Android中的网络管理源码分析--netd
http://www.voidcn.com/blog/a34140974/article/p-5033426.html
1 Netd簡介
Netd是Android的網(wǎng)絡(luò)守護進程。NetD是個網(wǎng)絡(luò)管家，封裝了復(fù)雜的底層各種類型的網(wǎng)絡(luò)(NAT，PLAN，PPP，SOFTAP，TECHER，ETHO，MDNS等)，隔離了底層網(wǎng)絡(luò)接口的差異，給Framework提供了統(tǒng)一調(diào)用接口，簡化了網(wǎng)絡(luò)的使用。NetD主要功能是:第一、接收Framework的網(wǎng)絡(luò)請求，處理請求，向Framework層反饋處理結(jié)果；第二、監(jiān)聽網(wǎng)絡(luò)事件(斷開/連接/錯誤等)，向Framework層上報。
2Netd的啟動過程
Netd作為后臺服務(wù)進程在Andriod系統(tǒng)啟動的init1階段就被啟動了，其在init.rc文件的配置如下：
|
service netd /system/bin/netd class main socket netd stream 0660 root system socket dnsproxyd stream 0660 root inet socket mdns stream 0660 root system socket fwmarkd stream 0660 root inet |
看一看到，這里為netd配置了4個socket(比老版本多了一個名字為“fwmakd”的socket)，根據(jù)配置可找到netd的入口函數(shù)為main():
|
int main() { CommandListener *cl; NetlinkManager *nm; DnsProxyListener *dpl; MDnsSdListener *mdnsl; FwmarkServer* fwmarkServer; ALOGI("Netd 1.0 starting"); remove_pid_file();//猜測為每次重啟時刪除舊的 blockSigpipe();//禁止SIGPIPE中斷 //創(chuàng)建NetlinkManager實例 if (!(nm = NetlinkManager::Instance())) { ALOGE("Unable to create NetlinkManager"); exit(1); }; //創(chuàng)建CommandListener實例，并將其設(shè)置為NetlinkManager的Broadcaster,之后啟動nm cl = new CommandListener(); nm->setBroadcaster((SocketListener *) cl); if (nm->start()) { ALOGE("Unable to start NetlinkManager (%s)", strerror(errno)); exit(1); } // Set local DNS mode, to prevent bionic from proxying // back to this service, recursively. setenv("ANDROID_DNS_MODE", "local", 1); //創(chuàng)建并開始監(jiān)聽“dnsproxyd”socket dpl = new DnsProxyListener(CommandListener::sNetCtrl); if (dpl->startListener()) { ALOGE("Unable to start DnsProxyListener (%s)", strerror(errno)); exit(1); } //創(chuàng)建并開始監(jiān)聽“mdns”socket mdnsl = new MDnsSdListener(); if (mdnsl->startListener()) { ALOGE("Unable to start MDnsSdListener (%s)", strerror(errno)); exit(1); } //創(chuàng)建并開始監(jiān)聽“fwmarkd”socket fwmarkServer = new FwmarkServer(CommandListener::sNetCtrl); if (fwmarkServer->startListener()) { ALOGE("Unable to start FwmarkServer (%s)", strerror(errno)); exit(1); } /* * Now that we're up, we can respond to commands */ //開始監(jiān)聽“netd”socket if (cl->startListener()) { ALOGE("Unable to start CommandListener (%s)", strerror(errno)); exit(1); } bool wrote_pid = write_pid_file(); while(1) { sleep(30); // 30 sec if (!wrote_pid) { wrote_pid = write_pid_file(); } } ALOGI("Netd exiting"); remove_pid_file(); exit(0); } |
從上面?zhèn)€可以看出netd的啟動并不復(fù)雜，主要是啟動了4個監(jiān)聽socket，后面的分析將會看到每個socket對應(yīng)這一個監(jiān)聽線程。首先來看NetlinkManage，NetlinkManager（以后簡稱NM）主要負責(zé)接收并解析來自Kernel的UEvent消息。如果對linux的socket特別熟悉的話，光從“NetlinkMananger”的名字就能推斷出此類的基本實現(xiàn)和作用：肯定使用了PF_NETLINK的socket。這種socket一般是在應(yīng)用層（相對于內(nèi)核）監(jiān)聽內(nèi)核事件的時候使用。例如USB的插拔等等。從main的代碼可以知道它的入口為start()函數(shù)。
|
int NetlinkManager::start() { //創(chuàng)建接收NETLINK_KOBJECT_UEVENT消息的socket，其值保存在mUeventSock中 //其中，NETLINK_FORMAT_ASCII代表UEvent消息的內(nèi)容為ASCII字符串 if ((mUeventHandler = setupSocket(&mUeventSock, NETLINK_KOBJECT_UEVENT, 0xffffffff, NetlinkListener::NETLINK_FORMAT_ASCII, false)) == NULL) { return -1; } //創(chuàng)建接收RTMGPR_LINK消息的socket，其值保存在mRouteSock中 //其中，NETLINK_FORMAT_BINARY代表UEvent消息的類型為結(jié)構(gòu)體，故需要進行二進制解析 if ((mRouteHandler = setupSocket(&mRouteSock, NETLINK_ROUTE, RTMGRP_LINK | RTMGRP_IPV4_IFADDR | RTMGRP_IPV6_IFADDR | RTMGRP_IPV6_ROUTE | (1 << (RTNLGRP_ND_USEROPT - 1)), NetlinkListener::NETLINK_FORMAT_BINARY, false)) == NULL) { return -1; } //創(chuàng)建接收NETLINK_NFLOG消息的socket，其值保存在mQuotaSock中 if ((mQuotaHandler = setupSocket(&mQuotaSock, NETLINK_NFLOG, NFLOG_QUOTA_GROUP, NetlinkListener::NETLINK_FORMAT_BINARY, false)) == NULL) { ALOGE("Unable to open quota socket");
} //創(chuàng)建接收NETLINK_NETFILTER消息的socket，其值保存在mQuotaSock中 if ((mStrictHandler = setupSocket(&mStrictSock, NETLINK_NETFILTER, 0, NetlinkListener::NETLINK_FORMAT_BINARY_UNICAST, true)) == NULL) { ALOGE("Unable to open strict socket");
} return 0; } |
start()四次調(diào)用了setupSocket函數(shù)，新建了4個PF_NETLINK類型的socket監(jiān)聽內(nèi)核的不同事件。查看函數(shù)setupSocket()。
|
NetlinkHandler *NetlinkManager::setupSocket(int *sock, int netlinkFamily, int groups, int format, bool configNflog) { struct sockaddr_nl nladdr; int sz = 64 * 1024; int on = 1; memset(&nladdr, 0, sizeof(nladdr)); nladdr.nl_family = AF_NETLINK; nladdr.nl_pid = getpid(); nladdr.nl_groups = groups; //新建socket,一定要注意這里的socket類型為SOCK_DGRAM,這句是整個Nm的關(guān)鍵 //netlinkFamily指定了soket監(jiān)聽的內(nèi)核事件 if ((*sock = socket(PF_NETLINK, SOCK_DGRAM | SOCK_CLOEXEC, netlinkFamily)) < 0) { ALOGE("Unable to create netlink socket: %s", strerror(errno)); return NULL; } //設(shè)置socket的屬性 if (setsockopt(*sock, SOL_SOCKET, SO_RCVBUFFORCE, &sz, sizeof(sz)) < 0) { ALOGE("Unable to set uevent socket SO_RCVBUFFORCE option: %s", strerror(errno)); close(*sock); return NULL; } if (setsockopt(*sock, SOL_SOCKET, SO_PASSCRED, &on, sizeof(on)) < 0) { SLOGE("Unable to set uevent socket SO_PASSCRED option: %s", strerror(errno)); close(*sock); return NULL; } //綁定 if (bind(*sock, (struct sockaddr *) &nladdr, sizeof(nladdr)) < 0) { ALOGE("Unable to bind netlink socket: %s", strerror(errno)); close(*sock); return NULL; } if (configNflog) {//只有mStrictSock對應(yīng)的為true if (android_nflog_send_config_cmd(*sock, 0, NFULNL_CFG_CMD_PF_UNBIND, AF_INET) < 0) { ALOGE("Failed NFULNL_CFG_CMD_PF_UNBIND: %s", strerror(errno)); return NULL; } if (android_nflog_send_config_cmd(*sock, 0, NFULNL_CFG_CMD_PF_BIND, AF_INET) < 0) { ALOGE("Failed NFULNL_CFG_CMD_PF_BIND: %s", strerror(errno)); return NULL; } if (android_nflog_send_config_cmd(*sock, 0, NFULNL_CFG_CMD_BIND, AF_UNSPEC) < 0) { ALOGE("Failed NFULNL_CFG_CMD_BIND: %s", strerror(errno)); return NULL; } } //將socket封裝成 NetLinkHandler，從而在socket有活動的時候處理 NetlinkHandler *handler = new NetlinkHandler(this, *sock, format); if (handler->start()) {//啟動NetlinkHandler,實際就是啟動監(jiān)聽 ALOGE("Unable to start NetlinkHandler: %s", strerror(errno)); close(*sock); return NULL; } return handler; } |
NetlinkHandler的start()函數(shù)轉(zhuǎn)調(diào)了this-> startListener()，此方法實際上是繼承自SocketListener類。這個類是一個比較通用的類，很多與socket的IO復(fù)用有關(guān)的模塊都會調(diào)用此類的相關(guān)方法。
|
int SocketListener::startListener(int backlog) { //注意這個變量是類的成員變量，實際上這里就是想方設(shè)法得到socket if (!mSocketName && mSock == -1) { SLOGE("Failed to start unbound listener"); errno = EINVAL; return -1; } else if (mSocketName) { if ((mSock = android_get_control_socket(mSocketName)) < 0) { SLOGE("Obtaining file descriptor socket '%s' failed: %s", mSocketName, strerror(errno)); return -1; } SLOGV("got mSock = %d for %s", mSock, mSocketName); fcntl(mSock, F_SETFD, FD_CLOEXEC); } //如果設(shè)置了mListen則監(jiān)聽socket，如果沒有設(shè)置則新建一個socketClient放入客戶端集合 //注意短路,對于NetlinkHandler,從其構(gòu)造函數(shù)可知mListen為false if (mListen && SLOGE("Unable to listen on socket (%s)", strerror(errno)); return -1; } else if (!mListen) mClients->push_back(new SocketClient(mSock, false, mUseCmdNum));//這里 if (pipe(mCtrlPipe)) { SLOGE("pipe failed (%s)", strerror(errno)); return -1; } //創(chuàng)建線程處理監(jiān)聽socket，這里其實并沒有所謂的“監(jiān)聽socket”，因為是NETLINK型的socket if (pthread_create(&mThread, NULL, SocketListener::threadStart, this)) { SLOGE("pthread_create (%s)", strerror(errno)); return -1; } return 0; } |
進入線程的入口函數(shù)SocketListener::threadStart()
|
void *SocketListener::threadStart(void *obj) { SocketListener *me = reinterpret_cast<SocketListener *>(obj); //注意obj為主線程傳遞進來的參數(shù)，就是SocketListener me->runListener(); pthread_exit(NULL); return NULL; } |
進入runListener
|
void SocketListener::runListener() { //此函數(shù)的主要邏輯就是select（） SocketClientCollection pendingList;//新建一個socketClientCollection存放活動fd while(1) { SocketClientCollection::iterator it; fd_set read_fds; int rc = 0; int max = -1; FD_ZERO(&read_fds); if (mListen) {//監(jiān)聽listenSocket的讀事件，前面已經(jīng)知道m(xù)Listen此時為fasle max = mSock; FD_SET(mSock, &read_fds); } FD_SET(mCtrlPipe[0], &read_fds);//這里的pipe什么作用？中斷循環(huán)標(biāo)志？ if (mCtrlPipe[0] > max) max = mCtrlPipe[0]; pthread_mutex_lock(&mClientsLock); //遍歷mClients集合 for (it = mClients->begin(); it != mClients->end(); ++it) { // NB: calling out to an other object with mClientsLock held (safe) int fd = (*it)->getSocket();//獲取與客戶通信的socket FD_SET(fd, &read_fds);//監(jiān)聽它 if (fd > max) { max = fd; } } pthread_mutex_unlock(&mClientsLock); SLOGV("mListen=%d, max=%d, mSocketName=%s", mListen, max, mSocketName); if ((rc = select(max + 1, &read_fds, NULL, NULL, NULL)) < 0) {//select if (errno == EINTR) continue; SLOGE("select failed (%s) mListen=%d, max=%d", strerror(errno), mListen, max); sleep(1); continue; } else if (!rc) continue; if (FD_ISSET(mCtrlPipe[0], &read_fds)) {//如果是pipe有活動 char c = CtrlPipe_Shutdown; TEMP_FAILURE_RETRY(read(mCtrlPipe[0], &c, 1));//讀取管道 if (c == CtrlPipe_Shutdown) { break;//難道這就是監(jiān)聽pipe的作用？ } continue; } //如果是監(jiān)聽socket,接收連接請求,當(dāng)然NETLINK不會走這里 if (mListen && FD_ISSET(mSock, &read_fds)) { struct sockaddr addr; socklen_t alen; int c; do { alen = sizeof(addr); c = accept(mSock, &addr, &alen); SLOGV("%s got %d from accept", mSocketName, c); } while (c < 0 && errno == EINTR); if (c < 0) { SLOGE("accept failed (%s)", strerror(errno)); sleep(1); continue; } fcntl(c, F_SETFD, FD_CLOEXEC); pthread_mutex_lock(&mClientsLock); //放入client集合 mClients->push_back(new SocketClient(c, true, mUseCmdNum)); pthread_mutex_unlock(&mClientsLock); } //將所有活動的fd都放入pendingList，貌似也只有一個 pendingList.clear(); pthread_mutex_lock(&mClientsLock); for (it = mClients->begin(); it != mClients->end(); ++it) { SocketClient* c = *it; // NB: calling out to an other object with mClientsLock held (safe) int fd = c->getSocket(); if (FD_ISSET(fd, &read_fds)) {//fd如果有活動 pendingList.push_back(c);//放入pendingList c->incRef(); } } pthread_mutex_unlock(&mClientsLock); //處理pendingList，這里的具體意思就是內(nèi)核有事件了，需要上層處理 while (!pendingList.empty()) { /* Pop the first item from the list */ it = pendingList.begin(); SocketClient* c = *it; pendingList.erase(it); /* Process it, if false is returned, remove from list */ if (!onDataAvailable(c)) { release(c, false); } c->decRef(); } } } |
從上面的函數(shù)可以看到，這里實際上是對3類fd作了監(jiān)聽處理。一類是監(jiān)聽socket，一類是client socket，并且這類socket被封裝成SocketClient集中在一個集合之內(nèi)。還有一個就是pipe。從NetlinkManager.start()中我們已經(jīng)知道啟動了四套這樣的結(jié)構(gòu)，其socket分別為mUeventSock ，mRouteSock，mQuotaSock，mStrictSock。這些Socket都是PF_NETLINK類型的sockegt，并不是監(jiān)聽socket，具體一點就是他們對應(yīng)的mListen均為false。也就是這四個socket被當(dāng)做SocketClient添加進了mClients（注意有四個實例）。等等，那么監(jiān)聽socket在哪呢？壓根就沒有監(jiān)聽socket，這里采用的是SOCK_DGRAM類型的socket！
當(dāng)檢測到這些socket有可讀事件發(fā)生時，也就是內(nèi)核有上層感興趣的事件發(fā)生時。相應(yīng)的onDataAvailable()被調(diào)用，這是一個虛函數(shù)。分析可知此時this的具體類型為NetlinkHandler，因此調(diào)用的是NetlinkHandler的onDataAvailable()。
|
bool NetlinkListener::onDataAvailable(SocketClient *cli) { int socket = cli->getSocket(); ssize_t count; uid_t uid = -1; bool require_group = true; if (mFormat == NETLINK_FORMAT_BINARY_UNICAST) { require_group = false; } //讀取數(shù)據(jù) count = TEMP_FAILURE_RETRY(uevent_kernel_recv(socket, mBuffer, sizeof(mBuffer), require_group, &uid)); if (count < 0) { if (uid > 0) LOG_EVENT_INT(65537, uid); SLOGE("recvmsg failed (%s)", strerror(errno)); return false; } NetlinkEvent *evt = new NetlinkEvent();//新建一個NetLinkEvent if (evt->decode(mBuffer, count, mFormat)) {//解碼 onEvent(evt);//調(diào)用了此函數(shù)，對event做了處理 } else if (mFormat != NETLINK_FORMAT_BINARY) { // Don't complain if parseBinaryNetlinkMessage returns false. That can // just mean that the buffer contained no messages we're interested in. SLOGE("Error decoding NetlinkEvent"); } delete evt; return true; } |
這里調(diào)用了onEvent()才是NetlinkHandler的入口。
|
void NetlinkHandler::onEvent(NetlinkEvent *evt) { const char *subsys = evt->getSubsystem(); if (!subsys) { ALOGW("No subsystem found in netlink event"); return; } if (!strcmp(subsys, "net")) { NetlinkEvent::Action action = evt->getAction(); const char *iface = evt->findParam("INTERFACE"); if (action == NetlinkEvent::Action::kAdd) { notifyInterfaceAdded(iface); } else if (action == NetlinkEvent::Action::kRemove) { notifyInterfaceRemoved(iface); } else if (action == NetlinkEvent::Action::kChange) { evt->dump(); notifyInterfaceChanged("nana", true); } else if (action == NetlinkEvent::Action::kLinkUp) { notifyInterfaceLinkChanged(iface, true); } else if (action == NetlinkEvent::Action::kLinkDown) { notifyInterfaceLinkChanged(iface, false); } else if (action == NetlinkEvent::Action::kAddressUpdated || action == NetlinkEvent::Action::kAddressRemoved) { const char *address = evt->findParam("ADDRESS"); const char *flags = evt->findParam("FLAGS"); const char *scope = evt->findParam("SCOPE"); if (action == NetlinkEvent::Action::kAddressRemoved && iface && address) { int resetMask = strchr(address, ':') ? RESET_IPV6_ADDRESSES : RESET_IPV4_ADDRESSES; resetMask |= RESET_IGNORE_INTERFACE_ADDRESS; if (int ret = ifc_reset_connections(iface, resetMask)) { ALOGE("ifc_reset_connections failed on iface %s for address %s (%s)", iface, address, strerror(ret)); } } if (iface && flags && scope) { notifyAddressChanged(action, address, iface, flags, scope); } } else if (action == NetlinkEvent::Action::kRdnss) { const char *lifetime = evt->findParam("LIFETIME"); const char *servers = evt->findParam("SERVERS"); if (lifetime && servers) { notifyInterfaceDnsServers(iface, lifetime, servers); } } else if (action == NetlinkEvent::Action::kRouteUpdated || action == NetlinkEvent::Action::kRouteRemoved) { const char *route = evt->findParam("ROUTE"); const char *gateway = evt->findParam("GATEWAY"); const char *iface = evt->findParam("INTERFACE"); if (route && (gateway || iface)) { notifyRouteChange(action, route, gateway, iface); } } } else if (!strcmp(subsys, "qlog")) { const char *alertName = evt->findParam("ALERT_NAME"); const char *iface = evt->findParam("INTERFACE"); notifyQuotaLimitReached(alertName, iface); } else if (!strcmp(subsys, "strict")) { const char *uid = evt->findParam("UID"); const char *hex = evt->findParam("HEX"); notifyStrictCleartext(uid, hex); } else if (!strcmp(subsys, "xt_idletimer")) { const char *label = evt->findParam("INTERFACE"); const char *state = evt->findParam("STATE"); const char *timestamp = evt->findParam("TIME_NS"); const char *uid = evt->findParam("UID"); if (state) notifyInterfaceClassActivity(label, !strcmp("active", state), timestamp, uid); #if !LOG_NDEBUG } else if (strcmp(subsys, "platform") && strcmp(subsys, "backlight")) { /* It is not a VSYNC or a backlight event */ ALOGV("unexpected event from subsystem %s", subsys); #endif } } |
可以看到，這里對不同的時間進行看了notifXxx所有的notifyXXXXX函數(shù)都會調(diào)用notify()函數(shù)
|
void NetlinkHandler::notify(int code, const char *format, ...) { char *msg; va_list args; va_start(args, format); if (vasprintf(&msg, format, args) >= 0) { //一定要注意這里的所使用的是cl的socket，名字為”netd”，而非之前的那四個 mNm->getBroadcaster()->sendBroadcast(code, msg, false); free(msg); } else { SLOGE("Failed to send notification: vasprintf: %s", strerror(errno)); } va_end(args); } |
mNm就是之前在main()中新建的NetworkMananger。其BroadCaster已經(jīng)設(shè)置為了cl（即CommandListener的一個實例）。CommandListener通過netd向NetworkManagementService發(fā)送消息。這里的消息可能有兩種：一種是底層主動上報的消息，另一種是上層請求的response。（這個和RILD很類似）
現(xiàn)在我們來整理一下上面的步驟：NetlinkManager新建了4個PF_NETLINK類型的socket，監(jiān)聽內(nèi)核發(fā)生的uEvent。當(dāng)內(nèi)核發(fā)生相應(yīng)的uEvent被對應(yīng)的 NetlinkManager檢測到。NetlinkManager將著個uEvent轉(zhuǎn)化為NetlinkEvent 通過CommandListener廣播到更上層。而這里的“更上層”指的是java層。可見，底層C/C++和上層java的聯(lián)系是通過socket聯(lián)系在一起的。
這里一定要清楚出兩點：之前的4個socket并不是這里的BroadCaster的socket；而且，個人覺得這個BroadCaster名字也容易讓人產(chǎn)生誤解，以為是廣播，廣播對應(yīng)的socket就應(yīng)該是UDP。而實際上這個socket是init.rc配置的名字為“netd”的socke所accept出來的clientSocket，是一個TCPsocket。而TCPsocket是無法廣播的。這里直接將sendBroadCast理解為sendMsg后面的就很好理解了。
接著分析CommandListener。這個類同樣繼承自SocketListener，與之前的4個Netlink socket所不同的是此類的mListen被設(shè)置為true。也就是“netd”為監(jiān)聽socket。CommandListener在之前的main函數(shù)中調(diào)用startListener開啟監(jiān)聽來自java層的連接。當(dāng)上層有連接時，select返回，accpet得到一個clientSocket，之后將其封裝成SocketClient添加經(jīng)list，并添加進select的監(jiān)聽隊列。當(dāng)java層下發(fā)命令，SocketClient的可讀事件被檢測到，從而做后續(xù)的處理，最后將底層處理結(jié)果response回上層，底層主動上報的消息也是通過此clientSocket上發(fā)到上層的。熟悉網(wǎng)絡(luò)編程的就應(yīng)該知道，這里是一個很典型的Select型的IO復(fù)用服務(wù)端模型。
除“netd”外，其他三個在init.rc中配置的socket：dnsproxyd mdns fwmarkd也構(gòu)建了幾乎一樣的服務(wù)端結(jié)構(gòu)。這里就不再贅述。以下為netd的大致框圖：
總結(jié)
以上是生活随笔為你收集整理的Android中的网络管理源码分析--netd的全部內(nèi)容，希望文章能夠幫你解決所遇到的問題。
- 上一篇： 抗美援朝老兵死亡有什么补偿 -法律知识？
- 下一篇： 低耦合的理解与作用