Re: [Xen-devel] [PATCH 3/4] xen: Add xenbusd device driver

[Ian Campbell <Ian.Campbell@xxxxxxxxxx>]

On Wed, 2011-12-07 at 21:13 +0000, Bastian Blank wrote:
> Access for xenstored to the event channel and pre-allocated ring is
> managed via xenfs.  This adds its own device driver featuring mmap for
> the ring and an ioctl for the event channel.
> 
> Signed-off-by: Bastian Blank <waldi@xxxxxxxxxx>
> ---
>  drivers/xen/xenbus/Makefile             |    1 +
>  drivers/xen/xenbus/xenbus_dev_backend.c |   79 
> +++++++++++++++++++++++++++++++
>  include/xen/xenbus_dev.h                |   41 ++++++++++++++++
>  3 files changed, 121 insertions(+), 0 deletions(-)
>  create mode 100644 drivers/xen/xenbus/xenbus_dev_backend.c
>  create mode 100644 include/xen/xenbus_dev.h
> 
> diff --git a/drivers/xen/xenbus/Makefile b/drivers/xen/xenbus/Makefile
> index a2ea363..31e2e90 100644
> --- a/drivers/xen/xenbus/Makefile
> +++ b/drivers/xen/xenbus/Makefile
> @@ -10,4 +10,5 @@ xenbus-objs += xenbus_probe.o
>  xenbus-be-objs-$(CONFIG_XEN_BACKEND) += xenbus_probe_backend.o
>  xenbus-objs += $(xenbus-be-objs-y)
>  
> +obj-$(CONFIG_XEN_BACKEND) += xenbus_dev_backend.o

In principal this could be its own module which does not depend on any
of the backend stuff but I think this is fine for now.

> [...]

> +static int __init xenbusd_init(void)
> +{
> +     int err;
> +
> +     if (!xen_initial_domain())
> +             return -ENODEV;
> +
> +     err = misc_register(&xenbusd_dev);
> +     if (err)
> +             printk(KERN_ERR "Could not register xenbus device\n");

"Could not register xenbus backend device" to distinguish from the f.e.
failure message, you could also add "frontend" there I suppose. Other
than that:

Acked-by: Ian Campbell <ian.campbell@xxxxxxxxxx>



_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel