From: Stefan Hajnoczi <stefanha@redhat.com>
To: dev@dpdk.org
Cc: maxime.coquelin@redhat.com, Yuanhan Liu <yliu@fridaylinux.org>,
wei.w.wang@intel.com, mst@redhat.com, zhiyong.yang@intel.com,
jasowang@redhat.com, Stefan Hajnoczi <stefanha@redhat.com>
Subject: [dpdk-dev] [RFC 05/24] vhost: move start_server/client() calls to trans_af_unix.c
Date: Fri, 19 Jan 2018 13:44:25 +0000 [thread overview]
Message-ID: <20180119134444.24927-6-stefanha@redhat.com> (raw)
In-Reply-To: <20180119134444.24927-1-stefanha@redhat.com>
Introduce a vhost_transport_ops->socket_start() interface so the
transport can begin establishing vhost-user connections. This is part
of the AF_UNIX transport refactoring and removes AF_UNIX code from
vhost.h and socket.c.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
lib/librte_vhost/vhost.h | 16 +++++++++++++---
lib/librte_vhost/socket.c | 5 +----
lib/librte_vhost/trans_af_unix.c | 16 ++++++++++++++--
3 files changed, 28 insertions(+), 9 deletions(-)
diff --git a/lib/librte_vhost/vhost.h b/lib/librte_vhost/vhost.h
index 3aefe6597..7cbef04ab 100644
--- a/lib/librte_vhost/vhost.h
+++ b/lib/librte_vhost/vhost.h
@@ -236,6 +236,19 @@ struct vhost_transport_ops {
*/
void (*socket_cleanup)(struct vhost_user_socket *vsocket);
+ /**
+ * Start establishing vhost-user connections. This function is
+ * asynchronous and connections may be established after it has
+ * returned. Call vhost_user_add_connection() to register new
+ * connections.
+ *
+ * @param vsocket
+ * vhost socket
+ * @return
+ * 0 on success, -1 on failure
+ */
+ int (*socket_start)(struct vhost_user_socket *vsocket);
+
/**
* Notify the guest that used descriptors have been added to the vring.
* The VRING_AVAIL_F_NO_INTERRUPT flag has already been checked so this
@@ -346,9 +359,6 @@ struct vhost_user {
extern struct vhost_user vhost_user;
-int vhost_user_start_server(struct vhost_user_socket *vsocket);
-int vhost_user_start_client(struct vhost_user_socket *vsocket);
-
extern pthread_t reconn_tid;
int vhost_user_reconnect_init(void);
diff --git a/lib/librte_vhost/socket.c b/lib/librte_vhost/socket.c
index 78f847ccc..f8a96ab5f 100644
--- a/lib/librte_vhost/socket.c
+++ b/lib/librte_vhost/socket.c
@@ -318,8 +318,5 @@ rte_vhost_driver_start(const char *path)
"failed to create fdset handling thread");
}
- if (vsocket->is_server)
- return vhost_user_start_server(vsocket);
- else
- return vhost_user_start_client(vsocket);
+ return vsocket->trans_ops->socket_start(vsocket);
}
diff --git a/lib/librte_vhost/trans_af_unix.c b/lib/librte_vhost/trans_af_unix.c
index cc8d7ccdc..6c22093a4 100644
--- a/lib/librte_vhost/trans_af_unix.c
+++ b/lib/librte_vhost/trans_af_unix.c
@@ -51,6 +51,8 @@ struct af_unix_socket {
};
static int create_unix_socket(struct vhost_user_socket *vsocket);
+static int vhost_user_start_server(struct vhost_user_socket *vsocket);
+static int vhost_user_start_client(struct vhost_user_socket *vsocket);
static void vhost_user_read_cb(int connfd, void *dat, int *remove);
/* return bytes# of read on success or negative val on failure. */
@@ -276,7 +278,7 @@ create_unix_socket(struct vhost_user_socket *vsocket)
return 0;
}
-int
+static int
vhost_user_start_server(struct vhost_user_socket *vsocket)
{
struct af_unix_socket *s =
@@ -433,7 +435,7 @@ vhost_user_reconnect_init(void)
return ret;
}
-int
+static int
vhost_user_start_client(struct vhost_user_socket *vsocket)
{
struct af_unix_socket *s =
@@ -523,6 +525,15 @@ af_unix_socket_cleanup(struct vhost_user_socket *vsocket)
}
}
+static int
+af_unix_socket_start(struct vhost_user_socket *vsocket)
+{
+ if (vsocket->is_server)
+ return vhost_user_start_server(vsocket);
+ else
+ return vhost_user_start_client(vsocket);
+}
+
static int
af_unix_vring_call(struct virtio_net *dev __rte_unused,
struct vhost_virtqueue *vq)
@@ -536,5 +547,6 @@ const struct vhost_transport_ops af_unix_trans_ops = {
.socket_size = sizeof(struct af_unix_socket),
.socket_init = af_unix_socket_init,
.socket_cleanup = af_unix_socket_cleanup,
+ .socket_start = af_unix_socket_start,
.vring_call = af_unix_vring_call,
};
--
2.14.3
next prev parent reply other threads:[~2018-01-19 13:45 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-01-19 13:44 [dpdk-dev] [RFC 00/24] vhost: add virtio-vhost-user transport Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 01/24] vhost: move vring_call() into trans_af_unix.c Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 02/24] vhost: move AF_UNIX code from socket.c to trans_af_unix.c Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 03/24] vhost: allocate per-socket transport state Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 04/24] vhost: move socket_fd and un sockaddr into trans_af_unix.c Stefan Hajnoczi
2018-01-19 13:44 ` Stefan Hajnoczi [this message]
2018-01-19 13:44 ` [dpdk-dev] [RFC 06/24] vhost: move vhost_user_connection to trans_af_unix.c Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 07/24] vhost: move vhost_user_reconnect_init() into trans_af_unix.c Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 08/24] vhost: move vhost_user.fdset to trans_af_unix.c Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 09/24] vhost: pass vhost_transport_ops through vhost_new_device() Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 10/24] vhost: embed struct virtio_net inside struct vhost_user_connection Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 11/24] vhost: extract vhost_user.c socket I/O into transport Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 12/24] vhost: move slave_req_fd field to AF_UNIX transport Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 13/24] vhost: move mmap/munmap " Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 14/24] vhost: move librte_vhost to drivers/ Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 15/24] vhost: add virtio pci framework Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 16/24] vhost: remember a vhost_virtqueue's queue index Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 17/24] vhost: add virtio-vhost-user transport Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 18/24] vhost: add RTE_VHOST_USER_VIRTIO_TRANSPORT flag Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 19/24] net/vhost: add virtio-vhost-user support Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 20/24] examples/vhost_scsi: add --socket-file argument Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 21/24] examples/vhost_scsi: add virtio-vhost-user support Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 22/24] usertools: add virtio-vhost-user devices to dpdk-devbind.py Stefan Hajnoczi
2018-01-19 13:44 ` [dpdk-dev] [RFC 23/24] WORKAROUND revert virtio-net mq vring deletion Stefan Hajnoczi
2018-01-30 17:52 ` Maxime Coquelin
2018-01-19 13:44 ` [dpdk-dev] [RFC 24/24] WORKAROUND examples/vhost_scsi: avoid broken EVENT_IDX Stefan Hajnoczi
2018-01-19 19:31 ` Michael S. Tsirkin
2018-01-31 10:02 ` [dpdk-dev] [RFC 00/24] vhost: add virtio-vhost-user transport Maxime Coquelin
[not found] ` <20180410093847.GA22081@stefanha-x1.localdomain>
2018-04-10 14:56 ` Wang, Wei W
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=20180119134444.24927-6-stefanha@redhat.com \
--to=stefanha@redhat.com \
--cc=dev@dpdk.org \
--cc=jasowang@redhat.com \
--cc=maxime.coquelin@redhat.com \
--cc=mst@redhat.com \
--cc=wei.w.wang@intel.com \
--cc=yliu@fridaylinux.org \
--cc=zhiyong.yang@intel.com \
/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).