http://bugzilla.opensuse.org/show_bug.cgi?id=1010973 http://bugzilla.opensuse.org/show_bug.cgi?id=1010973#c2 --- Comment #2 from Raf J --- (In reply to Raf J from comment #0)

I'm using an Intel E3-1245 v5 on a SuperMicro X11SAE-M.

Previously using Tumbleweed with later kernel versions and PC8 was achievable. Fresh install of Leap 42.2 (Kernel 4.4.27-2-default) and PC2 is the deepest package C-state (idle state) that is achieved.

I'm using integrated graphics and no other peripherals. Two SATA drives (No PCIe cards/drives or NVMe drives).

powertop was used to check Idle stats (version 2.8-1.5) -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1010973 http://bugzilla.opensuse.org/show_bug.cgi?id=1010973#c4 --- Comment #4 from Takashi Iwai --- I checked a Dell Latitude E7270 laptop with a Skylake chip, and it reached PC8 with 4.4.x kernel as well as 4.9-rc kernel. So it doesn't look like a generic problem, at least. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1010973 http://bugzilla.opensuse.org/show_bug.cgi?id=1010973#c6 --- Comment #6 from Raf J --- Created attachment 702984 --> http://bugzilla.opensuse.org/attachment.cgi?id=702984&action=edit hwinfo full output -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1010973 http://bugzilla.opensuse.org/show_bug.cgi?id=1010973#c8 --- Comment #8 from Raf J --- Just tried it with Kernel 4.8.9 and have the original state -- disabled PM on SATA devices and stuck in PC2. It sounds like a user space issue or version of PowerTOP as I have not controlled for that. Sorry I am not more helpful. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1010973 http://bugzilla.opensuse.org/show_bug.cgi?id=1010973#c10 Tino K changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |REOPENED CC| |dr.tino.k@gmail.com Resolution|FEATURE |--- --- Comment #10 from Tino K --- I'm observing exactly the same issue on a Dell Inspiron 3650 consumer PC with Intel i3-6100. I'm only ever reaching C2 (pc2) by default, and after enabling all power saving options in powertop under "Tunables" I'm reaching C3 (pc3) but not lower. Any hints on how to achieve lower states? # powertop PowerTOP 2.8 Overview Idle stats Frequency stats Device stats Tunab Package | Core | CPU 0 CPU 2 | | C0 active 0.5% 0.5% | | POLL 0.9% 4.4 ms 0.6% | | C1E-SKL 0.4% 0.2 ms 0.1% C2 (pc2) 60.5% | | C3 (pc3) 0.0% | C3 (cc3) 0.2% | C3-SKL 0.0% 0.1 ms 0.2% C6 (pc6) 0.0% | C6 (cc6) 1.7% | C6-SKL 1.7% 0.9 ms 0.5% C7 (pc7) 0.0% | C7 (cc7) 94.1% | C7s-SKL 0.0% 0.0 ms 0.0% C8 (pc8) 0.0% | | C8-SKL 95.6% 9.0 ms 97.6% 1 C9 (pc9) 0.0% | | C10 (pc10) 0.0% | | | Core | CPU 1 CPU 3 | | C0 active 1.3% 2.7% | | POLL 0.0% 0.0 ms 0.8% | | C1E-SKL 0.8% 0.4 ms 1.3% | | | C3 (cc3) 0.0% | C3-SKL 0.2% 0.3 ms 0.1% | C6 (cc6) 0.9% | C6-SKL 1.9% 0.6 ms 0.3% Then change "Tunables" to Bad Wireless Power Saving for interface wlan0 Bad Autosuspend for USB device Dell MS116 USB Optical Mouse [PixArt] Good NMI watchdog should be turned off Good Bluetooth device interface status Good Enable SATA link power management for host3 Good Enable SATA link power management for host1 Good Enable Audio codec power management Good VM writeback timeout Good Enable SATA link power management for host0 Good Enable SATA link power management for host2 Good Autosuspend for unknown USB device 1-5 (0cf3:e005) Good Autosuspend for USB device KAAN Advanced [KOBIL] Good Autosuspend for USB device xHCI Host Controller [usb2] Good Autosuspend for USB device Dell KB216 Wired Keyboard [1-8] Good Autosuspend for USB device USB2.0-CRW [Generic] Good Autosuspend for USB device xHCI Host Controller [usb1] Good Runtime PM for PCI Device Intel Corporation Sunrise Point-H LPC Controller Good Runtime PM for PCI Device Intel Corporation Skylake Host Bridge/DRAM Registers Good Runtime PM for PCI Device Intel Corporation Skylake Integrated Graphics Good Runtime PM for PCI Device Intel Corporation Sunrise Point-H Thermal subsystem Good Runtime PM for PCI Device Intel Corporation Sunrise Point-H USB 