Re: [dpdk-dev] [PATCH] vhost: fix crash if set vring num handling failed

[Tiwei Bie <tiwei.bie@intel.com>]

On Thu, Aug 30, 2018 at 09:54:50AM +0300, Ilya Maximets wrote:
> On 30.08.2018 09:06, Tiwei Bie wrote:
> > On Wed, Aug 29, 2018 at 05:00:47PM +0300, Ilya Maximets wrote:
> >> Any thoughts on this?
> > 
> > Hi Ilya,
> > 
> > Currently, the errors happened during messages handling
> > aren't handled properly. E.g. in vhost_user_set_mem_table(),
> > similar issues [1] may happen too. We might want to find
> > a way to fix this once and for all.
> > 
> > [1]
> > https://github.com/DPDK/dpdk/blob/3605968c2fa7/lib/librte_vhost/vhost_user.c#L826
> > https://github.com/DPDK/dpdk/blob/3605968c2fa7/lib/librte_vhost/vhost_user.c#L837
> > https://github.com/DPDK/dpdk/blob/3605968c2fa7/lib/librte_vhost/vhost_user.c#L887
> > 
> > Thanks
> 
> Hi. I understand your position position and, actually, I looked at
> other functions while preparing this fix. There is one difference
> between vhost_user_set_mem_table() and vhost_user_set_vring_num().
> The first one is able to reply bad status.
> But yes, you're right that we should handle all the errors.
> What about following patch (compile tested only):
> ----
> diff --git a/lib/librte_vhost/vhost_user.c b/lib/librte_vhost/vhost_user.c
> index 9aa1ce118..63d145b2d 100644
> --- a/lib/librte_vhost/vhost_user.c
> +++ b/lib/librte_vhost/vhost_user.c
> @@ -1014,7 +1014,7 @@ vhost_user_set_vring_call(struct virtio_net *dev, struct VhostUserMsg *pmsg)
>  	vq->callfd = file.fd;
>  }
>  
> -static void
> +static int
>  vhost_user_set_vring_kick(struct virtio_net **pdev, struct VhostUserMsg *pmsg)
>  {
>  	struct vhost_vring_file file;
> @@ -1032,7 +1032,7 @@ vhost_user_set_vring_kick(struct virtio_net **pdev, struct VhostUserMsg *pmsg)
>  	/* Interpret ring addresses only when ring is started. */
>  	dev = translate_ring_addresses(dev, file.index);
>  	if (!dev)
> -		return;
> +		return -1;
>  
>  	*pdev = dev;
>  
> @@ -1049,6 +1049,7 @@ vhost_user_set_vring_kick(struct virtio_net **pdev, struct VhostUserMsg *pmsg)
>  	if (vq->kickfd >= 0)
>  		close(vq->kickfd);
>  	vq->kickfd = file.fd;
> +	return 0;
>  }
>  
>  static void
> @@ -1172,14 +1173,19 @@ vhost_user_get_protocol_features(struct virtio_net *dev,
>  	msg->size = sizeof(msg->payload.u64);
>  }
>  
> -static void
> +static int
>  vhost_user_set_protocol_features(struct virtio_net *dev,
>  				 uint64_t protocol_features)
>  {
> -	if (protocol_features & ~VHOST_USER_PROTOCOL_FEATURES)
> -		return;
> +	if (protocol_features & ~VHOST_USER_PROTOCOL_FEATURES) {
> +		RTE_LOG(ERR, VHOST_CONFIG,
> +			"(%d) received invalid protocol features.\n",
> +			dev->vid);
> +		return -1;
> +	}
>  
>  	dev->protocol_features = protocol_features;
> +	return 0;
>  }
>  
>  static int
> @@ -1657,8 +1663,6 @@ vhost_user_msg_handler(int vid, int fd)
>  		break;
>  	case VHOST_USER_SET_FEATURES:
>  		ret = vhost_user_set_features(dev, msg.payload.u64);
> -		if (ret)
> -			return -1;
>  		break;
>  
>  	case VHOST_USER_GET_PROTOCOL_FEATURES:
