Linux Virtualization - May 2019 - [PATCH net 3/4] vhost: vsock: add weight support

Hi:

This series try to prvernt a guest triggerable CPU hogging through
vhost kthread. This is done by introducing and checking the weight
after each requrest. The patch has been tested with reproducer of
vsock and virtio-net. Only compile test is done for vhost-scsi.

Please review.

This addresses CVE-2019-3900.

Jason Wang (4):
  vhost: introduce vhost_exceeds_weight()
  vhost_net: fix possible infinite loop
  vhost: vsock: add weight support
  vhost: scsi: add weight support

 drivers/vhost/net.c   | 41 ++++++++++++++---------------------------
 drivers/vhost/scsi.c  | 21 ++++++++++++++-------
 drivers/vhost/vhost.c | 20 +++++++++++++++++++-
 drivers/vhost/vhost.h |  5 ++++-
 drivers/vhost/vsock.c | 28 +++++++++++++++++++++-------
 5 files changed, 72 insertions(+), 43 deletions(-)

-- 
1.8.3.1

We used to have vhost_exceeds_weight() for vhost-net to:

- prevent vhost kthread from hogging the cpu
- balance the time spent between TX and RX

This function could be useful for vsock and scsi as well. So move it
to vhost.c. Device must specify a weight which counts the number of
requests, or it can also specific a byte_weight which counts the
number of bytes that has been processed.

Signed-off-by: Jason Wang <jasowang at redhat.com>
---
 drivers/vhost/net.c   | 22 ++++++----------------
 drivers/vhost/scsi.c  |  9 ++++++++-
 drivers/vhost/vhost.c | 20 +++++++++++++++++++-
 drivers/vhost/vhost.h |  5 ++++-
 drivers/vhost/vsock.c | 12 +++++++++++-
 5 files changed, 48 insertions(+), 20 deletions(-)

diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index df51a35..061a06d 100644
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@ -604,12 +604,6 @@ static size_t init_iov_iter(struct vhost_virtqueue *vq,
struct iov_iter *iter,
 	return iov_iter_count(iter);
 }
 
-static bool vhost_exceeds_weight(int pkts, int total_len)
-{
-	return total_len >= VHOST_NET_WEIGHT ||
-	       pkts >= VHOST_NET_PKT_WEIGHT;
-}
-
 static int get_tx_bufs(struct vhost_net *net,
 		       struct vhost_net_virtqueue *nvq,
 		       struct msghdr *msg,
@@ -845,10 +839,8 @@ static void handle_tx_copy(struct vhost_net *net, struct
socket *sock)
 		vq->heads[nvq->done_idx].id = cpu_to_vhost32(vq, head);
 		vq->heads[nvq->done_idx].len = 0;
 		++nvq->done_idx;
-		if (vhost_exceeds_weight(++sent_pkts, total_len)) {
-			vhost_poll_queue(&vq->poll);
+		if (vhost_exceeds_weight(vq, ++sent_pkts, total_len))
 			break;
-		}
 	}
 
 	vhost_tx_batch(net, nvq, sock, &msg);
@@ -951,10 +943,9 @@ static void handle_tx_zerocopy(struct vhost_net *net,
struct socket *sock)
 		else
 			vhost_zerocopy_signal_used(net, vq);
 		vhost_net_tx_packet(net);
-		if (unlikely(vhost_exceeds_weight(++sent_pkts, total_len))) {
-			vhost_poll_queue(&vq->poll);
+		if (unlikely(vhost_exceeds_weight(vq, ++sent_pkts,
+						  total_len)))
 			break;
-		}
 	}
 }
 
@@ -1239,10 +1230,8 @@ static void handle_rx(struct vhost_net *net)
 			vhost_log_write(vq, vq_log, log, vhost_len,
 					vq->iov, in);
 		total_len += vhost_len;
-		if (unlikely(vhost_exceeds_weight(++recv_pkts, total_len))) {
-			vhost_poll_queue(&vq->poll);
+		if (unlikely(vhost_exceeds_weight(vq, ++recv_pkts, total_len)))
 			goto out;
-		}
 	}
 	if (unlikely(busyloop_intr))
 		vhost_poll_queue(&vq->poll);
@@ -1338,7 +1327,8 @@ static int vhost_net_open(struct inode *inode, struct file
*f)
 		vhost_net_buf_init(&n->vqs[i].rxq);
 	}
 	vhost_dev_init(dev, vqs, VHOST_NET_VQ_MAX,
-		       UIO_MAXIOV + VHOST_NET_BATCH);
+		       UIO_MAXIOV + VHOST_NET_BATCH,
+		       VHOST_NET_WEIGHT, VHOST_NET_PKT_WEIGHT);
 
 	vhost_poll_init(n->poll + VHOST_NET_VQ_TX, handle_tx_net, EPOLLOUT, dev);
 	vhost_poll_init(n->poll + VHOST_NET_VQ_RX, handle_rx_net, EPOLLIN, dev);
diff --git a/drivers/vhost/scsi.c b/drivers/vhost/scsi.c
index 618fb64..d830579 100644
--- a/drivers/vhost/scsi.c
+++ b/drivers/vhost/scsi.c
@@ -57,6 +57,12 @@
 #define VHOST_SCSI_PREALLOC_UPAGES 2048
 #define VHOST_SCSI_PREALLOC_PROT_SGLS 2048
 
+/* Max number of requests before requeueing the job.
+ * Using this limit prevents one virtqueue from starving others with
+ * request.
+ */
+#define VHOST_SCSI_WEIGHT 256
+
 struct vhost_scsi_inflight {
 	/* Wait for the flush operation to finish */
 	struct completion comp;
@@ -1622,7 +1628,8 @@ static int vhost_scsi_open(struct inode *inode, struct
file *f)
 		vqs[i] = &vs->vqs[i].vq;
 		vs->vqs[i].vq.handle_kick = vhost_scsi_handle_kick;
 	}
-	vhost_dev_init(&vs->dev, vqs, VHOST_SCSI_MAX_VQ, UIO_MAXIOV);
+	vhost_dev_init(&vs->dev, vqs, VHOST_SCSI_MAX_VQ, UIO_MAXIOV,
+		       VHOST_SCSI_WEIGHT, 0);
 
 	vhost_scsi_init_inflight(vs, NULL);
 
diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c
index 351af88..290d6e5 100644
--- a/drivers/vhost/vhost.c
+++ b/drivers/vhost/vhost.c
@@ -413,8 +413,24 @@ static void vhost_dev_free_iovecs(struct vhost_dev *dev)
 		vhost_vq_free_iovecs(dev->vqs[i]);
 }
 
+bool vhost_exceeds_weight(struct vhost_virtqueue *vq,
+			  int pkts, int total_len)
+{
+	struct vhost_dev *dev = vq->dev;
+
+	if ((dev->byte_weight && total_len >= dev->byte_weight) ||
+	    pkts >= dev->weight) {
+		vhost_poll_queue(&vq->poll);
+		return true;
+	}
+
+	return false;
+}
+EXPORT_SYMBOL_GPL(vhost_exceeds_weight);
+
 void vhost_dev_init(struct vhost_dev *dev,
-		    struct vhost_virtqueue **vqs, int nvqs, int iov_limit)
+		    struct vhost_virtqueue **vqs, int nvqs,
+		    int iov_limit, int weight, int byte_weight)
 {
 	struct vhost_virtqueue *vq;
 	int i;
@@ -428,6 +444,8 @@ void vhost_dev_init(struct vhost_dev *dev,
 	dev->mm = NULL;
 	dev->worker = NULL;
 	dev->iov_limit = iov_limit;
+	dev->weight = weight;
+	dev->byte_weight = byte_weight;
 	init_llist_head(&dev->work_list);
 	init_waitqueue_head(&dev->wait);
 	INIT_LIST_HEAD(&dev->read_list);
diff --git a/drivers/vhost/vhost.h b/drivers/vhost/vhost.h
index 9490e7d..27a78a9 100644
--- a/drivers/vhost/vhost.h
+++ b/drivers/vhost/vhost.h
@@ -171,10 +171,13 @@ struct vhost_dev {
 	struct list_head pending_list;
 	wait_queue_head_t wait;
 	int iov_limit;
+	int weight;
+	int byte_weight;
 };
 
