OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

virtio-comment message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [virtio-dev] RE: [virtio-comment] [PATCH v14 04/11] transport-pci: Avoid first vq index reference


On Tue, Apr 25, 2023 at 06:20:08PM +0200, Halil Pasic wrote:
> On Tue, 25 Apr 2023 09:15:14 -0400
> "Michael S. Tsirkin" <mst@redhat.com> wrote:
> 
> > On Tue, Apr 25, 2023 at 03:02:15PM +0200, Halil Pasic wrote:
> > > On Tue, 25 Apr 2023 04:10:00 +0000
> > > Parav Pandit <parav@nvidia.com> wrote:
> [..]
> > > > Sure. All you wrote is correct.
> > > >   
> > > 
> > > I'm happy we agree. All I say we may want to rewrite the 
> > > 
> > > "Each virtqueue is identified by a virtqueue index; virtqueue index
> > > range is from 0 to 65535 inclusive."
> > > as
> > > "Each virtqueue is uniquely identified by a virtqueue index. The number
> > > of supported virtqueues is device dependent, but can never exceed 65536.
> > > Thus 16 bit is sufficient to represent virtqueue indexes. If the number
> > > of virtqueues currently supported by some device is N, each of it is
> > > virtqueues is uniquely identified by a single index from the range
> > > [0..N-1]."  
> > 
> > Seems to be repeating same thing over and over.
> 
> Nod.
> 
> > This redundancy has cost, e.g. more places to change when we
> > talk about admin queues.
> > Yes it can not be any number 0 to 65535 but this kind of nitpicking
> > belongs in conformance statements not in general description.
> > 
> 
> I tend to agree. I would prefer to have it in some sort of conformance
> statement, but I would also prefer to have it in exactly one place (and
> not all over the place).
> 
> Regards,
> Halil
> 
> [..]

I don't feel there is much to say in one place. We did not in the past
specify what happens if drivers e.g. set queue select to a large number.
So some drivers might do just that for a variety of reasons.

In other words, given a queue there's an index
and i feel Parav's text explains what it is
pretty unambigously and also succintly.
I would even maybe drop the 0 to 65535 thing and just
say that it is 0 based.
What are the specific limitations on the index is already
explained elsewhere.


-- 
MST



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]