Dear Friends and Colleagues,
We are happy to announce Memory Profiler, a new profiler for Pharo. This profiler gives a nice output that indicates the memory consumption along the method call graph. Consider the following piece of code:
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
(MSP profile: [
Object browse
] onPackagesMatching: 'Nautilus*' ) inspect
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Consider that MSP uses instrumentation, therefore long execution may take really long. The code above takes about 19 seconds on my iMac.
It opens a visualization that indicates some interesting fact about memory consumption. For example, the expression “Object browse” consumes 14 851 kb with instances of the class Point. More than 1M of points are created.
Here is the blueprint obtained:
A description of the visualization is available here:
You need to load Spy from the Configuration Browser. Else the following incantation should work:
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Gofer it
smalltalkhubUser: 'ObjectProfile' project: 'S2py';
configurationOf: 'Spy';
loadDevelopment.
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Cheers,
Alejandro & Alexandre
--
_,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
Alexandre Bergel http://www.bergel.eu
^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.