
does not match this kernel/arch. This indicates that it should not be built.
run-parts: executing /etc/kernel/postinst.d/initramfs-tools 4.6.0-040600rc5-generic /boot/vmlinuz-4.6.0-040600rc5-generic
update-initramfs: Generating /boot/initrd.img-4.6.0-040600rc5-generic
run-parts: executing /etc/kernel/postinst.d/pm-utils 4.6.0-040600rc5-generic /boot/vmlinuz-4.6.0-040600rc5-generic
run-parts: executing /etc/kernel/postinst.d/unattended-upgrades 4.6.0-040600rc5-generic /boot/vmlinuz-4.6.0-040600rc5-generic
run-parts: executing /etc/kernel/postinst.d/update-notifier 4.6.0-040600rc5-generic /boot/vmlinuz-4.6.0-040600rc5-generic
run-parts: ex...

[Andrew (keen101)]

Interesting. I doubt that is a coincidence. I am testing now with a standard hard drive with a fresh install of 16.04. I will report back soon whether the freezing problems go away or not. Hopefully this will shed some light on this issue.

But assuming that this might be the case, then one has to wonder if this bug is the result of some strange interaction of the i915 intel graphics driver and SSDs. Seems odd, but not impossible.

[Thomas (th1001)]

Perhaps it is helpful to report my experiences to this topic. My Hardware: Acer Travelmate B115 with Pentium N3540, Kingston SSD SV300S37A240G.
One month ago I updated the system from 14.04 to 16.04. I got two problems. The first was a non working Broadcom WLAN. But this was the simple solution: The STA driver failed to load, if secure boot was activated in bios. The second problem was this terrible freezing after short uptime. I found only the solution with the kernel parameter „intel_idle.max_cstate=1". This option worked, but the battery lifetime decreased appreciably. Now I read the last entries in this thread, so I decided to make some experiments. I updated the firmware of the SSD drive with the tool from Kingston by using Windows. Unfortunately I didn't note the old version of the firmware. Afterward I removed the cstate boot option. I also not forgot to call update_grub. But - since I‘ve done this, I saw no freeze anymore.

[Andrew (keen101)]

Interesting. i tried removing my SSD completely and putting in a normal
Hard Drive. The freezing problems did NOT go away despite my best hopes
that they would.

On Sun, Jul 17, 2016 at 9:11 AM, Thomas <email address hidden> wrote:

> Perhaps it is helpful to report my experiences to this topic. My Hardware:
> Acer Travelmate B115 with Pentium N3540, Kingston SSD SV300S37A240G.
> One month ago I updated the system from 14.04 to 16.04. I got two
> problems. The first was a non working Broadcom WLAN. But this was the
> simple solution: The STA driver failed to load, if secure boot was
> activated in bios. The second problem was this terrible freezing after
> short uptime. I found only the solution with the kernel parameter
> „intel_idle.max_cstate=1". This option worked, but the battery lifetime
> decreased appreciably. Now I read the last entries in this thread, so I
> decided to make some experiments. I updated the firmware of the SSD drive
> with the tool from Kingston by using Windows. Unfortunately I didn't note
> the old version of the firmware. Afterward I removed the cstate boot
> option. I also not forgot to call update_grub. But - since I‘ve done this,
> I saw no freeze anymore.
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/1575467
>
> Title:
> [Dell Inspiron 3451] WARNING: CPU: 1 PID: 1297 at /build/linux-lts-
> utopic-vy2yyy/linux-lts-
> utopic-3.16.0/drivers/gpu/drm/i915/intel_display.c:1743
> vlv_wait_port_ready+0x126/0x170 [i915]()
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1575467/+subscriptions
>

[Victor Kulikov (kulikovv)]

