[Qemu-devel] [PATCH v8 14/17] Add the vhost-user netdev backend to the command line

[Antonios Motakis <a.motakis@virtualopensystems.com>]

The supplied chardev id will be inspected for supported options. Only
a socket backend, with a set path (i.e. a Unix socket) and optionally
the server parameter set, will be allowed. Other options (nowait, telnet)
will make the chardev unusable and the netdev will not be initialised.

Additional checks for validity:
  - requires `-mempath ...,share=on`
  - requires `-device virtio-net-*`

The `vhostforce` option is uset to force vhost-net when we deal with no-MSIX
guest.

Signed-off-by: Antonios Motakis <a.motakis@virtualopensystems.com>
Signed-off-by: Nikolay Nikolaev <n.nikolaev@virtualopensystems.com>
---
 hmp-commands.hx     |   4 +-
 hw/net/vhost_net.c  |   4 ++
 hw/net/virtio-net.c |   3 ++
 net/hub.c           |   1 +
 net/net.c           |  25 ++++++-----
 net/vhost-user.c    | 127 +++++++++++++++++++++++++++++++++++++++++++++++++++-
 qapi-schema.json    |  19 +++++++-
 qemu-options.hx     |  17 +++++++
 8 files changed, 184 insertions(+), 16 deletions(-)

diff --git a/hmp-commands.hx b/hmp-commands.hx
index f3fc514..68128c1 100644
--- a/hmp-commands.hx
+++ b/hmp-commands.hx
@@ -1195,7 +1195,7 @@ ETEXI
     {
         .name       = "host_net_add",
         .args_type  = "device:s,opts:s?",
-        .params     = "tap|user|socket|vde|netmap|dump [options]",
+        .params     = "tap|user|socket|vde|netmap|vhost-user|dump [options]",
         .help       = "add host VLAN client",
         .mhandler.cmd = net_host_device_add,
     },
@@ -1223,7 +1223,7 @@ ETEXI
     {
         .name       = "netdev_add",
         .args_type  = "netdev:O",
-        .params     = "[user|tap|socket|hubport|netmap],id=str[,prop=value][,...]",
+        .params     = "[user|tap|socket|hubport|netmap|vhost-user],id=str[,prop=value][,...]",
         .help       = "add host network device",
         .mhandler.cmd = hmp_netdev_add,
     },
diff --git a/hw/net/vhost_net.c b/hw/net/vhost_net.c
index cc27426..7a097fc 100644
--- a/hw/net/vhost_net.c
+++ b/hw/net/vhost_net.c
@@ -15,6 +15,7 @@
 
 #include "net/net.h"
 #include "net/tap.h"
+#include "net/vhost-user.h"
 
