On Wed, Dec 30, 2020 at 12:49:03PM +0100, Evgeni Golov wrote: > On Sun, Dec 27, 2020 at 03:46:27PM +0100, Evgeni Golov wrote: > > 1. qemu will happily continue booting cirros when you stop tuned *after* > > starting qemu > > 2. it boots fine if you use the "default" (which is not default anymore, > > but a legacy fallback) or the "balanced" profiles. > > 3. it hangs as you describe in the "throughput-performance", > > "virtual-guest" and "virtual-host" profiles -- I didn't test any others > > 4. using any of the problematic profiles on my laptop doesn't make it > > freeze in the non-nested case > > > > Still, not a fix, but at least a direction ;-) > > Digging deeper. > > This doesn't happen on the buster cloud images, and a few "apt install" > invocations later I could bisect the issue to be the upgrade of seabios > from 1.12.0-1 to 1.14.0-2. > > I also can't repro this in vagrant (by using the buster64 box and > upgrading it fully to testing) with libvirt/qemu as a backend instead of > plain qemu. > > So there seem to be more levels of weirdness to this.
And the tuned line that breaks the whole thing is [sysctl] kernel.sched_wakeup_granularity_ns = 15000000