+bool vhost_exceeds_weight(struct vhost_virtqueue *vq, int pkts, int total_len);
 void vhost_dev_init(struct vhost_dev *, struct vhost_virtqueue **vqs,
-		    int nvqs, int iov_limit);
+		    int nvqs, int iov_limit, int weight, int byte_weight);
 long vhost_dev_set_owner(struct vhost_dev *dev);
 bool vhost_dev_has_owner(struct vhost_dev *dev);
 long vhost_dev_check_owner(struct vhost_dev *);
diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c
index bb5fc0e..47c6d4d 100644
--- a/drivers/vhost/vsock.c
+++ b/drivers/vhost/vsock.c
@@ -21,6 +21,14 @@
 #include "vhost.h"
 
 #define VHOST_VSOCK_DEFAULT_HOST_CID	2
+/* Max number of bytes transferred before requeueing the job.
+ * Using this limit prevents one virtqueue from starving others. */
+#define VHOST_VSOCK_WEIGHT 0x80000
+/* Max number of packets transferred before requeueing the job.
+ * Using this limit prevents one virtqueue from starving others with
+ * small pkts.
+ */
+#define VHOST_VSOCK_PKT_WEIGHT 256
 
 enum {
 	VHOST_VSOCK_FEATURES = VHOST_FEATURES,
@@ -531,7 +539,9 @@ static int vhost_vsock_dev_open(struct inode *inode, struct
file *file)
 	vsock->vqs[VSOCK_VQ_TX].handle_kick = vhost_vsock_handle_tx_kick;
 	vsock->vqs[VSOCK_VQ_RX].handle_kick = vhost_vsock_handle_rx_kick;
 
-	vhost_dev_init(&vsock->dev, vqs, ARRAY_SIZE(vsock->vqs),
UIO_MAXIOV);
+	vhost_dev_init(&vsock->dev, vqs, ARRAY_SIZE(vsock->vqs),
+		       UIO_MAXIOV, VHOST_VSOCK_PKT_WEIGHT,
+		       VHOST_VSOCK_WEIGHT);
 
 	file->private_data = vsock;
 	spin_lock_init(&vsock->send_pkt_list_lock);
-- 
1.8.3.1

When the rx buffer is too small for a packet, we will discard the vq
descriptor and retry it for the next packet:

while ((sock_len = vhost_net_rx_peek_head_len(net, sock->sk,
					      &busyloop_intr))) {
...
	/* On overrun, truncate and discard */
	if (unlikely(headcount > UIO_MAXIOV)) {
		iov_iter_init(&msg.msg_iter, READ, vq->iov, 1, 1);
		err = sock->ops->recvmsg(sock, &msg,
					 1, MSG_DONTWAIT | MSG_TRUNC);
		pr_debug("Discarded rx packet: len %zd\n", sock_len);
		continue;
	}
...
}

This makes it possible to trigger a infinite while..continue loop
through the co-opreation of two VMs like:

1) Malicious VM1 allocate 1 byte rx buffer and try to slow down the
   vhost process as much as possible e.g using indirect descriptors or
   other.
2) Malicious VM2 generate packets to VM1 as fast as possible

Fixing this by checking against weight at the end of RX and TX
loop. This also eliminate other similar cases when:

- userspace is consuming the packets in the meanwhile
- theoretical TOCTOU attack if guest moving avail index back and forth
  to hit the continue after vhost find guest just add new buffers

This addresses CVE-2019-3900.

