aboutsummaryrefslogtreecommitdiff
path: root/net.c
diff options
context:
space:
mode:
authorLaszlo Ersek <lersek@redhat.com>2012-07-17 16:17:13 +0200
committerStefan Hajnoczi <stefanha@linux.vnet.ibm.com>2012-07-23 11:55:18 +0100
commit6687b79d636cd60ed9adb1177d0d946b58fa7717 (patch)
tree26380fc6c6a92c1774bbc6b2d3620a6a97c39dbe /net.c
parent2be64a68ed05c65fc510dc450a1eb1823edf9330 (diff)
convert net_client_init() to OptsVisitor
The net_client_init() prototype is kept intact. Based on "is_netdev", the QemuOpts-rooted QemuOpt-list is parsed as a Netdev or a NetLegacy. The original meat of net_client_init() is moved to and simplified in net_client_init1(): Fields not common between -net and -netdev are clearly separated. Getting the name for the init functions is cleaner: Netdev::id is mandatory, and all init functions handle a NULL NetLegacy::name. NetLegacy::vlan explicitly depends on -net (see below). Verifying the "type=" option for -netdev can be turned into a switch. Format validation with qemu_opts_validate() can be removed because the visitor covers it. Relatedly, the "net_client_types" array is reduced to an array of init functions that can be directly indexed by opts->kind. (Help text is available in the schema JSON.) The outermost negation in the condition around qemu_find_vlan() was flattened, because it expresses the dependent code's requirements more clearly. VLAN lookup is avoided if there's no init function to pass the VLAN to. Whenever the value of type=... is needed, we substitute NetClientOptionsKind_lookup[kind]. The individual init functions are not converted yet, thus the original QemuOpts instance is passed transparently. v1->v2: - NetLegacy::name is optional. Tracked it through all init functions: they all handle a NULL name. Updated commit message accordingly. v2->v3: - NetLegacy::id is allowed and takes precedence over NetLegacy::name. Signed-off-by: Laszlo Ersek <lersek@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Diffstat (limited to 'net.c')
-rw-r--r--net.c429
1 files changed, 96 insertions, 333 deletions
diff --git a/net.c b/net.c
index c46695f6b5..af544b2479 100644
--- a/net.c
+++ b/net.c
@@ -37,6 +37,9 @@
#include "qmp-commands.h"
#include "hw/qdev.h"
#include "iov.h"
+#include "qapi-visit.h"
+#include "qapi/opts-visitor.h"
+#include "qapi/qapi-dealloc-visitor.h"
/* Net bridge is currently not supported for W32. */
#if !defined(_WIN32)
@@ -745,7 +748,8 @@ int net_handle_fd_param(Monitor *mon, const char *param)
return fd;
}
-static int net_init_nic(QemuOpts *opts, const char *name, VLANState *vlan)
+static int net_init_nic(QemuOpts *opts, const NetClientOptions *new_opts,
+ const char *name, VLANState *vlan)
{
int idx;
NICInfo *nd;
@@ -802,371 +806,130 @@ static int net_init_nic(QemuOpts *opts, const char *name, VLANState *vlan)
return idx;
}
-#define NET_COMMON_PARAMS_DESC \
- { \
- .name = "type", \
- .type = QEMU_OPT_STRING, \
- .help = "net client type (nic, tap etc.)", \
- }, { \
- .name = "vlan", \
- .type = QEMU_OPT_NUMBER, \
- .help = "vlan number", \
- }, { \
- .name = "name", \
- .type = QEMU_OPT_STRING, \
- .help = "identifier for monitor commands", \
- }
-
-typedef int (*net_client_init_func)(QemuOpts *opts,
- const char *name,
- VLANState *vlan);
-
-/* magic number, but compiler will warn if too small */
-#define NET_MAX_DESC 20
-
-static const struct {
- const char *type;
- net_client_init_func init;
- QemuOptDesc desc[NET_MAX_DESC];
-} net_client_types[NET_CLIENT_OPTIONS_KIND_MAX] = {
- [NET_CLIENT_OPTIONS_KIND_NONE] = {
- .type = "none",
