On Wed, 13 Feb 2013, Jim Klimov wrote:

For clarity of experiment, it might help to disable cpupm completely
and/or to run the experiment with a larger threshold (and loop more
of your program's executions).

I disabled CPU power-management and cold-booted and this is what 'time' reports for successive runs with a 1 second sleep in between. Notice no more bouncy total time readings but also performance still seems inadequate (compared with Linux) with excessive system time. The total time reading is similar to the better readins when results were inconsistent.

./magick large.gif  5.81s user 2.91s system 86% cpu 10.127 total
./magick large.gif  5.81s user 2.86s system 87% cpu 9.918 total
./magick large.gif  5.81s user 2.89s system 87% cpu 9.940 total
./magick large.gif  5.81s user 2.92s system 87% cpu 9.976 total
./magick large.gif  5.81s user 2.74s system 91% cpu 9.387 total
./magick large.gif  5.85s user 2.71s system 88% cpu 9.628 total
./magick large.gif  6.16s user 3.13s system 87% cpu 10.592 total
./magick large.gif  6.14s user 3.10s system 90% cpu 10.195 total
./magick large.gif  6.07s user 3.07s system 87% cpu 10.446 total
./magick large.gif  5.99s user 3.05s system 87% cpu 10.339 total
./magick large.gif  6.00s user 3.07s system 87% cpu 10.396 total
./magick large.gif  6.03s user 3.07s system 87% cpu 10.439 total
./magick large.gif  6.01s user 3.08s system 87% cpu 10.398 total
./magick large.gif  6.01s user 3.08s system 87% cpu 10.403 total
./magick large.gif  6.03s user 3.09s system 87% cpu 10.441 total
./magick large.gif  6.02s user 3.09s system 87% cpu 10.452 total
./magick large.gif  6.02s user 3.09s system 87% cpu 10.449 total
./magick large.gif  6.01s user 2.81s system 90% cpu 9.700 total
./magick large.gif  6.01s user 2.81s system 89% cpu 9.870 total
./magick large.gif  6.02s user 2.83s system 89% cpu 9.910 total
./magick large.gif  6.01s user 2.77s system 90% cpu 9.668 total
./magick large.gif  6.06s user 2.85s system 89% cpu 9.983 total
./magick large.gif  5.97s user 2.87s system 89% cpu 9.913 total
./magick large.gif  6.15s user 2.91s system 89% cpu 10.146 total
./magick large.gif  6.13s user 2.91s system 89% cpu 10.120 total
./magick large.gif  6.02s user 2.83s system 89% cpu 9.862 total
./magick large.gif  6.02s user 2.85s system 89% cpu 9.940 total
./magick large.gif  6.01s user 2.87s system 89% cpu 9.947 total
./magick large.gif  6.03s user 3.10s system 87% cpu 10.466 total
./magick large.gif  6.02s user 3.12s system 86% cpu 10.506 total

Powertop says: 3483 Mhz(turbo)  100.0%

It also says: Wakeups-from-idle per second: 7193.2    interval: 5.0s
Most of the wakeups are due to DTrace probes from Powertop.

'hotkernel' says:

unix`mutex_exit                                           692   0.0%
unix`0xfffffffffb80                                      1395   0.1%
unix`page_lookup_create                                  1529   0.1%
unix`mutex_enter                                         4818   0.2%
unix`hwblkclr                                            6432   0.3%
unix`i86_mwait                                         109291   4.9%
unix`acpi_cpu_cstate                                  2072388  93.7%

Notice that unix`acpi_cpu_cstate percentage is even worse than before.

Bob
--
Bob Friesenhahn
[email protected], http://www.simplesystems.org/users/bfriesen/
GraphicsMagick Maintainer,    http://www.GraphicsMagick.org/


-------------------------------------------
illumos-discuss
Archives: https://www.listbox.com/member/archive/182180/=now
RSS Feed: https://www.listbox.com/member/archive/rss/182180/21175430-2e6923be
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=21175430&id_secret=21175430-6a77cda4
Powered by Listbox: http://www.listbox.com

Reply via email to