harryhaaren / fyprealtimecppprograming Goto Github PK

/*  Author: Harry van Haaren, 2013
 *          [email protected]
 * 
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 3 of the License, or
 * (at your option) any later version.
 * 
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 * 
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
 * MA 02110-1301, USA.
 */

==== Comparing two solutions to writing real-time audio applications in C++

This is the deliverable submission for the FYP project by Harry van Haaren, MMPT.
The entireity of the code (with one exception) contained within this folder has
been written as part of the project. The exception is imp2/ringbuffer.hpp, which
has been credited accordingly. The project uses various tools that should be
present on the system. The following is a list of necessary libraries and tools
to compile the code, and generate the statistics as presented in the report.

== Programs
-a C++ compiler
-GDB
-make
-valgrind
-GNUplot
-gprof2dot.py
-graphviz
-cpupower
-rtirq
-iwconfig
-convert
-general unix tools, ps, xterm etc
-massifplot (http://download.sarine.nl/massifplot/)

== Libraries
-JACK
-GTKmm
-libsndfile
-Boost headers


==== Compilation
To compile the code a "compile.sh" script is provided for convinience. This script
compiles 3 versions, and copies them to "testing/bin/<version>":
-optimized
-debug
-selfTesting


==== Running
Each implementation can be run from the subdirectory it was copied to, the
samples subdir will be scanned for .wav samples automatically.


==== Generating Statistics
To generate statistics, a "runTests.sh" script is provided for convienience.
It sets up the system (turns off wifi, amends the rtprio of the tasklets etc)
It proceeds to launch each selfTesting implementation using xterm, and collect stats.
Then graphs are generated from the profiling run, and written to the "output" dir.

To generate the massif graphs, run the self profiling implementations like so:
valgrind --tool=massif ./showcaseX

At the end, valgrind will output a massif.out.<PID> text file.
Convert that file to an image using GNUplot and massifplot:
./test massif.out.<PID>

Rotate the outputted graph 90 degree counter clockwise:
convert -rotate -90 massif.out.<PID>.png outputImage.png


== Comparing massif runs
The "testing" directory contains a gnuplot script that will plot the output of
the massifplot 'input.txt' files. In order to retain these files, comment the
rm input.txt 
line in test.sh (a script which is part of massifplot).

Run both implementations, calling one input1.txt and the other input2.txt.
To create the graph of these two input text files, simply run
gnuplot massifCompare.gnuplot
and an image 'massifCompare.png' will be created in the output directory.
Rotate as outlined previously.

If running valgrind with the -t ms option, it will be necessary to update the
label in the massifCompare.gnuplot file, and change the output filename.

==== Contact
If a query about this project arises, you can contact the author.
I hope this work has been useful, its certainly taught me a lot about
real-time programming.

 -Harry van Haaren
  [email protected]