Hi,
Where can I find the software for Phase Prediction and Dynamic
Clustering using Simpoint mentioned in this paper "Dynamic Phase
Analysis for Cycle-Close Trace Generation" at CODES+ISSS 2005 .
Simpoint's webpage hosts only the offline version.
Taj.
As the business war on the air is hotting up with so many low-budget
airlines dominating the sky, the passengers can now enjoy the best
deals. The webguide - http://air.myguideforlife.com - helps you find
the cheapest and even free air tickets to make your journey ore
happier and very less expensive.
Visit the website now to find out how to get the cheap and free air
tickets: http://air.myguideforlife.com
As the business war on the air is hotting up with so many low-budget
airlines dominating the sky, the passengers can now enjoy the best
deals. The webguide - http://air.myguideforlife.com - helps you find
the cheapest and even free air tickets to make your journey ore
happier and very less expensive.
Visit the website now to find out how to get the cheap and free air
tickets: http://air.myguideforlife.com
Hi All,
I am trying to use the BBtrackerSS to generate the Basic Block
Vector, the benchmark i used is the SPEC2000 Alpha binaries, and the
commond is:
$./sim-fast gcc00.peak.ev6 < 166.i
but the simulation stops after execute about 1 millon
instructions. Apparently, the benchmark hasnot been executed
completely.
Juring the execution, there is a warning says:
Warning: partially supported sigprocmask() call ....
Does anyone meet the same problem as me, how to solve this
problem, that will be so nice for you to help me!
Thank you for your suggestion in advance!
Pan
Hi Greg,
Thanks for the reply. This is just what I want.
Yu
--- In simpoint@yahoogroups.com, Greg Hamerly <Greg_Hamerly@...> wrote:
>
> Use the -saveLabels option. It will give a list of the cluster label
> for each interval (and the distance from that interval to the cluster
> center).
>
> This paper serves as a reference to SimPoint version 3:
> http://www.jilp.org/vol7/v7paper14.pdf
>
> - greg
>
> On Tue, 6 May 2008, nwuyuzhang wrote:
>
> > Hi,
> >
> > I'm new to SimPoint and I was trying to use it to do phase analysis on
> > some program. While, by just using the "-maxK" option, I can only get
> > the optimal number of clusters. How can I know which interval is in
> > which cluster? I checked with the "-saveVectorsTxtFmt" option and
> > generated an output file, how should I use that file?
> >
> > Thanks,
> > Yu
> >
> >
> > ------------------------------------
> >
> > Yahoo! Groups Links
> >
> >
> >
>
Use the -saveLabels option. It will give a list of the cluster label
for each interval (and the distance from that interval to the cluster
center).
This paper serves as a reference to SimPoint version 3:
http://www.jilp.org/vol7/v7paper14.pdf
- greg
On Tue, 6 May 2008, nwuyuzhang wrote:
> Hi,
>
> I'm new to SimPoint and I was trying to use it to do phase analysis on
> some program. While, by just using the "-maxK" option, I can only get
> the optimal number of clusters. How can I know which interval is in
> which cluster? I checked with the "-saveVectorsTxtFmt" option and
> generated an output file, how should I use that file?
>
> Thanks,
> Yu
>
>
> ------------------------------------
>
> Yahoo! Groups Links
>
>
>
Hi,
I'm new to SimPoint and I was trying to use it to do phase analysis on
some program. While, by just using the "-maxK" option, I can only get
the optimal number of clusters. How can I know which interval is in
which cluster? I checked with the "-saveVectorsTxtFmt" option and
generated an output file, how should I use that file?
Thanks,
Yu
Hi everybody,
I know this email seems a little off-topic but has anyone here
tried porting a Basic Block Vector Generator in M5 for use in
Simpoints? Would appreciate if you can provide diff files. Also, If I
use the BBV generator from SimpleScalar(sim-fast)[1] will I get the
same results(since it has been said in the documentation that M5 is
based on SimpleScalar,and also considering I use the ALPHA ISA)?
thanks,
-dean
[1]http://www.cs.ucsd.edu/~calder/simpoint/simplescalar-bbv.htm
Ccietutorial.com is your gateway to the best sites
on the Internet for Software! Looking for Computers,
Domain Names, or Education? Browse our recommended
resources or just try the Ccietutorial.com search
http://ccietutorial.com/
--- In simpoint@yahoogroups.com, "Alexandre Strube" <surak@...> wrote:
>
> 2007/8/20, xiejingsong <xiejingsong@...>:
> >
> > For example, and we assume the Simulation Point is 90, then I
start
> > simulating at instruction 90 * 100 million (9 billion) and stop
> > simulating at instruction 9.1 billion assuming a 100 million
interval
> > size was used to create the simulation points.
