| Date | Thu, 19 Dec 2013 08:09:29 -0800 | From | "H. Peter Anvin" <> | Subject | Re: [PATCH] x86 idle: repair large-server 50-watt idle-power regression |
| |
On 12/19/2013 08:02 AM, Ingo Molnar wrote: >> >> __monitor() currently doesn't, which is idiotic. > > Hm, __monitor() seems to take a void *: > > arch/x86/include/asm/processor.h:static inline void __monitor(const void *eax, unsigned long ecx, > > So writing: > > if (this_cpu_has(X86_FEATURE_CLFLUSH_MONITOR)) > clflush(¤t_thread_info()->flags); > > __monitor(¤t_thread_info()->flags, 0, 0); >
Yes, brain failure on my part.
-hpa
|