<div dir="ltr"><br><div class="gmail_extra"><br><br><div class="gmail_quote">On Tue, Jul 16, 2013 at 11:03 PM, shiek kaleem <span dir="ltr"><<a href="mailto:shiek.kaleem@gmail.com" target="_blank">shiek.kaleem@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">Hi Guys ,<div><br></div><div>Just the same question from my side is there any guidelines i can follow if I have created my own OS.(LFS) </div>
</div><div class="HOEnZb"><div class="h5"><div class="gmail_extra"><br><br><div class="gmail_quote">On Mon, Jul 15, 2013 at 12:26 PM, bill4carson <span dir="ltr"><<a href="mailto:bill4carson@gmail.com" target="_blank">bill4carson@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hello Steven<br>
<div><br>
On 2013年07月15日 10:56, Steven Zhou wrote:<br>
> Hi,<br>
><br>
> We have developed our private OS and we want to benchmark the performance of it, including time of context switch, interrupt latency, IO output and so on ...<br>
</div> ^^^^^^^^^^^^^^^<br>
To what degree does "private" goes? Is your private OS an newly created one from scratch or based on one of open source version?<br>
<div><br>
<br>
><br>
> We want to study from Linux firstly, so could you guys give me some guide of Linux benchmark testing, including test methodologies, test code and so on.<br>
><br>
> Thanks in advance.<br>
><br>
> --<br>
> Best Regards.<br>
><br>
><br>
><br>
</div><div>> _______________________________________________<br>
> Kernelnewbies mailing list<br>
> <a href="mailto:Kernelnewbies@kernelnewbies.org" target="_blank">Kernelnewbies@kernelnewbies.org</a><br>
> <a href="http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies" target="_blank">http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies</a><br>
<br>
</div><span><font color="#888888">--<br>
八百里秦川尘土飞扬,三千万老陕齐吼秦腔。<br>
<br>
--bill<br>
</font></span><div><div><br>
_______________________________________________<br>
Kernelnewbies mailing list<br>
<a href="mailto:Kernelnewbies@kernelnewbies.org" target="_blank">Kernelnewbies@kernelnewbies.org</a><br>
<a href="http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies" target="_blank">http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies</a><br>
</div></div></blockquote></div><br></div>
</div></div><br>_______________________________________________<br>
Kernelnewbies mailing list<br>
<a href="mailto:Kernelnewbies@kernelnewbies.org">Kernelnewbies@kernelnewbies.org</a><br>
<a href="http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies" target="_blank">http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies</a><br>
<br></blockquote></div><br></div><div class="gmail_extra">Hi,<br><br></div><div class="gmail_extra">Benchmarking is a difficult ask. You need to define the matrix of what all things you would be measure and for what kind of workload. This data should be directly influenced by the application of the OS, where and how your OS would be used, what are critical performance areas you need to focus on.<br>
</div><div class="gmail_extra"><br></div><div class="gmail_extra">First of all whats the type of kernel you guys have implemented. The main thing is, what were the design decisions taken while implementing the policies, how should they perform in theory. After listing all these things down, you need to design benchmarking software modules for that though you can have a look at some existing once for the other OSes.<br>
<br></div><div class="gmail_extra">For linux,<br><br></div><div class="gmail_extra">1. systemtap.<br></div><div class="gmail_extra">2. perf.<br></div><div class="gmail_extra">3. Vtunes from intel.<br></div><div class="gmail_extra">
4. ftrace.<br></div><div class="gmail_extra"><br></div><div class="gmail_extra">Also, you would be required to emulate the workload to the real load that would be on your OS and do benchmarking.<br><br></div><div class="gmail_extra">
- Rohan<br></div></div>