> -----Original Message-----
> From: Brian Brooks [mailto:[email protected]]
> Sent: Wednesday, April 19, 2017 9:37 PM
> To: Savolainen, Petri (Nokia - FI/Espoo) <[email protected]>
> Cc: Honnappa Nagarahalli <[email protected]>; lng-
> [email protected]
> Subject: Re: [lng-odp] [API-NEXT PATCH 8/8] linux-gen: time: use hw time
> counter when available
> 
> On 04/26 07:11:57, Savolainen, Petri (Nokia - FI/Espoo) wrote:
> >
> >
> > > > From coverletter:
> > > > "This patch set modifies time implementation to use TSC when running
> on
> > > a x86
> > > > CPU that has invarint TSC CPU flag set. Otherwise, the same Linux
> system
> > > time
> > > > is used as before. TSC is much more efficient both in performance
> and
> > > > latency/jitter wise than Linux system call. This can be seen also
> with
> > > > scheduler latency test which time stamps events with this API. All
> > > latency
> > > > measurements (min, ave, max) improved significantly."
> > >
> > > odp_sched_latency currently uses clock_gettime. It is my understanding
> > > that clock_gettime does not have the over head of the system call. Can
> > > you elaborate more on the 'improved significantly' part?
> > >
> >
> > clock_gettime() uses the same TSC, but when you profile it with perf you
> can see tens of kernel functions including system call entry, RCU
> maintenance, etc.
> >
> > E.g. in sched_latency test kernel consumed about 10% of all the cycles.
> Also latency measurement results improved like this:
> > * min >3x lower
> > * avg 2x lower
> > * max more stable and 50% lower
> 
> You might want to share more information on the environment
> where you're seeing such significant improvements because the
> results on Broadwell do not match the above interpretation.
> 
> PS - This patch series breaks the build on ARM.
> 

Use v2. It should build on ARM.

-Petri



Reply via email to