Arch Linux

fab

Member

Registered: 2012-10-17

Posts: 1

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

Affected, Lenovo X220 Tablet, Intel i5 Sandy Bridge, Intel 3000 HD, Linux-3.6.2

But only after resume from suspend. Normal behavior after boot.

ANOKNUSA

Member

Registered: 2010-10-22

Posts: 2,141

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

Two things that stand out to me here:

1) My boot parameters:

ro quiet elevator=bfq pcie_aspm=force i915.i915_enable_rc6=1 i915.i915_enable_fbc=1 i915.lvds_downclock=1

The section "pcie_aspm=force" was added to fix an earlier regression; it's my understanding that this was supposed to be fixed around 3.4 or 3.5 (not sure which), and I just forgot to remove it.  However, no one else here mentioned using it, so I there it is.

2) This only seems to affect folks with hybrid graphics, so: Is everyone here testing this with only the integrated graphics enabled (i.e. discrete chip is disabled in bios or through acpi)?

I'm bringing these things up because I'm not experiencing this issue with my setup; whether booting clean or waking from suspend, my temps are currently 33-41 degress celsius across four cores, all scaling properly. (i7-2720QM Sandy Bridge, Intel HD 3000 integrated graphics only, system completely up-to-date).

---

Linux != Windows  ::  Ask Questions the Smart Way  ::  Configs

trsk

Member

Registered: 2012-10-14

Posts: 38

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

2) This only seems to affect folks with hybrid graphics, so: Is everyone here testing this with only the integrated graphics enabled (i.e. discrete chip is disabled in bios or through acpi)?

The X220 is affected, but it doesn't have a discrete graphics chip, only an integrated Sandy Bridge GPU.

neycho

Member

Registered: 2012-10-16

Posts: 2

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

Same issue here - After upgrade, gnome-shell start to use 70-150% of CPU, temperatures are around 82 C - that's really hot! - my cpu is T5450, my video is Ati HD2600 mobile

jrk

Member

From: Nämberch

Registered: 2012-10-16

Posts: 37

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

Ok, some more data points from me (X220)
I've updated the BIOS to latest version (1.34).
I've also figured out that the issue is independent from being docked or undocked, on AC or on battery.
The issue seems also to be existent in 3.5.* which I didn't notice because I updated from 3.4 to 3.6 directly.

Sometimes a suspend/resume cycle seems to fix the issue.
I also get the CPU Power Package warnings. In addition there are warnings about the buggy ACPI table but I don't think it's relevant.
I can however, post a dmesg with suspend/resume if it's of interest.

In addition here's my (trimmed) pacman.log. If you see suspicious packages or want me to try something, please tell me.
I'm going to downgrade the kernel later to 3.4. FWIW, if I feel like, I'll try to bisect.