Hi!
I have recently bought ubuntu pc stick PCG02U from MeLE, and struggling with it since then.
It have z3735f atom processor and 32G eMMC drive.
Originally it was going with ubuntu desktop 14.04 and 3.16 kernel, but it had problems with some sporadic eMMC errors under high drive load.
Then I have upgraded it to ubuntu 14.04.4 with 3.19 kernel, and eMMC errors seems to be gone (not sure though), but now I faced freezes.
It have never freezed while being in use, but after I leave it running for about 6 hours or more, I always find is freezed then. Monitor doesn't show any signal (guess cause it gets freezed in sleep) and no reaction on keyboard and mouse options.
after reboot I see:
[ 7110.326234] RTL8723BS: nolinked power save enter
[ 7171.775617] RTL8723BS: nolinked power save leave
[ 7173.461854] RTL8723BS: nolinked power save enter
[ 7234.832616] RTL8723BS: nolinked power save leave
etc....
stops at some point of time, and normal boot messages then in kernel.log

Have tested 16.04 with original kernel then, and with upgraded to 4.7.0 kernel, all the same.

Trying intel_idle.max_cstate=1 option now but not much time have passed to see wether it helped. Will report it later.

[Victor Kulikov (kulikovv)]

intel_idle.max_cstate=1 didn't help - system hung after 3 hours of inactivity. Will try 4.1.29 which is the latest from 4.1.x

[Victor Kulikov (kulikovv)]

4.1.29 seems to be working, at least it didn't freeze for 2 days and still work now, but intel_idle.max_cstate=1 is enabled. I'll remove it and see will the system be stable.

[SirPecanGum (sirpecangum)]

Following the guide on https://wiki.ubuntu.com/Kernel/MainlineBuilds I have installed kernel 4.7.0-040700-generic. The system has currently been up for 3 days and 5 hours (suspended multiple times) without freezing. I've not installed Chrome yet.

I noticed the cursor would stutter just before a freeze on the default 16.04 kernel.

My device is an Asus E402MA (Intel 3540) with Xubuntu 16.04.

[ParkerMc (parkermc)]

I have the same bug with an Intel Pentium N3540. It seems to be freezing the most when I am running hard core programs. If there is any thing I can do to help let me know.

[yustnip (yustnip)]

May will be helpful - https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1490647

The similar old issue.

[Chris (blue-dragan+launchpad)]

I am hitting this same issue with an Asus E402MA with Intel N3540 running Ubuntu 16.04.

Setting intel_idle.max_cstate=1 does not help (confirmed it's active in kernel log), so there could be two different bugs there.

I have two ways in which the bug can be trivially and immediately reproduced:
1) One way to repro is to log in as one user, then switch to and log in as a second user.
2) Another way is to just launch Minecraft. It hangs immediately after loading a world.

Both of these repro cases happen immediately when performing these tasks.

The result is that the screen appears to be mostly frozen, the mouse cursor moves very little (unusable), the machine can be pinged from the outside, upon trying to log in over ssh it responds with a prompt but it's impossible to log in.

I think this is slightly different than in the case of the other issue related to intel_idle.max_cstate=1 (this WAR does not help anyway). In the case of the cstate issue, the hang is more rare, it still occurs relatively frequently when using Chrome on any website with GPU (HW) acceleration enabled. Turning off GPU acceleration in Chrome reduces repro rate of this type of hang. In this case when the hang occurs, mouse cursor is not movable. I haven't confirmed yet whether this issue goes away with intel_idle.max_cstate=1 or not (need more time).

[Chris (blue-dragan+launchpad)]

I fixed one of the issues I described above by reinstalling Ubuntu with a swap. Originally I installed it without a swap, with the rationale that this tiny laptop has only 32GB SSD, which is very little. I determined that vswapd was killing the CPU, that's why the mouse cursor was moving so slowly.

I reinstalled with full swap support. On top of that I installed zram-config and also reduced vm.swappiness to 10, which arguably seems more snappy.

So far it works like a charm. If I have other similar problems, I will report them here.

Note: I am still using intel_idle.cstate_max=1 in order to avoid random hangs. It would be nice if Intel got around to fixing this bug in their firmware...

