Re: [virtio-comment] [PATCH v1 1/2] virtio-mem: introduce VIRTIO_MEM_F_UNPLUGGED_INACCESSIBLE

[Cornelia Huck <cohuck@redhat.com>]

On Tue, Aug 17 2021, David Hildenbrand <david@redhat.com> wrote:

> On 17.08.21 11:23, Cornelia Huck wrote:
>> On Tue, Aug 17 2021, David Hildenbrand <david@redhat.com> wrote:
>> 
>>> On 17.08.21 10:50, Cornelia Huck wrote:
>>>> On Thu, Aug 12 2021, David Hildenbrand <david@redhat.com> wrote:
>>>>> +With VIRTIO_MEM_F_UNPLUGGED_INACCESSIBLE, the device MAY allow the CPU to
>>>>
>>>> "If VIRTIO_MEM_F_UNPLUGGED_INACCESSIBLE has been negotiated, ..."
>>>>
>>>
>>> ack
>>>
>>>>> +read memory of unplugged memory blocks inside \field{usable_region_size}.
>>>>
>>>> A compliant driver would not read that memory, would it?
>>>
>>> Indeed. The device could decide to allow for reading, but it's pretty
>>> much undefined behavior.
>> 
>> Maybe make it "SHOULD NOT"? Less strong than "MUST NOT", but still makes
>> clear that a driver reading it is pretty much doing the wrong thing.
>> 
>
> I was also wondering to just drop it completely, because it's actually 
> undefined behavior. What do you think?

Right, I'm not sure we need to specify this at all. Do others have an
opinion? The one thing it does is adding a nicely symmetrical statement.