[2012-10-15 20:02] starting full system upgrade
[2012-10-15 20:03] removed eject (2.1.5-7)
[2012-10-15 20:03] removed libsystemd (186-2)
[2012-10-15 20:03] removed systemd-tools (186-2)
[2012-10-15 20:03] warning: /etc/conf.d/cpufreq saved as /etc/conf.d/cpufreq.pacsave
[2012-10-15 20:03] removed cpufrequtils (008-2)
[2012-10-15 20:03] upgraded linux-api-headers (3.4.6-1 -> 3.5.5-1)
[2012-10-15 20:03] upgraded tzdata (2012d-1 -> 2012f-1)
[2012-10-15 20:03] Generating locales...
[2012-10-15 20:03]   de_DE.UTF-8... done
[2012-10-15 20:03]   de_DE.ISO-8859-1... done
[2012-10-15 20:03]   de_DE.ISO-8859-15@euro... done
[2012-10-15 20:03]   en_US.UTF-8... done
[2012-10-15 20:03]   en_US.ISO-8859-1... done
[2012-10-15 20:03] Generation complete.
[2012-10-15 20:03] upgraded glibc (2.16.0-2 -> 2.16.0-4)
[2012-10-15 20:03] upgraded cracklib (2.8.18-2 -> 2.8.19-1)
[2012-10-15 20:03] upgraded pam (1.1.5-4 -> 1.1.6-1)
[2012-10-15 20:03] upgraded coreutils (8.17-3 -> 8.19-1)
[2012-10-15 20:03] warning: /etc/protocols installed as /etc/protocols.pacnew
[2012-10-15 20:03] warning: /etc/services installed as /etc/services.pacnew
[2012-10-15 20:03] upgraded iana-etc (2.30-2 -> 2.30-3)
[2012-10-15 20:03] warning: /etc/group installed as /etc/group.pacnew
[2012-10-15 20:03] warning: /etc/passwd installed as /etc/passwd.pacnew
[2012-10-15 20:03] warning: /etc/gshadow installed as /etc/gshadow.pacnew
[2012-10-15 20:03] upgraded filesystem (2012.6-4 -> 2012.10-1)
[2012-10-15 20:03] upgraded dbus-core (1.6.4-1 -> 1.6.8-1)
[2012-10-15 20:03] upgraded kmod (9-2 -> 10-1)
[2012-10-15 20:03] upgraded hwids (20120711-1 -> 20121012-1)
[2012-10-15 20:03] upgraded util-linux (2.21.2-5 -> 2.22.1-1)
[2012-10-15 20:03] ln -s '/usr/lib/systemd/system/getty@.service' '/etc/systemd/system/getty.target.wants/getty@tty1.service'
[2012-10-15 20:03] :: Append 'init=/bin/systemd' to your kernel command line in your
[2012-10-15 20:03]    bootloader to replace sysvinit with systemd
[2012-10-15 20:03] installed systemd (194-3)
[2012-10-15 20:03] upgraded accountsservice (0.6.22-1 -> 0.6.25-1)
[2012-10-15 20:03] warning: /etc/acpi/handler.sh installed as /etc/acpi/handler.sh.pacnew
[2012-10-15 20:03] upgraded acpid (2.0.16-3 -> 2.0.17-3)
[2012-10-15 20:03] upgraded libmysqlclient (5.5.25.a-1 -> 5.5.28-1)
[2012-10-15 20:03] upgraded mysql-clients (5.5.25.a-1 -> 5.5.28-1)
[2012-10-15 20:03] upgraded mysql (5.5.25.a-1 -> 5.5.28-1)
[2012-10-15 20:03] upgraded libtiff (4.0.2-1 -> 4.0.3-1)
[2012-10-15 20:03] upgraded sqlite (3.7.13-1 -> 3.7.14.1-1)
[2012-10-15 20:03] upgraded dbus (1.6.4-1 -> 1.6.8-1)
[2012-10-15 20:03] upgraded libdrm (2.4.37-1 -> 2.4.39-1)
[2012-10-15 20:03] upgraded libglapi (8.0.4-2 -> 9.0-1)
[2012-10-15 20:03] upgraded lib32-gcc-libs (4.7.1-6 -> 4.7.2-1)
[2012-10-15 20:03] upgraded gcc-libs-multilib (4.7.1-6 -> 4.7.2-1)
[2012-10-15 20:03] upgraded libgl (8.0.4-2 -> 9.0-1)
[2012-10-15 20:03] upgraded randrproto (1.3.2-2 -> 1.4.0-1)
[2012-10-15 20:03] upgraded libxrandr (1.3.2-2 -> 1.4.0-1)
[2012-10-15 20:03] upgraded alsa-lib (1.0.25-1 -> 1.0.26-1)
[2012-10-15 20:03] upgraded qt (4.8.2-3 -> 4.8.3-4)
[2012-10-15 20:03] upgraded e2fsprogs (1.42.4-1 -> 1.42.5-1)
[2012-10-15 20:03] upgraded libldap (2.4.31-4 -> 2.4.33-1)
[2012-10-15 20:03] upgraded virtuoso-base (6.1.5.20120717-2 -> 6.1.6-1)
[2012-10-15 20:03] upgraded soprano (2.7.6-2 -> 2.8.0-2)
[2012-10-15 20:03] upgraded akonadi (1.7.2-2 -> 1.8.1-1)
[2012-10-15 20:03] upgraded alex (3.0.2-2 -> 3.0.2-26)
[2012-10-15 20:03] upgraded alsa-utils (1.0.25-3 -> 1.0.26-1)
[2012-10-15 20:03] ==> Appending keys from archlinux.gpg...
...
[2012-10-15 20:03] gpg: next trustdb check due at 2013-01-22
[2012-10-15 20:03] upgraded archlinux-keyring (20120622-1 -> 20120915-1)
[2012-10-15 20:03] upgraded attica (0.4.0-1 -> 0.4.1-1)
[2012-10-15 20:03] upgraded perl (5.16.0-2 -> 5.16.1-1)
[2012-10-15 20:03] upgraded automake (1.12.2-1 -> 1.12.4-1)
[2012-10-15 20:03] upgraded avahi (0.6.31-3 -> 0.6.31-4)
[2012-10-15 20:03] upgraded bison (2.6-1 -> 2.6.2-1)
[2012-10-15 20:03] upgraded mpfr (3.1.1-1 -> 3.1.1.p2-1)
[2012-10-15 20:03] upgraded libmpc (1.0-1 -> 1.0.1-1)
[2012-10-15 20:04] upgraded gcc-multilib (4.7.1-6 -> 4.7.2-1)
[2012-10-15 20:04] ==> Unregistering cabalized packages...
...
[2012-10-15 20:04] ==> See /usr/share/haskell/ and ghc-pkg list --user for a tentative list of affected packages.
[2012-10-15 20:04] upgraded ghc (7.4.2-1 -> 7.6.1-1)
[2012-10-15 20:04] upgraded cabal-install (0.14.0-2 -> 1.16.0-2)
[2012-10-15 20:04] upgraded cairo (1.12.2-2 -> 1.12.2-3)
[2012-10-15 20:04] upgraded libcap-ng (0.6.6-1 -> 0.7-1)
[2012-10-15 20:04] upgraded krb5 (1.10.2-2 -> 1.10.3-1)
[2012-10-15 20:04] upgraded libwbclient (3.6.6-1 -> 3.6.8-1)
[2012-10-15 20:04] upgraded cifs-utils (5.5-1 -> 5.6-2)
[2012-10-15 20:04] upgraded llvm (3.1-3 -> 3.1-4)
[2012-10-15 20:04] upgraded clang (3.1-3 -> 3.1-4)
[2012-10-15 20:04] upgraded colordiff (1.0.9-3 -> 1.0.10-1)
[2012-10-15 20:04] installed cpupower (3.6-1)
[2012-10-15 20:04] upgraded device-mapper (2.02.96-3 -> 2.02.97-1)
[2012-10-15 20:04] upgraded cryptsetup (1.4.3-1 -> 1.5.0-2)
[2012-10-15 20:04] upgraded cscope (15.8-1 -> 15.8.a-1)
[2012-10-15 20:04] upgraded libtasn1 (2.13-1 -> 2.14-1)
[2012-10-15 20:04] upgraded p11-kit (0.12-1 -> 0.13-1)
[2012-10-15 20:04] upgraded gnutls (3.0.21-2 -> 3.1.3-1)
[2012-10-15 20:04] upgraded libcups (1.5.3-6 -> 1.6.1-6)
[2012-10-15 20:04] upgraded lcms2 (2.3-2 -> 2.4-1)
[2012-10-15 20:04] upgraded poppler (0.20.2-1 -> 0.20.3-1)
[2012-10-15 20:04] installed qpdf (3.0.2-1)
[2012-10-15 20:04] upgraded cups-filters (1.0.20-1 -> 1.0.24-1)
[2012-10-15 20:04] upgraded libusbx (1.0.12-2 -> 1.0.14-1)
[2012-10-15 20:04] * avahi-daemon should now run before you start cupsd for
[2012-10-15 20:04]   printer Browsing support
[2012-10-15 20:04] * check your config with cupsd -t
[2012-10-15 20:04]   and fix all Browsing related settings
[2012-10-15 20:04] > It's now safe to load usblp kernel module.
[2012-10-15 20:04] > There's no more need to blacklist it.
[2012-10-15 20:04] upgraded cups (1.5.3-6 -> 1.6.1-6)
[2012-10-15 20:04] upgraded curl (7.26.0-1 -> 7.28.0-1)
[2012-10-15 20:04] upgraded cyrus-sasl (2.1.23-9 -> 2.1.23-11)
[2012-10-15 20:04] upgraded dhcpcd (5.5.6-1 -> 5.6.2-1)
[2012-10-15 20:04] upgraded djvulibre (3.5.24-3 -> 3.5.25.3-1)
[2012-10-15 20:04] upgraded dnsutils (9.9.1.P3-1 -> 9.9.2-1)
[2012-10-15 20:04] upgraded dri2proto (2.6-1 -> 2.8-1)
[2012-10-15 20:04] upgraded ethtool (1:3.4-1 -> 1:3.5-1)
[2012-10-15 20:04] upgraded fcron (3.0.6-7 -> 3.0.6-8)
[2012-10-15 20:04] upgraded xcb-proto (1.7.1-1 -> 1.8-1)
[2012-10-15 20:04] upgraded libxcb (1.8.1-1 -> 1.9-1)
[2012-10-15 20:04] installed libgbm (9.0-1)
[2012-10-15 20:04] upgraded khrplatform-devel (8.0.4-2 -> 9.0-1)
[2012-10-15 20:04] upgraded libegl (8.0.4-2 -> 9.0-1)
[2012-10-15 20:04] upgraded libva (1.1.0-1 -> 1.1.0-2)
[2012-10-15 20:04] upgraded sdl (1.2.15-1 -> 1.2.15-3)
[2012-10-15 20:04] upgraded ffmpeg (1:0.11.1-1 -> 1:1.0-1)
[2012-10-15 20:04] upgraded gdk-pixbuf2 (2.26.1-1 -> 2.26.3-1)
[2012-10-15 20:04] upgraded gtk-update-icon-cache (2.24.11-2 -> 2.24.13-1)
[2012-10-15 20:04] upgraded gtk2 (2.24.11-2 -> 2.24.13-1)
[2012-10-15 20:04] upgraded libevent (2.0.19-1 -> 2.0.20-1)
[2012-10-15 20:04] upgraded nspr (4.9.1-1 -> 4.9.2-1)
[2012-10-15 20:04] upgraded nss (3.13.5-1 -> 3.13.6-1)
[2012-10-15 20:04] upgraded hunspell (1.3.2-1 -> 1.3.2-2)
[2012-10-15 20:04] upgraded firefox (14.0.1-1 -> 16.0.1-1)
[2012-10-15 20:04] upgraded flashplugin (11.2.202.236-1 -> 11.2.202.243-1)
[2012-10-15 20:04] upgraded flex (2.5.36-1 -> 2.5.37-1)
[2012-10-15 20:04] upgraded freeglut (2.8.0-1 -> 2.8.0-2)
[2012-10-15 20:04] upgraded fribidi (0.19.2-2 -> 0.19.4-1)
[2012-10-15 20:04] upgraded fuse (2.9.0-1 -> 2.9.1-1)
[2012-10-15 20:04] upgraded gc (7.2-1 -> 7.2.d-1)
[2012-10-15 20:04] upgraded gcc-fortran (4.7.1-6 -> 4.7.2-1)
[2012-10-15 20:04] upgraded gegl (0.2.0-2 -> 0.2.0-4)
[2012-10-15 20:04] installed glu (9.0.0-1)
[2012-10-15 20:04] upgraded jasper (1.900.1-7 -> 1.900.1-8)
[2012-10-15 20:04] upgraded ghostscript (9.05-1 -> 9.06-1)
[2012-10-15 20:04] upgraded librsvg (2.36.1-1 -> 2.36.3-1)
[2012-10-15 20:04] upgraded libexif (0.6.20-2 -> 0.6.21-1)
[2012-10-15 20:04] upgraded gimp (2.8.0-2 -> 2.8.2-1)
[2012-10-15 20:04] upgraded git (1.7.11.3-1 -> 1.7.12.3-1)
[2012-10-15 20:04] upgraded gnome-keyring (3.4.1-2 -> 3.4.1-3)
[2012-10-15 20:04] upgraded grep (2.13-2 -> 2.14-1)
[2012-10-15 20:04] upgraded vim-runtime (7.3.600-1 -> 7.3.661-1)
[2012-10-15 20:04] upgraded ruby (1.9.3_p194-2 -> 1.9.3_p286-1)
[2012-10-15 20:04] upgraded gvim (7.3.600-1 -> 7.3.661-1)
[2012-10-15 20:04] upgraded happy (1.18.9-8 -> 1.18.10-25)
[lot of ghc noise..]
[2012-10-15 20:05] upgraded haskell-x11-xft (0.3.1-13 -> 0.3.1-25)
[2012-10-15 20:05] upgraded hdparm (9.39-1 -> 9.42-1)
[2012-10-15 20:05] upgraded hspell (1.1-2 -> 1.2-1)
[2012-10-15 20:05] upgraded libxml2 (2.7.8-2 -> 2.8.0-1)
[2012-10-15 20:05] upgraded hwloc (1.4.2-1 -> 1.5-1)
[2012-10-15 20:05] upgraded openjdk6 (6.b24_1.11.3-1 -> 6.b24_1.11.4-1)
[2012-10-15 20:05] upgraded icedtea-web (1.2-1 -> 1.3-1)
[2012-10-15 20:05] upgraded libltdl (2.4.2-6 -> 2.4.2-7)
