Kernel 3.6.2 Power Regressions (Sandy Bridge)

koz · 2012-11-25 07:02:10

I am no longer experiencing overheating since latest update 3.6.7-1?

mabl · 2012-11-25 10:55:24

koz wrote:

I am no longer experiencing overheating since latest update 3.6.7-1?

I still see it with 3.6.7-1-ARCH...

thesofty · 2012-11-25 16:41:11

spinacz wrote:

thesofty wrote:
I'm also affected by the sandy bridge - problem on my new lenovo X1.
Because it was a 1'st installation on the Laptop downgrading was not an option for me.
You can downgrading system with this repo . After downgrade i added this line to pacman.conf
IgnorePkg   = linux linux-headers linux-api-headers
now i am waiting for solution.
More you can read here

Thank you very much for the info.

I think, I'll keep my patched 3.6.6-kernel the next time. It's not perfect, but it works for me so far.

1126 · 2012-11-26 12:51:44

mabl wrote:

koz wrote:
I am no longer experiencing overheating since latest update 3.6.7-1?
I still see it with 3.6.7-1-ARCH...

I do, too.

Booting, starting firefox and writing this post heats my (docked) Thinkpad X220 up to 69°C.

I'll stick with patched mainline for a while, I guess.

EDIT: Rebooting with mainline, starting firefox and writing this post: 48°C.

Last edited by 1126 (2012-11-26 12:54:39)

KaiSforza · 2012-11-26 13:50:48

spinacz wrote:

thesofty wrote:
I'm also affected by the sandy bridge - problem on my new lenovo X1.
Because it was a 1'st installation on the Laptop downgrading was not an option for me.
You can downgrading system with this repo . After downgrade i added this line to pacman.conf
IgnorePkg   = linux linux-headers linux-api-headers
now i am waiting for solution.
More you can read here

Be warned, you may need to add more packages to the ignorepkg list. If you are using packages that depend on a certain kernel version, a `pacman -Syu` will fail every time. You will also need to download these extra packages.

Good luck.

(You may also start to see a lack of features in systemd as it starts to rely on more modern kernels for certain features. This should not break systemd, just not let you use the newest and shiniest features, possibly.)

altercation · 2012-11-26 17:10:27

koz wrote:

I am no longer experiencing overheating since latest update 3.6.7-1?

Same here. Knock on wood.

LordChaos73 · 2012-11-26 20:17:27

Finally found the bug I'm hitting - https://bugs.freedesktop.org/show_bug.cgi?id=54089
Looks like the fix is already in 3.7.0-rc6.

KaiSforza · 2012-11-26 21:57:37

LordChaos73 wrote:

Finally found the bug I'm hitting - https://bugs.freedesktop.org/show_bug.cgi?id=54089
Looks like the fix is already in 3.7.0-rc6.

Just a note, that was something from a drm-intel branch, not the official linux-mainline branch. (My guess is it'll make it's way in sooner rather than later.)

Using 3.7.0-rc6 from Linus won't fix this yet, sadly. I've quickly looked over the commit log and nothing is popping out at me. Correct me if I'm being retarded.

LordChaos73 · 2012-11-26 22:03:59

KaiSforza wrote:

Just a note, that was something from a drm-intel branch, not the official linux-mainline branch. (My guess is it'll make it's way in sooner rather than later.)
Using 3.7.0-rc6 from Linus won't fix this yet, sadly. I've quickly looked over the commit log and nothing is popping out at me. Correct me if I'm being retarded.

You're definitely not retarded, looks like I am for not reading correctly. However, I also found out I can't reproduce the GPU issues in that report, just the symptoms only.

KaiSforza · 2012-11-26 22:32:21

LordChaos73 wrote:

KaiSforza wrote:
Just a note, that was something from a drm-intel branch, not the official linux-mainline branch. (My guess is it'll make it's way in sooner rather than later.)
Using 3.7.0-rc6 from Linus won't fix this yet, sadly. I've quickly looked over the commit log and nothing is popping out at me. Correct me if I'm being retarded.
You're definitely not retarded, looks like I am for not reading correctly. However, I also found out I can't reproduce the GPU issues in that report, just the symptoms only.

That's what I've noticed, as well. I've only intermittently gotten the cpu scaling issue, which could possibly be tied in, but I haven't seen anything supporting that.

LordChaos73 · 2012-11-27 09:45:30

KaiSforza wrote:

LordChaos73 wrote:
KaiSforza wrote:
Just a note, that was something from a drm-intel branch, not the official linux-mainline branch. (My guess is it'll make it's way in sooner rather than later.)
Using 3.7.0-rc6 from Linus won't fix this yet, sadly. I've quickly looked over the commit log and nothing is popping out at me. Correct me if I'm being retarded.
You're definitely not retarded, looks like I am for not reading correctly. However, I also found out I can't reproduce the GPU issues in that report, just the symptoms only.
That's what I've noticed, as well. I've only intermittently gotten the cpu scaling issue, which could possibly be tied in, but I haven't seen anything supporting that.

This is getting quite frustrating, I really can't explain why my laptop is using ~5W more after a suspend/resume cycle. Powertop is showing my ethernet card is supposedly using more than 10W, but that looks like a bug. And as I said before, I can't reproduce the GPU issues. I have tested loads of different kernels so I'm thinking this may not be kernel related at all.

eduedix · 2012-11-27 19:23:09

well yeah, according to powertop my iwlwifi is also consuming like hell around 7-10 W

replax · 2012-11-27 21:38:52

curiously, it is my Ethernet adapter which seems to be using a lot of power, too. according to powertop at least. i am not even using the eth0 interface and no cable is attached, nevertheless it appears to be using more than 2W on average, only second in consumption to the LCD back light.

LordChaos73 · 2012-11-27 22:02:25

replax wrote:

curiously, it is my Ethernet adapter which seems to be using a lot of power, too. according to powertop at least. i am not even using the eth0 interface and no cable is attached, nevertheless it appears to be using more than 2W on average, only second in consumption to the LCD back light.

Well, have a look at the first few lines of powertop output on my laptop. I'm not using the eth0 interface either:

The battery reports a discharge rate of 9.83 W
The estimated remaining time is 4 hours, 28 minutes

Summary: 128.6 wakeups/second,  0.4 GPU ops/seconds, 0.0 VFS ops/sec and 1.5% CPU use

Power est.              Usage       Events/s    Category       Description
  10.0 W      0.0 pkts/s                Device         Network interface: eth0 (r8169)
  466 mW      1.0 ms/s      43.6        Interrupt      [41] i915
  130 mW    266.3 µs/s      12.2        Timer          tick_sched_timer
  125 mW      1.2 ms/s      11.4        Interrupt      [6] tasklet(softirq)
  118 mW    100.0%                      Device         Radio device: Sony Notebook Control Driver

eduedix · 2012-11-27 22:15:10

i can never get around 9.83W though, was chromium running when you had this LordChaos73?

LordChaos73 · 2012-11-27 22:22:20

eduedix wrote:

i can never get around 9.83W though, was chromium running when you had this LordChaos73?

The battery reports a discharge rate of 10.3 W
The estimated remaining time is 3 hours, 34 minutes

Summary: 81.6 wakeups/second,  0.1 GPU ops/seconds, 0.0 VFS ops/sec and 1.1% CPU use