- .desc = {
- NET_COMMON_PARAMS_DESC,
- { /* end of list */ }
- },
- },
- [NET_CLIENT_OPTIONS_KIND_NIC] = {
- .type = "nic",
- .init = net_init_nic,
- .desc = {
- NET_COMMON_PARAMS_DESC,
- {
- .name = "netdev",
- .type = QEMU_OPT_STRING,
- .help = "id of -netdev to connect to",
- },
- {
- .name = "macaddr",
- .type = QEMU_OPT_STRING,
- .help = "MAC address",
- }, {
- .name = "model",
- .type = QEMU_OPT_STRING,
- .help = "device model (e1000, rtl8139, virtio etc.)",
- }, {
- .name = "addr",
- .type = QEMU_OPT_STRING,
- .help = "PCI device address",
- }, {
- .name = "vectors",
- .type = QEMU_OPT_NUMBER,
- .help = "number of MSI-x vectors, 0 to disable MSI-X",
- },
- { /* end of list */ }
- },
- },
+
+static int (* const net_client_init_fun[NET_CLIENT_OPTIONS_KIND_MAX])(
+ QemuOpts *old_opts,
+ const NetClientOptions *new_opts,
+ const char *name,
+ VLANState *vlan) = {
+ [NET_CLIENT_OPTIONS_KIND_NIC] = net_init_nic,
#ifdef CONFIG_SLIRP
- [NET_CLIENT_OPTIONS_KIND_USER] = {
- .type = "user",
- .init = net_init_slirp,
- .desc = {
- NET_COMMON_PARAMS_DESC,
- {
- .name = "hostname",
- .type = QEMU_OPT_STRING,
- .help = "client hostname reported by the builtin DHCP server",
- }, {
- .name = "restrict",
- .type = QEMU_OPT_STRING,
- .help = "isolate the guest from the host (y|yes|n|no)",
- }, {
- .name = "ip",
- .type = QEMU_OPT_STRING,
- .help = "legacy parameter, use net= instead",
- }, {
- .name = "net",
- .type = QEMU_OPT_STRING,
- .help = "IP address and optional netmask",
- }, {
- .name = "host",
- .type = QEMU_OPT_STRING,
- .help = "guest-visible address of the host",
- }, {
- .name = "tftp",
- .type = QEMU_OPT_STRING,
- .help = "root directory of the built-in TFTP server",
- }, {
- .name = "bootfile",
- .type = QEMU_OPT_STRING,
- .help = "BOOTP filename, for use with tftp=",
- }, {
- .name = "dhcpstart",
- .type = QEMU_OPT_STRING,
- .help = "the first of the 16 IPs the built-in DHCP server can assign",
- }, {
- .name = "dns",
- .type = QEMU_OPT_STRING,
- .help = "guest-visible address of the virtual nameserver",
- }, {
- .name = "smb",
- .type = QEMU_OPT_STRING,
- .help = "root directory of the built-in SMB server",
- }, {
- .name = "smbserver",
- .type = QEMU_OPT_STRING,
- .help = "IP address of the built-in SMB server",
- }, {
- .name = "hostfwd",
- .type = QEMU_OPT_STRING,
- .help = "guest port number to forward incoming TCP or UDP connections",
- }, {
- .name = "guestfwd",
- .type = QEMU_OPT_STRING,
- .help = "IP address and port to forward guest TCP connections",
- },
- { /* end of list */ }
- },
- },
-#endif
- [NET_CLIENT_OPTIONS_KIND_TAP] = {
- .type = "tap",
- .init = net_init_tap,
- .desc = {
- NET_COMMON_PARAMS_DESC,
- {
- .name = "ifname",
- .type = QEMU_OPT_STRING,
- .help = "interface name",
- },
-#ifndef _WIN32
- {
- .name = "fd",
- .type = QEMU_OPT_STRING,
- .help = "file descriptor of an already opened tap",
- }, {
- .name = "script",
- .type = QEMU_OPT_STRING,
- .help = "script to initialize the interface",
- }, {
- .name = "downscript",
- .type = QEMU_OPT_STRING,
- .help = "script to shut down the interface",
- }, {
-#ifdef CONFIG_NET_BRIDGE
- .name = "helper",
- .type = QEMU_OPT_STRING,
- .help = "command to execute to configure bridge",
- }, {
+ [NET_CLIENT_OPTIONS_KIND_USER] = net_init_slirp,
#endif
- .name = "sndbuf",
- .type = QEMU_OPT_SIZE,
- .help = "send buffer limit"
- }, {
- .name = "vnet_hdr",
- .type = QEMU_OPT_BOOL,
- .help = "enable the IFF_VNET_HDR flag on the tap interface"
- }, {
- .name = "vhost",
- .type = QEMU_OPT_BOOL,
