Message-ID: <CAMB2axNx=nCh-B-=XLtto2nEsKsV0p+b7yzXRX9OKSgUbRzzWA@mail.gmail.com>
Date: Mon, 29 Jul 2024 17:35:01 -0700
From: Amery Hung <ameryhung@...il.com>
To: Stefano Garzarella <sgarzare@...hat.com>
Cc: stefanha@...hat.com, mst@...hat.com, jasowang@...hat.com, 
	xuanzhuo@...ux.alibaba.com, davem@...emloft.net, edumazet@...gle.com, 
	kuba@...nel.org, pabeni@...hat.com, kys@...rosoft.com, haiyangz@...rosoft.com, 
	wei.liu@...nel.org, decui@...rosoft.com, bryantan@...are.com, 
	vdasa@...are.com, pv-drivers@...are.com, dan.carpenter@...aro.org, 
	simon.horman@...igine.com, oxffffaa@...il.com, kvm@...r.kernel.org, 
	virtualization@...ts.linux-foundation.org, netdev@...r.kernel.org, 
	linux-kernel@...r.kernel.org, linux-hyperv@...r.kernel.org, 
	bpf@...r.kernel.org, bobby.eshleman@...edance.com, jiang.wang@...edance.com, 
	amery.hung@...edance.com, xiyou.wangcong@...il.com
Subject: Re: [RFC PATCH net-next v6 09/14] virtio/vsock: add common datagram
 recv path

On Tue, Jul 23, 2024 at 7:42 AM Stefano Garzarella <sgarzare@...hat.com> wrote:
>
> On Wed, Jul 10, 2024 at 09:25:50PM GMT, Amery Hung wrote:
> >From: Bobby Eshleman <bobby.eshleman@...edance.com>
> >
> >This commit adds the common datagram receive functionality for virtio
> >transports. It does not add the vhost/virtio users of that
> >functionality.
> >
> >This functionality includes:
> >- changes to the virtio_transport_recv_pkt() path for finding the
> >  bound socket receiver for incoming packets
> >- virtio_transport_recv_pkt() saves the source cid and port to the
> >  control buffer for recvmsg() to initialize sockaddr_vm structure
> >  when using datagram
> >
> >Signed-off-by: Bobby Eshleman <bobby.eshleman@...edance.com>
> >Signed-off-by: Amery Hung <amery.hung@...edance.com>
> >---
> > net/vmw_vsock/virtio_transport_common.c | 79 +++++++++++++++++++++----
> > 1 file changed, 66 insertions(+), 13 deletions(-)
> >
> >diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
> >index 46cd1807f8e3..a571b575fde9 100644
> >--- a/net/vmw_vsock/virtio_transport_common.c
> >+++ b/net/vmw_vsock/virtio_transport_common.c
> >@@ -235,7 +235,9 @@ EXPORT_SYMBOL_GPL(virtio_transport_deliver_tap_pkt);
> >
> > static u16 virtio_transport_get_type(struct sock *sk)
> > {
> >-      if (sk->sk_type == SOCK_STREAM)
> >+      if (sk->sk_type == SOCK_DGRAM)
> >+              return VIRTIO_VSOCK_TYPE_DGRAM;
> >+      else if (sk->sk_type == SOCK_STREAM)
> >               return VIRTIO_VSOCK_TYPE_STREAM;
> >       else
> >               return VIRTIO_VSOCK_TYPE_SEQPACKET;
> >@@ -1422,6 +1424,33 @@ virtio_transport_recv_enqueue(struct vsock_sock *vsk,
> >               kfree_skb(skb);
> > }
> >
> >+static void
> >+virtio_transport_dgram_kfree_skb(struct sk_buff *skb, int err)
> >+{
> >+      if (err == -ENOMEM)
> >+              kfree_skb_reason(skb, SKB_DROP_REASON_SOCKET_RCVBUFF);
> >+      else if (err == -ENOBUFS)
> >+              kfree_skb_reason(skb, SKB_DROP_REASON_PROTO_MEM);
> >+      else
> >+              kfree_skb(skb);
> >+}
> >+
> >+/* This function takes ownership of the skb.
> >+ *
> >+ * It either places the skb on the sk_receive_queue or frees it.
> >+ */
> >+static void
> >+virtio_transport_recv_dgram(struct sock *sk, struct sk_buff *skb)
> >+{
> >+      int err;
> >+
> >+      err = sock_queue_rcv_skb(sk, skb);
> >+      if (err) {
> >+              virtio_transport_dgram_kfree_skb(skb, err);
> >+              return;
> >+      }
> >+}
> >+
> > static int
> > virtio_transport_recv_connected(struct sock *sk,
> >                               struct sk_buff *skb)
> >@@ -1591,7 +1620,8 @@ virtio_transport_recv_listen(struct sock *sk, struct sk_buff *skb,
> > static bool virtio_transport_valid_type(u16 type)
> > {
> >       return (type == VIRTIO_VSOCK_TYPE_STREAM) ||
> >-             (type == VIRTIO_VSOCK_TYPE_SEQPACKET);
> >+             (type == VIRTIO_VSOCK_TYPE_SEQPACKET) ||
> >+             (type == VIRTIO_VSOCK_TYPE_DGRAM);
> > }
> >
> > /* We are under the virtio-vsock's vsock->rx_lock or vhost-vsock's vq->mutex
> >@@ -1601,44 +1631,57 @@ void virtio_transport_recv_pkt(struct virtio_transport *t,
> >                              struct sk_buff *skb)
> > {
> >       struct virtio_vsock_hdr *hdr = virtio_vsock_hdr(skb);
> >+      struct vsock_skb_cb *vsock_cb;
>
> This can be defined in the block where it's used.
>

Got it.