[2012-10-15 20:05] upgraded imagemagick (6.7.8.6-1 -> 6.7.9.8-2)
[2012-10-15 20:05] upgraded inetutils (1.9.1-2 -> 1.9.1-4)
[2012-10-15 20:05] upgraded iproute2 (3.4.0-2 -> 3.5.1-1)
[2012-10-15 20:05] installed sysvinit-tools (2.88-9)
[2012-10-15 20:05] warning: /etc/rc.conf installed as /etc/rc.conf.pacnew
[2012-10-15 20:05] warning: /etc/rc.local installed as /etc/rc.local.pacnew
[2012-10-15 20:05] ----
[2012-10-15 20:05] > initscripts now ships a compatibility layer to make systemd use DAEMONS
[2012-10-15 20:05] > from rc.conf; and run rc.local and rc.local.shutdown on boot and
[2012-10-15 20:05] > shutdown, respectively. Refer to arch-modules-load(8) and arch-daemons(8).
[2012-10-15 20:05] ----
[2012-10-15 20:05] ----
[2012-10-15 20:05] > systemd no longer reads MODULES from rc.conf.
[2012-10-15 20:05] ----
[2012-10-15 20:05] upgraded initscripts (2012.06.3-2 -> 2012.10.1-1)
[2012-10-15 20:05] upgraded intel-dri (8.0.4-2 -> 9.0-1)
[2012-10-15 20:05] upgraded iputils (20101006-4 -> 20101006-7)
[2012-10-15 20:05] upgraded iso-codes (3.34-1 -> 3.38-1)
[2012-10-15 20:05] upgraded itstool (1.1.2-1 -> 1.2.0-1)
[2012-10-15 20:05] upgraded jack (0.121.3-6 -> 0.121.3-7)
[2012-10-15 20:05] upgraded smbclient (3.6.6-1 -> 3.6.8-1)
[2012-10-15 20:05] upgraded oxygen-icons (4.8.4-1 -> 4.9.2-1)
[2012-10-15 20:05] upgraded kdebase-runtime (4.9.2-1 -> 4.9.2-2)
[2012-10-15 20:05] upgraded strigi (0.7.7-1 -> 0.7.7-2)
[2012-10-15 20:05] upgraded openexr (1.7.0-2 -> 1.7.1-1)
[2012-10-15 20:05] upgraded usbmuxd (1.0.7-3 -> 1.0.8-1)
[2012-10-15 20:05] upgraded libimobiledevice (1.1.1-3 -> 1.1.4-2)
[2012-10-15 20:05] upgraded upower (0.9.17-1 -> 0.9.18-2)
[2012-10-15 20:05] upgraded libatasmart (0.18-2 -> 0.19-1)
[2012-10-15 20:05] upgraded udisks (1.0.4-4 -> 1.0.4-5)
[2012-10-15 20:05] upgraded kdelibs (4.9.2-1 -> 4.9.2-2)
[2012-10-15 20:05] upgraded less (444-3 -> 451-1)
[2012-10-15 20:05] upgraded lib32-dbus-core (1.6.4-1 -> 1.6.8-1)
[2012-10-15 20:05] upgraded lib32-libglapi (8.0.4-4 -> 9.0-1)
[2012-10-15 20:05] upgraded lib32-libgl (8.0.4-4 -> 9.0-1)
[2012-10-15 20:05] upgraded lib32-libltdl (2.4.2-6 -> 2.4.2-7)
[2012-10-15 20:05] upgraded lib32-libtiff (4.0.2-1 -> 4.0.3-1)
[2012-10-15 20:05] upgraded lib32-libxcb (1.8.1-2 -> 1.9-1)
[2012-10-15 20:05] upgraded lib32-sqlite (3.7.14-2 -> 3.7.14.1-1)
[2012-10-15 20:05] upgraded libarchive (3.0.4-1 -> 3.0.4-2)
[2012-10-15 20:05] upgraded libbluray (0.2.2-1 -> 0.2.3-1)
[2012-10-15 20:05] upgraded libedit (20120311_3.0-1 -> 20120601_3.0-1)
[2012-10-15 20:05] upgraded libgssglue (0.3-1 -> 0.4-1)
[2012-10-15 20:05] upgraded libpipeline (1.2.1-1 -> 1.2.2-1)
[2012-10-15 20:05] upgraded libspectre (0.2.6-3 -> 0.2.7-1)
[2012-10-15 20:05] upgraded libthai (0.1.17-1 -> 0.1.18-1)
[2012-10-15 20:05] upgraded libtool-multilib (2.4.2-6 -> 2.4.2-7)
[2012-10-15 20:05] upgraded libvdpau (0.4.1-2 -> 0.5-1)
[2012-10-15 20:05] upgraded libwnck3 (3.4.2-1 -> 3.4.3-1)
[2012-10-15 20:05] upgraded mkinitcpio-busybox (1.20.1-1 -> 1.20.2-1)
[2012-10-15 20:05] upgraded mkinitcpio (0.9.2-2 -> 0.10-1)
[2012-10-15 20:05] >>> Updating module dependencies. Please wait ...
[2012-10-15 20:05] >>> Generating initial ramdisk, using mkinitcpio.  Please wait...
[2012-10-15 20:05] ==> Building image from preset: 'default'
[2012-10-15 20:05]   -> -k /boot/vmlinuz-linux -c /etc/mkinitcpio.conf -g /boot/initramfs-linux.img
[2012-10-15 20:05] ==> Starting build: 3.6.2-1-ARCH
[2012-10-15 20:05]   -> Running build hook: [base]
[2012-10-15 20:05]   -> Running build hook: [udev]
[2012-10-15 20:05]   -> Running build hook: [autodetect]
[2012-10-15 20:05]   -> Running build hook: [sata]
[2012-10-15 20:05]   -> Running build hook: [lvm2]
[2012-10-15 20:05]   -> Running build hook: [uresume]
[2012-10-15 20:05]   -> Running build hook: [filesystems]
[2012-10-15 20:05]   -> Running build hook: [fsck]
[2012-10-15 20:05] ==> Generating module dependencies
[2012-10-15 20:05] ==> Creating gzip initcpio image: /boot/initramfs-linux.img
[2012-10-15 20:05] ==> Image generation successful
[2012-10-15 20:05] ==> Building image from preset: 'fallback'
[2012-10-15 20:05]   -> -k /boot/vmlinuz-linux -c /etc/mkinitcpio.conf -g /boot/initramfs-linux-fallback.img -S autodetect
[2012-10-15 20:05] ==> Starting build: 3.6.2-1-ARCH
[2012-10-15 20:05]   -> Running build hook: [base]
[2012-10-15 20:05]   -> Running build hook: [udev]
[2012-10-15 20:05]   -> Running build hook: [sata]
[2012-10-15 20:05]   -> Running build hook: [lvm2]
[2012-10-15 20:05]   -> Running build hook: [uresume]
[2012-10-15 20:05]   -> Running build hook: [filesystems]
[2012-10-15 20:05]   -> Running build hook: [fsck]
[2012-10-15 20:05] ==> Generating module dependencies
[2012-10-15 20:05] ==> Creating gzip initcpio image: /boot/initramfs-linux-fallback.img
[2012-10-15 20:05] ==> Image generation successful
[2012-10-15 20:05] upgraded linux (3.4.6-1 -> 3.6.2-1)
[2012-10-15 20:05] upgraded lirc-utils (1:0.9.0-22 -> 1:0.9.0-31)
[2012-10-15 20:05] upgraded lm_sensors (3.3.2-3 -> 3.3.2-5)
[2012-10-15 20:05] upgraded logrotate (3.8.1-2 -> 3.8.2-1)
[2012-10-15 20:05] upgraded lvm2 (2.02.96-3 -> 2.02.97-1)
[2012-10-15 20:05] upgraded make (3.82-4 -> 3.82-5)
[2012-10-15 20:05] upgraded man-db (2.6.2-1 -> 2.6.3-1)
[2012-10-15 20:05] upgraded man-pages (3.41-1 -> 3.42-1)
[2012-10-15 20:05] upgraded mercurial (2.2.3-1 -> 2.3.2-1)
[2012-10-15 20:05] upgraded mesa (8.0.4-2 -> 9.0-1)
[2012-10-15 20:05] upgraded motion (3.2.12-6 -> 3.2.12-7)
[2012-10-15 20:05] upgraded mpg123 (1.14.3-1 -> 1.14.4-1)
[2012-10-15 20:05] upgraded mtdev (1.1.2-1 -> 1.1.3-1)
[2012-10-15 20:05] upgraded net-snmp (5.7.1-3 -> 5.7.1-4)
[2012-10-15 20:05] upgraded net-tools (1.60.20110819cvs-3 -> 1.60.20120804git-2)
[2012-10-15 20:05] upgraded ntp (4.2.6.p5-7 -> 4.2.6.p5-11)
[2012-10-15 20:05] upgraded xine-lib (1.2.2-1 -> 1.2.2-2)
[2012-10-15 20:05] upgraded opencv (2.4.2-2 -> 2.4.2-4)
[2012-10-15 20:05] upgraded openldap (2.4.31-4 -> 2.4.33-1)
[2012-10-15 20:05] warning: /etc/ssh/sshd_config installed as /etc/ssh/sshd_config.pacnew
[2012-10-15 20:05] upgraded openssh (6.0p1-3 -> 6.1p1-2)
[2012-10-15 20:05] upgraded openmpi (1.6.1-1 -> 1.6.2-1)
[2012-10-15 20:05] upgraded openslp (1.2.1-4 -> 1.2.1-5)
[2012-10-15 20:05] upgraded pacman-color (4.0.3-1 -> 4.0.3-3)
[2012-10-15 20:05] warning: /etc/pacman.d/mirrorlist installed as /etc/pacman.d/mirrorlist.pacnew
[2012-10-15 20:05] upgraded pacman-mirrorlist (20120626-1 -> 20120908-1)
[2012-10-15 20:05] upgraded patch (2.6.1-3 -> 2.7.1-1)
[2012-10-15 20:05] upgraded phonon-gstreamer (4.6.1-1 -> 4.6.2-1)
[2012-10-15 20:05] upgraded pinentry (0.8.1-4 -> 0.8.2-1)
[2012-10-15 20:05] upgraded pkg-config (0.27-1 -> 0.27.1-1)
[2012-10-15 20:05] upgraded poppler-glib (0.20.2-1 -> 0.20.3-1)
[2012-10-15 20:05] upgraded powertop (2.0-1 -> 2.1-1)
[2012-10-15 20:05] upgraded ppp (2.4.5-3 -> 2.4.5-4)
[2012-10-15 20:05] upgraded procps-ng (3.3.3-3 -> 3.3.3-6)
[2012-10-15 20:05] upgraded python2-urwid (1.0.1-1 -> 1.0.2-1)
[2012-10-15 20:05] upgraded qrencode (3.3.1-1 -> 3.4.0-1)
[2012-10-15 20:05] upgraded rfkill (0.4-4 -> 0.4-5)
[2012-10-15 20:05] upgraded rsync (3.0.9-3 -> 3.0.9-5)
[2012-10-15 20:05] upgraded run-parts (4.3.2-1 -> 4.3.4-1)
[2012-10-15 20:05] upgraded sane (1.0.22-9 -> 1.0.23-1)
[2012-10-15 20:05] upgraded smpeg (0.4.4-6 -> 0.4.4-7)
[2012-10-15 20:05] upgraded subversion (1.7.5-2 -> 1.7.7-1)
[2012-10-15 20:05] upgraded sudo (1.8.5.p2-1 -> 1.8.6.p3-1)
[2012-10-15 20:05] upgraded syslog-ng (3.3.5-2 -> 3.3.6-1)
[2012-10-15 20:05] upgraded sysvinit (2.88-6 -> 2.88-9)
[2012-10-15 20:05] >>> texlive: updating the filename database...
[2012-10-15 20:05] mktexlsr: Updating /etc/texmf/ls-R... 
[2012-10-15 20:05] mktexlsr: Updating /usr/share/texmf/ls-R... 
[2012-10-15 20:05] mktexlsr: Updating /usr/share/texmf-dist/ls-R... 
[2012-10-15 20:06] mktexlsr: Updating /var/lib/texmf/ls-R... 
[2012-10-15 20:06] mktexlsr: Done.
[2012-10-15 20:06] recreating all formats... done.
[2012-10-15 20:06]  (logs are under /var/lib/texmf/web2c/<engine>/<formatname>.log)
[2012-10-15 20:06] upgraded texlive-bin (2012.0-3 -> 2012.0-4)
[2012-10-15 20:06] upgraded thunderbird (14.0-1 -> 16.0.1-1)
[2012-10-15 20:06] upgraded tmux (1.6-2 -> 1.7-1)
[2012-10-15 20:06] upgraded tp_smapi (0.41-3 -> 0.41-9)
[2012-10-15 20:06] upgraded wget (1.13.4-1 -> 1.14-1)
[2012-10-15 20:06] upgraded wireless_tools (29-6 -> 29-7)
[2012-10-15 20:06] upgraded xf86-input-evdev (2.7.1-1 -> 2.7.3-2)
[2012-10-15 20:06] upgraded xf86-input-synaptics (1.6.2-1 -> 1.6.2-2)
[2012-10-15 20:06] upgraded xf86-video-intel (2.20.1-2 -> 2.20.9-2)
[2012-10-15 20:06] upgraded xkeyboard-config (2.6-1 -> 2.7-1)
[2012-10-15 20:06] upgraded xlockmore (5.38-1 -> 5.40-3)
[2012-10-15 20:06] upgraded xmobar (0.15-2 -> 0.16-1)
[2012-10-15 20:06] upgraded xorg-server-common (1.12.3-1 -> 1.13.0-2)
[2012-10-15 20:06] upgraded xorg-server (1.12.3-1 -> 1.13.0-2)
[2012-10-15 20:06] upgraded xorg-server-xnest (1.12.4-1 -> 1.13.0-2)
[2012-10-15 20:06] upgraded xorg-xbacklight (1.1.2-3 -> 1.2.0-1)
[2012-10-15 20:06] upgraded xterm (281-1 -> 283-1)
[2012-10-15 20:06] upgraded xulrunner (14.0.1-1 -> 15.0-1)