[mattias (mattias0114)]

Running ubuntu 16.04 kernel 4.4.0-38-generic tryed The intel_idle.cstate_max=1
It dont work bug still active

[mattias (mattias0114)]

Sorry i typed The same massage twice

[N. W. (nw9165-3201)]

Update from:

https://bugzilla.kernel.org/show_bug.cgi?id=109051

Apparently Intel has released a PDF which explains a hardware bug:

Quote:

> VLP52. EOI Transactions May Not be Sent if Software Enters
> Core C6 During an Interrupt Service Routine
>
> Problem: If core C6 is entered after the start of an interrupt
> service routine but before a write to the APIC EOI (End of Interrupt)
> register, and the core is woken up by an event other than a fixed
> interrupt source the core may drop the EOI transaction the next time
> APIC EOI register is written and further interrupts from the same or
> lower priority level will be blocked.
>
> Implication: EOI transactions may be lost and interrupts may be blocked
> when core C6 is used during interrupt service routines.
>
> Workaround: It is possible for the firmware to contain a workaround for this
> erratum.
>
> Status: For the steppings affected, see “Summary Table of Changes”."

Sources:
http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/pentium-n3520-j2850-celeron-n2920-n2820-n2815-n2806-j1850-j1750-spec-update.pdf
https://bugzilla.kernel.org/show_bug.cgi?id=109051#c425

Meanwhile, someone has created a script for it which seems to work around the problem, see:

https://bugzilla.kernel.org/show_bug.cgi?id=109051#c437

Could this please be implemented into Ubuntu 16.10?

Regards

[Andrew (keen101)]

@jsalisbury

Just wondering if you have a fix for this yet. This bug is still actively affecting me. Since there has been no input from you or any of the other kernel developers or other ubuntu bug people for some time perhaps i should suggest that they all file new bug reports so that at least one person can get some attention and traction on this bug issue. It's a pretty bad bug to be present in an STL release. Since it seems to be a widespread kernel bug i guess it is not specifically an Ubuntu Kernel problem, so perhaps my frustration is a bit misplaced, but still any response that someone is paying attention would be nice.

I have tried all the available upstream kernels and while the freezing problems have gotten better they have not completely gone away. At one point we (was it me and you?) were doing a collaborative kernel bisection in hopes of narrowing down where the bug was introduced and/or how to fix the bug. You seem to have given up on that with no explanation. As of now i assume you gave up on this bug. Is that correct?

[Joseph Salisbury (jsalisbury)]

According to comment #139, this looks like a hardware bug. @Andrew, your suggesting of opening a new bug is probably best, so the different issues experienced by folks can be uniquely reviewed.

[Josep Pujadas-Jubany (jpujades)]

Acer TM (TravelMate) B115M N2930 processor (manufactured on Jully-2014)
Acer TM (TravelMate) B115M N2940 processor (manufactured on Jully-2015)
Acer TM (TravelMate) B117M N3150 processor using SSD (manufactured on Jully-2016)

Lubuntu 14.04 LTS + LTSEnablementStack
(https://wiki.ubuntu.com/Kernel/LTSEnablementStack)

We have about 110 units of each model. Some of them are freezing.

Tested 2 machines of each model during more than 72 hours. Playing Youtube videos with Chrome at full screen, playing with MineCraft (Java application), using the webcam with guvcview, running powertop, pinging our gateway, showing iwevent...

Kernel version 4.4.0-36-generic 64bit

Using c6off+c7on.sh (see comment #139) seems to solve the problem. However, N3150 is seen has CHT (Cherry Trail) processor, not BYT (Bay Trail) processor. So, for this case it's necessary to modify c6off+c7on.sh for cases C6*-CHT and C7*-CHT. cstateInfo.sh (see comment #139) shows all the state-names, if doubts about it.

N2930, N2940 --> Bay Trail Processors
(http://ark.intel.com/products/codename/55844/Bay-Trail?q=bay%20trail#@All)

N3150 --> Braswell Processor
(http://ark.intel.com/products/codename/66094/Braswell#@All)
But kernel code sees as Cherry Trail Processor (???)
(http://ark.intel.com/products/codename/46629/Cherry-Trail)

Upgraded the machines to latest stable kernel (4.8.10). No script applied. It also seems to solve the problem. Solution found at https://forum.manjaro.org/t/intel-bay-trail-freezes-the-linux-kernel/1931/12

For B117M, Acer technical services changed the SSD unit. But the system continued freezing. At comment #126 one user solved his problem updating the SSD firmware. Our 2 tested B117M seem to be stable with 4.8.10 kernel.