 #include "hw/virtio/virtio-net.h"
 #include "net/vhost_net.h"
@@ -328,6 +329,9 @@ VHostNetState *get_vhost_net(NetClientState *nc)
     case NET_CLIENT_OPTIONS_KIND_TAP:
         vhost_net = tap_get_vhost_net(nc);
         break;
+    case NET_CLIENT_OPTIONS_KIND_VHOST_USER:
+        vhost_net = vhost_user_get_vhost_net(nc);
+        break;
     default:
         break;
     }
diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
index 72acd15..d49ee82 100644
--- a/hw/net/virtio-net.c
+++ b/hw/net/virtio-net.c
@@ -322,6 +322,9 @@ static void peer_test_vnet_hdr(VirtIONet *n)
     case NET_CLIENT_OPTIONS_KIND_TAP:
         n->has_vnet_hdr = tap_has_vnet_hdr(nc->peer);
         break;
+    case NET_CLIENT_OPTIONS_KIND_VHOST_USER:
+        n->has_vnet_hdr = 0;
+        break;
     default:
         break;
     }
diff --git a/net/hub.c b/net/hub.c
index 33a99c9..7e0f2d6 100644
--- a/net/hub.c
+++ b/net/hub.c
@@ -322,6 +322,7 @@ void net_hub_check_clients(void)
             case NET_CLIENT_OPTIONS_KIND_TAP:
             case NET_CLIENT_OPTIONS_KIND_SOCKET:
             case NET_CLIENT_OPTIONS_KIND_VDE:
+            case NET_CLIENT_OPTIONS_KIND_VHOST_USER:
                 has_host_dev = 1;
                 break;
             default:
diff --git a/net/net.c b/net/net.c
index 2c3af20..07bc42d 100644
--- a/net/net.c
+++ b/net/net.c
@@ -714,23 +714,24 @@ static int (* const net_client_init_fun[NET_CLIENT_OPTIONS_KIND_MAX])(
     const NetClientOptions *opts,
     const char *name,
     NetClientState *peer) = {
-        [NET_CLIENT_OPTIONS_KIND_NIC]       = net_init_nic,
+        [NET_CLIENT_OPTIONS_KIND_NIC]           = net_init_nic,
 #ifdef CONFIG_SLIRP
-        [NET_CLIENT_OPTIONS_KIND_USER]      = net_init_slirp,
+        [NET_CLIENT_OPTIONS_KIND_USER]          = net_init_slirp,
 #endif
-        [NET_CLIENT_OPTIONS_KIND_TAP]       = net_init_tap,
-        [NET_CLIENT_OPTIONS_KIND_SOCKET]    = net_init_socket,
+        [NET_CLIENT_OPTIONS_KIND_TAP]           = net_init_tap,
+        [NET_CLIENT_OPTIONS_KIND_SOCKET]        = net_init_socket,
 #ifdef CONFIG_VDE
-        [NET_CLIENT_OPTIONS_KIND_VDE]       = net_init_vde,
+        [NET_CLIENT_OPTIONS_KIND_VDE]           = net_init_vde,
 #endif
 #ifdef CONFIG_NETMAP
-        [NET_CLIENT_OPTIONS_KIND_NETMAP]    = net_init_netmap,
+        [NET_CLIENT_OPTIONS_KIND_NETMAP]        = net_init_netmap,
 #endif
-        [NET_CLIENT_OPTIONS_KIND_DUMP]      = net_init_dump,
+        [NET_CLIENT_OPTIONS_KIND_DUMP]          = net_init_dump,
 #ifdef CONFIG_NET_BRIDGE
-        [NET_CLIENT_OPTIONS_KIND_BRIDGE]    = net_init_bridge,
+        [NET_CLIENT_OPTIONS_KIND_BRIDGE]        = net_init_bridge,
 #endif
-        [NET_CLIENT_OPTIONS_KIND_HUBPORT]   = net_init_hubport,
+        [NET_CLIENT_OPTIONS_KIND_HUBPORT]       = net_init_hubport,
+        [NET_CLIENT_OPTIONS_KIND_VHOST_USER]    = net_init_vhost_user,
 };
 
 
@@ -764,6 +765,7 @@ static int net_client_init1(const void *object, int is_netdev, Error **errp)
         case NET_CLIENT_OPTIONS_KIND_BRIDGE:
 #endif
         case NET_CLIENT_OPTIONS_KIND_HUBPORT:
+        case NET_CLIENT_OPTIONS_KIND_VHOST_USER:
             break;
 
         default:
@@ -847,11 +849,12 @@ static int net_host_check_device(const char *device)
                                        , "bridge"
 #endif
 #ifdef CONFIG_SLIRP
-                                       ,"user"
+                                       , "user"
 #endif
 #ifdef CONFIG_VDE
-                                       ,"vde"
+                                       , "vde"
 #endif
+                                       , "vhost-user"
     };
     for (i = 0; i < ARRAY_SIZE(valid_param_list); i++) {
         if (!strncmp(valid_param_list[i], device,
diff --git a/net/vhost-user.c b/net/vhost-user.c
index 292be41..600f31f 100644
--- a/net/vhost-user.c
+++ b/net/vhost-user.c
@@ -12,6 +12,7 @@
 #include "net/vhost_net.h"
 #include "net/vhost-user.h"
 #include "sysemu/char.h"
+#include "qemu/config-file.h"
 #include "qemu/error-report.h"
 
 typedef struct VhostUserState {
@@ -22,9 +23,17 @@ typedef struct VhostUserState {
     unsigned long long features;
 } VhostUserState;
 
+typedef struct VhostUserChardevProps {
+    bool is_socket;
+    bool is_unix;
+    bool is_server;
+    bool has_unsupported;
+} VhostUserChardevProps;
+
 VHostNetState *vhost_user_get_vhost_net(NetClientState *nc)
 {
     VhostUserState *s = DO_UPCAST(VhostUserState, nc, nc);
+    assert(nc->info->type == NET_CLIENT_OPTIONS_KIND_VHOST_USER);
     return s->vhost_net;
 }
 
@@ -79,7 +88,7 @@ static void vhost_user_cleanup(NetClientState *nc)
 }
 
 static NetClientInfo net_vhost_user_info = {
-        .type = 0,
+        .type = NET_CLIENT_OPTIONS_KIND_VHOST_USER,
         .size = sizeof(VhostUserState),
         .cleanup = vhost_user_cleanup,
 };
@@ -143,8 +152,122 @@ static int net_vhost_user_init(NetClientState *peer, const char *device,
     return 0;
 }
 
+static int net_vhost_chardev_opts(const char *name, const char *value,
+        void *opaque)
+{
+    VhostUserChardevProps *props = opaque;
+
+    if (strcmp(name, "backend") == 0 && strcmp(value, "socket") == 0) {
+        props->is_socket = 1;
+    } else if (strcmp(name, "path") == 0) {
+        props->is_unix = 1;
+    } else if (strcmp(name, "server") == 0) {
+        props->is_server = 1;
+    } else {
+        error_report("vhost-user does not support a chardev"
+                     " with the following option:\n %s = %s",
+                     name, value);
+        props->has_unsupported = 1;
+        return -1;
+    }
+    return 0;
+}
+
+static CharDriverState *net_vhost_parse_chardev(
+        const NetdevVhostUserOptions *opts)
+{
+    CharDriverState *chr = qemu_chr_find(opts->chardev);
+    VhostUserChardevProps props;
+
+    if (chr == NULL) {
+        error_report("chardev \"%s\" not found\n", opts->chardev);
+        return 0;
+    }
+
+    /* inspect chardev opts */
+    memset(&props, 0, sizeof(props));
+    qemu_opt_foreach(chr->opts, net_vhost_chardev_opts, &props, false);
+
+    if (!props.is_socket || !props.is_unix) {
+        error_report("chardev \"%s\" is not a unix socket\n",
+                     opts->chardev);
+        return 0;
+    }
+
+    if (props.has_unsupported) {
+        error_report("chardev \"%s\" has an unsupported option\n",
+                opts->chardev);
+        return 0;
+    }
+
+    qemu_chr_fe_claim_no_fail(chr);
+
+    return chr;
+}
+
+static int net_vhost_check_net(QemuOpts *opts, void *opaque)
+{
+    const char *name = opaque;
+    const char *driver, *netdev;
+    const char virtio_name[] = "virtio-net-";
+
+    driver = qemu_opt_get(opts, "driver");
+    netdev = qemu_opt_get(opts, "netdev");
+
+    if (!driver || !netdev) {
+        return 0;
+    }
+
+    if ((strcmp(netdev, name) == 0)
+            && (strncmp(driver, virtio_name, strlen(virtio_name)) != 0)) {
+        error_report("vhost-user requires frontend driver virtio-net-*");
+        return -1;
+    }
+
+    return 0;
+}
+
 int net_init_vhost_user(const NetClientOptions *opts, const char *name,
                    NetClientState *peer)
 {
-    return net_vhost_user_init(peer, "vhost_user", 0, 0, 0);
+    const NetdevVhostUserOptions *vhost_user_opts;
+    CharDriverState *chr;
+    bool vhostforce;
+    QemuOpts *mem_opts;
+    unsigned int mem_share = 0;
+
+    assert(opts->kind == NET_CLIENT_OPTIONS_KIND_VHOST_USER);
+    vhost_user_opts = opts->vhost_user;
+
+    chr = net_vhost_parse_chardev(vhost_user_opts);
+    if (!chr) {
+        error_report("No suitable chardev found\n");
+        return -1;
+    }
+
+    /* verify mem-path is set and shared */
+    mem_opts = qemu_opts_find(qemu_find_opts("mem-path"), NULL);
+    if (mem_opts) {
+        mem_share = qemu_opt_get_bool(mem_opts, "share", 0);
+    }
+
+    if (!mem_share) {
+        error_report("vhost-user requires -mem-path /path,share=on");
+        return -1;
+    }
+
+    /* verify net frontend */
+    if (qemu_opts_foreach(qemu_find_opts("device"), net_vhost_check_net,
+                          (void *)name, true) == -1) {
+        return -1;
+    }
+
+    /* vhostforce for non-MSIX */
+    if (vhost_user_opts->has_vhostforce) {
+        vhostforce = vhost_user_opts->vhostforce;
+    } else {
+        vhostforce = false;
+    }
+
+    return net_vhost_user_init(peer, "vhost_user", name, chr, vhostforce);
 }
diff --git a/qapi-schema.json b/qapi-schema.json
index 05ced9d..568f3e7 100644
--- a/qapi-schema.json
+++ b/qapi-schema.json
@@ -3104,6 +3104,22 @@
     '*devname':    'str' } }
 