trsk

Member

Registered: 2012-10-14

Posts: 38

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

Ok, some more data points from me (X220)
I've updated the BIOS to latest version (1.34).
I've also figured out that the issue is independent from being docked or undocked, on AC or on battery.
The issue seems also to be existent in 3.5.* which I didn't

This is strange... I've been using linux 3.5 with my x220 since arch fed it into the repo and haven't noticed this issue. However, it is present with linux 3.6. As I've downgraded to 3.5, I'll try to confirm if it affects me or not.

donniezazen

Member

From: Salt Lake City

Registered: 2011-06-24

Posts: 671

Website

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

Two things that stand out to me here:

1) My boot parameters:

ro quiet elevator=bfq pcie_aspm=force i915.i915_enable_rc6=1 i915.i915_enable_fbc=1 i915.lvds_downclock=1

The section "pcie_aspm=force" was added to fix an earlier regression; it's my understanding that this was supposed to be fixed around 3.4 or 3.5 (not sure which), and I just forgot to remove it.  However, no one else here mentioned using it, so I there it is.

2) This only seems to affect folks with hybrid graphics, so: Is everyone here testing this with only the integrated graphics enabled (i.e. discrete chip is disabled in bios or through acpi)?

I'm bringing these things up because I'm not experiencing this issue with my setup; whether booting clean or waking from suspend, my temps are currently 33-41 degress celsius across four cores, all scaling properly. (i7-2720QM Sandy Bridge, Intel HD 3000 integrated graphics only, system completely up-to-date).

ASPM Power Regression was fixed long time ago and ported back and forth to probably all stable kernels. It is not required anymore.

I started noticing new power regression around 3.4 which have continued to increase and Sandy Bridge seems to be main hardware affected. I do have Optimus but I have disabled it in BIOS and been using power saving Intel HD 3000.

My current power usage is 12W and temp is 50C. Last night while copying some 80GB files to an external hard drive my system temperature spiked to 97C. I had a kernel panic. This seems to be very erratic.

---

github | blog

jrk

Member

From: Nämberch

Registered: 2012-10-16

Posts: 37

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

Did a naive bisect.
I fetched linux-3.4.7 to linux-3.5.5 from arm.
Tested 3.4.7, 3.5.3, 3.5.5 and 3.5.6 with multiple reboots and suspend/resume cycles.
They all behave fine.
There are also kernel version 3.5.7, 3.6.0 and 3.6.1 for which I do not have packages (3.5.7 and 3.6.0 never existed as pkg I think).
Right now I'm compiling 3.6.0. *sigh*

Seriously, stuff like this makes think about switching to Mac (which I would never do) or *cough* Windows *cough*. Sorry, just venting off a bit.

[edit]
diff between kernel config 3.5.6 and 3.6.* (it says 3.6.0 but I fetched the latest abs pkg)
There's nothing that jumps to my eye specifically..

--- /proc/self/fd/11	2012-10-17 20:43:41.443399251 +0200
+++ config.x86_64	2012-10-17 06:05:12.000000000 +0200
@@ -1,6 +1,6 @@
 #
 # Automatically generated file; DO NOT EDIT.
-# Linux/x86_64 3.5.6-1 Kernel Configuration
+# Linux/x86_64 3.6.0 Kernel Configuration
 #
 CONFIG_64BIT=y
 # CONFIG_X86_32 is not set
@@ -136,10 +136,8 @@
 CONFIG_PROC_PID_CPUSET=y
 CONFIG_CGROUP_CPUACCT=y
 CONFIG_RESOURCE_COUNTERS=y
-CONFIG_CGROUP_MEM_RES_CTLR=y
-CONFIG_CGROUP_MEM_RES_CTLR_SWAP=y
-# CONFIG_CGROUP_MEM_RES_CTLR_SWAP_ENABLED is not set
-CONFIG_CGROUP_MEM_RES_CTLR_KMEM=y
+# CONFIG_MEMCG is not set
+# CONFIG_CGROUP_HUGETLB is not set
 # CONFIG_CGROUP_PERF is not set
 CONFIG_CGROUP_SCHED=y
 CONFIG_FAIR_GROUP_SCHED=y
@@ -154,7 +152,6 @@
 CONFIG_PID_NS=y
 CONFIG_NET_NS=y
 CONFIG_SCHED_AUTOGROUP=y
-CONFIG_MM_OWNER=y
 # CONFIG_SYSFS_DEPRECATED is not set
 CONFIG_RELAY=y
 CONFIG_BLK_DEV_INITRD=y
@@ -231,6 +228,7 @@
 CONFIG_HAVE_ALIGNED_STRUCT_PAGE=y
 CONFIG_HAVE_CMPXCHG_LOCAL=y
 CONFIG_HAVE_CMPXCHG_DOUBLE=y
+CONFIG_ARCH_WANT_COMPAT_IPC_PARSE_VERSION=y
 CONFIG_ARCH_WANT_OLD_COMPAT_IPC=y
 CONFIG_HAVE_ARCH_SECCOMP_FILTER=y
 CONFIG_SECCOMP_FILTER=y
@@ -429,6 +427,7 @@
 CONFIG_HAVE_MEMBLOCK=y
 CONFIG_HAVE_MEMBLOCK_NODE_MAP=y
 CONFIG_ARCH_DISCARD_MEMBLOCK=y
+CONFIG_MEMORY_ISOLATION=y
 CONFIG_MEMORY_HOTPLUG=y
 CONFIG_MEMORY_HOTPLUG_SPARSE=y
 CONFIG_MEMORY_HOTREMOVE=y
@@ -505,7 +504,7 @@
 CONFIG_PM_DEBUG=y
 CONFIG_PM_ADVANCED_DEBUG=y
 # CONFIG_PM_TEST_SUSPEND is not set
-CONFIG_CAN_PM_TRACE=y
+CONFIG_PM_SLEEP_DEBUG=y
 # CONFIG_PM_TRACE_RTC is not set
 CONFIG_ACPI=y
 CONFIG_ACPI_SLEEP=y
@@ -577,6 +576,7 @@
 CONFIG_CPU_IDLE=y
 CONFIG_CPU_IDLE_GOV_LADDER=y
 CONFIG_CPU_IDLE_GOV_MENU=y
+# CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED is not set
 CONFIG_INTEL_IDLE=y
 
 #
@@ -712,6 +712,7 @@
 CONFIG_IP_PIMSM_V2=y
 # CONFIG_ARPD is not set
 CONFIG_SYN_COOKIES=y
+CONFIG_NET_IPVTI=m
 CONFIG_INET_AH=m
 CONFIG_INET_ESP=m
 CONFIG_INET_IPCOMP=m
@@ -803,6 +804,8 @@
 CONFIG_NF_CONNTRACK_TFTP=m
 CONFIG_NF_CT_NETLINK=m
 CONFIG_NF_CT_NETLINK_TIMEOUT=m
+CONFIG_NF_CT_NETLINK_HELPER=m
+CONFIG_NETFILTER_NETLINK_QUEUE_CT=y
 CONFIG_NETFILTER_TPROXY=m
 CONFIG_NETFILTER_XTABLES=m
 
@@ -1255,6 +1258,7 @@
 CONFIG_BT_HCIUART_BCSP=y
 CONFIG_BT_HCIUART_ATH3K=y
 CONFIG_BT_HCIUART_LL=y
+CONFIG_BT_HCIUART_3WIRE=y
 CONFIG_BT_HCIBCM203X=m
 CONFIG_BT_HCIBPA10X=m
 CONFIG_BT_HCIBFUSB=m
@@ -1278,14 +1282,13 @@
 CONFIG_WEXT_SPY=y
 CONFIG_WEXT_PRIV=y
 CONFIG_CFG80211=m
-CONFIG_NL80211_TESTMODE=y
+# CONFIG_NL80211_TESTMODE is not set
 # CONFIG_CFG80211_DEVELOPER_WARNINGS is not set
 # CONFIG_CFG80211_REG_DEBUG is not set
 CONFIG_CFG80211_DEFAULT_PS=y
 # CONFIG_CFG80211_DEBUGFS is not set
 # CONFIG_CFG80211_INTERNAL_REGDB is not set
 CONFIG_CFG80211_WEXT=y
-CONFIG_WIRELESS_EXT_SYSFS=y
 CONFIG_LIB80211=m
 CONFIG_LIB80211_CRYPT_WEP=m
 CONFIG_LIB80211_CRYPT_CCMP=m
@@ -1300,6 +1303,7 @@
 CONFIG_MAC80211_MESH=y
 CONFIG_MAC80211_LEDS=y
 # CONFIG_MAC80211_DEBUGFS is not set
+# CONFIG_MAC80211_MESSAGE_TRACING is not set
 # CONFIG_MAC80211_DEBUG_MENU is not set
 CONFIG_WIMAX=m
 CONFIG_WIMAX_DEBUG_LEVEL=8
@@ -1351,6 +1355,7 @@
 # CONFIG_GENERIC_CPU_DEVICES is not set
 CONFIG_REGMAP=y
 CONFIG_REGMAP_I2C=m
+CONFIG_REGMAP_IRQ=y
 CONFIG_DMA_SHARED_BUFFER=y
 CONFIG_CONNECTOR=y
 CONFIG_PROC_EVENTS=y
@@ -1575,9 +1580,6 @@
 CONFIG_CB710_CORE=m
 # CONFIG_CB710_DEBUG is not set
 CONFIG_CB710_DEBUG_ASSUMPTIONS=y
-CONFIG_IWMC3200TOP=m
-# CONFIG_IWMC3200TOP_DEBUG is not set
-# CONFIG_IWMC3200TOP_DEBUGFS is not set
 
