[Libav-user] gcc auto-vectorisation

["René J.V. Bertin"]

On Feb 27, 2013, at 17:56, Claudio Freire wrote:

> I think it's just academic. It would be one mystery less.

In the end I decided to go a step further than disabling HT - I forced the use of a single thread. And a mystery less? Actually, no:

> time /usr/local/FFmpeg/trunk/bin/ffmpeg-rjvb -benchmark_most -threads 1 -y -v 0 -i ~/Desktop/Downloads/SOA4ep11.flv -pix_fmt argb -vcodec rawvideo -acodec pcm_f32le -f mov /dev/null ; time /usr/local/FFmpeg/trunk.vect/bin/ffmpeg-rjvb -benchmark_most -threads 1 -y -v 0 -i ~/Desktop/Downloads/SOA4ep11.flv -pix_fmt argb -vcodec rawvideo -acodec pcm_f32le -f mov /dev/null ; time /usr/local/FFmpeg/trunk.O0/bin/ffmpeg-rjvb -benchmark_most -threads 1 -y -v 0 -i ~/Desktop/Downloads/SOA4ep11.flv -pix_fmt argb -vcodec rawvideo -acodec pcm_f32le -f mov /dev/null ; time /usr/local/FFmpeg/trunk.O0vect/bin/ffmpeg-rjvb -benchmark_most -threads 1 -y -v 0 -i ~/Desktop/Downloads/SOA4ep11.flv -pix_fmt argb -vcodec rawvideo -acodec pcm_f32le -f mov /dev/null 
Detailed benchmark results:
                   samples          user t        kernel t          real t           CPU %
Video decode  :      85162         85.4759s        0.11533s        85.7097s        99.8618%
Audio decode  :     152971         3.66945s       0.095735s        3.74859s        100.443%
Video encode  :      85164         14.9526s       0.090322s        15.0694s        99.8241%
Audio encode  :     152969        0.341693s        0.08129s       0.392368s        107.803%
Failed loops  :          1               0s          1e-06s     7.03003e-07s       142.247%
Weighed totals:   476267/5         19.2462s      0.0936312s        19.3505s        99.9444%
Overall execution timing:
             :           1         152.645s         2.5378s        155.699s        99.6688%
152.655 user_cpu 2.553 kernel_cpu 2:36.18 total_time 99.3%CPU {0W 0X 0D 0K 12974080M 604F 7590R 10I 0O 0r 0s 0k 36w 1676c}
Detailed benchmark results:
                   samples          user t        kernel t          real t           CPU %
Video decode  :      85162         87.1021s       0.118153s        87.3188s        99.8872%
Audio decode  :     152971          3.7146s       0.095249s        3.79602s        100.364%
Video encode  :      85164         15.2192s       0.091017s        15.3272s        99.8888%
Audio encode  :     152969        0.338267s        0.08176s       0.390411s        107.586%
Failed loops  :          1           1e-06s          1e-06s     1.21299e-06s       164.882%
Weighed totals:   476267/5          19.598s      0.0942551s         19.699s        99.9659%
Overall execution timing:
             :           1         155.574s        2.53883s        158.133s        99.9877%
155.580 user_cpu 2.544 kernel_cpu 2:38.14 total_time 99.9%CPU {0W 0X 0D 0K 12974080M 50F 8144R 0I 0O 0r 0s 0k 1w 1592c}
Detailed benchmark results:
                   samples          user t        kernel t          real t           CPU %
Video decode  :      85162         198.603s       0.194156s        198.966s        99.9148%
Audio decode  :     152971         4.72838s       0.091876s        4.80426s        100.333%
Video encode  :      85164         15.6997s       0.084267s        15.8032s        99.8786%
Audio encode  :     152969        0.316537s        0.07774s       0.364985s        108.026%
Failed loops  :          1               0s              0s     6.47007e-07s             0%
Weighed totals:   476267/5         39.9402s       0.104264s        40.0636s        99.9521%
Overall execution timing:
             :           1         254.731s        2.65197s        258.002s          99.76%
254.736 user_cpu 2.662 kernel_cpu 4:18.40 total_time 99.6%CPU {0W 0X 0D 0K 13543424M 624F 7796R 8I 0O 0r 0s 0k 21w 2877c}
Detailed benchmark results:
                   samples          user t        kernel t          real t           CPU %
Video decode  :      85162         208.447s       0.196297s        208.806s        99.9218%
Audio decode  :     152971         4.89258s       0.094659s        4.97187s        100.309%
Video encode  :      85164         15.7334s       0.090704s        15.8499s        99.8372%
Audio encode  :     152969        0.326942s       0.077216s       0.378373s        106.815%
Failed loops  :          1           1e-06s          1e-06s     7.37025e-07s       271.361%
Weighed totals:   476267/5         41.7625s       0.106523s        41.8896s        99.9509%
Overall execution timing:
             :           1         265.295s        2.69075s        268.599s        99.7717%
265.301 user_cpu 2.699 kernel_cpu 4:28.80 total_time 99.6%CPU {0W 0X 0D 0K 13570048M 630F 7804R 9I 0O 0r 0s 0k 18w 2904c}