On 25.01.2014, at 03:04, Scott Wood <scottwood@xxxxxxxxxxxxx> wrote: > On Fri, 2014-01-24 at 17:56 -0800, Christoffer Dall wrote: >> On Sat, Jan 25, 2014 at 12:24:08AM +0000, Peter Maydell wrote: >>> Finally, I think it's a bit confusing in that "as if the guest had >>> accessed memory" is assigning implicit semantics to memory >>> in the emulated system, when memory is actually kind of outside >>> KVM's purview because it's not part of the CPU. >>> >> >> The "as if the guest had accessed memory", would imply that user space >> needs to look at the settings of the CPU (in the case of ARM the E-bit) >> to understand which value the CPU intended for the memory transaction. > > No, it doesn't, just as userspace doesn't need to do that to interpret > DMA descriptors. DMA descriptors are handled by specific pieces of hardware that access memory. Guests usually push data to memory in a layout so that it's "natural" to the device they want to talk to. MMIO is more tricky. Alex -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