 #
 # Texas Instruments shared transport line discipline
@@ -1619,7 +1621,6 @@
 # CONFIG_SCSI_CONSTANTS is not set
 # CONFIG_SCSI_LOGGING is not set
 # CONFIG_SCSI_SCAN_ASYNC is not set
-CONFIG_SCSI_WAIT_SCAN=m
 
 #
 # SCSI Transports
@@ -1854,7 +1855,6 @@
 CONFIG_DM_SNAPSHOT=m
 CONFIG_DM_THIN_PROVISIONING=m
 # CONFIG_DM_DEBUG_BLOCK_STACK_TRACING is not set
-# CONFIG_DM_DEBUG_SPACE_MAPS is not set
 CONFIG_DM_MIRROR=m
 CONFIG_DM_RAID=m
 CONFIG_DM_LOG_USERSPACE=m
@@ -1914,6 +1914,7 @@
 CONFIG_IEEE802154_FAKELB=m
 CONFIG_IFB=m
 CONFIG_NET_TEAM=m
+CONFIG_NET_TEAM_MODE_BROADCAST=m
 CONFIG_NET_TEAM_MODE_ROUNDROBIN=m
 CONFIG_NET_TEAM_MODE_ACTIVEBACKUP=m
 CONFIG_NET_TEAM_MODE_LOADBALANCE=m
@@ -2170,6 +2171,7 @@
 CONFIG_VITESSE_PHY=m
 CONFIG_SMSC_PHY=m
 CONFIG_BROADCOM_PHY=m
+CONFIG_BCM87XX_PHY=m
 CONFIG_ICPLUS_PHY=m
 CONFIG_REALTEK_PHY=m
 CONFIG_NATIONAL_PHY=m
@@ -2330,13 +2332,13 @@
 CONFIG_LIBIPW=m
 # CONFIG_LIBIPW_DEBUG is not set
 CONFIG_IWLWIFI=m
+CONFIG_IWLDVM=m
 
 #
 # Debugging Options
 #
 # CONFIG_IWLWIFI_DEBUG is not set
 CONFIG_IWLWIFI_DEVICE_TRACING=y
-CONFIG_IWLWIFI_DEVICE_TESTMODE=y
 # CONFIG_IWLWIFI_P2P is not set
 # CONFIG_IWLWIFI_EXPERIMENTAL_MFP is not set
 CONFIG_IWLEGACY=m
@@ -2347,9 +2349,6 @@
 # iwl3945 / iwl4965 Debugging Options
 #
 # CONFIG_IWLEGACY_DEBUG is not set
-CONFIG_IWM=m
-# CONFIG_IWM_DEBUG is not set
-CONFIG_IWM_TRACING=y
 CONFIG_LIBERTAS=m
 CONFIG_LIBERTAS_USB=m
 CONFIG_LIBERTAS_CS=m
@@ -2378,6 +2377,7 @@
 CONFIG_RT2800PCI_RT33XX=y
 CONFIG_RT2800PCI_RT35XX=y
 CONFIG_RT2800PCI_RT53XX=y
+CONFIG_RT2800PCI_RT3290=y
 CONFIG_RT2500USB=m
 CONFIG_RT73USB=m
 CONFIG_RT2800USB=m
@@ -2404,6 +2404,7 @@
 CONFIG_WL1251=m
 CONFIG_WL1251_SDIO=m
 CONFIG_WL12XX=m
+CONFIG_WL18XX=m
 CONFIG_WLCORE=m
 CONFIG_WLCORE_SDIO=m
 CONFIG_WL12XX_PLATFORM_DATA=y
@@ -2419,8 +2420,6 @@
 #
 CONFIG_WIMAX_I2400M=m
 CONFIG_WIMAX_I2400M_USB=m
-CONFIG_WIMAX_I2400M_SDIO=m
-# CONFIG_WIMAX_IWMC3200_SDIO is not set
 CONFIG_WIMAX_I2400M_DEBUG_LEVEL=8
 # CONFIG_WAN is not set
 CONFIG_XEN_NETDEV_FRONTEND=m
@@ -2680,10 +2679,12 @@
 CONFIG_TOUCHSCREEN_WACOM_I2C=m
 CONFIG_TOUCHSCREEN_MAX11801=m
 CONFIG_TOUCHSCREEN_MCS5000=m
+CONFIG_TOUCHSCREEN_MMS114=m
 CONFIG_TOUCHSCREEN_MTOUCH=m
 CONFIG_TOUCHSCREEN_INEXIO=m
 CONFIG_TOUCHSCREEN_MK712=m
 CONFIG_TOUCHSCREEN_PENMOUNT=m
+CONFIG_TOUCHSCREEN_EDT_FT5X06=m
 CONFIG_TOUCHSCREEN_TOUCHRIGHT=m
 CONFIG_TOUCHSCREEN_TOUCHWIN=m
 CONFIG_TOUCHSCREEN_UCB1400=m
@@ -2978,6 +2979,7 @@
 #
 # I2C GPIO expanders:
 #
+CONFIG_GPIO_ARIZONA=m
 # CONFIG_GPIO_MAX7300 is not set
 # CONFIG_GPIO_MAX732X is not set
 # CONFIG_GPIO_PCA953X is not set
@@ -2988,6 +2990,7 @@
 # PCI GPIO expanders:
 #
 CONFIG_GPIO_CS5535=m
+CONFIG_GPIO_AMD8111=m
 CONFIG_GPIO_LANGWELL=y
 # CONFIG_GPIO_PCH is not set
 # CONFIG_GPIO_ML_IOH is not set
@@ -3024,6 +3027,7 @@
 # CONFIG_CHARGER_LP8727 is not set
 # CONFIG_CHARGER_GPIO is not set
 # CONFIG_CHARGER_SMB347 is not set
+CONFIG_POWER_AVS=y
 CONFIG_HWMON=y
 CONFIG_HWMON_VID=m
 # CONFIG_HWMON_DEBUG_CHIP is not set
@@ -3062,6 +3066,7 @@
 CONFIG_SENSORS_GL518SM=m
 CONFIG_SENSORS_GL520SM=m
 # CONFIG_SENSORS_GPIO_FAN is not set
+CONFIG_SENSORS_HIH6130=m
 CONFIG_SENSORS_CORETEMP=m
 CONFIG_SENSORS_IBMAEM=m
 CONFIG_SENSORS_IBMPEX=m
@@ -3238,12 +3243,13 @@
 CONFIG_BCMA_BLOCKIO=y
 CONFIG_BCMA_HOST_PCI_POSSIBLE=y
 CONFIG_BCMA_HOST_PCI=y
+CONFIG_BCMA_DRIVER_GMAC_CMN=y
 # CONFIG_BCMA_DEBUG is not set
 
 #
 # Multifunction device drivers
 #
-CONFIG_MFD_CORE=m
+CONFIG_MFD_CORE=y
 CONFIG_MFD_SM501=m
 CONFIG_MFD_SM501_GPIO=y
 CONFIG_HTC_PASIC3=m
@@ -3254,6 +3260,10 @@
 CONFIG_TPS6507X=m
 CONFIG_MFD_TPS65217=m
 # CONFIG_MFD_TMIO is not set
+CONFIG_MFD_ARIZONA=y
+CONFIG_MFD_ARIZONA_I2C=m
+CONFIG_MFD_WM5102=y
+CONFIG_MFD_WM5110=y
 CONFIG_MFD_PCF50633=m
 CONFIG_PCF50633_ADC=m
 CONFIG_PCF50633_GPIO=m
@@ -3275,22 +3285,28 @@
 #
 # Multimedia core support
 #
+CONFIG_MEDIA_CAMERA_SUPPORT=y
+CONFIG_MEDIA_ANALOG_TV_SUPPORT=y
+CONFIG_MEDIA_DIGITAL_TV_SUPPORT=y
+CONFIG_MEDIA_RADIO_SUPPORT=y
+CONFIG_MEDIA_RC_SUPPORT=y
 CONFIG_MEDIA_CONTROLLER=y
 CONFIG_VIDEO_DEV=m
 CONFIG_VIDEO_V4L2_COMMON=m
 CONFIG_VIDEO_V4L2_SUBDEV_API=y
 CONFIG_DVB_CORE=m
 CONFIG_DVB_NET=y
-CONFIG_VIDEO_MEDIA=m
 
