16.6: Counting Messages

Last updated
Save as PDF

Page ID: 43762

\( \newcommand{\vecs}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} } \) \( \newcommand{\vecd}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash {#1}}} \)\(\newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\) \( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\) \( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\) \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\) \( \newcommand{\Span}{\mathrm{span}}\) \(\newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\) \( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\) \( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\) \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\) \( \newcommand{\Span}{\mathrm{span}}\)\(\newcommand{\AA}{\unicode[.8,0]{x212B}}\)

The profiling we have realized so far is focused on method execution time. The advantage of method call stack sampling is that it has a relatively low impact on the execution. The disadvantage is the relatively imprecision of the result. Even though the obtained results are sufficient in most of the case, they are always an approximation of the real execution.

MessageTally allows for a profiling based on program interpretation. The idea is to use a bytecode interpreter instead of execution sampler. The main advantage is the exactness of the result. The information obtained with the message tallySends: indicates the amount of time each method involved in a computation has been executed. Figure \(\PageIndex{1}\) gives the result obtained by executing

MessageTally tallySends:[ 1000 timesRepeat: [3.14159 printString]].

Spy Results showing executed messages. — Figure \(\PageIndex{1}\): All executed messages during an execution.

The downside of tallySend: is the time taken to execute the provided block. The block to profile is executed by an interpreter written in Pharo, which is slower then the one of the virtual machine. A piece of code profiled by tallySends: is about 200 times slower. The interpreter is available from the method ContextPart»runSimulated: aBlock contextAtEachStep: block2.