https://bugzilla.novell.com/show_bug.cgi?id=687998 https://bugzilla.novell.com/show_bug.cgi?id=687998#c3 Stefan Seyfried <seife@novell.slipkontur.de> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |oneukum@novell.com, | |vbotka@novell.com --- Comment #3 from Stefan Seyfried <seife@novell.slipkontur.de> 2011-04-30 15:28:12 CEST --- Still there with 2.6.39-rc5-14, both -desktop and -vanilla (KOTD of today). I found out that it is most probably the bluetooth stuff that's to blame. If I "killall bluetoothd" before suspend (which brings btusb usage count to zero and would allow its unloading), then it never crashes. Otherwise, I get a pretty reproducible kernel panic (visible by blinking capslock, nothing on console). But the bug is not that simple ;-) * unloading btusb, then loading it and starting bluetoothd, then suspending => no crash * then virtually un-/replugging the bluetooth adapter (via thinkpad-acpi module facilities), then suspending => crash Basically the workaround for me is to make sure that bluetooth is disabled (I have configured a hotkey for that), disabling via the gnome bluetooth applet is enough, before suspending. I'm suspecting that something in the btusb autosuspend code fails during suspend to RAM. Unfortunately, I'm not able to get any backtrace or similar at all. This is all with the vanilla kernel, with the SUSE desktop kernel I don't even get the blinking capslock led... To add another data point: if I disable intel modesetting (i915.modeset=0 vga=0), which I tried in order to get a backtrace on plain VGA, then the VESA driver is used by X - and then it does not crash. So I really suspect it's a nasty race condition. Again - this is all with a vanilla kernel, so it is a bad upstream regression. Should I also complain loudly on LKML to avoid 2.6.39 being a show stopper on x200s? Adding Oliver to CC who might know something about btusb suspend code :-) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.