http://bugzilla.opensuse.org/show_bug.cgi?id=1039768 http://bugzilla.opensuse.org/show_bug.cgi?id=1039768#c1 --- Comment #1 from Michael Paton --- Additional Information: UEFI booting is enabled storage is 256GB SSD The update containing the 4.11.0 kernel exhibits the same behaviour on initial boot. However after the update, the system locks up early in the boot from hibernate state. The update containing the 4.11.1 kernel is a partial improvement in that once the initial kernel workload subsides to around the 2% mark, the frequency will reduce to a nominal 800MHz (not as low as the 500MHz in earlier kernels). This kernel will also not resume from hibernation. After holding in the power button and rebooting to the 4.11.1 kernel in recovery more, the frequency remains at Max Turbo. However this kernel has some minor difference in the contents of /sys/devices/system/cpu/intel_pstate/* file status is new and shows "active", the value in min_perf_pct goes from 19 to 18, and the value in turbo_pct changes from 38 to 69. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1039768 http://bugzilla.opensuse.org/show_bug.cgi?id=1039768#c4 --- Comment #4 from Michael Paton --- With kernel 4.11.1-1, prior to the application of the updates in Changes_20170522.txt, a freshly booted kernel would settle at a nominal idle frequency of 800MHz, and a suspend and resume would settle at a nominal idle frequency of 500MHz. I then applied the updates from Changes_20170522.txt and now a fresh reboot causes the idle frequency to settle at a nominal 2GHz. A suspend and resume now causes the idle frequency to settle at a nominal 500MHz. Interestingly the first suspend did not complete properly. It blanked the screen, turned the power light off, and when I would have expected a brief flash from the power light (after a couple of seconds), the power light came on steadily, the screen was blank apart from the KDE cursor. When I pressed the power button (as if to resume from suspend) the idle frequency settled at a nominal 500MHz. I wonder if this is indication that the enabler for correct power management and lower frequencies is executing as the system goes down for suspend (or hibernate) rather than when it resumes? One of the early messages displayed very briefly on the console on the unsuccessful resume from hibernation is "sparse file not allowed" I'll file a separate bug report on this when I have a little more information. I don't find this text string in the kernel tree though. -- You are receiving this mail because: You are on the CC list for the bug.