- .help = "enable vhost-net network accelerator",
- }, {
- .name = "vhostfd",
- .type = QEMU_OPT_STRING,
- .help = "file descriptor of an already opened vhost net device",
- }, {
- .name = "vhostforce",
- .type = QEMU_OPT_BOOL,
- .help = "force vhost on for non-MSIX virtio guests",
- },
-#endif /* _WIN32 */
- { /* end of list */ }
- },
- },
- [NET_CLIENT_OPTIONS_KIND_SOCKET] = {
- .type = "socket",
- .init = net_init_socket,
- .desc = {
- NET_COMMON_PARAMS_DESC,
- {
- .name = "fd",
- .type = QEMU_OPT_STRING,
- .help = "file descriptor of an already opened socket",
- }, {
- .name = "listen",
- .type = QEMU_OPT_STRING,
- .help = "port number, and optional hostname, to listen on",
- }, {
- .name = "connect",
- .type = QEMU_OPT_STRING,
- .help = "port number, and optional hostname, to connect to",
- }, {
- .name = "mcast",
- .type = QEMU_OPT_STRING,
- .help = "UDP multicast address and port number",
- }, {
- .name = "localaddr",
- .type = QEMU_OPT_STRING,
- .help = "source address and port for multicast and udp packets",
- }, {
- .name = "udp",
- .type = QEMU_OPT_STRING,
- .help = "UDP unicast address and port number",
- },
- { /* end of list */ }
- },
- },
+ [NET_CLIENT_OPTIONS_KIND_TAP] = net_init_tap,
+ [NET_CLIENT_OPTIONS_KIND_SOCKET] = net_init_socket,
#ifdef CONFIG_VDE
- [NET_CLIENT_OPTIONS_KIND_VDE] = {
- .type = "vde",
- .init = net_init_vde,
- .desc = {
- NET_COMMON_PARAMS_DESC,
- {
- .name = "sock",
- .type = QEMU_OPT_STRING,
- .help = "socket path",
- }, {
- .name = "port",
- .type = QEMU_OPT_NUMBER,
- .help = "port number",
- }, {
- .name = "group",
- .type = QEMU_OPT_STRING,
- .help = "group owner of socket",
- }, {
- .name = "mode",
- .type = QEMU_OPT_NUMBER,
- .help = "permissions for socket",
- },
- { /* end of list */ }
- },
- },
+ [NET_CLIENT_OPTIONS_KIND_VDE] = net_init_vde,
#endif
- [NET_CLIENT_OPTIONS_KIND_DUMP] = {
- .type = "dump",
- .init = net_init_dump,
- .desc = {
- NET_COMMON_PARAMS_DESC,
- {
- .name = "len",
- .type = QEMU_OPT_SIZE,
- .help = "per-packet size limit (64k default)",
- }, {
- .name = "file",
- .type = QEMU_OPT_STRING,
- .help = "dump file path (default is qemu-vlan0.pcap)",
- },
- { /* end of list */ }
- },
- },
+ [NET_CLIENT_OPTIONS_KIND_DUMP] = net_init_dump,
#ifdef CONFIG_NET_BRIDGE
- [NET_CLIENT_OPTIONS_KIND_BRIDGE] = {
- .type = "bridge",
- .init = net_init_bridge,
- .desc = {
- NET_COMMON_PARAMS_DESC,
- {
- .name = "br",
- .type = QEMU_OPT_STRING,
- .help = "bridge name",
- }, {
- .name = "helper",
- .type = QEMU_OPT_STRING,
- .help = "command to execute to configure bridge",
- },
- { /* end of list */ }
- },
- },
-#endif /* CONFIG_NET_BRIDGE */
+ [NET_CLIENT_OPTIONS_KIND_BRIDGE] = net_init_bridge,
+#endif
};
-int net_client_init(QemuOpts *opts, int is_netdev, Error **errp)
+
+static int net_client_init1(const void *object, int is_netdev,
+ QemuOpts *old_opts, Error **errp)
{
+ union {
+ const Netdev *netdev;
+ const NetLegacy *net;
+ } u;
+ const NetClientOptions *opts;
const char *name;
- const char *type;
- int i;
-
- type = qemu_opt_get(opts, "type");
- if (!type) {
- error_set(errp, QERR_MISSING_PARAMETER, "type");
- return -1;
- }
if (is_netdev) {
- if (strcmp(type, "tap") != 0 &&
-#ifdef CONFIG_NET_BRIDGE
- strcmp(type, "bridge") != 0 &&
-#endif
+ u.netdev = object;
+ opts = u.netdev->opts;
+ name = u.netdev->id;
+
+ switch (opts->kind) {
#ifdef CONFIG_SLIRP
- strcmp(type, "user") != 0 &&
+ case NET_CLIENT_OPTIONS_KIND_USER:
#endif
+ case NET_CLIENT_OPTIONS_KIND_TAP:
+ case NET_CLIENT_OPTIONS_KIND_SOCKET:
#ifdef CONFIG_VDE
- strcmp(type, "vde") != 0 &&
