Re: [dpdk-dev] [PATCH] vhost: add interface name to virtio-net struct

["Loftus, Ciara" <ciara.loftus@intel.com>]

Hi Konstantin,

Thank you for the feedback. 

As regards the following:
"So if sizeof(ifr.ifr_name) > sizeof(dev->ifname) then we can endup with dev->ifname not being null-termianted?"
This should never be the case as now both dev->ifname and ifr.ifr_name are of size IFNAMSIZ in v2 of the patch.

Thanks,
Ciara

-----Original Message-----
From: Ananyev, Konstantin 
Sent: Thursday, December 18, 2014 5:03 PM
To: Loftus, Ciara; dev@dpdk.org
Cc: Anthony Fee
Subject: RE: [dpdk-dev] [PATCH] vhost: add interface name to virtio-net struct



> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of 
> ciara.loftus@intel.com
> Sent: Thursday, December 18, 2014 2:55 PM
> To: dev@dpdk.org
> Cc: Anthony Fee
> Subject: [dpdk-dev] [PATCH] vhost: add interface name to virtio-net 
> struct
> 
> From: Ciara Loftus <ciara.loftus@intel.com>
> 
> This patch fixes the issue whereby when using userspace vhost ports in 
> the context of vSwitching, the name provided to the hypervisor/QEMU of 
> the vhost tap device needs to be exposed in the library, in order for 
> the vSwitch to be able to direct packets to the correct device.
> This patch introduces an 'ifname' member to the virtio-net structure 
> which is populated with the tap device name when QEMU is brought up 
> with a vhost device.
> 
> Signed-off-by: Ciara Loftus <ciara.loftus@intel.com>
> Signed-off-by: Anthony Fee <anthonyx.fee@intel.com>
> Acked-by: Huawei Xie <huawei.xie@intel.com>
> ---
>  lib/librte_vhost/rte_virtio_net.h |    1 +
>  lib/librte_vhost/virtio-net.c     |   41 ++++++++++++++++++++++++++++++++++++-
>  2 files changed, 41 insertions(+), 1 deletions(-)
> 
> diff --git a/lib/librte_vhost/rte_virtio_net.h 
> b/lib/librte_vhost/rte_virtio_net.h
> index 00b1328..aebb4b5 100644
> --- a/lib/librte_vhost/rte_virtio_net.h
> +++ b/lib/librte_vhost/rte_virtio_net.h
> @@ -96,6 +96,7 @@ struct virtio_net {
>  	uint64_t		features;	/**< Negotiated feature set. */
>  	uint64_t		device_fh;	/**< device identifier. */
>  	uint32_t		flags;		/**< Device flags. Only used to check if device is running on data core. */
> +	char			ifname[32];	/** Name of the tap device **/
>  	void			*priv;		/**< private context */
>  } __rte_cache_aligned;
> 
> diff --git a/lib/librte_vhost/virtio-net.c 
> b/lib/librte_vhost/virtio-net.c index 852b6d1..7f90ecf 100644
> --- a/lib/librte_vhost/virtio-net.c
> +++ b/lib/librte_vhost/virtio-net.c
> @@ -43,6 +43,10 @@
>  #include <sys/mman.h>
>  #include <unistd.h>
> 
> +#include <sys/socket.h>
> +#include <linux/if_tun.h>
> +#include <linux/if.h>
> +
>  #include <rte_ethdev.h>
>  #include <rte_log.h>
>  #include <rte_string_fns.h>
> @@ -1000,6 +1004,39 @@ set_vring_kick(struct vhost_device_ctx ctx, 
> struct vhost_vring_file *file)  }
> 
>  /*
> + * Function to get the tap device name from the provided file 
> +descriptor and
> + * save it in the device structure.
> + */
> +static int
> +get_ifname(struct virtio_net *dev, int tap_fd, int pid) {
> +	struct eventfd_copy fd_tap;
> +	struct ifreq ifr;
> +	uint32_t size;
> +
> +	fd_tap.source_fd = eventfd(0, EFD_NONBLOCK | EFD_CLOEXEC);
> +	fd_tap.target_fd = tap_fd;
> +	fd_tap.target_pid = pid;
> +
> +	if (eventfd_copy(dev, &fd_tap))
> +		return -1;
> +
> +	ioctl(fd_tap.source_fd, TUNGETIFF, &ifr);

Shouldn't we check that ioctl() returns with success here, and if it fails, don't copy stuff over?

> +
> +	if (close(fd_tap.source_fd) < 0)
> +		RTE_LOG(ERR, VHOST_CONFIG,
> +			"(%"PRIu64") fd close failed\n",
> +			dev->device_fh);
> +
> +	size = strnlen(ifr.ifr_name, sizeof(ifr.ifr_name)) > sizeof(dev->ifname)?
> +			sizeof(dev->ifname): strnlen(ifr.ifr_name, sizeof(ifr.ifr_name));
> +

So if sizeof(ifr.ifr_name) > sizeof(dev->ifname) then we can endup with dev->ifname not being null-termianted?
Another nit: there is no need to call strnlen() twice.

Konstantin

> +	strncpy(dev->ifname, ifr.ifr_name, size);
> +
> +	return 0;
> +}
> +
> +/*
>   * Called from CUSE IOCTL: VHOST_NET_SET_BACKEND
>   * To complete device initialisation when the virtio driver is loaded,
>   * we are provided with a valid fd for a tap device (not used by us).
> @@ -1026,8 +1063,10 @@ set_backend(struct vhost_device_ctx ctx, struct vhost_vring_file *file)
>  	 */
>  	if (!(dev->flags & VIRTIO_DEV_RUNNING)) {
>  		if (((int)dev->virtqueue[VIRTIO_TXQ]->backend != VIRTIO_DEV_STOPPED) &&
> -			((int)dev->virtqueue[VIRTIO_RXQ]->backend != VIRTIO_DEV_STOPPED))
> +			((int)dev->virtqueue[VIRTIO_RXQ]->backend != VIRTIO_DEV_STOPPED)) {
> +			get_ifname(dev, file->fd, ctx.pid);
>  			return notify_ops->new_device(dev);
> +		}
>  	/* Otherwise we remove it. */
>  	} else
>  		if (file->fd == VIRTIO_DEV_STOPPED)
> --
> 1.7.4.1

--------------------------------------------------------------
Intel Shannon Limited
Registered in Ireland
Registered Office: Collinstown Industrial Park, Leixlip, County Kildare
Registered Number: 308263
Business address: Dromore House, East Park, Shannon, Co. Clare

This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.