Fixes: d8316f3991d20 ("vhost: fix total length when packets are too
short")
Fixes: 3a4d5c94e9593 ("vhost_net: a kernel-level virtio server")
Signed-off-by: Jason Wang <jasowang at redhat.com>
---
The patch is needed for stable.
---
 drivers/vhost/net.c | 29 +++++++++++++----------------
 1 file changed, 13 insertions(+), 16 deletions(-)

diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index 061a06d..2d9df78 100644
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@ -773,7 +773,7 @@ static void handle_tx_copy(struct vhost_net *net, struct
socket *sock)
 	int sent_pkts = 0;
 	bool sock_can_batch = (sock->sk->sk_sndbuf == INT_MAX);
 
-	for (;;) {
+	do {
 		bool busyloop_intr = false;
 
 		if (nvq->done_idx == VHOST_NET_BATCH)
@@ -839,9 +839,7 @@ static void handle_tx_copy(struct vhost_net *net, struct
socket *sock)
 		vq->heads[nvq->done_idx].id = cpu_to_vhost32(vq, head);
 		vq->heads[nvq->done_idx].len = 0;
 		++nvq->done_idx;
-		if (vhost_exceeds_weight(vq, ++sent_pkts, total_len))
-			break;
-	}
+	} while (likely(!vhost_exceeds_weight(vq, ++sent_pkts, total_len)));
 
 	vhost_tx_batch(net, nvq, sock, &msg);
 }
@@ -866,7 +864,7 @@ static void handle_tx_zerocopy(struct vhost_net *net, struct
socket *sock)
 	bool zcopy_used;
 	int sent_pkts = 0;
 
-	for (;;) {
+	do {
 		bool busyloop_intr;
 
 		/* Release DMAs done buffers first */
@@ -943,10 +941,7 @@ static void handle_tx_zerocopy(struct vhost_net *net,
struct socket *sock)
 		else
 			vhost_zerocopy_signal_used(net, vq);
 		vhost_net_tx_packet(net);
-		if (unlikely(vhost_exceeds_weight(vq, ++sent_pkts,
-						  total_len)))
-			break;
-	}
+	} while (likely(!vhost_exceeds_weight(vq, ++sent_pkts, total_len)));
 }
 
 /* Expects to be always run from workqueue - which acts as
@@ -1144,8 +1139,11 @@ static void handle_rx(struct vhost_net *net)
 		vq->log : NULL;
 	mergeable = vhost_has_feature(vq, VIRTIO_NET_F_MRG_RXBUF);
 
-	while ((sock_len = vhost_net_rx_peek_head_len(net, sock->sk,
-						      &busyloop_intr))) {
+	do {
+		sock_len = vhost_net_rx_peek_head_len(net, sock->sk,
+						      &busyloop_intr);
+		if (!sock_len)
+			break;
 		sock_len += sock_hlen;
 		vhost_len = sock_len + vhost_hlen;
 		headcount = get_rx_bufs(vq, vq->heads + nvq->done_idx,
@@ -1230,12 +1228,11 @@ static void handle_rx(struct vhost_net *net)
 			vhost_log_write(vq, vq_log, log, vhost_len,
 					vq->iov, in);
 		total_len += vhost_len;
-		if (unlikely(vhost_exceeds_weight(vq, ++recv_pkts, total_len)))
-			goto out;
-	}
+	} while (likely(!vhost_exceeds_weight(vq, ++recv_pkts, total_len)));
+
 	if (unlikely(busyloop_intr))
 		vhost_poll_queue(&vq->poll);
-	else
+	else if (!sock_len)
 		vhost_net_enable_vq(net, vq);
 out:
 	vhost_net_signal_used(nvq);
@@ -1328,7 +1325,7 @@ static int vhost_net_open(struct inode *inode, struct file
*f)
 	}
 	vhost_dev_init(dev, vqs, VHOST_NET_VQ_MAX,
 		       UIO_MAXIOV + VHOST_NET_BATCH,
-		       VHOST_NET_WEIGHT, VHOST_NET_PKT_WEIGHT);
+		       VHOST_NET_PKT_WEIGHT, VHOST_NET_WEIGHT);
 
 	vhost_poll_init(n->poll + VHOST_NET_VQ_TX, handle_tx_net, EPOLLOUT, dev);
 	vhost_poll_init(n->poll + VHOST_NET_VQ_RX, handle_rx_net, EPOLLIN, dev);
-- 
1.8.3.1

This patch will check the weight and exit the loop if we exceeds the
weight. This is useful for preventing vsock kthread from hogging cpu
which is guest triggerable. The weight can help to avoid starving the
request from on direction while another direction is being processed.

The value of weight is picked from vhost-net.

This addresses CVE-2019-3900.

Cc: Stefan Hajnoczi <stefanha at redhat.com>
Fixes: 433fc58e6bf2 ("VSOCK: Introduce vhost_vsock.ko")
Signed-off-by: Jason Wang <jasowang at redhat.com>
---
 drivers/vhost/vsock.c | 16 ++++++++++------
 1 file changed, 10 insertions(+), 6 deletions(-)

diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c
index 47c6d4d..1fa9deb 100644
--- a/drivers/vhost/vsock.c
+++ b/drivers/vhost/vsock.c
@@ -86,6 +86,7 @@ static struct vhost_vsock *vhost_vsock_get(u32 guest_cid)
 			    struct vhost_virtqueue *vq)
 {
 	struct vhost_virtqueue *tx_vq = &vsock->vqs[VSOCK_VQ_TX];
+	int pkts = 0, total_len = 0;
 	bool added = false;
 	bool restart_tx = false;
 
@@ -97,7 +98,7 @@ static struct vhost_vsock *vhost_vsock_get(u32 guest_cid)
 	/* Avoid further vmexits, we're already processing the virtqueue */
 	vhost_disable_notify(&vsock->dev, vq);
 
-	for (;;) {
+	do {
 		struct virtio_vsock_pkt *pkt;
 		struct iov_iter iov_iter;
 		unsigned out, in;
@@ -183,7 +184,8 @@ static struct vhost_vsock *vhost_vsock_get(u32 guest_cid)
 		virtio_transport_deliver_tap_pkt(pkt);
 
 		virtio_transport_free_pkt(pkt);
-	}
+		total_len += pkt->len;
+	} while(likely(!vhost_exceeds_weight(vq, ++pkts, total_len)));
 	if (added)
 		vhost_signal(&vsock->dev, vq);
 
@@ -358,7 +360,7 @@ static void vhost_vsock_handle_tx_kick(struct vhost_work
*work)
 	struct vhost_vsock *vsock = container_of(vq->dev, struct vhost_vsock,
 						 dev);
 	struct virtio_vsock_pkt *pkt;
-	int head;
+	int head, pkts = 0, total_len = 0;
 	unsigned int out, in;
 	bool added = false;
 
@@ -368,7 +370,7 @@ static void vhost_vsock_handle_tx_kick(struct vhost_work
*work)
 		goto out;
 
 	vhost_disable_notify(&vsock->dev, vq);
-	for (;;) {
+	do {
 		u32 len;
 
 		if (!vhost_vsock_more_replies(vsock)) {
@@ -409,9 +411,11 @@ static void vhost_vsock_handle_tx_kick(struct vhost_work
*work)
 		else
 			virtio_transport_free_pkt(pkt);
 
-		vhost_add_used(vq, head, sizeof(pkt->hdr) + len);
+		len += sizeof(pkt->hdr);
+		vhost_add_used(vq, head, len);
+		total_len += len;
 		added = true;
-	}
+	} while(likely(!vhost_exceeds_weight(vq, ++pkts, total_len)));
 
 no_more_replies:
 	if (added)
-- 
1.8.3.1

This patch will check the weight and exit the loop if we exceeds the
weight. This is useful for preventing scsi kthread from hogging cpu
which is guest triggerable.

This addresses CVE-2019-3900.

Cc: Paolo Bonzini <pbonzini at redhat.com>
Cc: Stefan Hajnoczi <stefanha at redhat.com>
Fixes: 057cbf49a1f0 ("tcm_vhost: Initial merge for vhost level target
fabric driver")
Signed-off-by: Jason Wang <jasowang at redhat.com>
---
 drivers/vhost/scsi.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/vhost/scsi.c b/drivers/vhost/scsi.c
index d830579..3a59f47 100644
--- a/drivers/vhost/scsi.c
+++ b/drivers/vhost/scsi.c
@@ -918,7 +918,7 @@ static void vhost_scsi_submission_work(struct work_struct
*work)
 	struct iov_iter in_iter, prot_iter, data_iter;
 	u64 tag;
 	u32 exp_data_len, data_direction;
-	int ret, prot_bytes;
+	int ret, prot_bytes, c = 0;
 	u16 lun;
 	u8 task_attr;
 	bool t10_pi = vhost_has_feature(vq, VIRTIO_SCSI_F_T10_PI);
@@ -938,7 +938,7 @@ static void vhost_scsi_submission_work(struct work_struct
*work)
 
 	vhost_disable_notify(&vs->dev, vq);
 