Power est.              Usage       Events/s    Category       Description
  10.1 W      0.0 pkts/s                Device         Network interface: eth0 (r8169)
  147 mW    284.2 µs/s      13.9        Interrupt      [41] i915
  123 mW    249.1 µs/s      11.7        Timer          tick_sched_timer
  110 mW      0.8 ms/s      10.2        Interrupt      [6] tasklet(softirq)
  105 mW    252.9 µs/s       9.9        Process        [rtsx-polling]
 52.7 mW    119.6 µs/s       5.0        kWork          ieee80211_iface_work
 50.5 mW      2.4 ms/s       3.9        Process        /usr/lib/chromium/chromium
 33.3 mW     86.7 µs/s       3.1        Timer          hrtimer_wakeup
 18.6 mW    193.5 µs/s       1.7        Process        /usr/lib/upower/upowerd
 17.9 mW     24.8 µs/s       1.7        Timer          ehci_hrtimer_func
 17.6 mW     92.0 µs/s       1.6        Process        kdeinit4: krunner [kdeinit]
 15.8 mW     34.2 µs/s       1.5        Process        [usb-storage]
 15.1 mW      0.0 µs/s       1.4        kWork          disk_events_workfn
 14.5 mW      1.1 ms/s       0.9        Process        kdeinit4: konsole [kdeinit]
 13.9 mW    325.0 µs/s       1.2        Process        kdeinit4: kded4 [kdeinit]
 13.7 mW     31.4 µs/s       1.3        Interrupt      [20] ehci_hcd:usb4
 11.8 mW    209.2 µs/s       1.0        Interrupt      [7] sched(softirq)
 11.3 mW     83.1 µs/s       1.0        Process        /usr/bin/ntpd -g -u ntp:ntp
 10.5 mW    521.2 µs/s       0.8        Process        kdeinit4: plasma-desktop [kdeinit]
 10.5 mW     12.5 µs/s       1.0        kWork          pci_pme_list_scan
 7.59 mW    193.1 µs/s       0.6        Process        /usr/sbin/NetworkManager --no-daemon
 7.40 mW     21.8 µs/s       0.7        Timer          posix_timer_fn
 7.01 mW     53.0 µs/s       0.6        Interrupt      [4] block(softirq)
 6.84 mW     11.0 µs/s       0.6        Process        [rts_pstor]
 6.80 mW      1.9 µs/s       0.6        kWork          ieee80211_dynamic_ps_enable_wor
 6.07 mW     79.1 µs/s       0.6        Process        kwin
 6.02 mW    193.0 µs/s       0.5        Process        /usr/bin/virtuoso-t +foreground +configfile /tmp/virtuoso_hX1769.ini +wait
 5.37 mW     36.4 µs/s       0.5        kWork          iwl_bg_run_time_calib_work
 5.23 mW      2.4 µs/s       0.5        kWork          mei_timer
 4.94 mW     57.0 µs/s       0.4        Interrupt      [9] acpi
 4.37 mW     45.8 µs/s       0.4        Process        /usr/lib/chromium/chromium --type=renderer --lang=en-US --force-fieldtrials=AsyncDns/disabled/ConnCountImpact/conn_count_6/Conn

donniezazen · 2012-11-28 06:41:55

Any of you booting into a system that is already consuming 25-30W of power and heated upto 90C? Or is it just me.

LordChaos73 · 2012-11-28 10:05:56

donniezazen wrote:

Any of you booting into a system that is already consuming 25-30W of power and heated upto 90C? Or is it just me.

donniezazen, have a look at this bug report. See if you can reproduce some of the cpufreq artifacts in the report.

https://bugs.freedesktop.org/show_bug.cgi?id=54089

Last edited by LordChaos73 (2012-11-28 10:30:29)

KaiSforza · 2012-11-28 19:42:13

I've just opened up a pacman repository with the linux-mainline packages at http://sakuscans.com/pacmanpkg/x86_64/ . Please feel free to use it. I've added in the headers and docs from the vanilla mainline package, so it may prove useful and time-saving. (If there are configuration options that don't work, tell me. I'll try to fix that nex time I release.)

This repo's linux-mainline has the msleep patch.

Thanks!

johnLate · 2012-11-28 20:51:21

I tried drm-linux-next today, and it seems to fix the issue on my X220 Tablet (I verified 2 reboots and about 20 suspends).

