On Tue, Nov 11, 2008 at 10:54:52AM -0600, Elliot Shank wrote:
> Basically, the program is doing a report upon a couple of log
> files. If you look at the nytprofhtml output, you'll see lots of
> complaints about it not being able to find the source of methods
> created by Moose.
~630 source files but 1687097 fids. That's a lot of string evals.
The timing for base_fid_2_eval_fids_map seems odd.
It's showing 4405s for this simple loop:
247 630 4405s 6.99s while ( my ($eval_fid, $base_fid) = each %$e2b ) {
248 push @{ $b2e->{$base_fid} }, $eval_fid;
249 }
It's not showing any count or time for the push, which is odd. I'd guess
it got optimized to "push ... while ..." and the optimizer ate the
OP_NEXTSTATE. (Right Nick?)
Overall it looks like the C code needs to do a little more of the data
model creation so the perl doesn't need to go digging for details during
report generation. Specifically, the data that base_fid_2_eval_fids_map
creates could be setup in C during the load.
Tim.
--~--~---------~--~----~------------~-------~--~----~
You've received this message because you are subscribed to
the Devel::NYTProf Development User group.
Group hosted at: http://groups.google.com/group/develnytprof-dev
Project hosted at: http://perl-devel-nytprof.googlecode.com
CPAN distribution: http://search.cpan.org/dist/Devel-NYTProf
To post, email: [email protected]
To unsubscribe, email: [EMAIL PROTECTED]
-~----------~----~----~----~------~----~------~--~---