-	for (;;) {
+	do {
 		ret = vhost_scsi_get_desc(vs, vq, &vc);
 		if (ret)
 			goto err;
@@ -1118,7 +1118,7 @@ static void vhost_scsi_submission_work(struct work_struct
*work)
 			break;
 		else if (ret == -EIO)
 			vhost_scsi_send_bad_target(vs, vq, vc.head, vc.out);
-	}
+	} while (likely(!vhost_exceeds_weight(vq, ++c, 0)));
 out:
 	mutex_unlock(&vq->mutex);
 }
@@ -1177,7 +1177,7 @@ static void vhost_scsi_submission_work(struct work_struct
*work)
 	} v_req;
 	struct vhost_scsi_ctx vc;
 	size_t typ_size;
-	int ret;
+	int ret, c = 0;
 
 	mutex_lock(&vq->mutex);
 	/*
@@ -1191,7 +1191,7 @@ static void vhost_scsi_submission_work(struct work_struct
*work)
 
 	vhost_disable_notify(&vs->dev, vq);
 
-	for (;;) {
+	do {
 		ret = vhost_scsi_get_desc(vs, vq, &vc);
 		if (ret)
 			goto err;
@@ -1270,7 +1270,7 @@ static void vhost_scsi_submission_work(struct work_struct
*work)
 			break;
 		else if (ret == -EIO)
 			vhost_scsi_send_bad_target(vs, vq, vc.head, vc.out);
-	}
+	} while (likely(!vhost_exceeds_weight(vq, ++c, 0)));
 out:
 	mutex_unlock(&vq->mutex);
 }
-- 
1.8.3.1

On Thu, May 16, 2019 at 03:47:41AM -0400, Jason Wang
wrote:
> @@ -183,7 +184,8 @@ static struct vhost_vsock *vhost_vsock_get(u32
guest_cid)
>  		virtio_transport_deliver_tap_pkt(pkt);
>  
>  		virtio_transport_free_pkt(pkt);
> -	}
> +		total_len += pkt->len;
Please increment total_len before virtio_transport_free_pkt(pkt) to
avoid use-after-free.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 488 bytes
Desc: not available
URL:
<http://lists.linuxfoundation.org/pipermail/virtualization/attachments/20190516/06ae5670/attachment.sig>

On Thu, May 16, 2019 at 03:47:38AM -0400, Jason Wang
wrote:
> Hi:
> 
> This series try to prvernt a guest triggerable CPU hogging through
> vhost kthread. This is done by introducing and checking the weight
> after each requrest. The patch has been tested with reproducer of
> vsock and virtio-net. Only compile test is done for vhost-scsi.
> 
> Please review.
> 
> This addresses CVE-2019-3900.
> 
> Jason Wang (4):
>   vhost: introduce vhost_exceeds_weight()
>   vhost_net: fix possible infinite loop
>   vhost: vsock: add weight support
>   vhost: scsi: add weight support
> 
>  drivers/vhost/net.c   | 41 ++++++++++++++---------------------------
>  drivers/vhost/scsi.c  | 21 ++++++++++++++-------
>  drivers/vhost/vhost.c | 20 +++++++++++++++++++-
>  drivers/vhost/vhost.h |  5 ++++-
>  drivers/vhost/vsock.c | 28 +++++++++++++++++++++-------
>  5 files changed, 72 insertions(+), 43 deletions(-)
> 
> -- 
> 1.8.3.1
> 
Looks good aside from the use-after-free in the vsock patch.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 488 bytes
Desc: not available
URL:
<http://lists.linuxfoundation.org/pipermail/virtualization/attachments/20190516/0156fb13/attachment.sig>