On Tue, Nov 19, 2013 at 01:24:44PM +0100, Mike Galbraith wrote: > On Tue, 2013-11-19 at 12:01 +0100, Peter Zijlstra wrote: > > > What idle function does the core2 use? Something like: > > > > perf record -a sleep 1; perf report > > > > Should show you the idle function on top if ran on a mostly idle system. > > I would be expecting mwait_idle_with_hints() for core2. And I think I've > > covered all paths to it with the right magic, but how knows, these acpi > > people never reacted to any of these patches. > > My Toshiba Satellite lappy does use mwait_idle_with_hints, which works > fine unless you boot box with processor.max_cstate=1, which scares hell > outta mwait_idle_with_hints, and default_idle is then used. I started > tracing it out, and there's a hard coded check (don't recall var name) > for cstate 2 being available for some odd reason. The old mwait_idle > didn't care, worked fine on Q6600 box for years on end, no idea why the > acpi folks made mwait_idle_with_hints get all upset and bail. > > While tracing out my Q6600 box, it didn't even get to the early probing > I had trace_printk laced leading to mwait_idle_with_hints use/fail, then > rainy day went away, and I haven't gotten back to it, so default_idle is > now all there is for (crusty ole beloved) desktop box. So acpi_processor_get_power_info_cst() can set C1 to FFH, but I'm completely in the dark as to what the conditions under which that can happen mean etc. Per default it uses HALT for C1, which is what you're seeing. Best take this up with the ACPI folks. -- To unsubscribe from this list: send the line "unsubscribe stable" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html