NMI hardlock stacktrace deadlock [was Re: Linux 5.2-rc5]

From: Chris Wilson
Date: Wed Jun 19 2019 - 08:45:10 EST

I haven't bisected this, but with the merge of rc5 into our CI we
started hitting an issue that resulted in a oops and the NMI watchdog
firing as we dumped the ftrace. This NMI watchdog locks up prior to the
backtraces being printed, preventing the machine from rebooting, and can
be avoided with hardlockup_all_cpu_backtrace=0. Running on arch/x86.