kvm.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Stefano Garzarella <sgarzare@redhat.com>
To: Arseny Krasnov <arseny.krasnov@kaspersky.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>,
	"Michael S. Tsirkin" <mst@redhat.com>,
	Jason Wang <jasowang@redhat.com>,
	"David S. Miller" <davem@davemloft.net>,
	Jakub Kicinski <kuba@kernel.org>,
	Jorgen Hansen <jhansen@vmware.com>,
	Andra Paraschiv <andraprs@amazon.com>,
	Colin Ian King <colin.king@canonical.com>,
	Norbert Slusarek <nslusarek@gmx.net>,
	Jeff Vander Stoep <jeffv@google.com>,
	Alexander Popov <alex.popov@linux.com>, kvm <kvm@vger.kernel.org>,
	Linux Virtualization <virtualization@lists.linux-foundation.org>,
	netdev <netdev@vger.kernel.org>,
	kernel list <linux-kernel@vger.kernel.org>,
	stsp <stsp2@yandex.ru>, Krasnov Arseniy <oxffffaa@gmail.com>
Subject: Re: [RFC PATCH v8 11/19] virtio/vsock: dequeue callback for SOCK_SEQPACKET
Date: Wed, 21 Apr 2021 10:56:45 +0200	[thread overview]
Message-ID: <CAGxU2F7pCfVow7B5KG4hSYjxyH2LcL3wriRvrgURTxeqzn8M9A@mail.gmail.com> (raw)
In-Reply-To: <20210413124443.3403382-1-arseny.krasnov@kaspersky.com>

On Tue, Apr 13, 2021 at 03:44:40PM +0300, Arseny Krasnov wrote:
>This adds transport callback and it's logic for SEQPACKET dequeue.
>Callback fetches RW packets from rx queue of socket until whole record
>is copied(if user's buffer is full, user is not woken up). This is done
>to not stall sender, because if we wake up user and it leaves syscall,
>nobody will send credit update for rest of record, and sender will wait
>for next enter of read syscall at receiver's side. So if user buffer is
>full, we just send credit update and drop data.
>
>Signed-off-by: Arseny Krasnov <arseny.krasnov@kaspersky.com>
>---
>v7 -> v8:
> - Things like SEQ_BEGIN, SEQ_END, 'msg_len' and 'msg_id' now removed.
>   This callback fetches and copies RW packets to user's buffer, until
>   last packet of message found(this packet is marked in 'flags' field
>   of header).
>
> include/linux/virtio_vsock.h            |  5 ++
> net/vmw_vsock/virtio_transport_common.c | 73 +++++++++++++++++++++++++
> 2 files changed, 78 insertions(+)
>
>diff --git a/include/linux/virtio_vsock.h b/include/linux/virtio_vsock.h
>index dc636b727179..02acf6e9ae04 100644
>--- a/include/linux/virtio_vsock.h
>+++ b/include/linux/virtio_vsock.h
>@@ -80,6 +80,11 @@ virtio_transport_dgram_dequeue(struct vsock_sock *vsk,
>                              struct msghdr *msg,
>                              size_t len, int flags);
>
>+ssize_t
>+virtio_transport_seqpacket_dequeue(struct vsock_sock *vsk,
>+                                 struct msghdr *msg,
>+                                 int flags,
>+                                 bool *msg_ready);
> s64 virtio_transport_stream_has_data(struct vsock_sock *vsk);
> s64 virtio_transport_stream_has_space(struct vsock_sock *vsk);
>
>diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
>index 833104b71a1c..8492b8bd5df5 100644
>--- a/net/vmw_vsock/virtio_transport_common.c
>+++ b/net/vmw_vsock/virtio_transport_common.c
>@@ -393,6 +393,67 @@ virtio_transport_stream_do_dequeue(struct vsock_sock *vsk,
>       return err;
> }
>
>+static int virtio_transport_seqpacket_do_dequeue(struct vsock_sock *vsk,
>+                                               struct msghdr *msg,
>+                                               int flags,
>+                                               bool *msg_ready)
>+{
>+      struct virtio_vsock_sock *vvs = vsk->trans;
>+      struct virtio_vsock_pkt *pkt;
>+      int err = 0;
>+      size_t user_buf_len = msg->msg_iter.count;
>+
>+      *msg_ready = false;
>+      spin_lock_bh(&vvs->rx_lock);
>+
>+      while (!*msg_ready && !list_empty(&vvs->rx_queue) && err >= 0) {
>+              pkt = list_first_entry(&vvs->rx_queue, struct virtio_vsock_pkt, list);
>+
>+              if (le16_to_cpu(pkt->hdr.op) == VIRTIO_VSOCK_OP_RW) {

Is this check still necessary, should they all be RW?