> >       struct sockaddr_vm src, dst;
> >       struct vsock_sock *vsk;
> >       struct sock *sk;
> >       bool space_available;
> >+      u16 type;
> >
> >       vsock_addr_init(&src, le64_to_cpu(hdr->src_cid),
> >                       le32_to_cpu(hdr->src_port));
> >       vsock_addr_init(&dst, le64_to_cpu(hdr->dst_cid),
> >                       le32_to_cpu(hdr->dst_port));
> >
> >+      type = le16_to_cpu(hdr->type);
> >+
> >       trace_virtio_transport_recv_pkt(src.svm_cid, src.svm_port,
> >                                       dst.svm_cid, dst.svm_port,
> >                                       le32_to_cpu(hdr->len),
> >-                                      le16_to_cpu(hdr->type),
> >+                                      type,
> >                                       le16_to_cpu(hdr->op),
> >                                       le32_to_cpu(hdr->flags),
> >                                       le32_to_cpu(hdr->buf_alloc),
> >                                       le32_to_cpu(hdr->fwd_cnt));
> >
> >-      if (!virtio_transport_valid_type(le16_to_cpu(hdr->type))) {
> >+      if (!virtio_transport_valid_type(type)) {
> >               (void)virtio_transport_reset_no_sock(t, skb);
> >               goto free_pkt;
> >       }
> >
> >-      /* The socket must be in connected or bound table
> >-       * otherwise send reset back
> >+      /* For stream/seqpacket, the socket must be in connected or bound table
> >+       * otherwise send reset back.
> >+       *
> >+       * For datagrams, no reset is sent back.
> >        */
> >       sk = vsock_find_connected_socket(&src, &dst);
> >       if (!sk) {
> >-              sk = vsock_find_bound_socket(&dst);
> >-              if (!sk) {
> >-                      (void)virtio_transport_reset_no_sock(t, skb);
> >-                      goto free_pkt;
> >+              if (type == VIRTIO_VSOCK_TYPE_DGRAM) {
> >+                      sk = vsock_find_bound_dgram_socket(&dst);
> >+                      if (!sk)
> >+                              goto free_pkt;
> >+              } else {
> >+                      sk = vsock_find_bound_socket(&dst);
> >+                      if (!sk) {
> >+                              (void)virtio_transport_reset_no_sock(t, skb);
> >+                              goto free_pkt;
> >+                      }
> >               }
> >       }
> >
> >-      if (virtio_transport_get_type(sk) != le16_to_cpu(hdr->type)) {
> >-              (void)virtio_transport_reset_no_sock(t, skb);
> >+      if (virtio_transport_get_type(sk) != type) {
> >+              if (type != VIRTIO_VSOCK_TYPE_DGRAM)
> >+                      (void)virtio_transport_reset_no_sock(t, skb);
> >               sock_put(sk);
> >               goto free_pkt;
> >       }
> >@@ -1654,12 +1697,21 @@ void virtio_transport_recv_pkt(struct virtio_transport *t,
> >
> >       /* Check if sk has been closed before lock_sock */
> >       if (sock_flag(sk, SOCK_DONE)) {
> >-              (void)virtio_transport_reset_no_sock(t, skb);
> >+              if (type != VIRTIO_VSOCK_TYPE_DGRAM)
> >+                      (void)virtio_transport_reset_no_sock(t, skb);
> >               release_sock(sk);
> >               sock_put(sk);
> >               goto free_pkt;
> >       }
> >
> >+      if (sk->sk_type == SOCK_DGRAM) {
> >+              vsock_cb = vsock_skb_cb(skb);
> >+              vsock_cb->src_cid = src.svm_cid;
> >+              vsock_cb->src_port = src.svm_port;
> >+              virtio_transport_recv_dgram(sk, skb);
>
>
> What about adding an API that transports can use to hide this?
>
> I mean something that hide vsock_cb creation and queue packet in the
> socket receive queue. I'd also not expose vsock_skb_cb in an header, but
> I'd handle it internally in af_vsock.c. So I'd just expose API to
> queue/dequeue them.
>

Got it. I will move vsock_skb_cb to af_vsock.c and create an API:

vsock_dgram_skb_save_src_addr(struct sk_buff *skb, u32 cid, u32 port)

Different dgram implementations will call this API instead of the code
block above to save the source address information into the control
buffer.

A side note on why this is a vsock API instead of a member function in
transport: As we move to support multi-transport dgram, different
transport implementations can place skb into the sk->sk_receive_queue.
Therefore, we cannot call transport-specific function in
vsock_dgram_recvmsg() to initialize struct sockaddr_vm. Hence, the
receiving paths of different transports need to call this API to save
source address.

> Also why VMCI is using sk_receive_skb(), while we are using
> sock_queue_rcv_skb()?
>

I _think_ originally we referred to UDP and UDS when designing virtio
dgram, and ended up with placing skb into sk_receive_queue directly. I
will look into this to provide better justification.

Thank you,
Amery

> Thanks,
> Stefano
>
> >+              goto out;
> >+      }
> >+
> >       space_available = virtio_transport_space_update(sk, skb);
> >
> >       /* Update CID in case it has changed after a transport reset event */
> >@@ -1691,6 +1743,7 @@ void virtio_transport_recv_pkt(struct virtio_transport *t,
> >               break;
> >       }
> >
> >+out:
> >       release_sock(sk);
> >
> >       /* Release refcnt obtained when we fetched this socket out of the
> >--
> >2.20.1
> >
>

Powered by blists - more mailing lists