On Tue, Jul 3, 2012 at 9:53 AM, HU, Yang <[email protected]> wrote:

> Hi everyone,
>
> Now I have some questions on Marss86 and PTLSim 3.0. I know marss takes
> some advantage of PTLSim, and as far as I know there is a version of PTLSim
> named PTLSim 3.0 in which the full system environment is constructed based
> on KVM while no longer Xen. And now marss is developed based on qemu. So I
> am wondering what's the difference in terms of developing between qemu or
> kvm? And how does marss implement the kvm-enable feature? Is there any
> document that I can read?
>
> First of all I am not sure if PTLsim 3.0 is being maintained and if it
came out of 'alpha' stage.  Major difference between PTLsim 3.0 and MARSS
is how simulation mode is integrated into QEMU/KVM environment.

In MARSS, we have placed simulation model integrated into Qemu to get full
advantage of Qemu's all emulation models. This makes it little easy to hack
device emulation models and communicate it with core/cache models.

PTLsim 3.0 approach is to insert simulation model as 'ROM/Firmware' into
KVM's memory space that takes over KVM/real-hardware when simulation mode
is triggered. Biggest issue with this 'firmware' based design is ability to
easily communicate between different emulated devices and simulation
models.  Also older Xen based design's biggest limitation was lack of using
any standard libraries, I am not sure if that limitation was carried over
to KVM based design or not.

- Avadh

Best,
>
> Yang
> --
>
>
> _______________________________________________
> 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