 ##
+# @NetdevVhostUserOptions
+#
+# Vhost-user network backend
+#
+# @path: control socket path
+#
+# @vhostforce: #optional vhost on for non-MSIX virtio guests
+#
+# Since 2.0
+##
+{ 'type': 'NetdevVhostUserOptions',
+  'data': {
+    'chardev':		'str',
+    '*vhostforce':	'bool' } }
+
+##
 # @NetClientOptions
 #
 # A discriminated record of network device traits.
@@ -3121,7 +3137,8 @@
     'dump':     'NetdevDumpOptions',
     'bridge':   'NetdevBridgeOptions',
     'hubport':  'NetdevHubPortOptions',
-    'netmap':   'NetdevNetmapOptions' } }
+    'netmap':   'NetdevNetmapOptions',
+    'vhost-user': 'NetdevVhostUserOptions' } }
 
 ##
 # @NetLegacy
diff --git a/qemu-options.hx b/qemu-options.hx
index 60ecc95..0c6bfdb 100644
--- a/qemu-options.hx
+++ b/qemu-options.hx
@@ -1435,6 +1435,7 @@ DEF("netdev", HAS_ARG, QEMU_OPTION_netdev,
 #ifdef CONFIG_NETMAP
     "netmap|"
 #endif
+    "vhost-user|"
     "socket|"
     "hubport],id=str[,option][,option][,...]\n", QEMU_ARCH_ALL)
 STEXI
@@ -1766,6 +1767,22 @@ The hubport netdev lets you connect a NIC to a QEMU "vlan" instead of a single
 netdev.  @code{-net} and @code{-device} with parameter @option{vlan} create the
 required hub automatically.
 
+@item -netdev vhost-user,chardev=@var{id}[,vhostforce=on|off]
+
+Establish a vhost-user netdev, backed by a chardev @var{id}. The chardev should
+be a unix domain socket backed one. The vhost-user uses a specifically defined
+protocol to pass vhost ioctl replacement messages to an application on the other
+end of the socket. On non-MSIX guests, the feature can be forced with
+@var{vhostforce}.
+
+Example:
+@example
+qemu -m 1024 -mem-path /hugetlbfs,prealloc=on,share=on \
+     -chardev socket,path=/path/to/socket \
+     -netdev type=vhost-user,id=net0,chardev=chr0 \
+     -device virtio-net-pci,netdev=net0
+@end example
+
 @item -net dump[,vlan=@var{n}][,file=@var{file}][,len=@var{len}]
 Dump network traffic on VLAN @var{n} to file @var{file} (@file{qemu-vlan0.pcap} by default).
 At most @var{len} bytes (64k by default) per packet are stored. The file format is
-- 
1.8.3.2