[Vincent Gerris (vgerris)]

thank you for sharing. I installed that kernel and again the builtin Broadcom wireless of my Lenovo Yoga 2 11 is not working, so I use a USB stick.
The performance is amazing!
I played 2 different videos with the soud going over bluetooth (HD profile) and copied a file and it all ran smoothly like windows can only dream of.

I removed the kernel option and rebooted and I do not have any freezes so far.

Looks promising. I will try and get the wireless to work and see what happens.
thank you

[Vincent Gerris (vgerris)]

Seems I forgot to update is. Unfortunately I still get freezes (N3520 Baytrail).
There are some promising developments that seem to actually fix the issue:
see:
https://patchwork.ozlabs.org/patch/704786/
and
https://bugzilla.kernel.org/attachment.cgi?id=241811

It would be great if people have time to test those.

I am in the process of building kernels with these patches to test and I will report when testing is done.

[Vincent Gerris (vgerris)]

Hi all,

It seems the patch actually fixes the free.
I runn 16.10 now with a patched 4.8.0-30 kernel and I have no freezes yet.

I attached the patch for the 4.8 series, there is a change in it.

Can you please try this on your Bay trail based processors?

The other link is for Cherry trail : to test start to patch from patch 1 and upwards to test (may need some manual fixing).

Tips:
REMOVE original 4.8.0-30 if you for example patch that according to https://wiki.ubuntu.com/Kernel/BuildYourOwnKernel (I didn't before and kept the freeze) so install for example 4.8.0-22 with synaptic and then completely remove ALL 4.8.0-30 packages.

Please test and report if this does or does not help you!
Thank you

[Vincent Gerris (vgerris)]

To facilitate testing, I put up the kernels I compiled on dropbox:
https://www.dropbox.com/sh/c39et4hr6tgp60q/AAC35c56aOEOwkmhjdvtG6dsa?dl=0

Mind that the 4.8.0-30 kernel seems to get updated by the default kernel of the same version.
If you want to test it, please remove all 4.8.0-30 based packages and install 4.8.0-22 as fallback.

Sources are from the Ubuntu git repo:
https://wiki.ubuntu.com/Kernel/Dev/KernelGitGuide

The 4.8 is yakkety, 4.9 zesty with the 4.9.0-12.11 tag

Note: these are only for Bay trail freeze issues. For Cherry trail, see my post about the kernel patches from Hans de Goede.

Can anyone with Bay trail based CPU's please test and give feedback?
For me this seems to fix it, when we get more people confirming we can get it into the main line kernel.
Thank you!

[Andrew (keen101)]

I guess i will attempt to test your patch since you are so invested in working on it. I honestly had just given up and decided to live with the freezing as i had heard it was more of a hardware bug.

I am now running Ubuntu 16.10 Yakkety Yak. The kernels have gotten better over time and i dont experience freezes nearly as much as i used to, but they are still there.

Currently i am running this kernel:
Linux dell-Inspiron-3451 4.8.0-30-generic #32-Ubuntu SMP Fri Dec 2 03:43:27 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux

I will try to do as you suggest and remove my current kernel and replace it with your dropbox patched kernel. I will report back when i can.

[Vincent Gerris (vgerris)]

Great! There are many reasons why that is important.
You are right, this is a hardware bug, you can read more on :
https://bugzilla.kernel.org/show_bug.cgi?id=109051

It should have been fixed in the 1.93 bios from Lenovo, but since they only support Windows, it may have been a combination of that and the driver, not sure.

Personally I think any freeze is unacceptable, I would rather use a kernel parameter.
But I find it very important that people who just install Linux do not encounter this bug, hence the big effort.
I like Linux a lot and despite the challenges, the power of open source and motivation of people that make it awesome.
The patch keeps the powermanagement better because c7 is still used.

I will await your feedback because the kernel folks that push upstream will need some testing to get it in to the mainline kernel. Thanks for your contribution!