> @@ -1666,14 +1670,14 @@ vhost_user_msg_handler(int vid, int fd)
>  		send_vhost_reply(fd, &msg);
>  		break;
>  	case VHOST_USER_SET_PROTOCOL_FEATURES:
> -		vhost_user_set_protocol_features(dev, msg.payload.u64);
> +		ret = vhost_user_set_protocol_features(dev, msg.payload.u64);
>  		break;
>  
>  	case VHOST_USER_SET_OWNER:
> -		vhost_user_set_owner();
> +		ret = vhost_user_set_owner();
>  		break;
>  	case VHOST_USER_RESET_OWNER:
> -		vhost_user_reset_owner(dev);
> +		ret = vhost_user_reset_owner(dev);
>  		break;
>  
>  	case VHOST_USER_SET_MEM_TABLE:
> @@ -1681,8 +1685,9 @@ vhost_user_msg_handler(int vid, int fd)
>  		break;
>  
>  	case VHOST_USER_SET_LOG_BASE:
> -		vhost_user_set_log_base(dev, &msg);
> -
> +		ret = vhost_user_set_log_base(dev, &msg);
> +		if (ret)
> +			goto skip_to_reply;
>  		/* it needs a reply */
>  		msg.size = sizeof(msg.payload.u64);
>  		send_vhost_reply(fd, &msg);
> @@ -1693,23 +1698,25 @@ vhost_user_msg_handler(int vid, int fd)
>  		break;
>  
>  	case VHOST_USER_SET_VRING_NUM:
> -		vhost_user_set_vring_num(dev, &msg);
> +		ret = vhost_user_set_vring_num(dev, &msg);
>  		break;
>  	case VHOST_USER_SET_VRING_ADDR:
> -		vhost_user_set_vring_addr(&dev, &msg);
> +		ret = vhost_user_set_vring_addr(&dev, &msg);
>  		break;
>  	case VHOST_USER_SET_VRING_BASE:
> -		vhost_user_set_vring_base(dev, &msg);
> +		ret = vhost_user_set_vring_base(dev, &msg);
>  		break;
>  
>  	case VHOST_USER_GET_VRING_BASE:
> -		vhost_user_get_vring_base(dev, &msg);
> +		ret = vhost_user_get_vring_base(dev, &msg);
> +		if (ret)
> +			goto skip_to_reply;
>  		msg.size = sizeof(msg.payload.state);
>  		send_vhost_reply(fd, &msg);
>  		break;
>  
>  	case VHOST_USER_SET_VRING_KICK:
> -		vhost_user_set_vring_kick(&dev, &msg);
> +		ret = vhost_user_set_vring_kick(&dev, &msg);
>  		break;
>  	case VHOST_USER_SET_VRING_CALL:
>  		vhost_user_set_vring_call(dev, &msg);
> @@ -1728,10 +1735,10 @@ vhost_user_msg_handler(int vid, int fd)
>  		break;
>  
>  	case VHOST_USER_SET_VRING_ENABLE:
> -		vhost_user_set_vring_enable(dev, &msg);
> +		ret = vhost_user_set_vring_enable(dev, &msg);
>  		break;
>  	case VHOST_USER_SEND_RARP:
> -		vhost_user_send_rarp(dev, &msg);
> +		ret = vhost_user_send_rarp(dev, &msg);
>  		break;
>  
>  	case VHOST_USER_NET_SET_MTU:
> @@ -1752,7 +1759,7 @@ vhost_user_msg_handler(int vid, int fd)
>  	}
>  
>  skip_to_post_handle:
> -	if (dev->extern_ops.post_msg_handle) {
> +	if (!ret && dev->extern_ops.post_msg_handle) {
>  		uint32_t need_reply;
>  
>  		ret = (*dev->extern_ops.post_msg_handle)(
> @@ -1772,6 +1779,10 @@ vhost_user_msg_handler(int vid, int fd)
>  		msg.payload.u64 = !!ret;
>  		msg.size = sizeof(msg.payload.u64);
>  		send_vhost_reply(fd, &msg);
> +	} else if (ret) {
> +		RTE_LOG(ERR, VHOST_CONFIG,
> +			"vhost message handling failed.\n");
> +		return -1;
>  	}

The basic idea behind above code is to drop the connection
if we can't report the error to QEMU when error happens. It
looks good to me.

Thanks

>  
>  	if (!(dev->flags & VIRTIO_DEV_RUNNING) && virtio_is_ready(dev)) {
> 
> ----
> 
> Best regards, Ilya Maximets.