It seems like you are not actually running the checkpoint after you create
it. Once the checkpoint is created, append the flag

-loadvm test

to your marss run string.


On Mon, May 20, 2013 at 2:39 AM, Sungchan Kim <[email protected]>wrote:

> Dear all,
>
> I'm having difficulties in execution cycle measurement and using
> checkpoint on MARSS.
>
> First of all, I want to measure "ls" command by using start_sim, stop_sim,
> and kill_sim utilities on MARSS as follows.
>
> > ./start_sim; ls -la; /stop_sim; ./kill_sim;
>
> Regarding this, I have the following questions.
> - Where can I get the execution cycle for the "ls -la" command after
> getting the above command line done? I know MARSS provides detailed
> statistical information of simulation. But the meanings of numbers are
> sometimes hard to understand. I am not sure which number I  should look
> into.
> - What actually happens inside MARSS if I execute stop_sim? I recognize
> the simulation becomes slow after start_sim begins. But nothing
> recognizable happens with stop_sim. (I expected the simulation speed would
> get back to before executing start_sim)
>
> Second, I want to avoid long waiting time until Linux cold boot is done
> whenever I launch MARSS. My understanding is that checkpoint would help for
> that purpose. I tried to run the simple checkpoint example in the MICRO12
> tutorial.
>
> >./create_checkpoint test; ./FFT; ./stop_sim
>
> No further messages are out on guest OS afterwards, and in host console,
> the following message were displayed only.
>
> > PTLCALL type PTLCALL_CHECKPOINT
> > MARSSx86::Creating checkpoint test
>
> It seems that the simulator seems to get stuck after giving the command
> line above on the guest OS. Any other messages did not appear after then
> and key input didn't get accepted. I used Ubuntu12.04 LTS, which has been
> upgraded from the original disk image on MARSS web site.
> - How can I confirm that checkpoint has been successfully created?
> - How can I use the checkpoint afterwards?
>
> The questions may be stupid. Unfortunately, it seems that the details on
> the above issues are missing on  MARSS website. It would be highly
> appreciated if anyone can provide some instructions to follow easily.
>
> Best regards,
> Sungchan Kim
>
> _______________________________________________
> http://www.marss86.org
> Marss86-Devel mailing list
> [email protected]
> https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel
>
>
_______________________________________________
http://www.marss86.org
Marss86-Devel mailing list
[email protected]
https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel

Reply via email to