Dell Inspiron 7559 fans do not shut off once started, even when cool

[Occams Beard]

Running Ubuntu 16.04 on a Dell Inspiron 7559 laptop there appears to be an issue with the fan control. When I first boot up the machine the fans are off. However, after the laptop warms up under normal usage, the fans kick on and then never shut off again. Even when the CPU cools down to below 35c, the fans will continue to run.

If I suspend the the laptop, and then immediately wake it up again then fans are once again off, until it warms up again. So essentially, once the fans start they don't seem to ever shut off.

Fans seem to work normally under Windows 10. They turn on as needed, and then turn off once the laptop cools down.

I've tried the various BIOS updates that Dell has released, without success. I am currently running Dell's latest BIOS for this laptop (v1.1.8)

Below is the output of lm-sensors showing the fans running when the CPU is at a cool 35c

dell_smm-virtual-0
Adapter: Virtual device
Processor Fan: 1925 RPM
Processor Fan: 2143 RPM
CPU: +35.0°C
GPU: +1.0°C
SODIMM: +49.0°C

ProblemType: Bug
DistroRelease: Ubuntu 16.04
Package: linux-image-4.4.0-28-generic 4.4.0-28.47
ProcVersionSignature: Ubuntu 4.4.0-28.47-generic 4.4.13
Uname: Linux 4.4.0-28-generic x86_64
ApportVersion: 2.20.1-0ubuntu2.1
Architecture: amd64
AudioDevicesInUse:
 USER PID ACCESS COMMAND
 /dev/snd/controlC0: bob 1709 F.... pulseaudio
CurrentDesktop: Unity
Date: Wed Jul 13 19:18:45 2016
HibernationDevice: RESUME=UUID=14de580b-76f3-47e4-9850-fe7637f9ff77
InstallationDate: Installed on 2016-07-13 (0 days ago)
InstallationMedia: Ubuntu 16.04 LTS "Xenial Xerus" - Release amd64 (20160420.1)
MachineType: Dell Inc. Inspiron 7559
ProcFB: 0 inteldrmfb
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-4.4.0-28-generic.efi.signed root=UUID=0b3dbcf3-0bdb-4c35-9200-7bb303055a81 ro quiet splash vt.handoff=7
RelatedPackageVersions:
 linux-restricted-modules-4.4.0-28-generic N/A
 linux-backports-modules-4.4.0-28-generic N/A
 linux-firmware 1.157.2
SourcePackage: linux
UpgradeStatus: No upgrade log present (probably fresh install)
dmi.bios.date: 04/17/2016
dmi.bios.vendor: Dell Inc.
dmi.bios.version: 1.1.8
dmi.board.name: 0H0CC0
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.:bvr1.1.8:bd04/17/2016:svnDellInc.:pnInspiron7559:pvr1.1.8:rvnDellInc.:rn0H0CC0:rvrA00:cvnDellInc.:ct8:cvrNotSpecified:
dmi.product.name: Inspiron 7559
dmi.product.version: 1.1.8
dmi.sys.vendor: Dell Inc.

[Brad Figg (brad-figg)]

This change was made by a bot.

[Max (splokow)]

A lot of Dell Inspiron 7559 users are facing this annoying bug.
It doesn't matter if we're using the Intel Graphic card or the Nvidia one.
I have the problem under Linux Mint and Kubuntu.

[penalvch (penalvch)]

Occams Beard, thank you for reporting this and helping make Ubuntu better.

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.

[Occams Beard (d-bob-1)]

After installing the v4.7-rc7 mainline build kernel I can confirm that this bug still exists.

I installed the following 3 debs with sudo dpkg -i *.deb:

linux-headers-4.7.0-040700rc7_4.7.0-040700rc7.201607110032_all.deb
linux-headers-4.7.0-040700rc7-generic_4.7.0-040700rc7.201607110032_amd64.deb
linux-image-4.7.0-040700rc7-generic_4.7.0-040700rc7.201607110032_amd64.deb

Next, I rebooted the laptop and confirmed I was running on 4.7 with uname -r:
4.7.0-040700rc7-generic

I used the laptop for a few minutes until the fans came on. The fans continued to run, and are still running 30 minutes later. Here's the output of lm-sensors showing that the cpu is cool, yet the fans are spinning away.

dell_smm-virtual-0
Adapter: Virtual device
Processor Fan: 1935 RPM
Processor Fan: 2134 RPM
CPU: +33.0°C
GPU: +1.0°C
SODIMM: +45.0°C

Other thoughts - I don't know if this is relevant to the fan issue, but in powertop under the "idle stats" tab, the CPU package never gets past C2. This is the case with all kernels I've tried. Perhaps the fan issue is related to the screwy skylake power management? I've attached a text file of the powertop idle stats with the 4.7 kernel.

[Peter (peter-biely)]

Guys, by installing kernel 4.8-RC2 I've noticed (until I've rebooted) that fans were working fine, once the CPU was heated up they started to spin and then once cooled down to about 41C they stopped. Only one thing I've noticed in psensor utility was that the GPU temperature min/current/max was still 1C! I was running on Intel GPU and powertop package state was only C2.

So there pops up the question:

Is there a possibility that the fans problem is related to sensor from Intel/nVidia GPU? Unfortunately I am not able to reproduce this again so I would take at least a screenshot.
Because after reboot the issue is presented on Kernel 4.8.0-040800rc2 and PSensor shows correct GPU temperature again.

http://imgur.com/pCSfqUi

[Occams Beard (d-bob-1)]

I've noticed that without the nvidia driver installed the GPU temp sensor always reads 1c. After the nvidia driver is installed it reports the correct temp. I think probably not related to the fan issue. Also it seems strange that the fans stopping didn't survive a reboot. I've noticed that if the laptop is cool sometimes the fans will stop on their own. But this happens very very rarely, and only if the ambient temp is cool as well.

The best explanation/theory I've heard for the fan issue is #14 on this thread:
https://answers.launchpad.net/ubuntu-certification/+question/293705

In any case I think this bug needs to be reported upstream to the Linux kernel team. I'm not sure where or to which maintainer though. I've had to resort to using Windows 10 with cygwin (Yuck) and unfortunately had to remove Ubuntu to free up drive space :(

[Peter (peter-biely)]

I've created a bug for kernel (hopefully correctly)
https://bugzilla.kernel.org/show_bug.cgi?id=153281

[Z.H. (zeehaa)]

Perhaps a workaround: try to install i8kutils!
It worked for me, but I'm not sure if it is a real solution.
(after installing that package, both fans stopped, but they not started when CPU's temperature were over 60C...)

[Peter (peter-biely)]

Already tried the i8kutils, but this is just workaround, not a real solution for this problem. Also I see that however the CPU is on kernel 4.8 quite effective with power management the whole package can't reach lower than C3 state.

[Petr Soukup (soukup77)]

Dell has introduced a successor of Inspiron 7559 - Inspiron 7566 at the end of October 2016 - http://www.dell.com/sg/p/inspiron-15-7566-laptop/pd?oc=w51755724sgw10 .It has got the same CPU, GPU and display but comes with DDR4 RAM and SSD disk. It also has a new look, much uglier than 7559 in my opinion ☺, but in the end something I can probably live with ... Since it has been around for only a few days, there is virtually no information on how it works with linux, so I guess I can only extrapolate whether it's going to have the same issues as the 7559 and potetially higher risks since it has not been tested. Nevertheless, I welcome all your inputs and tips ...

[Peter (peter-biely)]

A friend of mine he has i5 model and he claims that the fans are shutting off, but not as good (often) as on Windows. On the other hand, the i7 model has this fans problem, see my bug I've created on bugzilla. Also battery lasts less on Linux than on Windows. From mine point of view, buying this model was a big mistake for Linux usage.

[Occams Beard (d-bob-1)]

Disabling nvidia with bbswitch makes no difference for me. Fans keep on spinning. I have an i7.

I am not convinced the fan issue even has anything to do with the graphics.

[Occams Beard (d-bob-1)]

> Dell has introduced a successor of Inspiron 7559 - Inspiron 7566

Stay far away from it if you are a Linux user. It probably has the same crappy Dell bios/ACPI. Also Dell is no help and is generally clueless about everything.

Actually, I would advise staying away from any nvidia optimus laptop. Linux support is a kludge at best.

[Cuate (enayunas)]

Has anyone tried changing the EFI .aml files to try and fix some of the issues in ubuntu?

I saw a hackingtosh tutorial for el capital (which is an os that was not made for skylake) on 7559 i7 UHD where they modify efi .aml files to make it work. The result was a i7 7559 uhd laptop running macos with the same battery time as windows and also with the same fan behaviour as windows.

In the tutorial they disabled the nvidia card (which I don't need in linux) and they also made some patches to for the power management and procesor turbo modes all in the .aml files.

I understabd it is posible to downlaod the efi files from the bios, uncompile them and modify them in linux, and to test you can use the e option in grub to indicate wehre the new file that has the modified aml is located.

Unfortunately I have no time to play with this setting to see if I can get battery life and the fan under control, but if anyone can it would be great.

[Occams Beard (d-bob-1)]

cat /proc/acpi/bbswitch says the discrete nvidia card is OFF.

Fans are still running.

I do not use bumblebee/optirun, only bbswitch. I will try bumblebee when i get some free time and see if that makes any difference.

[Petr Soukup (soukup77)]

Thank you so much for your inputs, everyone ...

> If Nvidia is properly shutted down, the only big issue lasted is a bad Skylake power management in Linux (going no deeper than package c-state 3), but then almost any Skylake laptop is not the best choice.

Wasn't Skylake bad power management supposed to be fixed with kernel 4.8? I'm not saying it is, but I saw the announcement by the mighty Linus himself quoted and reposted everywhere ...

> Stay far away from it if you are a Linux user. It probably has the same crappy Dell bios/ACPI. Also Dell is no help and is generally clueless about everything.

Well, that's my dilema - I've read about the issues and the shitty Dell support, on the other hand 7566 is a real price breaker - i7 with 8 threads, nVidia 960 4gb,8gb ddr4, 1920x1080 display and 128 gb ssd + another 500gb hdd for the price of 1100$, you won't find much competion there, at least not where I live, and I want to also use it for occasional Win gaming, although I can't stand Win for more than game launching. I could get the same configuration Linux certified, but it's about extra 200$, and I would have to ship it from Germany as there is only a few shops that sell Linux certified laptops around Europe, I live in Prague, and it's not exactly a manufacturer with a shop around the corner, so in case of warranty repair it's going back to Germany. With the Dell, I can buy in my favorite internet computer store. Then some people say it sucks with Linux, others say it actually works pretty fine for them, especially with the latest kernel version. Most of the issues looks solvable, more or less, except of the fan issue, but again, some people don't experience it, one guy was actually claimimg the other day that 7559 worked for him almost out of the box with Debian-testing, no issues whatsoever, so I'm kind off being torn appart. In the end, it's a little gamble, I guess ...

Alex, how is your battery life?

[Occams Beard (d-bob-1)]

> one guy was actually claimimg the other day that 7559 worked for him almost out of the box with Debian-testing

"Working" can be very subjective depending on who you ask. This bug report and the kernel bugzilla report show that many users are having the fan issue.

In any case, let's not clutter up this bug report with issues/opinions/etc not related to the fan bug.

[turoyo dee (lpturoyo)]

i'm also experiencing this issue
 - only running on integrated graphics

getting 3.5-4 hrs. average battery life and hoping to get more if and when this issue is fixed.

[Petr Soukup (soukup77)]

Occams Beard, have you tested the fan issue with kernel 4.8? Here, you only mention kernel 4.7, but 4.8 was supposed to fix Skylake power management issues. I have noticed people having fan issues on other laptop models with Skylake, mostly i7 6700hq, (Asus, Lenovo), as well in Linux, so this points to the fans problem being a broader issue. On the other hand I see people here and the other kernel thread reporting that 4.8 alone has not fixed the problem for them (I have seen 2 so far). It would be nice to have more data in order to be able to clearly distinguish whether the 7759 fan issue is a Skylake problem alone, or if there is something else in play. I haven't found any reports from those other folks outside the 7759 linux community who previously had fan issues too, to see whether the 4.8 has had any improvement for them yet. 4.8 has only been out for about a month, so I guess there is not enough data ...

[Petr Soukup (soukup77)]

Correction, I mean 7559,not 7759 ... sorry

[Peter (peter-biely)]

I've read (but I do not remember where) that Lenovo had the same problem with the fans on some models with 6700HQ CPU, but they released BIOS update which resolved this problem, so it seems like a problem of ACPI/BIOS (maybe I am just wrong and the problem is really somewhere else). On the other hand XPS 15 models seems working correctly so I'd blame DELL incompetency (or lack of willing) to fix this problem.

[turoyo dee (lpturoyo)]

additional info:

running on integrated graphics HD 530
with 16.04 LTS + 4.4 kernel

I just updated to BIOS version: 1.2.0 [released: 31 Oct 2016]

issue persists.
[I will install and test 16.10+4.8 next week and post results]

[Peter (peter-biely)]

Fan issue persists on 4.8.6 and 4.9-RC3 with BIOS 1.2.0.

[filsd (filsdd)]

Hi!

I have this Laptop (i5 model) and I bought it from Dell (Brazil) already with Ubuntu (14.04) certified.
Using the Dell tools I can make an ISO of this version.

I have been using this version of Ubuntu for months now and it doesn't have any problems. ANY. (I even updated the Nvidia drivers using a PPA, it works!).

Are any of you interested in this ISO? I can upload it for you to download (using wetransfer or something). It could help to find the real issues.

[Peter (peter-biely)]

count me in for the iso!

[Occams Beard (d-bob-1)]

Count me in as well, I likely won't have time this week to try it, but I would be very interested to know what Dell does differently with their version of Ubuntu

[Petr Soukup (soukup77)]

Just to clarify this for me, because I'm getting kind of lost in reading all the posts, both Ubuntu developers in this thread, and kernel developers in the kernel thread where this issue is reported are more or less blaming this on Dell firmware, am I getting this correctly?

Concerning the Dell certified version of Ubuntu - Dell does not allow other users whose machine did not originally come with it to download, i.e. making it public to download, am I correct?

I also have a question about the i8kutils - does it remedy the issue completely without side effects and is it dangerous to use it? I mean, it's not vendor certified software and it meddles with temperature settings for the fans, so if I really screw up the settings, can I potentially fry my CPU?

[Peter (peter-biely)]

Petr:
ad 1: Possibly yes
ad 2: yes (I haven't seen anywhere the ISO for download)
ad 3: i8kutils seems working, I've tried that. But yes, in case of wrong temp. setting you can damage your CPU/GPU, I'm using it as workaround but I am not satisfied (not sleeping very well) by using that.

[filsd (filsdd)]

Would it be illegal to release it??
Its open source anyway. :/

I imagined that Dell wound not "guarantee" that would work or give any support to those other users. (But we know it works)

[turoyo dee (lpturoyo)]

peter-biely, could you share your i8kutils settings?

[Peter (peter-biely)]

This is what I am using, but I am not taking any responsibility of damage caused on your laptop. :) Also you can play with the temps to get the fan speeds to you satisfaction...

# /etc/i8kmon.conf
# Run as daemon, override with --daemon option
set config(daemon) 0

# Automatic fan control, override with --auto option
set config(auto) 1

# Report status on stdout, override with --verbose option
set config(verbose) 1

# Status check timeout (seconds), override with --timeout option
set config(timeout) 20

# For computer with 2 fans, use a variant of this instead:
# Temperature thresholds: {fan_speeds low_ac high_ac low_batt high_batt}
set config(0) {{0 0} -1 55 -1 55}
set config(1) {{1 1} 36 65 36 65}
set config(2) {{1 1} 50 75 50 75}
set config(3) {{2 2} 70 128 70 128}

# end of file

[turoyo dee (lpturoyo)]

thanks, I'm still determined to try a few more things.

If nothing works, I have no choice but to try i8kutils anyway. And I feel better having a reference config that works from someone with the same machine.

[Petr Soukup (soukup77)]

Peter, thank for clarifying things for me, I appreciate it.
Also, have you tried the http://linux.dell.com/ ? It's kind of unofficial dell Linux support site. Or so they say ...

>> Would it be illegal to release it?? Its open source anyway. :/
Sure, Ubuntu is open source, not sure if you can take parts of Ubuntu, change it and claim it proprietary. After all, I never read the license ... :) Anyway, I don't think they would make much fuss about it. So, just give it to us, please ... ;)

[Wei Sun (562241377-6)]

same issue. Annoying

[Julo Petko (julo.petko)]

I do not understand why the bug is closed. I have Ubuntu 17.04 with the latest updates on my Dell Inspiron 7559 with the latest BIOS 1.2.2 and the problem still persists. Fans are still on, but the CPU temperature is about 30deg. Celsius. If others having the same problem, we should reopen this bug.

[Occams Beard (d-bob-1)]

I've changed the status of the bug to "Confirmed". It lets me change the status for some reason. I guess because I originally submitted the bug.

A while back someone from canonical set it as "Fix Released," but like everyone else, I still have the bug with latest versions of everything. So I have no idea what the supposed fix is, or even if there is one.

For everyone else with this bug, the better bug report is probably here on the Linux Kernel bug tracker: https://bugzilla.kernel.org/show_bug.cgi?id=153281

[Davi Dagostin (zankuro)]

I'm on Kubuntu 17.04 and this bug persists on my Dell 7759 with bios 1.2.2. When fan is turned on never more is turned off.

[penalvch (penalvch)]

Occams Beard, as per your comment https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1602888/comments/73 the issue "fans do not shut off once started, even when cool" was resolved. Did you reopen this report because it stopped happening after an update? If so, it's best to file a new report targetting a regression after a fix, versus reopen a closed report.

[Occams Beard (d-bob-1)]

I do not consider it resolved. In my comment I said there was *some* improvement after making many archaic tweaks. However it's still not working right even after making those tweaks. I'm not sure why the status was set to "Fix Released" as I do not think there ever was a "fix"

[Julo Petko (julo.petko)]

I can confirm, the bug still persists even with the most recent kernel 4.11.0-041100 and the most recent BIOS 1.2.2. It's pretty annoying since it reduces my battery life a lot compared to Windows and also the fan noise disturbs during the evenings when I am leaving my laptop idle.

[Taulant Y (taulant01)]

Guys, I erased the old Ubuntu (16.04) and installed Ubuntu 17.10. I am currently setting up the environment, installing many things, and until now there isnt any fan noise. I think they fixed the bug.... I will report later again if the problem with the fan noise appears.

[Minas Karamanis (minaskar)]

Good to hear that. How is your battery life?

[Occams Beard (d-bob-1)]

I'm also interested in hearing if the fan issue is solved on 17.10. Please keep us updated.

I think I've found the reason the fans turn on, and I believe I found a work-around under Arch Linux.

After a good bit of research, I think the issue is caused by the CPU package not being able to get to lower power states. Under powertop, you'll notice that this laptop never gets below C3. Thereby causing increased CPU heat and poor battery life.

However, I've found the reason for this is the realtek network card driver not enabling Active State Power Management (ASPM). Setting most tunables to GOOD under powertop and replacing r8169 kernel module with this patched one https://github.com/KastB/r8169 allows the CPU package to reach C7 low power state. This -appears- to solve the fan problem. The fan still runs, but much less frequently. The laptop also seems to run cooler and improved battery life (Although still not as good as under Windows).

Arch Linux users can install this patched module from the AUR https://aur.archlinux.org/packages/r8169aspm-dkms/

I'd still be interested to know if this is fixed on Ubuntu.

[Julo Petko (julo.petko)]

No change with fans on Ubuntu 17.10 for me. By default lowest state for package is C2. After setting all tunables to good C3, but does not solve the issue though.

[Occams Beard (d-bob-1)]

Give the patched r8169 driver a try:
https://github.com/KastB/r8169

You should be able to reach C7, and the fan issue should be improved.

[Kai-Heng Feng (kaihengfeng)]

Hopefully the author will send patch to upstream.

https://github.com/KastB/r8169/issues/5

[Minas Karamanis (minaskar)]

Does the patch work in your case Kai-Heng Feng?

[Julo Petko (julo.petko)]

Please update the bug as the bios was updated but it did not resolve the issue.
Also I've tried to compile patched driver according the suggested how to from Occams Beard, but it did not help - maybe I did something wrong.

[Taulant Y (taulant01)]

I apologize for before. I thought it was fixed.... but its not fixed :(. The fans didn't turn on for a while, but today the problem occurred again :(

[Occams Beard (d-bob-1)]

Julo Petko,

Strange, the patched driver works perfectly for me. I'm on kernel 4.13.8, maybe that has something to do with it.

Also, even with the patched driver you MUST set the tunables to GOOD in powertop, otherwise the CPU package won't get to C7.

Another thing to check - did you do a cold boot after you installed the patched driver? E.g power off the laptop completely, and then power it back on? Apparently, this must be done according to the github page (If I understand it correctly)

[Julo Petko (julo.petko)]

Even shutdown of laptop didn't help after compiling the driver. Maybe I'm doing something wrong. Tunables set all to good but the package lowest state is C3. I'll wait for the upstream patch. I could wait for almost 2 year anyway, so couple of months I can survive.

[Bryce Frank (bfrank70)]

A tentative success story:

Using the patched r8169 driver & powertop tunables set to "GOOD", my CPU was able to reach the C7 package state. I noticed C7 is only reached when the laptop is unplugged, otherwise only C3 is reached. I am not sure if this behavior is expected, but may be leading to the confusion above.

CPU temperatures seem to idle around 40-60 degrees celsius. This is slightly higher than Windows 10 (where I checked it at about 40-50 degrees celsius). The fan kicks on every once in a while, but goes away as expected ~30 seconds later.

I am using Ubuntu 17.10 with the 4.13.0-16-generic kernel.

[Minas Karamanis (minaskar)]

Bryce Frank,

How was your power consumption when idle and how is your battery life?

Julo Petko, I guess you also need to upgrade to 17.10.

[Occams Beard (d-bob-1)]

I'm able to reach C7 even if the laptop is plugged in.

Although I'm running arch linux here. I also have the acpi_osi="!Windows 2015" boot parameter as mentioned earlier. Not sure if that matters.

[Bryce Frank (bfrank70)]

Minas,

Idle power consumption seems a little higher than Windows 10. After idling for 5 minutes on a full charge, the laptop reports temperatures < 50 degrees celsius, and a projected battery life of 3 hours and 25 minutes w/screen brightness of about 70%. This seems a little low, but not unreasonable.

Admittedly, I don't pay attention much to my battery life on Windows 10, but I can perform a similar test later this weekend for something to compare to.

At idle, powertop reports (under Package):
C2 - ~20%
C7 - ~40%

All others at 0.

In conclusion, power consumption seems normal at idle.

