Re: [PATCH v4 10/14] gunyah: sysfs: Add node to describe supported features

[Elliot Berman]

On 9/30/2022 5:06 AM, Greg Kroah-Hartman wrote:
> On Wed, Sep 28, 2022 at 12:56:29PM -0700, Elliot Berman wrote:
> > Add a sysfs node to list the features that the Gunyah hypervisor and
> > Linux supports. For now, Linux support cspace (capability IDs) and
> > message queues, so only report those..
> >
> > Signed-off-by: Elliot Berman <quic_eberman@xxxxxxxxxxx>
> > ---
> >   Documentation/ABI/testing/sysfs-hypervisor-gunyah | 15 +++++++++++++++
> >   drivers/virt/gunyah/sysfs.c                       | 15 +++++++++++++++
> >   2 files changed, 30 insertions(+)
> >
> > diff --git a/Documentation/ABI/testing/sysfs-hypervisor-gunyah b/Documentation/ABI/testing/sysfs-hypervisor-gunyah
> > index 7d74e74e9edd..6d0cde30355a 100644
> > --- a/Documentation/ABI/testing/sysfs-hypervisor-gunyah
> > +++ b/Documentation/ABI/testing/sysfs-hypervisor-gunyah
> > @@ -1,3 +1,18 @@
> > +What:		/sys/hypervisor/gunyah/features
> > +Date:		October 2022
> > +KernelVersion:	6.1
> > +Contact:	linux-arm-msm@xxxxxxxxxxxxxxx
> > +Description:	If running under Gunyah:
> > +		Space separated list of features supported by Linux and Gunyah:
> > +		"cspace": Gunyah devices
> > +		"doorbell": Sending/receiving virtual interrupts via Gunyah doorbells
> > +		"message-queue": Sending/receiving messages via Gunyah message queues
> > +		"vic": Interrupt lending
> > +		"vpm": Virtual platform management
> > +		"vcpu": Virtual CPU management
> > +		"memextent": Memory lending/management
> > +		"trace": Gunyah hypervisor tracing
>
> Please no.  Why do you really need this type of list?  What hypervisor
> will NOT have them all present already?  Who will use this file and what
> will it be used for?
>
> sysfs files should just be 1 value and not need to be parsed.  Yes, we
> have lists of features at times, but really, you need a very very good
> reason why this is the only way this information can be exposed and who
> is going to use it in order to be able to have this accepted.


We're currently at phase where all the features implemented so far as 
considered part of the "base" featureset. We're thinking of future 
features implemented in Gunyah: userspace might need to know that some 
hypervisor feature is present and that it should make use of the feature 
instead of some fallback behavior.

I can drop this and it should be OK IMO to introduce it later if needed. 
The lack of the "gunyah/features" node would be sufficient for a 
userspace program to know that some new feature isn't present.

> thanks,
>
> greg k-h