On Monday, April 23, 2018 11:16:33 AM CEST Peter Zijlstra wrote: > On Sat, Apr 21, 2018 at 11:19:55PM -0300, Diego Viola wrote: > > > The problem only happens *after* resuming from suspend, in normal S0 > > state it never happens. > > What kind of suspend are you doing? s2disk, s2ram ? > > > My machine is a desktop and has a dual core E5500 CPU with 2GB of RAM, > > I run Arch Linux and I use the i3wm. > > That's a Core2 era chip; does it actually have stable TSC ? > > > aa83c45762a242acce9b35020363225a7b59d7c9 is the first bad commit > > That is most curious; that was supposed to only affect early boot. I'll > try and figure out of there's some intersection with suspend.
Well, FWIW, I wonder if the flags that were set in init_tsc_clocksource() for unstable TSC before the above commit may be involved in this somehow.