> > How can I calculate when the interval is 250M: we assume the
simulation
> > point also is 90. do I start simulating at instruction 90*100M or
> > 90*250M?
> > .
> >
> >
> >
> If you collected the points using a 250m interval, you must then
> multiplicate your point number by that interval - in this case,
250. Notice
> that the point 90 taken with an interval of 100M instructions is
entirely
> different that the point 90 gathered with an interval of 250M.
> How did you collect your points?
>
>
> --
> []
> Alexandre Strube
> surak@...
>
I use single simulation point.
Thanks a lot.
For example, and we assume the Simulation Point is 90, then I start
simulating at instruction 90 * 100 million (9 billion) and stop
simulating at instruction 9.1 billion assuming a 100 million interval
size was used to create the simulation points.
How can I calculate when the interval is 250M: we assume the simulation
point also is 90. do I start simulating at instruction 90*100M or
90*250M?
.
If you collected the points using a 250m interval, you must then multiplicate your point number by that interval - in this case, 250. Notice that the point 90 taken with an interval of 100M instructions is entirely different that the point 90 gathered with an interval of 250M.
For example, and we assume the Simulation Point is 90, then I start
simulating at instruction 90 * 100 million (9 billion) and stop
simulating at instruction 9.1 billion assuming a 100 million interval
size was used to create the simulation points.
How can I calculate when the interval is 250M: we assume the simulation
point also is 90. do I start simulating at instruction 90*100M or
90*250M?
has anyone ever tried to compare the results obtained from simpoints with the full execution on a real machine?
I mean, instead of using some fast-forward technique on a simulator, use a real machine with, say, a checkpoint on a phase and instrumentation at the end of this phase to measure time and compare the performance indexes (with PAPI, for instance).
All the papers I found are related to simulation, and it's quite harder to do on a real machine (as checkpoints on real hardware are more difficult, the file handlers go away, there's no fast-forward, etc).
I am following this path, and if someone else has done it before, please let me know, let's exchange some ideas!
You need to be much, much more specific on where you are having
problems.
If you are having problems getting started, read the documentation
that comes with the BBV Tracker (README.txt in BBTrackerSS.tar) and
SimpleScalar (http://www.simplescalar.com/docs.html).
If you are having problems compiling SimpleScalar under cygwin, do a
Google search for "SimpleScalar cygwin" for some starting points. We
have not attempted to collect BBVs with SimpleScalar under cygwin, so
you may run into some problems here. We will try to help you by trying
to answer any specific questions you may have, but can't figure this
out for you. We routinely collect BBVs on x86 Linux, so you may want
to consider running on that platform.
Please keep in mind that SimPoint is a free research tool, and not a
commercial product. We are busy graduate students and professors - we
will do our best to answer questions, but we do not have the time or
money to provide commercial-grade technical support or customer
service. :)
-jeremy
On Tue, Jul 10, 2007 at 12:58:00AM +0200, sahbi ayed wrote:
> hi
> i m tryinig to use the Sim-Fast BBV Tracker to generate Basic block vectors.
> I downloaded the cygwin to compile them on windows.
> Could you please help me to generate some BBVs
> thanks
>
>
>
> ---------------------------------
> Nouveau : téléphonez moins cher avec Yahoo! Messenger ! Découvez les tarifs
exceptionnels pour appeler la France et l'international.Téléchargez la version
beta.
On Thu, Jul 12, 2007 at 01:47:55PM -0000, sahbi_ayed wrote:
> hi,
> I've been trying to generate the basic block vectors using the
> code provided at
> http://www-cse.ucsd.edu/~calder/simpoint/simplescalar-bbv.htm.
> the three files are: bbtracker.c bbtracker.h and sim-fast.c .
> could you please help me to generate some BB vectors.
> thanks
>
>
>
>
> Yahoo! Groups Links
>
>
>
On Tue, Jul 17, 2007 at 01:53:34PM -0400, ayed sahbi wrote:
> hi
> i m tryinig to use the Sim-Fast BBV Tracker to generate Basic block vectors.
> I downloaded the cygwin to compile them on windows and the code
> provided at :http://www-cse. ucsd.edu/ ~calder/simpoint /simplescalar-
> bbv.htm
> the three files are: bbtracker.c bbtracker.h and sim-fast.c.
>
> Could you please help me to generate some BBVs
> thanks
>
>
>
>
>
> **********************************************
> Tel : 514 690 1822
> Fax: 514 274 8313
hi,
I've been trying to generate the basic block vectors using the
code provided at
http://www-cse.ucsd.edu/~calder/simpoint/simplescalar-bbv.htm.
the three files are: bbtracker.c bbtracker.h and sim-fast.c .
could you please help me to generate some BB vectors.
thanks