Messages in this thread | | | Date | Wed, 11 Dec 2013 18:47:10 +0100 | From | Ingo Molnar <> | Subject | Re: 50 Watt idle power regression bisected to Linux-3.10 |
| |
* Peter Zijlstra <peterz@infradead.org> wrote:
> On Wed, Dec 11, 2013 at 03:42:38PM +0100, Ingo Molnar wrote: > > Another thing that is required I think is to issue a write barrier > > before CLFLUSH instruction. By my (possibly incorrect ...) reading of > > the documentation CLFLUSH does not appear to be ordered (at all), so > > it might execute before the modification to the affected memory? > > > > > > So something like: > > > > if (this_cpu_has(X86_FEATURE_CLFLUSH_MONITOR)) { > > smp_wmb(); /* order CLFLUSH */ > > clflush(¤t_thread_info()->flags); > > } > > smp_wmb() is a NO-OP on x86 remember :-)
Well, it's a compiler barrier but yes - I suspect a smp_mb() might be needed - at least according to the CLFLUSH documentation it has no implicit guaranteed ordering wrt. preceding writes.
Thanks,
Ingo
| |