Messages in this thread | | | Date | Wed, 2 May 2012 15:44:41 +0200 | From | Borislav Petkov <> | Subject | Re: [PATCH 2/3] x86/flush_tlb: try flush_tlb_single one by one in flush_tlb_range |
| |
On Wed, May 02, 2012 at 07:38:47PM +0800, Alex Shi wrote: > > Are you saying you want to have this setting per family? > > Set it according to CPU type is more precise, but looks ugly.
By "CPU type" do you mean microarchitecture here?
> I am wondering if it worth to do. Maybe conservative selection is > acceptable?
Well, as I said earlier, I'd run it on a couple of different machines and make FLUSHALL_BAR configurable from userspace - this way you have real, solid data instead of guessing the exact number.
> > Also, have you run your patches with other benchmarks beside your > > microbenchmark, say kernbench, SPEC<something>, i.e. some other > > multithreaded benchmark touching shared memory? Are you seeing any > > improvement there? > > I tested oltp reading and specjbb2005 with openjdk. They should not much > flush_tlb_range calling. So, no clear improvement. > Do you know benchmarks which cause enough flush_tlb_range?
Not really. Probably get a couple of benchmarks and count flush_tlb_range calls with trace_printk or perf probe? :-)
[..]
> Believe we didn't need to know this, much more thread number just > weaken and cover the improvement. When the thread number goes down, > the performance gain appears. So, don't need care this.
Ok, this is also what the data showed, much higher gain with smaller thread counts.
> Any more comments for this patchset?
Nope, thanks.
-- Regards/Gruss, Boris.
Advanced Micro Devices GmbH Einsteinring 24, 85609 Dornach GM: Alberto Bozzo Reg: Dornach, Landkreis Muenchen HRB Nr. 43632 WEEE Registernr: 129 19551
| |