This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [RFC] Lock elision "test suite", first draft


On Mon, Sep 02, 2013 at 10:46:09AM +0200, OndÅej BÃlka wrote:
> On Fri, Aug 09, 2013 at 10:18:46AM +0200, Dominik Vogt wrote:
> > On Wed, Aug 07, 2013 at 02:20:10PM +0200, OndÅej BÃlka wrote:
> snip
> > > > elision on:
> > > > -----------
> > > > pct_best 593.48...%
> > > > pct_mindev: 5.73...%
> > > > pct_maxdev: 4.09...%
> > > > pct_stddev: 3.43...%
> > > > 
> > > Do you know why elision is 6 times slower on this test?
> > 
> > The throughput is six times of the throughput without elision.
> > The percentages of pct_best are relative performance compared with
> > glibc-2.15.
> > 
> > Vice versa, an explanation is needed for the performance of the
> > tests 1, 2, 4 and 5.  I'm working on that and will post individual
> > analyses of the tests soon.
> > 
> We are waiting for analysis,

I have looked at the aborts of all of these tests, but I now have
to wait for more information from the IBM hardware developers -
hopefully this week.  I'll present detailed analyses then.

Ciao

Dominik ^_^  ^_^

-- 

Dominik Vogt
IBM Germany


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]