[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [f-cpu] More Dark and Dusty Corners

Ben Franchuk schrieb:
>  Why is a virtual cpu needed on the first revision? I don't mind
> slow interrupt service providing hardware is defined for it -
> dma access and fifo's where needed. The serial port on the PC (
> the whole pc) is a bad example of doing interrupts. What is
> needed is smart - dumb hardware i/o controllers. If you you
> define a virtual machine lets define virtual hardware that is
> clean and works well. Bring back I/O control blocks? Pesudo
> fixed memory frame buffers? More real time smarts.
> Ben. 
> -- 
> Standard Disclaimer : 97% speculation 2% bad grammar 1% facts.
> "Pre-historic Cpu's" http://www.jetnet.ab.ca/users/bfranchuk
> Now with schematics.

I do not know whether anyone will need a virtual CPU some day in the future.
I just suggested to make it possible, because someone might need it.

If an application can determine whether it runs under a VM or not, the VM is not 
a very good VM. The VMs of Windows are not good VMs from this point of view. And 
they are an acceptable solution to problem of using applications and drivers for 
another OS (DOS) which a compromise at the best.

I am thinking of VMs like VMware or Bochs here, where you have several virtual 
machines each including its OS on a single machine. There are people who need 
this, network simulations are one example.

At www.vmware.com there is a paper somewhere, which explains that the i386 
architecture does not really allow a VM. VMware has to replace drivers for the 
OS inside the VM to make it really work. 

To unsubscribe, send an e-mail to majordomo@seul.org with
unsubscribe f-cpu       in the body. http://f-cpu.seul.org/