[Andrew (keen101)]

Vincent, i just realized i never reported back! I've been using your patched kernel ever since my last reply. I don't think i have experienced even one freeze since then! Thank you for the patch! Yes, please work on getting this patch into the mainline kernel!

And everyone else please test this patch and see if it also works for you!!

-Andrew

[Vincent Gerris (vgerris)]

Hi Andrew,

That is great to read, thank you for reporting back!
In the mean time, this patch is in the discussion list with some kernel developers.
It may be that although the patch works, it is not the best way to work around the issue and it may not even have something to do with the errata reported.

In the mean time someone from intel added a patch here:
https://bugzilla.kernel.org/show_bug.cgi?id=109051
to re-enable auto-demotion.

At least one user reported success already.
I highly recommend everyone reading this to give that patch a try.
I will try to put a kernel up in the coming weekend that people can test.

If that is a fix, it is very likely to end up in the kernel main line.
Please share your experience and thank you :)
 - Vincent

[penalvch (penalvch)]

Andrew (keen101):

1) In order to better confirm applicability, could you please undo your WORKAROUND, and test the kernel parameter:
intel_idle.max_cstate=1

what is most helpful is you personally testing this (not relying on someone else, somewhere else, with some other/similar hardware).

2) If the above does indeed provide a WORKAROUND, as per Len Brown (upstream) he needs you to test a debug patch, and then use a script that increases the probability of invoking the bug:
1) Debug patch: https://bugzilla.kernel.org/attachment.cgi?id=248751
2) Script: https://bugzilla.kernel.org/attachment.cgi?id=248841

Could you please temporarily remove your WORKAROUND, and use this patch to provide the requested information?

[Andrew (keen101)]

ok, i suppose i could. Though i must say that i had avoided trying the intel_idle.max_cstate=1 as it seemed like a hack rather than a true fix, though i guess since the root of the problem is a hardware bug a hack is the best one can hope for.

I can remove my current patched kernel and go back to an older one, but how do i actually enable intel_idle.max_cstate=1 ??

And if it does work, how do i then compile the debug patch and install that? The links you provided just look like text files to me.

[penalvch (penalvch)]

Andrew (keen101):

>"I can remove my current patched kernel and go back to an older one, but how do i actually enable intel_idle.max_cstate=1 ??"

One may add this temporarily following https://wiki.ubuntu.com/DebuggingKernelBoot.

>"how do i then compile the debug patch and install that?"

One will undo all WORKAROUNDs and custom patches, and then install the debug patch following https://wiki.ubuntu.com/Kernel/KernelBisection#Testing_a_newly_released_patch_from_upstream.

[Andrew (keen101)]

Okay i have just removed Vincent's patched kernel. I will try and proceed and test Christopher's request to follow his steps to test his stuff.

Vincent i will add that your patch seemed to work well, though i did have it lock up once. Only once that i can remember though.

[Andrew (keen101)]

Christopher, the latest kernels don't freeze as much as the older ones. Do you want me to go back to an older kernel that froze up a lot to test the intel_idle.max_cstate=1 ?

[Andrew (keen101)]

Looking at the file Christopher linked to earlier states this:

"Organization: Intel Open Source Technology Center

The Baytrail SOC has a built-in feature to demote core
and module C6 to shallower idle states.

Based on work in the Android copy of the Linux kernel tree,
this feature was disabled starting in Linux-3.17-rc1:"

...So that looks promising as that is the exact thing we found in this bug report early on. Kernel 3.17 was the kernel i started getting huge freezes on, and when originally we were going through the bisecting rc1 was the one we had stopped at. So cool to see someone else has confirmed that portion.

...now to see if i can figure all this stuff Christopher wants me to do. Christopher i'm going to be honest and say that your really bad at explaining stuff since you just post links that are somewhat technical. I think i can figure it out, but my brain doesn't compute as fast as it used to. Right now i gather that i need to just edit my grub on bootup and test the intel_idle.max_cstate=1 parameter. (not sure if that will remain on future bootups or if that is persistent and will remain). Then if that works well i test the debug patch and script. (not sure if i need to disable intel_idle.max_cstate=1 when i do this).