>+                      size_t bytes_to_copy;
>+                      size_t pkt_len;
>+
>+                      pkt_len = (size_t)le32_to_cpu(pkt->hdr.len);
>+                      bytes_to_copy = min(user_buf_len, pkt_len);
>+

If bytes_to_copy == 0, we can avoid the next steps (release the lock try 
to copy 0 bytes, reacquire the lock)

>+                      /* sk_lock is held by caller so no one else can dequeue.
>+                       * Unlock rx_lock since memcpy_to_msg() may sleep.
>+                       */
>+                      spin_unlock_bh(&vvs->rx_lock);
>+
>+                      if (memcpy_to_msg(msg, pkt->buf, bytes_to_copy)) {
>+                              err = -EINVAL;

Here we should reacquire the lock or prevent it from being released out
of cycle.

>+                              break;
>+                      }
>+
>+                      spin_lock_bh(&vvs->rx_lock);
>+

As mentioned before, I think we could move this part into the core and 
here always return the real dimension.

>+                      /* If user sets 'MSG_TRUNC' we return real 
>length
>+                       * of message.
>+                       */
>+                      if (flags & MSG_TRUNC)
>+                              err += pkt_len;
>+                      else
>+                              err += bytes_to_copy;
>+
>+                      user_buf_len -= bytes_to_copy;
>+
>+                      if (pkt->hdr.flags & VIRTIO_VSOCK_SEQ_EOR)
                                     ^
We should use le32_to_cpu() to read the flags.


>+                              *msg_ready = true;
>+              }
>+
>+              virtio_transport_dec_rx_pkt(vvs, pkt);
>+              list_del(&pkt->list);
>+              virtio_transport_free_pkt(pkt);
>+      }
>+
>+      spin_unlock_bh(&vvs->rx_lock);
>+
>+      virtio_transport_send_credit_update(vsk);
>+
>+      return err;
>+}
>+
> ssize_t
> virtio_transport_stream_dequeue(struct vsock_sock *vsk,
>                               struct msghdr *msg,
>@@ -405,6 +466,18 @@ virtio_transport_stream_dequeue(struct vsock_sock *vsk,
> }
> EXPORT_SYMBOL_GPL(virtio_transport_stream_dequeue);
>
>+ssize_t
>+virtio_transport_seqpacket_dequeue(struct vsock_sock *vsk,
>+                                 struct msghdr *msg,
>+                                 int flags, bool *msg_ready)
>+{
>+      if (flags & MSG_PEEK)
>+              return -EOPNOTSUPP;
>+
>+      return virtio_transport_seqpacket_do_dequeue(vsk, msg, flags,
>msg_ready);
>+}
>+EXPORT_SYMBOL_GPL(virtio_transport_seqpacket_dequeue);
>+
> int
> virtio_transport_dgram_dequeue(struct vsock_sock *vsk,
>                              struct msghdr *msg,
>--
>2.25.1
>


  parent reply	other threads:[~2021-04-21  8:56 UTC|newest]

