On 6/19/2024 7:00 AM, Eric Farman wrote:
> On Wed, 2024-06-19 at 12:32 +0200, Halil Pasic wrote:
>> On Tue, 18 Jun 2024 16:11:33 -0400
>> Eric Farman <farman@xxxxxxxxxxxxx> wrote:
>>
>>>>> +MODULE_DESCRIPTION("VFIO based Physical Subchannel device
>>>>> driver");
>>>>
>>>> Halil/Mathew/Eric,
>>>> Could you please comment on this ?
>>>
>>> That's what is in the prologue, and is fine.
>>
>> Eric can you explain it to me why is the attribute "physical"
>> appropriate
>> here? I did a quick grep for "Physical Subchannel" only turned up
>> hits
>> in vfio-ccw.
>
> One hit, in the prologue comment of this module. "Physical device" adds
> three to the tally, but only one of those is in vfio-ccw so we should
> expand your query regarding "physical" vs "emulated" vs "virtual" in
> the context of, say, tape devices.
>
>>
>> My best guess is that "physical" was somehow intended to mean the
>> opposite of "virtual". But actually it does not matter if our
>> underlying
>> subchannel is emulated or not, at least AFAIU.
>
> I also believe this was intended to mean "not virtual," regardless of
> whether there's emulation taking place underneath. That point is moot
> since I don't see that information being surfaced, such that the driver
> can only work with "physical" subchannels.
>
> I'm fine with removing it if it bothers you, but I don't see it as an
> issue.
Since I'm not the domain expert here I just copied what was in the prologue.
If someone can supply a suitable description, I'll update the patch to use it :)
I'm hoping to have these issued cleaned up tree-wide before the 6.11 merge window.
/jeff