 #
-# Multimedia drivers
+# Media drivers
 #
 CONFIG_VIDEO_SAA7146=m
 CONFIG_VIDEO_SAA7146_VV=m
 CONFIG_RC_CORE=m
-CONFIG_LIRC=m
 CONFIG_RC_MAP=m
+CONFIG_RC_DECODERS=y
+CONFIG_LIRC=m
+CONFIG_IR_LIRC_CODEC=m
 CONFIG_IR_NEC_DECODER=m
 CONFIG_IR_RC5_DECODER=m
 CONFIG_IR_RC6_DECODER=m
@@ -3299,7 +3315,7 @@
 CONFIG_IR_RC5_SZ_DECODER=m
 CONFIG_IR_SANYO_DECODER=m
 CONFIG_IR_MCE_KBD_DECODER=m
-CONFIG_IR_LIRC_CODEC=m
+CONFIG_RC_DEVICES=y
 CONFIG_RC_ATI_REMOTE=m
 CONFIG_IR_ENE=m
 CONFIG_IR_IMON=m
@@ -3310,6 +3326,7 @@
 CONFIG_IR_REDRAT3=m
 CONFIG_IR_STREAMZAP=m
 CONFIG_IR_WINBOND_CIR=m
+CONFIG_IR_IGUANA=m
 CONFIG_RC_LOOPBACK=m
 CONFIG_IR_GPIO_CIR=m
 CONFIG_MEDIA_ATTACH=y
@@ -3337,6 +3354,8 @@
 CONFIG_MEDIA_TUNER_MAX2165=m
 CONFIG_MEDIA_TUNER_TDA18218=m
 CONFIG_MEDIA_TUNER_FC0011=m
+CONFIG_MEDIA_TUNER_FC0012=m
+CONFIG_MEDIA_TUNER_FC0013=m
 CONFIG_MEDIA_TUNER_TDA18212=m
 CONFIG_MEDIA_TUNER_TUA9001=m
 CONFIG_VIDEO_V4L2=m
@@ -3419,6 +3438,7 @@
 CONFIG_VIDEO_ADV7170=m
 CONFIG_VIDEO_ADV7175=m
 CONFIG_VIDEO_ADV7343=m
+CONFIG_VIDEO_ADV7393=m
 CONFIG_VIDEO_AK881X=m
 
 #
@@ -3457,7 +3477,10 @@
 CONFIG_VIDEO_M52790=m
 CONFIG_VIDEO_VIVI=m
 CONFIG_V4L_USB_DRIVERS=y
-CONFIG_VIDEO_AU0828=m
+
+#
+# Webcam devices
+#
 CONFIG_USB_VIDEO_CLASS=m
 CONFIG_USB_VIDEO_CLASS_INPUT_EVDEV=y
 CONFIG_USB_GSPCA=m
@@ -3507,15 +3530,32 @@
 CONFIG_USB_GSPCA_VICAM=m
 CONFIG_USB_GSPCA_XIRLINK_CIT=m
 CONFIG_USB_GSPCA_ZC3XX=m
+CONFIG_USB_PWC=m
+# CONFIG_USB_PWC_DEBUG is not set
+CONFIG_USB_PWC_INPUT_EVDEV=y
+CONFIG_VIDEO_CPIA2=m
+CONFIG_USB_ZR364XX=m
+CONFIG_USB_STKWEBCAM=m
+CONFIG_USB_S2255=m
+CONFIG_USB_SN9C102=m
+
+#
+# Webcam and/or TV USB devices
+#
+CONFIG_VIDEO_EM28XX=m
+CONFIG_VIDEO_EM28XX_ALSA=m
+CONFIG_VIDEO_EM28XX_DVB=m
+CONFIG_VIDEO_EM28XX_RC=m
+
+#
+# TV USB devices
+#
+CONFIG_VIDEO_AU0828=m
 CONFIG_VIDEO_PVRUSB2=m
 CONFIG_VIDEO_PVRUSB2_SYSFS=y
 CONFIG_VIDEO_PVRUSB2_DVB=y
 # CONFIG_VIDEO_PVRUSB2_DEBUGIFC is not set
 CONFIG_VIDEO_HDPVR=m
-CONFIG_VIDEO_EM28XX=m
-CONFIG_VIDEO_EM28XX_ALSA=m
-CONFIG_VIDEO_EM28XX_DVB=m
-CONFIG_VIDEO_EM28XX_RC=m
 CONFIG_VIDEO_TLG2300=m
 CONFIG_VIDEO_CX231XX=m
 CONFIG_VIDEO_CX231XX_RC=y
@@ -3525,14 +3565,6 @@
 CONFIG_VIDEO_TM6000_ALSA=m
 CONFIG_VIDEO_TM6000_DVB=m
 CONFIG_VIDEO_USBVISION=m
-CONFIG_USB_SN9C102=m
-CONFIG_USB_PWC=m
-# CONFIG_USB_PWC_DEBUG is not set
-CONFIG_USB_PWC_INPUT_EVDEV=y
-CONFIG_VIDEO_CPIA2=m
-CONFIG_USB_ZR364XX=m
-CONFIG_USB_STKWEBCAM=m
-CONFIG_USB_S2255=m
 CONFIG_V4L_PCI_DRIVERS=y
 CONFIG_VIDEO_BT848=m
 CONFIG_VIDEO_BT848_DVB=y
@@ -3578,6 +3610,8 @@
 CONFIG_USB_MR800=m
 CONFIG_USB_DSBR=m
 CONFIG_RADIO_MAXIRADIO=m
+CONFIG_RADIO_SHARK=m
+CONFIG_RADIO_SHARK2=m
 CONFIG_I2C_SI4713=m
 CONFIG_RADIO_SI4713=m
 CONFIG_USB_KEENE=m
@@ -3712,7 +3746,6 @@
 #
 # Supported DVB Frontends
 #
-# CONFIG_DVB_FE_CUSTOMISE is not set
 
 #
 # Multistandard (satellite) frontends
@@ -3779,6 +3812,7 @@
 CONFIG_DVB_STV0367=m
 CONFIG_DVB_CXD2820R=m
 CONFIG_DVB_RTL2830=m
+CONFIG_DVB_RTL2832=m
 
 #
 # DVB-C (cable) frontends
@@ -3876,6 +3910,7 @@
 CONFIG_DRM_VIA=m
 CONFIG_DRM_SAVAGE=m
 CONFIG_DRM_VMWGFX=m
+CONFIG_DRM_VMWGFX_FBCON=y
 CONFIG_DRM_GMA500=m
 CONFIG_DRM_GMA600=y
 CONFIG_DRM_GMA3600=y
@@ -4165,6 +4200,7 @@
 #
 CONFIG_HID=m
 CONFIG_HIDRAW=y
+CONFIG_UHID=m
 CONFIG_HID_GENERIC=m
 
 #
@@ -4195,6 +4231,7 @@
 CONFIG_HID_TWINHAN=m
 CONFIG_HID_KENSINGTON=m
 CONFIG_HID_LCPOWER=m
+CONFIG_HID_LENOVO_TPKBD=m
 CONFIG_HID_LOGITECH=m
 CONFIG_HID_LOGITECH_DJ=m
 CONFIG_LOGITECH_FF=y
@@ -4583,13 +4620,16 @@
 CONFIG_LEDS_DELL_NETBOOKS=m
 CONFIG_LEDS_MC13783=m
 CONFIG_LEDS_TCA6507=m
+CONFIG_LEDS_LM3556=m
 CONFIG_LEDS_OT200=m
+CONFIG_LEDS_BLINKM=m
 CONFIG_LEDS_TRIGGERS=y
 
 #
 # LED Triggers
 #
 CONFIG_LEDS_TRIGGER_TIMER=m
+CONFIG_LEDS_TRIGGER_ONESHOT=m
 CONFIG_LEDS_TRIGGER_HEARTBEAT=m
 CONFIG_LEDS_TRIGGER_BACKLIGHT=m
 CONFIG_LEDS_TRIGGER_GPIO=m
@@ -4607,6 +4647,7 @@
 #
 # Reporting subsystems
 #
+CONFIG_EDAC_LEGACY_SYSFS=y
 # CONFIG_EDAC_DEBUG is not set
 CONFIG_EDAC_DECODE_MCE=m
 CONFIG_EDAC_MCE_INJ=m
@@ -4722,6 +4763,7 @@
 CONFIG_UIO_SERCOS3=m
 CONFIG_UIO_PCI_GENERIC=m
 CONFIG_UIO_NETX=m
+# CONFIG_VFIO is not set
 CONFIG_VIRTIO=m
 CONFIG_VIRTIO_RING=m
 