Thread overview: 43+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-04-13 12:39 [RFC PATCH v8 00/19] virtio/vsock: introduce SOCK_SEQPACKET support Arseny Krasnov
2021-04-13 12:41 ` [RFC PATCH v8 01/19] af_vsock: update functions for connectible socket Arseny Krasnov
2021-04-13 12:42 ` [RFC PATCH v8 02/19] af_vsock: separate wait data loop Arseny Krasnov
2021-04-13 12:42 ` [RFC PATCH v8 03/19] af_vsock: separate receive " Arseny Krasnov
2021-04-13 12:42 ` [RFC PATCH v8 04/19] af_vsock: implement SEQPACKET receive loop Arseny Krasnov
2021-04-21  8:37   ` Stefano Garzarella
2021-04-21 14:47     ` Arseny Krasnov
2021-04-13 12:43 ` [RFC PATCH v8 05/19] af_vsock: implement send logic for SEQPACKET Arseny Krasnov
2021-04-13 12:43 ` [RFC PATCH v8 06/19] af_vsock: rest of SEQPACKET support Arseny Krasnov
2021-04-13 12:43 ` [RFC PATCH v8 07/19] af_vsock: update comments for stream sockets Arseny Krasnov
2021-04-13 12:43 ` [RFC PATCH v8 08/19] virtio/vsock: set packet's type in virtio_transport_send_pkt_info() Arseny Krasnov
2021-04-13 12:44 ` [RFC PATCH v8 09/19] virtio/vsock: simplify credit update function API Arseny Krasnov
2021-04-13 12:44 ` [RFC PATCH v8 10/19] virtio/vsock: defines and constants for SEQPACKET Arseny Krasnov
2021-04-13 12:44 ` [RFC PATCH v8 11/19] virtio/vsock: dequeue callback for SOCK_SEQPACKET Arseny Krasnov
2021-04-14  5:21   ` Arseny Krasnov
2021-04-21  8:56   ` Stefano Garzarella [this message]
2021-04-21 14:49     ` Arseny Krasnov
2021-04-13 12:45 ` [RFC PATCH v8 12/19] virtio/vsock: add SEQPACKET receive logic Arseny Krasnov
2021-04-13 12:45 ` [RFC PATCH v8 13/19] virtio/vsock: rest of SOCK_SEQPACKET support Arseny Krasnov
2021-04-21  9:12   ` Stefano Garzarella
2021-04-21 14:50     ` Arseny Krasnov
2021-04-13 12:45 ` [RFC PATCH v8 14/19] virtio/vsock: enable SEQPACKET for transport Arseny Krasnov
2021-04-21  9:22   ` Stefano Garzarella
2021-04-13 12:46 ` [RFC PATCH v8 15/19] vhost/vsock: " Arseny Krasnov
2021-04-21  9:31   ` Stefano Garzarella
2021-04-13 12:46 ` [RFC PATCH v8 16/19] vsock/loopback: " Arseny Krasnov
2021-04-21  9:34   ` Stefano Garzarella
2021-04-13 12:46 ` [RFC PATCH v8 17/19] vsock_test: add SOCK_SEQPACKET tests Arseny Krasnov
2021-04-21  9:35   ` Stefano Garzarella
2021-04-21 15:00     ` Arseny Krasnov
2021-04-13 12:47 ` [RFC PATCH v8 18/19] virtio/vsock: update trace event for SEQPACKET Arseny Krasnov
2021-04-13 12:47 ` [RFC PATCH v8 19/19] af_vsock: serialize writes to shared socket Arseny Krasnov
2021-04-14 10:51   ` Arseny Krasnov
2021-04-21  9:38     ` Stefano Garzarella
2021-04-21 15:02       ` Arseny Krasnov
2021-04-21  9:52 ` [RFC PATCH v8 00/19] virtio/vsock: introduce SOCK_SEQPACKET support Stefano Garzarella
2021-04-21 15:06   ` Arseny Krasnov
2021-04-22  8:46     ` Stefano Garzarella
2021-04-22  9:40       ` Arseny Krasnov
2021-04-22 10:02         ` Stefano Garzarella
2021-04-22 10:29           ` Arseny Krasnov
2021-04-22 10:48             ` Stefano Garzarella
2021-04-22 11:41               ` Arseny Krasnov

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=CAGxU2F7pCfVow7B5KG4hSYjxyH2LcL3wriRvrgURTxeqzn8M9A@mail.gmail.com \
    --to=sgarzare@redhat.com \
    --cc=alex.popov@linux.com \
    --cc=andraprs@amazon.com \
    --cc=arseny.krasnov@kaspersky.com \
    --cc=colin.king@canonical.com \
    --cc=davem@davemloft.net \
    --cc=jasowang@redhat.com \
    --cc=jeffv@google.com \
    --cc=jhansen@vmware.com \
    --cc=kuba@kernel.org \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mst@redhat.com \
    --cc=netdev@vger.kernel.org \
    --cc=nslusarek@gmx.net \
    --cc=oxffffaa@gmail.com \
    --cc=stefanha@redhat.com \
    --cc=stsp2@yandex.ru \
    --cc=virtualization@lists.linux-foundation.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).