+ case NET_CLIENT_OPTIONS_KIND_VDE:
+#endif
+#ifdef CONFIG_NET_BRIDGE
+ case NET_CLIENT_OPTIONS_KIND_BRIDGE:
#endif
- strcmp(type, "socket") != 0) {
+ break;
+
+ default:
error_set(errp, QERR_INVALID_PARAMETER_VALUE, "type",
"a netdev backend type");
return -1;
}
+ } else {
+ u.net = object;
+ opts = u.net->opts;
+ /* missing optional values have been initialized to "all bits zero" */
+ name = u.net->has_id ? u.net->id : u.net->name;
+ }
- if (qemu_opt_get(opts, "vlan")) {
- error_set(errp, QERR_INVALID_PARAMETER, "vlan");
- return -1;
- }
- if (qemu_opt_get(opts, "name")) {
- error_set(errp, QERR_INVALID_PARAMETER, "name");
- return -1;
+ if (net_client_init_fun[opts->kind]) {
+ VLANState *vlan = NULL;
+
+ /* Do not add to a vlan if it's a -netdev or a nic with a netdev=
+ * parameter. */
+ if (!is_netdev &&
+ (opts->kind != NET_CLIENT_OPTIONS_KIND_NIC ||
+ !opts->nic->has_netdev)) {
+ vlan = qemu_find_vlan(u.net->has_vlan ? u.net->vlan : 0, true);
}
- if (!qemu_opts_id(opts)) {
- error_set(errp, QERR_MISSING_PARAMETER, "id");
+
+ if (net_client_init_fun[opts->kind](old_opts, opts, name, vlan) < 0) {
+ /* TODO push error reporting into init() methods */
+ error_set(errp, QERR_DEVICE_INIT_FAILED,
+ NetClientOptionsKind_lookup[opts->kind]);
return -1;
}
}
+ return 0;
+}
+
- name = qemu_opts_id(opts);
- if (!name) {
- name = qemu_opt_get(opts, "name");
+static void net_visit(Visitor *v, int is_netdev, void **object, Error **errp)
+{
+ if (is_netdev) {
+ visit_type_Netdev(v, (Netdev **)object, NULL, errp);
+ } else {
+ visit_type_NetLegacy(v, (NetLegacy **)object, NULL, errp);
}
+}
- for (i = 0; i < NET_CLIENT_OPTIONS_KIND_MAX; i++) {
- if (net_client_types[i].type != NULL &&
- !strcmp(net_client_types[i].type, type)) {
- Error *local_err = NULL;
- VLANState *vlan = NULL;
- int ret;
- qemu_opts_validate(opts, &net_client_types[i].desc[0], &local_err);
- if (error_is_set(&local_err)) {
- error_propagate(errp, local_err);
- return -1;
- }
+int net_client_init(QemuOpts *opts, int is_netdev, Error **errp)
+{
+ void *object = NULL;
+ Error *err = NULL;
+ int ret = -1;
- /* Do not add to a vlan if it's a -netdev or a nic with a
- * netdev= parameter. */
- if (!(is_netdev ||
- (strcmp(type, "nic") == 0 && qemu_opt_get(opts, "netdev")))) {
- vlan = qemu_find_vlan(qemu_opt_get_number(opts, "vlan", 0), 1);
- }
+ {
+ OptsVisitor *ov = opts_visitor_new(opts);
- ret = 0;
- if (net_client_types[i].init) {
- ret = net_client_types[i].init(opts, name, vlan);
- if (ret < 0) {
- /* TODO push error reporting into init() methods */
- error_set(errp, QERR_DEVICE_INIT_FAILED, type);
- return -1;
- }
- }
- return ret;
- }
+ net_visit(opts_get_visitor(ov), is_netdev, &object, &err);
+ opts_visitor_cleanup(ov);
}
- error_set(errp, QERR_INVALID_PARAMETER_VALUE, "type",
- "a network client type");
- return -1;
+ if (!err) {
+ ret = net_client_init1(object, is_netdev, opts, &err);
+ }
+
+ if (object) {
+ QapiDeallocVisitor *dv = qapi_dealloc_visitor_new();
+
+ net_visit(qapi_dealloc_get_visitor(dv), is_netdev, &object, NULL);
+ qapi_dealloc_visitor_cleanup(dv);
+ }
+
+ error_propagate(errp, err);
+ return ret;
}
+
static int net_host_check_device(const char *device)
{
int i;
@@ -1286,7 +1049,7 @@ void qmp_netdev_del(const char *id, Error **errp)
static void print_net_client(Monitor *mon, VLANClientState *vc)
{
monitor_printf(mon, "%s: type=%s,%s\n", vc->name,
- net_client_types[vc->info->type].type, vc->info_str);
+ NetClientOptionsKind_lookup[vc->info->type], vc->info_str);
}
void do_info_network(Monitor *mon)