Re: c++/8675: 30% slower code using gcc3.2 vs. gcc3.0.1

Re: c++/8675: 30% slower code using gcc3.2 vs. gcc3.0.1

[neroden]

Synopsis: 30% slower code using gcc3.2 vs. gcc3.0.1

State-Changed-From-To: feedback->closed
State-Changed-By: neroden
State-Changed-When: Tue Mar 25 18:33:49 2003
State-Changed-Why:
    Can't track down without better-isolated test case.  No feedback 3 months.  It probably *is* due to the thread model anyway, since compiling in 'thread-safe' versions of the builtins certainly has runtime overhead.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=8675

Re: c++/8675: 30% slower code using gcc3.2 vs. gcc3.0.1

[bangerth]

Synopsis: 30% slower code using gcc3.2 vs. gcc3.0.1

State-Changed-From-To: open->feedback
State-Changed-By: bangerth
State-Changed-When: Sat Dec 21 10:52:45 2002
State-Changed-Why:
    It is very hard to track down where within 20,000 lines of
    code something becomes slower if you are not familiar with
    what the code does, where call-paths are, etc. I think you
    need to provide us with more information about what goes
    wrong, i.e. try to identify functions that become significantly
    slower and possibly show us the differences in the generated
    assembler code.
    
    Sorry for not being of more help -- I guess for performance
    regressions we really need cooperation from the people who
    file the reports...
    
    Thanks
      Wolfgang

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=8675