[Bryce Frank (bfrank70)]

After some more experimentation it is clear that I cannot reach C7 power states when plugged in (the lowest I reach is C3), even with tunables set to good. I tried the acpi_osi="!Windows 2015" boot parameter with no luck.

[Occams Beard (d-bob-1)]

With the patched r8169 driver Everything seems to be working properly on Arch when plugged in, so it must be something Ubuntu specific.

I am using bumblebee/bbswitch to disable the nvidia gpu. Where as Ubuntu uses something called "prime-select" to switch between GPUs.

I am using kernel 4.13.11. Not sure what kernel Ubuntu includes.

I think Ubuntu has some sort of "laptop mode tools," maybe this affects power management some how?

[Kai-Heng Feng (kaihengfeng)]

Please try the kernel, I enabled ALDPS and ASPM for r8169.

http://people.canonical.com/~khfeng/lp1602888/

[Julo Petko (julo.petko)]

@Kai-Heng Feng (kaihengfeng): tried your kernel and no change, even with all tunables set to good I got only PCakage C7 state. This machine seems to be enchanted or something.

[Occams Beard (d-bob-1)]

Package C7 is the best I can achive as well with the patched r8169 driver & all tunables set to GOOD. But the fan issue seems to be solved for me with it. The fan rarely even comes on, when it does, it shuts off a few seconds later, as expected.

[Kai-Heng Feng (kaihengfeng)]

Before I resend the patch to reinstate ALDPS and ASPM for r8169, I want to make sure the change bits I make can work.
The C7 status is exactly what I was looking for.

[Kai-Heng Feng (kaihengfeng)]

Hi, can you guys try the kernel [1] and attach dmesg here?

[1] http://people.canonical.com/~khfeng/lp1602888-8168g-aspm/

[Julo Petko (julo.petko)]

Here is the dmesg, fans are running even with tunables all set to good.
Also I had to set i915.modeset=1 and nouveau.modeset=0 to be able to boot this kernel.

[Julo Petko (julo.petko)]

Update: fans were switched off after a while, but powertop tunables still says package sate only C3

[Kai-Heng Feng (kaihengfeng)]

So the old one can achieve C7 but not the new one? They are pretty much the same.

[Bryce Frank (bfrank70)]

I was not able to achieve C7 (pc7) with your old kernel either (I just tried a few days ago). The new kernel does not achieve C7 (pc7) either. It gets down to C3 at best

[Occams Beard (d-bob-1)]

Happy to supply any info that might help. On Arch based Antergos with the patched realtek driver, and powertop tunables set to good, CPU package reaches C7 and fan issue is resolved.

Only other difference I can think of is that I have acpi_osi="!Windows 2015" as a kernel boot parameter.

[Julo Petko (julo.petko)]

Boot with acpi_osi="!Windows 2015" and all tunables set manually to good, powertop still shows lowest package state C3 but fans turned off properly. Attached new dmesg output anyway, but problem seems to be "workarounded" (not resolved) as I have to set tunables to good manually or via script.

[Kai-Heng Feng (kaihengfeng)]

Occams,

Can you try if the Linux source tree here works for you?
https://github.com/khfeng/linux/tree/r8169

[Kai-Heng Feng (kaihengfeng)]

Tried APSM enable r8169, the deepest package C state is C3.

The good news is, out-of-tree r8168-dkms can make it to C7.

Before Realtek folks put ASPM back to mainline's r8169, I think use r8168-dkms is the best compromise here.

Just do `sudo apt install r8168-dkms`.

[Bryce Frank (bfrank70)]

Installing r8168 gets me to C7 package state. Thanks for the help kaihengfeng, hopefully a more proper solution is on its way!

[oguz kagan (oguzkagan)]

I use Manjaro. Two days ago, I installed kernel 4.16. The fan is more stable on the kernel, almost like on windows 10.

[Minas Karamanis (minaskar)]

Sounds great! Does your package reach C7 when idle?