Comment 27 for bug 1774150

Unable to give apport info, but wakeup on a console also hangs, writing more and more infrequently messages containing "perf samples too long" with numbers in the 1.5 million and "lowering kernel.perf.event_max_sample_rate" from 500,000 then 250,000 all the way down to 250.

In the beginning I also saw "perf_event_nmi_handler took to long to run 208.309 ms".

Hope this helps.