@@ -4759,6 +4801,7 @@
 CONFIG_XEN_PCIDEV_BACKEND=m
 CONFIG_XEN_PRIVCMD=m
 CONFIG_XEN_ACPI_PROCESSOR=m
+# CONFIG_XEN_MCE_LOG is not set
 CONFIG_STAGING=y
 CONFIG_ET131X=m
 CONFIG_SLICOSS=m
@@ -4773,14 +4816,14 @@
 # CONFIG_COMEDI_DEBUG is not set
 CONFIG_COMEDI_DEFAULT_BUF_SIZE_KB=2048
 CONFIG_COMEDI_DEFAULT_BUF_MAXSIZE_KB=20480
-CONFIG_COMEDI_MISC_DRIVERS=m
+CONFIG_COMEDI_MISC_DRIVERS=y
 CONFIG_COMEDI_KCOMEDILIB=m
 CONFIG_COMEDI_BOND=m
 CONFIG_COMEDI_TEST=m
 CONFIG_COMEDI_PARPORT=m
 CONFIG_COMEDI_SERIAL2002=m
 CONFIG_COMEDI_SKEL=m
-CONFIG_COMEDI_PCI_DRIVERS=m
+CONFIG_COMEDI_PCI_DRIVERS=y
 CONFIG_COMEDI_ADDI_APCI_035=m
 CONFIG_COMEDI_ADDI_APCI_1032=m
 CONFIG_COMEDI_ADDI_APCI_1500=m
@@ -4832,13 +4875,16 @@
 CONFIG_COMEDI_NI_65XX=m
 CONFIG_COMEDI_NI_660X=m
 CONFIG_COMEDI_NI_670X=m
+CONFIG_COMEDI_NI_LABPC=m
 CONFIG_COMEDI_NI_PCIDIO=m
 CONFIG_COMEDI_NI_PCIMIO=m
 CONFIG_COMEDI_RTD520=m
 CONFIG_COMEDI_S526=m
 CONFIG_COMEDI_S626=m
 CONFIG_COMEDI_SSV_DNP=m
-CONFIG_COMEDI_PCMCIA_DRIVERS=m
+CONFIG_COMEDI_MITE=m
+CONFIG_COMEDI_NI_TIOCMD=m
+CONFIG_COMEDI_PCMCIA_DRIVERS=y
 CONFIG_COMEDI_CB_DAS16_CS=m
 CONFIG_COMEDI_DAS08_CS=m
 CONFIG_COMEDI_NI_DAQ_700_CS=m
@@ -4846,22 +4892,19 @@
 CONFIG_COMEDI_NI_LABPC_CS=m
 CONFIG_COMEDI_NI_MIO_CS=m
 CONFIG_COMEDI_QUATECH_DAQP_CS=m
-CONFIG_COMEDI_USB_DRIVERS=m
+CONFIG_COMEDI_USB_DRIVERS=y
 CONFIG_COMEDI_DT9812=m
 CONFIG_COMEDI_USBDUX=m
 CONFIG_COMEDI_USBDUXFAST=m
 CONFIG_COMEDI_USBDUXSIGMA=m
 CONFIG_COMEDI_VMK80XX=m
-CONFIG_COMEDI_NI_COMMON=m
-CONFIG_COMEDI_MITE=m
-CONFIG_COMEDI_NI_TIO=m
-CONFIG_COMEDI_NI_LABPC=m
 CONFIG_COMEDI_8255=m
 CONFIG_COMEDI_FC=m
 CONFIG_COMEDI_AMPLC_DIO200=m
 CONFIG_COMEDI_AMPLC_PC236=m
 CONFIG_COMEDI_AMPLC_PC263=m
 CONFIG_COMEDI_DAS08=m
+CONFIG_COMEDI_NI_TIO=m
 CONFIG_ASUS_OLED=m
 # CONFIG_PANEL is not set
 CONFIG_R8187SE=m
@@ -4973,6 +5016,7 @@
 CONFIG_WIMAX_GDM72XX_USB=y
 # CONFIG_WIMAX_GDM72XX_SDIO is not set
 CONFIG_WIMAX_GDM72XX_USB_PM=y
+CONFIG_CSR_WIFI=m
 CONFIG_X86_PLATFORM_DEVICES=y
 CONFIG_ACER_WMI=m
 CONFIG_ACERHDF=m
@@ -5052,9 +5096,11 @@
 # Extcon Device Drivers
 #
 CONFIG_EXTCON_GPIO=m
+CONFIG_EXTCON_ARIZONA=m
 CONFIG_MEMORY=y
 # CONFIG_IIO is not set
 # CONFIG_VME_BUS is not set
+# CONFIG_PWM is not set
 
 #
 # Firmware Drivers
@@ -5226,6 +5272,8 @@
 # CONFIG_QNX6FS_FS is not set
 # CONFIG_ROMFS_FS is not set
 CONFIG_PSTORE=y
+# CONFIG_PSTORE_CONSOLE is not set
+# CONFIG_PSTORE_FTRACE is not set
 CONFIG_PSTORE_RAM=m
 # CONFIG_SYSV_FS is not set
 CONFIG_UFS_FS=m
@@ -5236,10 +5284,11 @@
 CONFIG_ORE=m
 CONFIG_NETWORK_FILESYSTEMS=y
 CONFIG_NFS_FS=m
-CONFIG_NFS_V2=y
-CONFIG_NFS_V3=y
+CONFIG_NFS_V2=m
+CONFIG_NFS_V3=m
 CONFIG_NFS_V3_ACL=y
-CONFIG_NFS_V4=y
+CONFIG_NFS_V4=m
+# CONFIG_NFS_SWAP is not set
 # CONFIG_NFS_V4_1 is not set
 CONFIG_NFS_FSCACHE=y
 # CONFIG_NFS_USE_LEGACY_DNS is not set
@@ -5416,7 +5465,7 @@
 # CONFIG_DEBUG_FORCE_WEAK_PER_CPU is not set
 # CONFIG_DEBUG_PER_CPU_MAPS is not set
 CONFIG_LKDTM=m
-CONFIG_CPU_NOTIFIER_ERROR_INJECT=m
+# CONFIG_NOTIFIER_ERROR_INJECTION is not set
 # CONFIG_FAULT_INJECTION is not set
 # CONFIG_LATENCYTOP is not set
 # CONFIG_DEBUG_PAGEALLOC is not set
@@ -5482,6 +5531,7 @@
 # CONFIG_DEBUG_RODATA_TEST is not set
 # CONFIG_DEBUG_SET_MODULE_RONX is not set
 # CONFIG_DEBUG_NX_TEST is not set
+# CONFIG_DEBUG_TLBFLUSH is not set
 # CONFIG_IOMMU_DEBUG is not set
 # CONFIG_IOMMU_STRESS is not set
 CONFIG_HAVE_MMIOTRACE_SUPPORT=y
@@ -5567,6 +5617,8 @@
 CONFIG_CRYPTO_CRYPTD=m
 CONFIG_CRYPTO_AUTHENC=m
 CONFIG_CRYPTO_TEST=m
+CONFIG_CRYPTO_ABLK_HELPER_X86=m
+CONFIG_CRYPTO_GLUE_HELPER_X86=m
 
 #
 # Authenticated Encryption with Associated Data
@@ -5637,11 +5689,13 @@
 CONFIG_CRYPTO_SEED=m
 CONFIG_CRYPTO_SERPENT=m
 CONFIG_CRYPTO_SERPENT_SSE2_X86_64=m
+CONFIG_CRYPTO_SERPENT_AVX_X86_64=m
 CONFIG_CRYPTO_TEA=m
 CONFIG_CRYPTO_TWOFISH=m
 CONFIG_CRYPTO_TWOFISH_COMMON=m
 CONFIG_CRYPTO_TWOFISH_X86_64=m
 CONFIG_CRYPTO_TWOFISH_X86_64_3WAY=m
+CONFIG_CRYPTO_TWOFISH_AVX_X86_64=m
 
 #
 # Compression
@@ -5674,6 +5728,7 @@
 CONFIG_KVM_AMD=m
 CONFIG_KVM_MMU_AUDIT=y
 CONFIG_VHOST_NET=m
+CONFIG_TCM_VHOST=m
 CONFIG_BINARY_PRINTF=y
 
 #
@@ -5737,6 +5792,7 @@
 CONFIG_CPU_RMAP=y
 CONFIG_DQL=y
 CONFIG_NLATTR=y
+CONFIG_ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE=y
 CONFIG_LRU_CACHE=m
 CONFIG_AVERAGE=y
 CONFIG_CORDIC=m

Last edited by jrk (2012-10-17 18:47:57)

Gusar

Member

Registered: 2009-08-25

Posts: 3,605

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

Seriously, stuff like this makes think about switching to Mac (which I would never do) or *cough* Windows *cough*. Sorry, just venting off a bit.

You're using a bleeding edge distro. Instead of going to a whole different OS, wouldn't it make much more sense to stay where you are, but dial back the bleeding-edge-ness by using a LTS kernel? There's two LTS kernels now (3.4 and 3.0), though Arch only packages the latter.

ANOKNUSA

Member

Registered: 2010-10-22

Posts: 2,141

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

Two things that stand out to me here:

1) My boot parameters:

ro quiet elevator=bfq pcie_aspm=force i915.i915_enable_rc6=1 i915.i915_enable_fbc=1 i915.lvds_downclock=1

The section "pcie_aspm=force" was added to fix an earlier regression; it's my understanding that this was supposed to be fixed around 3.4 or 3.5 (not sure which), and I just forgot to remove it.  However, no one else here mentioned using it, so I there it is.

2) This only seems to affect folks with hybrid graphics, so: Is everyone here testing this with only the integrated graphics enabled (i.e. discrete chip is disabled in bios or through acpi)?

I'm bringing these things up because I'm not experiencing this issue with my setup; whether booting clean or waking from suspend, my temps are currently 33-41 degress celsius across four cores, all scaling properly. (i7-2720QM Sandy Bridge, Intel HD 3000 integrated graphics only, system completely up-to-date).

ASPM Power Regression was fixed long time ago and ported back and forth to probably all stable kernels. It is not required anymore.

I started noticing new power regression around 3.4 which have continued to increase and Sandy Bridge seems to be main hardware affected. I do have Optimus but I have disabled it in BIOS and been using power saving Intel HD 3000.

My current power usage is 12W and temp is 50C. Last night while copying some 80GB files to an external hard drive my system temperature spiked to 97C. I had a kernel panic. This seems to be very erratic.

Hmm, I can't see why else I might not be affected by this.  And I figured ASPM probably wasn't a cause (haven't changed my syslinux config in a long time, obviously), I'm just trying to figure out what made my setup different from all of yours, since I'm using the same kernel and much of the same hardware everyone else is.

---

Linux != Windows  ::  Ask Questions the Smart Way  ::  Configs

donniezazen

Member

From: Salt Lake City

Registered: 2011-06-24

Posts: 671

Website

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

Seriously, stuff like this makes think about switching to Mac (which I would never do) or *cough* Windows *cough*. Sorry, just venting off a bit.

You're using a bleeding edge distro. Instead of going to a whole different OS, wouldn't it make much more sense to stay where you are, but dial back the bleeding-edge-ness by using a LTS kernel? There's two LTS kernels now (3.4 and 3.0), though Arch only packages the latter.

1. 3.0 is sluggish.
2. Systemd fails or doesn't play nice on 3.0. Haven't done anything to remedy this failure because of point #1.
3. New power regression bug are introduced some where around 3.4 as reported in forum and arch bug tracker.

---

github | blog

cfr

Member

From: Cymru

Registered: 2011-11-27

Posts: 7,148

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

You know all those i915 parameters you two are using?  Check out

# systool -m i915 -av

I set all those things at first, because I wasn't sure what the setting of -1 meant in those cases.  Apparently those things are defaulting to being on.  For instance, the i915_enable_rc6 can be set to 0, 1, 3, or 7.  1 will enable rc6, 3 will enable rc6p also, and 7 will enable rc6pp in addition to the other two.  But without setting it to 1 on my own, it actually sits in rc6p most of the time (which is a lower power state than rc6, and rc6pp is even lower). 

So I think that all the rc6 setting will actually set it to use ore power than the default.  Frame buffer compression appears to be on by default, as does semaphores.  The only one I think you are having a real effect with is the lvds_downclock setting.

