Bug 153361

Summary: kernel panic on s2ram
Product: Power Management Reporter: Elmar Stellnberger (estellnb)
Component: Hibernation/SuspendAssignee: Rafael J. Wysocki (rjw)
Status: RESOLVED CODE_FIX    
Severity: high    
Priority: P1    
Hardware: x86-64   
OS: Linux   
Kernel Version: 4.8.0-rc1 Subsystem:
Regression: No Bisected commit-id:
Attachments: screenshot of the backtrace

Description Elmar Stellnberger 2016-08-18 19:42:38 UTC 
With 4.8.0-rc1 the kernel always panics on s2ram. Before with kernel 4.7.0-ARCH-13902-gd491e80 it was suspending and awaking o.k. (if not to speak of the screen staying sometimes black).
  Tested successfully with 4.8.0-rc1:
echo freezer, devices, platform, processor, core >pm_test

  Evokes the backtrace:
echo mem >/sys/power/state

booted with nouveau.blacklist=1 and no_console_suspend into /bin/bash.
Comment 1 Elmar Stellnberger 2016-08-18 19:45:38 UTC 
Created attachment 229271 [details]
screenshot of the backtrace
Comment 2 Rafael J. Wysocki 2016-08-18 23:30:49 UTC 
On Thursday, August 18, 2016 07:45:38 PM bugzilla-daemon@bugzilla.kernel.org wrote:
> https://bugzilla.kernel.org/show_bug.cgi?id=153361
> 
> --- Comment #1 from Elmar Stellnberger <estellnb@elstel.org> ---
> Created attachment 229271 [details]
>   --> https://bugzilla.kernel.org/attachment.cgi?id=229271&action=edit
> screenshot of the backtrace

Looks like a CPU offline/online problem.

Does it panic during suspend or during resume?
Comment 3 Elmar Stellnberger 2016-08-19 04:20:23 UTC 
It panics before going into suspend.
Comment 4 Elmar Stellnberger 2016-08-20 11:02:27 UTC 
good: 4.7.0-ARCH-13902-gd491e80  
bad: 4.8.0-rc1
bad: g1c88e19
Comment 5 Elmar Stellnberger 2016-08-20 19:06:07 UTC 
Today we all have reason to celebrate! This bug as well as kernel bugs  153371, 117581 and freedesktop bug 93887 (https://bugs.freedesktop.org/show_bug.cgi?id=93887) have in deed been resolved with commit 65ea11ec6a82b1d44aba62b59e9eb20247e57c6e (Ville Syrjälä - x86/hweight: Don't clobber %rdi) - and these were just the issues reported by me - not to count numerous others (of which Mark Asselstine has been talking about to me) ...