www
/
www
8
0
Fork 0
Code Issues Pull Requests 1 Projects Releases Wiki Activity

Tim Brecht Talk: linux

This commit is contained in:
Kyle Spaans 2011-06-28 10:00:22 -04:00
parent de834b2b15
commit 9428b8a8dc
1 changed files with 45 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

46
events.xml
View File

@ -49,6 +49,50 @@
More details to follow!
</abstract>
</eventitem>
<eventitem date="2011-07-04" time="1:30 PM" room="MC 5158" title="Our Troubles with Linux and Why You Should Care">
<short>
<p>
A joint work between Professors Tim Brecht, Ashif Harji, and
Peter Buhr, this talk describes experiences using the Linux
kernel as a platform for conducting performance evaluations.
</p>
</short>
<abstract>
<p>
Linux provides researchers with a full-fledged operating system that is
widely used and open source. However, due to its complexity and rapid
development, care should be exercised when using Linux for performance
experiments, especially in systems research. The size and continual
evolution of the Linux code-base makes it difficult to understand, and
as a result, decipher and explain the reasons for performance
improvements. In addition, the rapid kernel development cycle means
that experimental results can be viewed as out of date, or meaningless,
very quickly. We demonstrate that this viewpoint is incorrect because
kernel changes can and have introduced both bugs and performance
degradations.
</p>
<p>
This talk describes some of our experiences using the Linux kernel as a
platform for conducting performance evaluations and some performance
regressions we have found. Our results show, these performance
regressions can be serious (e.g., repeating identical experiments
results in large variability in results) and long lived despite having
a large negative impact on performance (one problem appears to have
existed for more than 3 years). Based on these experiences, we argue
that it is often reasonable to use an older kernel version,
experimental results need careful analysis to explain why a change in
performance occurs, and publishing papers that validate prior research
is essential.
</p>
<p>
This is joint work with Ashif Harji and Peter Buhr.
</p>
<p>
This talk will be about 20-25 minutes long with lots of time for
questions and discussion afterwards.
</p>
</abstract>
</eventitem>
<eventitem date="2011-06-24" time="7 PM" room="Comfy Lounge" title="Code Party 2">
<short>
<p>
@ -99,7 +143,7 @@
</p>
</abstract>
</eventitem>
<eventitem date="2011-06-09" time="4:30 PM" room="MC 2054"
<eventitem date="2011-06-09" time="4:30 PM" room="MC 2054"
title="General Purpose Computing on Graphics Cards">
<short>
In the first of our member talks for the term, Katie Hyatt will give a