Here's what I did:

mkdir drm-intel-next
cd drm-intel-next/
git remote add -t drm-intel-next -f origin git://people.freedesktop.org/~danvet/drm-intel
git checkout drm-intel-next

zcat /proc/config.gz > .config
make oldconfig
make -j4

sudo cp arch/x86/boot/bzImage /boot/vmlinuz-custom
sudo make modules_install
sudo mkinitcpio -k 3.7.0-rc4-ARCH-00324-g70b12bb -g /boot/initramfs-custom.img # Kernel Version from make modules_install output

vim /boot/grub/grub.conf
reboot

Edit: removed first command, was from an aborted attempt

Last edited by johnLate (2012-11-28 23:25:39)

KaiSforza · 2012-11-28 22:06:01

johnLate wrote:

I tried drm-linux-next today, and it seems to fix the issue on my X220 Tablet (I verified 2 reboots and about 20 suspends).

Here's what I did:

mkdir drm-intel-next
cd drm-intel-next/
git remote add -t drm-intel-next -f origin git://people.freedesktop.org/~danvet/drm-intel
git checkout drm-intel-next

zcat /proc/config.gz > .config
make oldconfig
make -j4

sudo cp arch/x86/boot/bzImage /boot/vmlinuz-custom
sudo make modules_install
sudo mkinitcpio -k 3.7.0-rc4-ARCH-00324-g70b12bb -g /boot/initramfs-custom.img # Kernel Version from make modules_install output

vim /boot/grub/grub.conf
reboot

Please put this in a pkgbuild so people don't get random untracked files floating around their systems.

Last edited by KaiSforza (2012-11-29 00:39:04)

Rolinh · 2012-11-29 18:15:04

@ KaiSforza: I added your repo and tried your msleep patched kernel. I just booted up, launched KDE and just a browser and here are my temperatures:

[robin@thor ~] % uname -r
3.7.0-2-rc7-mainline
[robin@thor ~] % uptime
 19:06:24 up 5 min,  2 users,  load average: 0.38, 0.82, 0.43
[robin@thor ~] % sensors
thinkpad-isa-0000
Adapter: ISA adapter
fan1:        3536 RPM

acpitz-virtual-0
Adapter: Virtual device
temp1:        +73.0°C  (crit = +98.0°C)

coretemp-isa-0000
Adapter: ISA adapter
Physical id 0:  +74.0°C  (high = +86.0°C, crit = +100.0°C)
Core 0:         +66.0°C  (high = +86.0°C, crit = +100.0°C)
Core 1:         +68.0°C  (high = +86.0°C, crit = +100.0°C)

jc42-i2c-6-19
Adapter: SMBus I801 adapter at efa0
temp1:        +50.5°C  (low  =  +0.0°C)                  ALARM (HIGH, CRIT)
                       (high =  +0.0°C, hyst =  +0.0°C)
                       (crit =  +0.0°C, hyst =  +0.0°C)

And here is what powertop shows me:

Device Power Report

Usage	Device name
3533 rpm	Laptop fan
5.6%	CPU use
100.0%	Audio codec alsa:hwC0D0: thinkpad (Conexant)
0.2 ops/s	GPU
437.7 pkts/s	Network interface: wlan0 (iwlwifi)
100.0%	USB device: EHCI Host Controller
100.0%	USB device: Integrated Camera (Chicony Electronics Co., Ltd.)
100.0%	Radio device: iwlwifi
100.0%	USB device: usb-device-8087-0024
100.0%	Display backlight
100.0%	Display backlight
100.0%	PCI Device: Ricoh Co Ltd PCIe SDXC/MMC Host Controller
100.0%	PCI Device: Intel Corporation Centrino Advanced-N 6205 [Taylor Peak]
100.0%	PCI Device: Intel Corporation 6 Series/C200 Series Chipset Family SMBus Controller
100.0%	PCI Device: Intel Corporation 6 Series/C200 Series Chipset Family 6 port SATA AHCI Controller
100.0%	PCI Device: Intel Corporation QM67 Express Chipset Family LPC Controller
100.0%	PCI Device: Intel Corporation 6 Series/C200 Series Chipset Family USB Enhanced Host Controller #1
100.0%	PCI Device: Intel Corporation 6 Series/C200 Series Chipset Family PCI Express Root Port 5
100.0%	PCI Device: Intel Corporation 6 Series/C200 Series Chipset Family PCI Express Root Port 2
100.0%	PCI Device: Intel Corporation 6 Series/C200 Series Chipset Family PCI Express Root Port 1
100.0%	PCI Device: Intel Corporation 6 Series/C200 Series Chipset Family High Definition Audio Controller
100.0%	PCI Device: Intel Corporation 6 Series/C200 Series Chipset Family USB Enhanced Host Controller #2
100.0%	PCI Device: Intel Corporation 82579LM Gigabit Network Connection
100.0%	PCI Device: Intel Corporation 6 Series/C200 Series Chipset Family KT Controller
100.0%	PCI Device: Intel Corporation 6 Series/C200 Series Chipset Family MEI Controller #1
100.0%	PCI Device: Intel Corporation 2nd Generation Core Processor Family Integrated Graphics Controller
100.0%	USB device: usb-device-8087-0024
100.0%	USB device: USB Optical Mouse (Logitech)
100.0%	USB device: EHCI Host Controller
0.0 pkts/s	Network interface: eth0 (e1000e)
0.0%	Thinkpad light