[Vincent Gerris (vgerris)]

Hi Andrew,

Thank you for updating. Which kernel did you test, can you do a uname -a?
I supplied a kernel with both an auto-demotion patch and the C6 disable patch.

Christopher asks you to confirm if you are actually experiencing the exact bug that Len Brown is trying to fix : https://bugzilla.kernel.org/show_bug.cgi?id=109051.
That bug will be closed when a patch is found that will make the kernel stable on Baytrail without the kernel parameter.

To be sure you have this issue, we all would like to see if an unpatched kernel with that parameter gives you no freezes, then we know you have that issue and your system is valuable to test with.
If it still freezes, you may have another problem (one of them is audio so try to test without it).

To apply the kernel parameter, in a bash terminal do (replace vim with any editor if needed):
sudo vim /etc/default/grub
comment out the line with
GRUB_CMDLINE_LINUX_DEFAULT="quiet splash"
and copy it and add the parameter so you will have:
GRUB_CMDLINE_LINUX_DEFAULT="quiet splash intel_idle.max_cstate=1"

then save and exit and run:
sudo update-grub

That will regenerate the grub entries with the parameter added.

[Vincent Gerris (vgerris)]

I forgot to add, if your BIOS is out of date, please update it. It's important because some manufacturers made fixes there related to freeze issues. It's also required for testing to have the same results as others. Thank you

[Richard Baka (bakarichard91)]

There is an extremely non official patch for this problem that works very well for me. Not perfectly but well. It's worth a try. Please use it at your own risk.

I copy this from here: https://forum.manjaro.org/t/intel-bay-trail-freezes-the-linux-kernel/1931/10

https://github.com/TheDrHax/t100ta-scripts/blob/master/cstate/c6off%2Bc7on.sh

1. Drop the script at "/usr/bin/c6off+c7on.sh"
->sudo nano /usr/bin/c6off+c7on.sh
->Paste the content of the script
->ctrl+o

2. Then create a file "/etc/systemd/system/cstatefix.service"
->sudo nano /etc/systemd/system/cstatefix.service
->Paste the content
->ctrl+o
with following contents:

[Unit]
Description=My script

[Service]
ExecStart=/usr/bin/c6off+c7on.sh

[Install]
WantedBy=multi-user.target

3. Then enable service with command

sudo systemctl enable cstatefix.service

[Richard Baka (bakarichard91)]

4. You should remove cstate kernel parameter

[Andrew (keen101)]

Hi Vincent i have been updated to the latest DELL/Intel bios since this began. Did not help. Dell nor Intel ever fixed the issue. I will never buy a Dell Ubuntu product ever again. Probably not any dell product.though i realize this also affected HP and others who used the same chip. This laptop is now starting to fail hardware wise and it is near death so it doesent even matter. I gave up on it. I also switched this machine over to Windows 7. Windows 7 ironically does not freeze. The battery failed and now the screen is going out slowly. My next laptop will be a System76.

I would close this bug if i could without saying that it's not a true bug cause it is. So instead i will leave it forever in the state it is in as "incomplete"

[Richard Baka (bakarichard91)]

Andrew do you not try the patch?

[Andrew (keen101)]

No Richard. The machine is currently running Windows 7.

On Sun, Nov 26, 2017 at 11:20 PM, Richard Baka <email address hidden>
wrote:

> Andrew do you not try the patch?
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/1575467
>
> Title:
> [Dell Inspiron 3451] WARNING: CPU: 1 PID: 1297 at /build/linux-lts-
> utopic-vy2yyy/linux-lts-
> utopic-3.16.0/drivers/gpu/drm/i915/intel_display.c:1743
> vlv_wait_port_ready+0x126/0x170 [i915]()
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/ubuntu/+source/linux/+bug/
> 1575467/+subscriptions
>

[penalvch (penalvch)]

Won't Fix for now given issue correlated to Intel CPU errata, and final errata WORKAROUND patch is to come from, and being worked on by upstream kernel developers.