Re: advice for clock settings

>So is the internal beat the IPS, so that the bochs machine
>will try to emulate that number of instructions per second?
>But on a physical CPU, don't different types of instructions
>take different numbers of cycles? And isn't possible that
>the same instruction may take a different number of cycles
>depending on, say, where the operands are located?

Everything is possible, it is even possible to write a full cycle accurate
simulator of modern out-of-order processor around Bochs.
But what is the reason to do so ?
One instruction as one cycle is the simplest way to increment internal
time, any other would significantly slow down the emulation and not
required at all for common usage.

Thanks,
Stanislav

On Sun, Dec 30, 2012 at 2:44 PM, Ariel Burton <[email protected]>wrote:

> Thank you for this.
>
> On Sat 29 Dec 03:17, Volker Ruppert wrote:
> > Hi!
> >
> > > What are the bochs clock settings that will most closely
> > > match this?  I am not really concerned by the actual time
> > > taken to run the experiments, that is, I don't care if
> > > a second inside the bochs machine takes 3 seconds to elapse
> > > on the `host'.  What I do want is a metric that allows me
> > > to compare different runs using the times reported inside
> > > the machine.  (As I understand it, clock set to none will
> > > have time inside the machine run at a rate that depends
> > > on the number of instructions executed, which I don't think
> > > does quite what I want.)
> >
> > The internal Bochs timing is always based on the specified number of
> > instructions per emulated second configured with the "cpu: ips="
> > directive. The "clock: sync=" directive is only used to have some kind
> > of synchronization between host and guest timing. If this parameter is
> > set to "none", the emulator runs at maximum speed from the host's point
> > of view and the timer intervals are stable from the guest OS's point of
> > view.
>
> So is the internal beat the IPS, so that the bochs machine
> will try to emulate that number of instructions per second?
> But on a physical CPU, don't different types of instructions
> take different numbers of cycles?  And isn't possible that
> the same instruction may take a different number of cycles
> depending on, say, where the operands are located?
>
> Thanks,
>
> Ariel Burton
>
>
>
> ------------------------------------------------------------------------------
> Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
> MVC, Windows 8 Apps, JavaScript and much more. Keep your skills current
> with LearnDevNow - 3,200 step-by-step video tutorials by Microsoft
> MVPs and experts. ON SALE this month only -- learn more at:
> http://p.sf.net/sfu/learnmore_123012
> _______________________________________________
> bochs-developers mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/bochs-developers
>

------------------------------------------------------------------------------
Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
MVC, Windows 8 Apps, JavaScript and much more. Keep your skills current
with LearnDevNow - 3,200 step-by-step video tutorials by Microsoft
MVPs and experts. ON SALE this month only -- learn more at:
http://p.sf.net/sfu/learnmore_123012

_______________________________________________
bochs-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bochs-developers