Sjur Brændeland <sjurb...@gmail.com> writes:
> On Fri, Feb 22, 2013 at 1:42 AM, Rusty Russell <ru...@rustcorp.com.au> wrote:
>> Ohad Ben-Cohen <o...@wizery.com> writes:
>>> On Thu, Feb 21, 2013 at 8:37 AM, Rusty Russell <ru...@rustcorp.com.au>
>>> wrote:
>>> What do you think about creating some virtio-level wrappers for the
>>> vringh handlers?
>>>
>>> I don't think we're going to stop with caif as the only vringh user,
>>> and it could be nice if we follow the virtio spirit of decoupling the
>>> drivers from the low level implementation. It sure did prove itself
>>> when the remoteproc use cases started showing up, and it's neat.
>>
>> The problem space is a bit different. My immediate concern is getting
>> vhost (and thus vhost_net/blk) to use vringh: I wanted to unify the
>> in-userspace and in-kernelspace ring implementations. We don't have
>> that issue in virtqueue.c.
>>
>> vhost is (will be) the higher abstraction for in-userspace rings,
>> perhaps we want an equivalent for in-kernelspace rings. I'm happy to
>> look at patches, but I don't immediately see what it would look like...
>
> I'm not sure if the tight binding between vringh and remoteproc is
> a big problem. I think the most obvious use-case for kernel vringh is
> when they are instantiated from remoteproc.
>
> But if we where to make wrappers, how about something like this?
BTW, I'm leaving Ohad and you to battle it out. There's no huge hurry,
so make sure you're both happy.
> diff --git a/include/linux/virtio_config.h b/include/linux/virtio_config.h
> index 29b9104..ca257d8 100644
> --- a/include/linux/virtio_config.h
> +++ b/include/linux/virtio_config.h
> @@ -53,0 +54,8 @@
> + * @find_vrh: find the host vrings and instantiate them
> + * vdev: the virtio_device
> + * nhvrs: the number of host vrings to find
> + * hvrs: on success, includes new host vrings
> + * callbacks: array of driver callbacks, for each host vring
> + * include a NULL entry for vqs that do not need a callback
> + * Returns 0 on success or error status
> + * @del_vrh: free the host vrings found by find_vrh().
> @@ -55,0 +64 @@ typedef void vq_callback_t(struct virtqueue *);
> +typedef void vrh_callback_t(struct virtio_device *, struct vringh *);
> @@ -72,0 +82,4 @@ struct virtio_config_ops {
> + int (*find_vrh) (struct virtio_device *, unsigned nhvrs,
> + struct vringh *vrhs[],
> + vrh_callback_t *callbacks[]);
> + int (*del_vrhs)(struct virtio_device *);
> diff --git a/include/linux/vringh.h b/include/linux/vringh.h
> index 4c4c918..78aecc9 100644
> --- a/include/linux/vringh.h
> +++ b/include/linux/vringh.h
> @@ -52,0 +53,3 @@ struct vringh {
> +
> + /* The function to call when buffers are available */
> + void (*notify)(struct vringh *);
This will work for vhost, too, so no problems here.
Cheers,
Rusty.
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization
