3
0
mirror of https://git.kernel.org/pub/scm/network/wireless/iwd.git synced 2024-11-17 17:39:28 +01:00
iwd/src/manager.c

767 lines
18 KiB
C
Raw Normal View History

/*
*
* Wireless daemon for Linux
*
* Copyright (C) 2019 Intel Corporation. All rights reserved.
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with this library; if not, write to the Free Software
* Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
*
*/
#ifdef HAVE_CONFIG_H
#include <config.h>
#endif
#include <stdlib.h>
#include <stdio.h>
#include <errno.h>
#include <linux/if_ether.h>
#include <fnmatch.h>
#include <ell/ell.h>
#include "linux/nl80211.h"
#include "src/iwd.h"
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
#include "src/netdev.h"
#include "src/wiphy.h"
#include "src/util.h"
#include "src/common.h"
#include "src/nl80211cmd.h"
static struct l_genl_family *nl80211 = NULL;
static char **whitelist_filter;
static char **blacklist_filter;
static bool randomize;
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
struct wiphy_setup_state {
uint32_t id;
struct wiphy *wiphy;
struct l_timeout *setup_timeout;
unsigned int pending_cmd_count;
bool aborted;
/*
* Data we may need if the driver does not seem to support interface
* manipulation and we fall back to using the driver-created default
* interface.
*/
bool use_default;
struct l_genl_msg *default_if_msg;
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
};
static struct l_queue *pending_wiphys;
/* With these drivers don't even try creating our interfaces */
static const char *default_if_driver_list[] = {
/*
* The out-of-tree rtl88x2bu crashes the kernel hard. Seemingly
* many other drivers are built from the same source code so
* blacklist all of them. Unfortunately there are in-tree drivers
* that also match these names and may be fine. Use
* use_default_interface to override.
*/
"rtl81*",
"rtl87*",
"rtl88*",
"rtw_*",
NULL,
};
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
static void wiphy_setup_state_free(void *data)
{
struct wiphy_setup_state *state = data;
if (state->setup_timeout)
l_timeout_remove(state->setup_timeout);
if (state->default_if_msg)
l_genl_msg_unref(state->default_if_msg);
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
L_WARN_ON(state->pending_cmd_count);
l_free(state);
}
static void wiphy_setup_state_destroy(struct wiphy_setup_state *state)
{
l_queue_remove(pending_wiphys, state);
wiphy_setup_state_free(state);
}
static bool manager_use_default(struct wiphy_setup_state *state)
{
l_debug("");
if (!state->default_if_msg) {
l_error("No default interface for wiphy %u",
(unsigned int) state->id);
return false;
}
netdev_create_from_genl(state->default_if_msg, randomize);
return true;
}
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
static void manager_new_interface_cb(struct l_genl_msg *msg, void *user_data)
{
struct wiphy_setup_state *state = user_data;
l_debug("");
if (state->aborted)
return;
if (l_genl_msg_get_error(msg) < 0) {
l_error("NEW_INTERFACE failed: %s",
strerror(-l_genl_msg_get_error(msg)));
/*
* Nothing we can do to use this wiphy since by now we
* will have successfully deleted any default interface
* there may have been.
*/
return;
}
netdev_create_from_genl(msg, randomize &&
!wiphy_has_feature(state->wiphy,
NL80211_FEATURE_MAC_ON_CREATE));
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
}
static void manager_new_interface_done(void *user_data)
{
struct wiphy_setup_state *state = user_data;
state->pending_cmd_count--;
wiphy_setup_state_destroy(state);
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
}
static void manager_create_interfaces(struct wiphy_setup_state *state)
{
struct l_genl_msg *msg;
char ifname[10];
uint32_t iftype = NL80211_IFTYPE_STATION;
unsigned cmd_id;
if (state->aborted)
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
return;
if (state->use_default) {
manager_use_default(state);
return;
}
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
/*
* Current policy: we maintain one netdev per wiphy for station,
* AP and Ad-Hoc modes, one optional p2p-device and zero or more
* p2p-GOs or p2p-clients. The P2P-related interfaces will be
* created on request.
*/
/* To be improved */
snprintf(ifname, sizeof(ifname), "wlan%i", (int) state->id);
l_debug("creating %s", ifname);
msg = l_genl_msg_new(NL80211_CMD_NEW_INTERFACE);
l_genl_msg_append_attr(msg, NL80211_ATTR_WIPHY, 4, &state->id);
l_genl_msg_append_attr(msg, NL80211_ATTR_IFTYPE, 4, &iftype);
l_genl_msg_append_attr(msg, NL80211_ATTR_IFNAME,
strlen(ifname) + 1, ifname);
l_genl_msg_append_attr(msg, NL80211_ATTR_4ADDR, 1, "\0");
l_genl_msg_append_attr(msg, NL80211_ATTR_SOCKET_OWNER, 0, "");
if (randomize && wiphy_has_feature(state->wiphy,
NL80211_FEATURE_MAC_ON_CREATE)) {
uint8_t random_addr[6];
wiphy_generate_random_address(state->wiphy, random_addr);
l_debug("Creating interface on phy: %s with random addr: "MAC,
wiphy_get_name(state->wiphy),
MAC_STR(random_addr));
l_genl_msg_append_attr(msg, NL80211_ATTR_MAC, 6, random_addr);
}
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
cmd_id = l_genl_family_send(nl80211, msg,
manager_new_interface_cb, state,
manager_new_interface_done);
if (!cmd_id) {
l_error("Error sending NEW_INTERFACE for %s", ifname);
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
return;
}
state->pending_cmd_count++;
}
static void manager_setup_cmd_done(void *user_data)
{
struct wiphy_setup_state *state = user_data;
state->pending_cmd_count--;
if (state->pending_cmd_count)
return;
manager_create_interfaces(state);
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
if (!state->pending_cmd_count)
wiphy_setup_state_destroy(state);
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
}
static void manager_del_interface_cb(struct l_genl_msg *msg, void *user_data)
{
struct wiphy_setup_state *state = user_data;
l_debug("");
if (state->aborted)
return;
if (l_genl_msg_get_error(msg) < 0) {
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
l_error("DEL_INTERFACE failed: %s",
strerror(-l_genl_msg_get_error(msg)));
state->use_default = true;
}
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
}
static void manager_get_interface_cb(struct l_genl_msg *msg, void *user_data)
{
struct wiphy_setup_state *state = user_data;
struct l_genl_attr attr;
uint16_t type, len;
const void *data;
const uint32_t *ifindex = NULL, *iftype = NULL;
const uint64_t *wdev_idx = NULL;
const char *ifname = NULL;
struct l_genl_msg *del_msg;
unsigned cmd_id;
char *pattern;
unsigned int i;
bool whitelisted = false, blacklisted = false;
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
l_debug("");
if (state->aborted)
return;
if (!l_genl_attr_init(&attr, msg))
return;
while (l_genl_attr_next(&attr, &type, &len, &data)) {
switch (type) {
case NL80211_ATTR_IFINDEX:
if (len != sizeof(uint32_t)) {
l_warn("Invalid interface index attribute");
return;
}
ifindex = data;
break;
case NL80211_ATTR_WDEV:
if (len != sizeof(uint64_t)) {
l_warn("Invalid wdev index attribute");
return;
}
wdev_idx = data;
break;
case NL80211_ATTR_WIPHY:
if (len != sizeof(uint32_t) ||
*((uint32_t *) data) != state->id) {
l_warn("Invalid wiphy attribute");
return;
}
break;
case NL80211_ATTR_IFTYPE:
if (len != sizeof(uint32_t)) {
l_warn("Invalid interface type attribute");
return;
}
iftype = data;
break;
case NL80211_ATTR_IFNAME:
if (len < 1 || !memchr(data + 1, 0, len - 1)) {
l_warn("Invalid interface name attribute");
return;
}
ifname = data;
break;
}
}
if (!wdev_idx || !iftype)
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
return;
if (ifindex) {
if (!ifname)
return;
if (whitelist_filter) {
for (i = 0; (pattern = whitelist_filter[i]); i++) {
if (fnmatch(pattern, ifname, 0) != 0)
continue;
whitelisted = true;
break;
}
}
if (blacklist_filter) {
for (i = 0; (pattern = blacklist_filter[i]); i++) {
if (fnmatch(pattern, ifname, 0) != 0)
continue;
blacklisted = true;
break;
}
}
}
/*
* If this interface is usable as our default netdev in case the
* driver does not support interface manipulation, save the message
* just in case.
*/
if ((*iftype == NL80211_IFTYPE_ADHOC ||
*iftype == NL80211_IFTYPE_STATION ||
*iftype == NL80211_IFTYPE_AP) &&
ifindex && *ifindex != 0 &&
!state->default_if_msg &&
(!whitelist_filter || whitelisted) &&
!blacklisted)
state->default_if_msg = l_genl_msg_ref(msg);
if (state->use_default)
return;
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
del_msg = l_genl_msg_new(NL80211_CMD_DEL_INTERFACE);
if (ifindex)
l_genl_msg_append_attr(del_msg, NL80211_ATTR_IFINDEX, 4, ifindex);
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
l_genl_msg_append_attr(del_msg, NL80211_ATTR_WDEV, 8, wdev_idx);
l_genl_msg_append_attr(del_msg, NL80211_ATTR_WIPHY, 4, &state->id);
cmd_id = l_genl_family_send(nl80211, del_msg,
manager_del_interface_cb, state,
manager_setup_cmd_done);
if (!cmd_id) {
l_error("Sending DEL_INTERFACE for %s failed",
ifname ?: "unnamed interface");
state->use_default = true;
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
return;
}
l_debug("");
state->pending_cmd_count++;
}
static void manager_wiphy_dump_interfaces(struct wiphy_setup_state *state)
{
struct l_genl_msg *msg;
unsigned cmd_id;
if (state->setup_timeout) {
l_timeout_remove(state->setup_timeout);
state->setup_timeout = NULL;
}
/*
* As the first step after new wiphy is detected we will query
* the initial interface setup, delete the default interfaces
* and create interfaces for our own use with NL80211_ATTR_SOCKET_OWNER
* on them. After that if new interfaces are created outside of
* IWD, or removed outside of IWD, we don't touch them and will
* try to minimally adapt to handle the removals correctly. It's
* a very unlikely situation in any case but it wouldn't make
* sense to try to continually enforce our setup fighting against
* some other process, and it wouldn't make sense to try to
* manage and use additional interfaces beyond the one or two
* we need for our operations.
*/
msg = l_genl_msg_new(NL80211_CMD_GET_INTERFACE);
l_genl_msg_append_attr(msg, NL80211_ATTR_WIPHY, 4, &state->id);
cmd_id = l_genl_family_dump(nl80211, msg,
manager_get_interface_cb, state,
manager_setup_cmd_done);
if (!cmd_id) {
l_error("Querying interface information for wiphy %u failed",
(unsigned int) state->id);
wiphy_setup_state_destroy(state);
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
return;
}
l_debug("");
state->pending_cmd_count++;
}
static struct wiphy_setup_state *manager_rx_cmd_new_wiphy(
struct l_genl_msg *msg)
{
struct wiphy_setup_state *state = NULL;
struct wiphy *wiphy;
struct l_genl_attr attr;
uint32_t id;
const char *name;
const char *driver, **driver_bad;
bool use_default;
const struct l_settings *settings = iwd_get_config();
if (!l_genl_attr_init(&attr, msg))
return NULL;
if (!wiphy_parse_id_and_name(&attr, &id, &name))
return NULL;
/*
* A Wiphy split dump can generate many (6+) NEW_WIPHY messages
* We need to parse attributes from all of them, but only perform
* initialization steps once for each new wiphy detected
*/
wiphy = wiphy_find(id);
if (wiphy)
goto done;
wiphy = wiphy_create(id, name);
if (!wiphy)
return NULL;
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
/*
* We've got a new wiphy, flag it as new and wait for a
* NEW_INTERFACE event for this wiphy's default driver-created
* interface. That event's handler will check the flag and
* finish setting up the interfaces for this new wiphy and then
* clear the flag. In some corner cases there may be no
* default interface on this wiphy and no user-space created
* interfaces from before IWD started, so set a 1-second timeout
* for the event. The timeout pointer is also used as the flag.
*/
state = l_new(struct wiphy_setup_state, 1);
state->id = id;
state->wiphy = wiphy;
l_queue_push_tail(pending_wiphys, state);
driver = wiphy_get_driver(wiphy);
for (driver_bad = default_if_driver_list; *driver_bad; driver_bad++)
if (fnmatch(*driver_bad, driver, 0) == 0)
state->use_default = true;
/*
* If whitelist/blacklist were given only try to use existing
* interfaces same as when the driver does not support NEW_INTERFACE
* or DEL_INTERFACE, otherwise the interface names will become
* meaningless after we've created our own interface(s). Optimally
* phy name white/blacklists should be used.
*/
if (whitelist_filter || blacklist_filter)
state->use_default = true;
/* The setting overrides our attempts to ensure things work */
if (l_settings_get_bool(settings, "General",
"use_default_interface", &use_default))
state->use_default = use_default;
if (state->use_default)
l_info("Wiphy %s will only use the default interface", name);
done:
wiphy_update_from_genl(wiphy, msg);
return state;
}
static void manager_wiphy_setup_timeout(struct l_timeout *timeout,
void *user_data)
{
struct wiphy_setup_state *state = user_data;
manager_wiphy_dump_interfaces(state);
}
static void manager_new_wiphy_event(struct l_genl_msg *msg)
{
struct wiphy_setup_state *state;
if (!pending_wiphys)
return;
state = manager_rx_cmd_new_wiphy(msg);
if (!state)
return;
wiphy_create_complete(state->wiphy);
/* Setup a timer just in case a default interface is not created */
state->setup_timeout = l_timeout_create(1, manager_wiphy_setup_timeout,
state, NULL);
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
}
static bool manager_wiphy_state_match(const void *a, const void *b)
{
const struct wiphy_setup_state *state = a;
uint32_t id = L_PTR_TO_UINT(b);
return (state->id == id);
}
static struct wiphy_setup_state *manager_find_pending(uint32_t id)
{
return l_queue_find(pending_wiphys, manager_wiphy_state_match,
L_UINT_TO_PTR(id));
}
static uint32_t manager_parse_ifindex(struct l_genl_attr *attr)
{
uint16_t type, len;
const void *data;
while (l_genl_attr_next(attr, &type, &len, &data)) {
if (type != NL80211_ATTR_IFINDEX)
continue;
if (len != sizeof(uint32_t))
break;
return *((uint32_t *) data);
}
return -1;
}
static uint32_t manager_parse_wiphy_id(struct l_genl_attr *attr)
{
uint16_t type, len;
const void *data;
while (l_genl_attr_next(attr, &type, &len, &data)) {
if (type != NL80211_ATTR_WIPHY)
continue;
if (len != sizeof(uint32_t))
break;
return *((uint32_t *) data);
}
return -1;
}
static void manager_del_wiphy_event(struct l_genl_msg *msg)
{
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
struct wiphy_setup_state *state;
struct wiphy *wiphy;
struct l_genl_attr attr;
uint32_t id;
if (!l_genl_attr_init(&attr, msg))
return;
id = manager_parse_wiphy_id(&attr);
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
state = manager_find_pending(id);
if (state) {
if (state->pending_cmd_count)
state->aborted = true;
else
wiphy_setup_state_destroy(state);
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
}
wiphy = wiphy_find(id);
if (wiphy)
wiphy_destroy(wiphy);
}
static void manager_config_notify(struct l_genl_msg *msg, void *user_data)
{
uint8_t cmd;
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
struct wiphy_setup_state *state;
struct l_genl_attr attr;
struct netdev *netdev;
cmd = l_genl_msg_get_command(msg);
l_debug("Notification of command %s(%u)",
nl80211cmd_to_string(cmd), cmd);
switch (cmd) {
case NL80211_CMD_NEW_WIPHY:
manager_new_wiphy_event(msg);
break;
case NL80211_CMD_DEL_WIPHY:
manager_del_wiphy_event(msg);
break;
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
case NL80211_CMD_NEW_INTERFACE:
/*
* If we have a NEW_INTERFACE for a freshly detected wiphy
* assume we can now query for the default or pre-created
* interfaces, remove any we don't need and create our own.
*/
if (!l_genl_attr_init(&attr, msg))
break;
state = manager_find_pending(manager_parse_wiphy_id(&attr));
if (!state || !state->setup_timeout)
break;
manager_wiphy_dump_interfaces(state);
break;
case NL80211_CMD_DEL_INTERFACE:
if (!l_genl_attr_init(&attr, msg))
break;
netdev = netdev_find(manager_parse_ifindex(&attr));
if (!netdev)
break;
netdev_destroy(netdev);
break;
}
}
static void manager_interface_dump_callback(struct l_genl_msg *msg,
void *user_data)
{
struct wiphy_setup_state *state;
struct l_genl_attr attr;
l_debug("");
if (!l_genl_attr_init(&attr, msg))
return;
state = manager_find_pending(manager_parse_wiphy_id(&attr));
if (!state || state->setup_timeout)
return;
manager_get_interface_cb(msg, state);
}
static bool manager_check_create_interfaces(const void *a, const void *b)
{
struct wiphy_setup_state *state = (void *) a;
/* phy might have been detected after the initial dump */
if (state->setup_timeout)
return false;
wiphy_create_complete(state->wiphy);
if (state->pending_cmd_count)
return false;
/* If we are here, then there are no interfaces for this phy */
manager_create_interfaces(state);
if (state->pending_cmd_count)
return false;
wiphy_setup_state_free(state);
return true;
}
static void manager_interface_dump_done(void *user_data)
{
l_queue_remove_if(pending_wiphys,
manager_check_create_interfaces, NULL);
}
static void manager_wiphy_dump_callback(struct l_genl_msg *msg, void *user_data)
{
l_debug("");
manager_rx_cmd_new_wiphy(msg);
}
bool manager_init(struct l_genl_family *in,
const char *if_whitelist, const char *if_blacklist)
{
const struct l_settings *config = iwd_get_config();
struct l_genl_msg *msg;
unsigned int wiphy_dump;
unsigned int interface_dump;
const char *randomize_str;
nl80211 = in;
if (if_whitelist)
whitelist_filter = l_strsplit(if_whitelist, ',');
if (if_blacklist)
blacklist_filter = l_strsplit(if_blacklist, ',');
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
pending_wiphys = l_queue_new();
if (!l_genl_family_register(nl80211, "config", manager_config_notify,
NULL, NULL)) {
l_error("Registering for config notifications failed");
return false;
}
msg = l_genl_msg_new_sized(NL80211_CMD_GET_WIPHY, 128);
2019-08-21 01:04:27 +02:00
l_genl_msg_append_attr(msg, NL80211_ATTR_SPLIT_WIPHY_DUMP, 0, NULL);
wiphy_dump = l_genl_family_dump(nl80211, msg,
manager_wiphy_dump_callback,
NULL, NULL);
if (!wiphy_dump) {
l_error("Initial wiphy information dump failed");
l_genl_msg_unref(msg);
return false;
}
msg = l_genl_msg_new(NL80211_CMD_GET_INTERFACE);
interface_dump = l_genl_family_dump(nl80211, msg,
manager_interface_dump_callback,
NULL,
manager_interface_dump_done);
if (!interface_dump) {
l_error("Initial interface information dump failed");
l_genl_msg_unref(msg);
l_genl_family_cancel(nl80211, wiphy_dump);
return false;
}
randomize_str =
l_settings_get_value(config, "General", "mac_randomize");
if (randomize_str && !strcmp(randomize_str, "once"))
randomize = true;
return true;
}
void manager_exit(void)
{
l_strfreev(whitelist_filter);
l_strfreev(blacklist_filter);
2019-06-21 19:22:48 +02:00
l_queue_destroy(pending_wiphys, wiphy_setup_state_free);
manager: Create an interface on each wiphy and register netdevs This is probably the trickiest part in this patchset. I'm introducing a new logic where instead of using the interfaces that we find present when a wiphy is detected, which would normally be the one default interface per wiphy but could be 0 or more than one, we create one ourselves with the socket owner attribute and use exactly one for Station, AP and Ad-Hoc modes. When IWD starts we delete all the interfaces on existing wiphys that we're going to use (as determined by the wiphy white/blacklists) or freshly hotplugged ones, and only then we register the interface we're going to use meaning that the wiphy's limits on the number of concurrent interfaces of each type should be at 0. Otherwise we'd be unlikely to be abe to create the station interface as most adapters only allow one. After that we ignore any interfaces that may be created by other processes as we have no use for multiple station interfaces. At this point manager.c only keeps local state for wiphys during the interface setup although when we start adding P2P code we will be creating and removing interfaces multiple times during the wiphy's runtime and may need to track it here or in wiphy.c. We do not specifically check the interface number limits received during the wiphy dump, if we need to create any interfaces and we're over the driver's maximum for that specific iftype we'll still attempt it and report error if it fails. I tested this and it seems to work with my laptop's intel card and some USB hotplug adapters.
2019-04-11 03:10:26 +02:00
pending_wiphys = NULL;
nl80211 = NULL;
randomize = false;
}