Anyone seen this?

2 messages Options
Embed this post
Permalink
Mark Bolstad

Anyone seen this?

Reply Threaded More More options
Print post
Permalink
Assume that we have a heirarchy of classes, A through D, A defines a
default_create, and every descendent redefines default_create and uses the
Precursor construct. Now suppose we create 4 instances of D.

When I profile something like this under workbench with the release version
of 6.4 (OS X under MacPorts) I will see the count for the number of calls
for D.default_create to not be 4, but 16. The profiler is counting all of
the inherited default_create(s) under class D's.

I assume that this is not the intended behavior? I'll file a bug report, but
just wanted to confirm I'm not nuts.

Mark


[Non-text portions of this message have been removed]

Emmanuel Stapf

RE: Anyone seen this?

Reply Threaded More More options
Print post
Permalink
> I assume that this is not the intended behavior? I'll file a bug report,
> but just wanted to confirm I'm not nuts.

Currently profiling is done based on the dynamic type of objects, not the static
type. One could argue that dynamic is better than static, but both offer different
kind of information. This is why calling Precursor is counted twice since it
usually shares the same name.

Regards,
Manu

------------------------------------------------------------------------  
Eiffel Software
805-685-1006
http://www.eiffel.com       
Customer support: http://support.eiffel.com       
User group: http://groups.eiffel.com/join       
------------------------------------------------------------------------