Re: Benchmarking with --enable-atomic-refcount

[Dmitri A. Sergatskov]

On Fri, Jan 10, 2014 at 4:14 PM, Rik <address@hidden> wrote:

1/10/14

...

 

% change: +12.1%

Since we have conflicting results maybe somebody else needs to run the
tests as well.  I ran 'make clean' in between re-configuring Octave.  It
shouldn't be necessary, but you never know what kind of cruft accumulates.

I get slightly less than 10% difference (built with default flags).

On i2600k system:

--disable-atomic

287(u)+24.50(s)

--enable-atomic

294(u)+24.8(s)

On AMD fx-8350

--disable-atomic

362.7(u)+47.2(s)

--enable-atomic

388(u) + 47.7(s)

gcc version 4.8.2 20131212 (Red Hat 4.8.2-7) (GCC)  Fedora 20 / x86_64

 

--Rik

Dmitri.
--