Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

C/C++ Program Perfometer: Tool for measuring comparative performance of the C/C++ code

40 views
Skip to first unread message

Alex Vinokur

unread,
Aug 9, 2004, 1:48:36 AM8/9/04
to
############################################
========= C/C++ Program Perfometer =========
############################################

The C/C++ Program Perfometer is a tool for measuring comparative performance of the C/C++ code.

=====================================
BRIEF DESCRIPTION

-----------------------------------
The Perfometer enables to get performance
of C/C++ program and separated pieces of the code for arbitrary metrics
(for instance : clocks, uclocks, rusage-metrics, metrics defined by user etc.).

The measurement results are represented in detailed/summary reports.
The detailed report holds results of individual tests,
the summary report holds average cost and its analysis.
The comparison results are represented in comparative tables
for separate comparison groups.

The user may set various parameters
in order to control the measurement/comparison process :
* measurement report and detailed measurement report flags,
* total tests,
* total repetitions,
* measurement scale,
* confidence threshold.

The Perfometer contains various testsuites for measuring comparative performance.


=====================================
LATEST RELEASE
Release 2.8.0-1.18
Note. Release number N1-N2 consists of :
* N1 : Version of C++ Perfometer itself; current version is 2.8.0,
* N2 : Version of Testsuites; current version is 1.18.

=====================================
AVAILABILITY

-----------------------------------
### Web page ###
* http://alexvn.freeservers.com/s1/perfometer.html


-----------------------------------
### Download ###
* http://sourceforge.net/projects/cpp-perfometer/
* http://alexvn.freeservers.com/s1/perfometer.html
* http://www.planet-source-code.com/vb/scripts/ShowCode.asp?txtCodeId=3512&lngWId=3


-----------------------------------
### Mailing List ###
* Web Page : http://lists.sourceforge.net/lists/listinfo/cpp-perfometer-users
* Posting : mailto:cpp-perfom...@lists.sourceforge.net
* Archives : http://sourceforge.net/mailarchive/forum.php?forum=cpp-perfometer-users


-----------------------------------
### Newsgroup ###
* News : news://news.gmane.org/gmane.comp.lang.c++.perfometer
* Archives : http://news.gmane.org/gmane.comp.lang.c++.perfometer


-----------------------------------
### Tests' Raw Logs ###
* http://groups.google.com/groups?th=fb0e11d9e7107414
* http://groups.google.com/groups?threadm=bcs4nc%24lsdbq%241%40ID-79865.news.dfncis.de

-----------------------------------
### Tests' Summary ###
* http://alexvn.freeservers.com/s1/perfo/tests/pftests.htm

-----------------------------------
Testsuites (current status)

- access to an element
- argument passing
- argument passing (built-in types)
- argument passing (containers-1)
- argument passing (containers-2)
- argument passing (containers-3)
- calling functions : templated vs. non-templated
- container performance (Stepanov & Stroustrup)
- copying files : input to output
- dynamic_cast vs. switch
- endl vs. "\n", '\n' [! stdout/stderr > tmp]
- exception handling vs. if-statement
- find method vs. find algorithm
- for-loop vs. for_each
- for_each vs. transform
- iterators
- multiple inheritance vs. single inheritance
- primitive operators
- pthreads
- reading contents from file into one string
- recursion vs. iteration vs. accumulate
- str-functions vs. mem-functions (C)
- string initialization (C vs. STL)
- string length (C vs. STL)
- string processing (C vs. STL)
- stub tested functions
- tech report : D1 - overhead of class operations
- tech report : D3 - the Stepanov abstraction penalty benchmark
- tech report : D4 - comparing function objects to function pointers
- tech report : D5 - measuring the cost of synchronized I/O
- virtual vs. ordinary methods

=====================================
SOURCES' SIZE
The C++ Perfometer 2.8.0-1.18 contains 42 program files (*.h, *.cpp)

--- Summary Report (numbers) ---
=====================================
| : Lines : Chars |
|-----------------------------------|
| Code Only : 11337 : 264666 |
| Code+Comment : 252 : - |
| Comment Only : 3845 : 135833 |
| Empty : 4205 : 66492 |
|...................................|
| * Total : 19639 : 466991 |
=====================================

--- Summary Report (percents) ---
=====================================
| : Lines : Chars |
|-----------------------------------|
| Code Only : 57.7 % : 56.7 % |
| Code+Comment : 1.3 % : - |
| Comment Only : 19.6 % : 29.1 % |
| Empty : 21.4 % : 14.2 % |
|...................................|
| * Total : 100 % : 100 % |
=====================================

Note. Perfometer sources' size has been counted with using cncc - "Counter of C/C++ code"
http://groups.google.com/groups?selm=bvtt1p%2410rbd7%241%40ID-79865.news.uni-berlin.de

--


=====================================
Alex Vinokur
mailto:ale...@connect.to
http://mathforum.org/library/view/10978.html
=====================================


0 new messages