3.0 xHCI Controller Good Runtime PM for PCI Device Intel Corporation Sunrise Point-H CSME HECI #1 Good Runtime PM for PCI Device Intel Corporation Sunrise Point-H SATA controller [AHCI mode] Good Runtime PM for PCI Device Intel Corporation Sunrise Point-H PCI Express Root Port #5 Good Runtime PM for PCI Device Intel Corporation Sunrise Point-H PCI Express Root Port #6 Good Runtime PM for PCI Device Intel Corporation Sunrise Point-H PMC Good Runtime PM for PCI Device Intel Corporation Sunrise Point-H HD Audio Good Runtime PM for PCI Device Intel Corporation Sunrise Point-H SMBus Good Runtime PM for PCI Device Qualcomm Atheros QCA9565 / AR9565 Wireless Network Adapter Good Runtime PM for PCI Device Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller Good Wake-on-lan status for device wlan0 Good Wake-on-lan status for device p1p1 This results in: # powertop PowerTOP 2.8 Overview Idle stats Frequency stats Device stats Tunab Package | Core | CPU 0 CPU 2 | | C0 active 1.4% 1.3% | | POLL 0.0% 0.3 ms 0.1% | | C1E-SKL 2.1% 1.0 ms 2.3% C2 (pc2) 10.1% | | C3 (pc3) 37.4% | C3 (cc3) 0.3% | C3-SKL 0.2% 0.3 ms 0.1% C6 (pc6) 0.0% | C6 (cc6) 5.0% | C6-SKL 4.8% 1.2 ms 1.7% C7 (pc7) 0.0% | C7 (cc7) 79.6% | C7s-SKL 0.0% 0.0 ms 0.0% C8 (pc8) 0.0% | | C8-SKL 86.8% 4.9 ms 89.8% C9 (pc9) 0.0% | | C10 (pc10) 0.0% | | | Core | CPU 1 CPU 3 | | C0 active 1.7% 2.1% | | POLL 0.1% 0.4 ms 0.1% | | C1E-SKL 1.1% 0.6 ms 4.2% | | | C3 (cc3) 0.5% | C3-SKL 0.5% 0.5 ms 0.2% | C6 (cc6) 3.8% | C6-SKL 3.3% 0.8 ms 2.1% $ rpm -q powertop powertop-2.8-1.5.x86_64 $ cat /proc/cpuinfo processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 94 model name : Intel(R) Core(TM) i3-6100 CPU @ 3.70GHz stepping : 3 microcode : 0x9e cpu MHz : 800.110 cache size : 3072 KB physical id : 0 siblings : 4 core id : 0 cpu cores : 2 apicid : 0 initial apicid : 0 fpu : yes fpu_exception : yes cpuid level : 22 wp : yes ... $ uname -a Linux pc 4.4.36-8-default #1 SMP Fri Dec 9 16:18:38 UTC 2016 (3ec5648) x86_64 x86_64 x86_64 GNU/Linux $ cat /etc/os-release NAME="openSUSE Leap" VERSION="42.2" ... -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1010973 http://bugzilla.opensuse.org/show_bug.cgi?id=1010973#c12 Takashi Iwai changed: What |Removed |Added ---------------------------------------------------------------------------- Flags| |needinfo?(dr.tino.k@gmail.c | |om) --- Comment #12 from Takashi Iwai --- Try with the latest kernel in OBS Kernel:stable repo. If the problem persists, it means that something still missing in user-space side -- or it's disabled intentionally. As I already mentioned, the kernel doesn't allow to reach to the deep C state on some Skyalke models for avoiding the hardware bugs. Yours might be one of them. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1010973 http://bugzilla.opensuse.org/show_bug.cgi?id=1010973#c14 --- Comment #14 from Takashi Iwai --- (In reply to Tino K from comment #13)

The new kernel fails to load with an "... has invalid signature" error message.

I guess you're using secure boot. The kernel isn't signed with the right key, so it won't work as is. Disable the secure boot.

My CPU (i3-6100) is a Skylake-S, how can I check whether it contains the bug and whether the linux kernel intentionally disables deeper states?

It's CPU ID 0x5e, and it's the exact model the exact model the kernel disables C8/C9/C10. But a deep PC-level should be still achievable. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1010973 http://bugzilla.opensuse.org/show_bug.cgi?id=1010973#c16 --- Comment #16 from Takashi Iwai --- (In reply to Tino K from comment #15)

With the new kernel I get the same behaviour as before, i.e. max C2 (pc2) by default and max C3 (pc3) after tuning all entries to "Good" in powertop.

Then please report to upstream. There is nothing we can help here, so far.

Which output is the CPU ID?

I meant the CPU model number, shown in /proc/cpuinfo. It's 94 = 0x5e. -- You are receiving this mail because: You are on the CC list for the bug.