WoodyZ wrote:
Since you're a VMware employee I'm going to ask this because I'm not sure whether or not you were joking with that last comment... So are you saying that when you originally said "Your physical optical drive is not directly accessible by the guest operating system." that you were mistaken? I'll assume so since Windows can actually see it the same as the Host does although because it's in a Virtual Machine does this present any limits to what functionality is available in the Virtual Machine vs its native functionality?
It is true that your physical drive is not directly accessible by the guest operating system. The physical drive is going to do DMA to/from buffers in physical memory, and the guest operating system can't be responsible for specifying the physical addresses, because the guest operating system knows nothing about where its buffers reside in host physical memory. VT-d offers remapping of guest physical addresses to host physical addresses so that this can work seamlessly, but Fusion does not make use of VT-d.
However, I was surprised to see that the physical drive identification is passed through by the IDE/ATAPI emulation layers. I really don't know enough about our IDE/ATAPI emulation to say what special features of the physical drive may be emulated in the guest. I had assumed that only generic optical drive features were emulated, but then there would be no reason to pass through the physical drive's identification to the guest. Apparently, our virtual device is more sophisticated than I thought. Hence, "I don't know what I'm talking about."
All of this aside, USB device emulation goes through a completely different path, so none of this is relevant to the original question.