I am not 100% on this.  But I noticed my Ivy Bridge machine actually uses less power when I don't set all that crap, though I have not tried the lvds_downclock parameter.

I don't understand the output from systool, to be honest. The output from modinfo suggests that "-1" is "use chip default" which I assumed meant that it would depend on the particular chip and therefore might or might not be enabled for my particular machine.

Note that I'm not questioning your conclusions - I'd just like to understand the nature of the evidence they are based on.

I can see why you *might* get better savings by not setting this stuff - if your chip defaults to something more economical than 1 for i915_enable_rc6, for example. (Or you set it to 7 - are you still doing that?)

EDIT: When I first set this stuff up, I definitely got power-savings. But that doesn't mean it is still the best option. Nor does it mean that nothing I did is detrimental...

Last edited by cfr (2012-10-17 22:24:05)

---

CLI Paste | How To Ask Questions

Arch Linux | x86_64 | GPT | EFI boot | refind | stub loader | systemd | LVM2 on LUKS
Lenovo x270 | Intel(R) Core(TM) i5-7200U CPU @ 2.50GHz | Intel Wireless 8265/8275 | US keyboard w/ Euro | 512G NVMe INTEL SSDPEKKF512G7L

hendry

Member

From: Singapore

Registered: 2009-09-08

Posts: 157

Website

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

I'm also suffering with this heat issue on 3.6.2-1-ARCH. My X220 can't be safely called a "laptop" now since it's running much hotter. I'm using no related /proc/cmdline arguments, but maybe I should revisit them.

jrk

Member

From: Nämberch

Registered: 2012-10-16

Posts: 37

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

Seriously, stuff like this makes think about switching to Mac (which I would never do) or *cough* Windows *cough*. Sorry, just venting off a bit.

You're using a bleeding edge distro. Instead of going to a whole different OS, wouldn't it make much more sense to stay where you are, but dial back the bleeding-edge-ness by using a LTS kernel? There's two LTS kernels now (3.4 and 3.0), though Arch only packages the latter.

1. 3.0 is sluggish.
2. Systemd fails or doesn't play nice on 3.0. Haven't done anything to remedy this failure because of point #1.
3. New power regression bug are introduced some where around 3.4 as reported in forum and arch bug tracker.

Didn't want to make this thread ot.
But donniezazen points it out quite well. 3.4 LTS would probably be nice. Or some kind of "stable" checkpoints for pacman, where you could just say "ok, the last checkpoint from one month ago, that was fine, so I'll just install all packages covered by that checkpoint".

I'm currently bisecting. It's a PITA though, compiling the whole kernel takes roughly one hour. I've thrown out the majority of device drivers where I'm sure I don't need it. Still takes an eternity to build.

git bisect start
# good: [73b6448a7705298b2b10367a50fd063b27cdbeb8] Linux 3.5.6
git bisect good 73b6448a7705298b2b10367a50fd063b27cdbeb8
# bad: [a0d271cbfed1dd50278c6b06bead3d00ba0a88f9] Linux 3.6
git bisect bad a0d271cbfed1dd50278c6b06bead3d00ba0a88f9
# good: [28a33cbc24e4256c143dce96c7d93bf423229f92] Linux 3.5
git bisect good 28a33cbc24e4256c143dce96c7d93bf423229f92
# good: [b13bc8dda81c54a66a1c84e66f60b8feba659f28] Merge tag 'staging-3.6-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging
git bisect good b13bc8dda81c54a66a1c84e66f60b8feba659f28
# good: [991b3137f21e13db4711f313edbe67d49bed795b] [media] media: soc_camera: don't clear pix->sizeimage in JPEG mode
git bisect good 991b3137f21e13db4711f313edbe67d49bed795b
# bad: [c67fe3752abe6ab47639e2f9b836900c3dc3da84] mm: compaction: Abort async compaction if locks are contended or taking too long
git bisect bad c67fe3752abe6ab47639e2f9b836900c3dc3da84
# skip: [2e26c73a1e410448fbd2c0fbd34f06d98eaf8e48] Merge branch 'drm-nouveau-fixes' of git://git.freedesktop.org/git/nouveau/linux-2.6 into drm-fixes
git bisect skip 2e26c73a1e410448fbd2c0fbd34f06d98eaf8e48

KaiSforza

Member

Registered: 2012-04-22

Posts: 133

Website

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

Has anyone noticed anything besides the unchanging clock speeds?

(Sorry, neycho, but the I don't think the gnome shell is related to this. Possibly something with your DE having updated.)

---

Thinkpad T420 | Intel 3000 | systemd {,--user}
PKGBUILDs I use | pywer AUR helper

dontbugme

Banned

Registered: 2011-11-04

Posts: 166

Website

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

Hmm maybe I should try bisecting this too. I've customized the kernel config for my asus NB (built in modules at least for the . builtin hardware), and compiling this kernel will probably take about 5 min. or so on my core i7 desktop.

ulke

Member

Registered: 2011-04-29

Posts: 50

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

FYI, there's a kernel bugzilla entry.

dontbugme

Banned

Registered: 2011-11-04

Posts: 166

Website

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

ANOKNUSA is right regarding optimus. I bisected the kernel and (idk why) tried powerstat with 3.6.2. Atm I'm back again to ~8.5W. I think I ran after a red herring. The power regression I'm experiening has something to do with bumblebee/nvidia driver. Disabling the nvidia card with bbswitch and no nvidia kernel module installed seems to help. Guess the nvidia driver is at fault in my case. Sorry for the confusion.

KaiSforza

Member

Registered: 2012-04-22

Posts: 133

Website

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

Everyone, here is an Arch bug report (FS#32025) about the same thing. Please give all the help you can.

---

Thinkpad T420 | Intel 3000 | systemd {,--user}
PKGBUILDs I use | pywer AUR helper

jrk

Member

From: Nämberch

Registered: 2012-10-16

Posts: 37

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

Hi,

finished my bisect, results are here.

Seems to be an i915 issue. From my understanding this won't/can't be fixed quickly..

altercation

Member

From: Seattle

Registered: 2011-05-15

Posts: 136

Website

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

ThinkPad x220 tablet Core i5 here. Affected. Apparently (in multiple tests) solved by changing kernel command line arguments from:

i915_enable_rc6=1

to

i915_enable_rc6=7

Note that the first option (rc6=1) was fine under 3.5.6 but under 3.6.2 only rc6=7 stops power draw from spiking up to 25W+ (operating at ~10W now).

Still watching this as I've noticed other strange behavior under 3.6.2.

(EDIT: not totally fixed... see below: some random power consumption spikes still)

Last edited by altercation (2012-10-18 22:14:32)

---

Ethan Schoonover
Precision Colors - http://ethanschoonover.com/solarized

KaiSforza

Member

Registered: 2012-04-22

Posts: 133

Website

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

ThinkPad x220 tablet Core i5 here. Affected. Apparently (in multiple tests) solved by changing kernel command line arguments from:

i915_enable_rc6=1

to

i915_enable_rc6=7

Note that the first option (rc6=1) was fine under 3.5.6 but under 3.6.2 only rc6=7 stops power draw from spiking up to 25W+ (operating at ~10W now).

Still watching this as I've noticed other strange behavior under 3.6.2.

I've done that, actually currently have that in my kernel line, but it's random whether it works or not. Sometimes it will work on boot and after a suspend cycle or two, not work on boot, but afte a suspend magically work, or just not work at all.

---

Thinkpad T420 | Intel 3000 | systemd {,--user}
PKGBUILDs I use | pywer AUR helper

KaiSforza

Member

Registered: 2012-04-22

Posts: 133

Website

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

More news:

This was posted in the Arch bug report, and it seems to be related to what we are seeing. If anyone here has experience using the patch on Arch, post something confirming or denying it's effectiveness. I'm going to try this out myself when I get to an outlet. Good luck, everyone.

---

Thinkpad T420 | Intel 3000 | systemd {,--user}
PKGBUILDs I use | pywer AUR helper

altercation

Member

From: Seattle

Registered: 2011-05-15

Posts: 136

Website

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

ThinkPad x220 tablet Core i5 here. Affected. Apparently (in multiple tests) solved by changing kernel command line arguments from:

i915_enable_rc6=1

to

i915_enable_rc6=7

Note that the first option (rc6=1) was fine under 3.5.6 but under 3.6.2 only rc6=7 stops power draw from spiking up to 25W+ (operating at ~10W now).

Still watching this as I've noticed other strange behavior under 3.6.2.

I've done that, actually currently have that in my kernel line, but it's random whether it works or not. Sometimes it will work on boot and after a suspend cycle or two, not work on boot, but afte a suspend magically work, or just not work at all.

I'm confirming this too... I've had at least one power consumption spike again post resume. Can't consistently replicate but it's definitely not a solid fix.

Blech. Back to 3.5 it is then.

---

Ethan Schoonover
Precision Colors - http://ethanschoonover.com/solarized

KaiSforza

Member

Registered: 2012-04-22

Posts: 133

Website

Re: Kernel 3.6.2 Power Regressions (Sandy Bridge)

ThinkPad x220 tablet Core i5 here. Affected. Apparently (in multiple tests) solved by changing kernel command line arguments from:

i915_enable_rc6=1

to

i915_enable_rc6=7

Note that the first option (rc6=1) was fine under 3.5.6 but under 3.6.2 only rc6=7 stops power draw from spiking up to 25W+ (operating at ~10W now).

Still watching this as I've noticed other strange behavior under 3.6.2.

I've done that, actually currently have that in my kernel line, but it's random whether it works or not. Sometimes it will work on boot and after a suspend cycle or two, not work on boot, but afte a suspend magically work, or just not work at all.

I'm confirming this too... I've had at least one power consumption spike again post resume. Can't consistently replicate but it's definitely not a solid fix.

Blech. Back to 3.5 it is then.

Sadly, I think that's our only real option now.
Time to add an IgnorePkg list...

---

Thinkpad T420 | Intel 3000 | systemd {,--user}
PKGBUILDs I use | pywer AUR helper