Just had a hard lockup on my haswell (Macbook Air 2013). Probably really hard to reproduce, but I thought I would report it anyway. Attached is the dmesg, it spans across 24 hours, including around 23 hours of suspend, you can see a lot backtraces before the suspend and so much after the wakeup that it's probably what triggered the lockup. It looks like pretty much all write to registers trigger a backtrace. I thought it would be interesting for the record, even if nobody can reproduce. Userspace stack : libdrm-intel1: 2.4.58 xserver-xorg-video-intel: 2.21.15 libva: 1.4.1 intel-libva-driver: 1.4.1
Created attachment 112371 [details] /var/log/messages.gz
I'm guessing this is the hsw pm refcounting issue, closing. Please reopen if the problem reappears with latest kernels.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.