Just for information, here is my system information:

PowerTOP Version	v2.1
Kernel Version	Linux version 3.7.0-2-rc7-mainline (wgiokas@WST420) (gcc version 4.7.2 (GCC) ) #1 SMP PREEMPT Wed Nov 28 13:18:11 CST 2012
System Name	LENOVO 4238A71 ThinkPad T420
CPU Information	4x Intel(R) Core(TM) i7-2620M CPU @ 2.70GHz
OS Information	Arch Linux

Not so good, hu.

/me reboots on 3.5

KaiSforza · 2012-11-29 19:41:11

Rolinh wrote:

@ KaiSforza: I added your repo and tried your msleep patched kernel. I just booted up, launched KDE and just a browser and here are my temperatures:
<--snip-->
Not so good, hu.
/me reboots on 3.5

I get similart numbers on a cold boot. I forgot to mention that it takes a suspend-resume cycle for it to all work and get normal temps/power usage.

(And thanks for trying it out.)

Also, might I ask for a powertop output if you try again, to see the cpu frequencies?

Last edited by KaiSforza (2012-11-29 19:42:22)

Rolinh · 2012-11-29 20:07:23

Sure. Right now I am working on stuff but I'll do it later and post the results.
Thanks for your package and repo by the way. ;-)

johnLate · 2012-11-29 20:59:07

johnLate wrote:

I tried drm-linux-next today, and it seems to fix the issue on my X220 Tablet (I verified 2 reboots and about 20 suspends).
[...]

I have bad news: Just got out of suspend and GPU was 100% active (powertop 2.1, Idle stats). That means even drm-linux-next (from yesterday) doesn't fix the issue.

Since switching to that kernel, 2 resumes out of about 4 reboots and 25 suspend-resume-cycles were bad.

Arch Linux

#176 2012-11-25 07:02:10

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#177 2012-11-25 10:55:24

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#178 2012-11-25 16:41:11

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#179 2012-11-26 12:51:44

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#180 2012-11-26 13:50:48

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#181 2012-11-26 17:10:27

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#182 2012-11-26 20:17:27

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#183 2012-11-26 21:57:37

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#184 2012-11-26 22:03:59

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#185 2012-11-26 22:32:21

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#186 2012-11-27 09:45:30

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#187 2012-11-27 19:23:09

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#188 2012-11-27 21:38:52

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#189 2012-11-27 22:02:25

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#190 2012-11-27 22:15:10

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#191 2012-11-27 22:22:20

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#192 2012-11-28 06:41:55

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#193 2012-11-28 10:05:56

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#194 2012-11-28 19:42:13

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#195 2012-11-28 20:51:21

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#196 2012-11-28 22:06:01

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#197 2012-11-29 18:15:04

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#198 2012-11-29 19:41:11

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#199 2012-11-29 20:07:23

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

#200 2012-11-29 20:59:07

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

Board footer