David C. Rankin wrote:
On 06/07/2018 12:40 AM, Per Jessen wrote:
perf: interrupt took too long (10078 > 2500), lowering kernel.perf_event_max_sample_rate to 19800
I see that all the time ? When I googled it very recently, I think the consensus was not to worry.
Well, generally I don't, and LKML doesn't seem to be concerned with it, the only problem for me is the set of dumps and checks adds about 2 min.+ to the init of the box as looping though all cores sending NMI to every other core.
You are sure they are related? I don't get any dumps, and the "perf: interrupt took too long" messages usually turn up a while after a reboot - [ 3.835386] usbcore: registered new interface driver usbhid [ 501.792135] perf: interrupt took too long (2520 > 2500), lowering kernel.perf_event_max_sample_rate to 79250 [ 534.657898] perf: interrupt took too long (3154 > 3150), lowering kernel.perf_event_max_sample_rate to 63250 [ 575.439441] perf: interrupt took too long (3956 > 3942), lowering kernel.perf_event_max_sample_rate to 50500 [ 661.743477] perf: interrupt took too long (4957 > 4945), lowering kernel.perf_event_max_sample_rate to 40250 [ 850.194665] perf: interrupt took too long (6214 > 6196), lowering kernel.perf_event_max_sample_rate to 32000 [ 1172.381354] perf: interrupt took too long (7773 > 7767), lowering kernel.perf_event_max_sample_rate to 25500 [snip]
[ 28.138850] watchdog: BUG: soft lockup - CPU#0 stuck for 23s! ^^^^^^^^^^^^^^^^
Try searching bugzilla for that one. Very annoying, I have had it a few times over the years. -- Per Jessen, Zürich (17.7°C) http://www.dns24.ch/ - your free DNS host, made in Switzerland. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org