From: Douglas Gregor (doug.gregor_at_[hidden])
Date: 2008-03-12 21:26:36
On Mar 12, 2008, at 6:38 PM, Michael Fawcett wrote:
> On Wed, Mar 5, 2008 at 4:22 PM, Sebastian Redl
> <sebastian.redl_at_[hidden]> wrote:
>> Michael Fawcett wrote:
>>> I forgot to mention before that this is using VS 2005 as the
>>> I'm pretty sure results for compilers with different exception
>>> implementations would vary greatly.
>> I imagine that a Linux/GCC build would indeed show quite different
>> results. Do you have the platform available, or a an easily
>> compilable/runnable test case for someone else to run?
> So I finally got it compiling and running under Cygwin/gcc. Here are
> the results (averaged over 5 runs):
> 1280 x 1280
> at_goal() - 23.45162 seconds
> exception - 23.39538 seconds
> 128 x 128
> at_goal() - 0.20304038 seconds
> exception - 0.1919092 seconds
> So exception beat at_goal() in both cases here, but the margins are
> much smaller. I suspect that I wasn't compiling to gcc's fullest
> optimization potential, but I did use -O3.
Very interesting. Thanks for running these experiments!
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk