[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: move constructors likely a requirement
From: |
Rik |
Subject: |
Re: move constructors likely a requirement |
Date: |
Fri, 23 Aug 2019 11:56:42 -0700 |
On 08/22/2019 10:34 PM, John W. Eaton wrote:
> I pushed the following changesets
>
> http://hg.savannah.gnu.org/hgweb/octave/rev/7335ebd4c798
> http://hg.savannah.gnu.org/hgweb/octave/rev/b00ddc40b89a
>
> to add move constructors and move assignment operators for the
> dim_vector, Array, string_vector, and octave_value classes. For me,
> these changes and the others that have been made for performance since
> bug #56752 was reported drops the time to run the bm_toeplitz.m file from
> about 39 seconds down to about 26 seconds. With version 4.4.1, I see a
> time of about 31 seconds.
>
This is a good result. I'm getting an 18% improvement on the dev branch
from 13.3 to 10.9 seconds. This is about equivalent historically to
version 4.2.1 which took 10.5 seconds.
I also tried compiling with '--disable-atomic-refcount'. The benchmark
then declines to 9.7 seconds which is a -11% change. This suggests there
is still some room for improvement, but that we will need to look elsewhere
for major improvements.
The absolute best performer is still version 3.4.3 at 4.5 seconds which is
more than double the performance of the current result of 10.9 seconds.
--Rik
- move constructors likely a requirement, Rik, 2019/08/20
- Re: move constructors likely a requirement, Carlo De Falco, 2019/08/21
- Re: move constructors likely a requirement, John W. Eaton, 2019/08/21
- Re: move constructors likely a requirement, Rik, 2019/08/21
- Re: move constructors likely a requirement, John W. Eaton, 2019/08/22
- Re: move constructors likely a requirement, John W. Eaton, 2019/08/23
- Re: move constructors likely a requirement,
Rik <=
- Re: move constructors likely a requirement, John W. Eaton, 2019/08/23
- Re: move constructors likely a requirement, Nicholas Jankowski, 2019/08/23
- Re: move constructors likely a requirement, Nicholas Jankowski, 2019/08/23
- Re: Longitudinal Octave Performance, Rik, 2019/08/24
- Re: move constructors likely a requirement, Rik, 2019/08/24
- Re: move constructors likely a requirement, Lars Winterfeld, 2019/08/26
- Re: move constructors likely a requirement, John W. Eaton, 2019/08/29
- Re: move constructors likely a requirement, Rik, 2019/08/30
- Re: move constructors likely a requirement, Daniel J Sebald, 2019/08/30
- Re: move constructors likely a requirement, ederag, 2019/08/30