-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Friday 23 February 2007 10:29 am, you wrote:
On Friday 23 February 2007 09:53 am, you wrote:
On Friday 23 February 2007 09:24 am, Peter Dimov wrote:
Have you actually run a test and observed a 50x slowdown? You can use libs/smart_ptr/test/shared_ptr_timing_test.cpp.
I get:
fhess@humidor:~/cvs/boost/libs/smart_ptr/test$ g++ -Wall -O3 shared_ptr_timing_test.cpp fhess@humidor:~/cvs/boost/libs/smart_ptr/test$ ./a.out 0.22 fhess@humidor:~/cvs/boost/libs/smart_ptr/test$ g++ -Wall -O3 shared_ptr_timing_test.cpp -pthread fhess@humidor:~/cvs/boost/libs/smart_ptr/test$ ./a.out 1.87
With boost 1.32, $ gcc --version gcc (GCC) 3.3.5 (Debian 1:3.3.5-13) on a 3.2GHz pentium d.
For further comparison, I modified the test to use an ordinary int* instead of a shared_ptr<int>, and got:
g++ -Wall -O3 raw_ptr_timing_test.cpp fhess@humidor:~/cvs/boost/libs/smart_ptr/test$ ./a.out 0.08
So it's about x3 slower using shared_ptr without -pthread, and about x20 slower using shared_ptr with -pthread.
And to round it out, compiling against a boost cvs checkout gives a x10 slowdown, with the -pthread option making no difference: fhess@humidor:~/cvs/boost/libs/smart_ptr/test$ g++ -Wall -O3 -I ~/cvs/boost shared_ptr_timing_test.cpp fhess@humidor:~/cvs/boost/libs/smart_ptr/test$ ./a.out 0.81 - -- Frank -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQFF3w5q5vihyNWuA4URAr7KAKCiTALn+p0wPXCg+MCLtZJkFk0B6wCgpJPp 8xpW0DS2MxByXTHKTw4KT74= =cLYX -----END PGP SIGNATURE-----