Android IPC 通讯机制源码分析

Client A与Binder kernel通信:
kernel\drivers\android\Binder.c）
static int binder_open(struct inode *nodp, struct file *filp)
{
struct binder_proc *proc;

if (binder_debug_mask & BINDER_DEBUG_OPEN_CLOSE)
printk(KERN_INFO "binder_open: %d:%d\n", current->group_leader->pid, current->pid);

proc = kzalloc(sizeof(*proc), GFP_KERNEL);
if (proc == NULL)
return -ENOMEM;
get_task_struct(current);
proc->tsk = current; //保存打开/dev/binder驱动的当前进程任务数据结构
INIT_LIST_HEAD(&proc->todo);
init_waitqueue_head(&proc->wait);
proc->default_priority = task_nice(current);
mutex_lock(&binder_lock);
binder_stats.obj_created[BINDER_STAT_PROC]++;
hlist_add_head(&proc->proc_node, &binder_procs);
proc->pid = current->group_leader->pid;
INIT_LIST_HEAD(&proc->delivered_death);
filp->private_data = proc;
mutex_unlock(&binder_lock);

if (binder_proc_dir_entry_proc) {
char strbuf[11];
snprintf(strbuf, sizeof(strbuf), "%u", proc->pid);
create_proc_read_entry(strbuf, S_IRUGO, binder_proc_dir_entry_proc, binder_read_proc_proc, proc); //为当前进程创建一个process入口结构信息
}
return 0;
}
从这里可以知道每一个打开/dev/binder的进程的信息都保存在binder kernel中，因而当一个进程调用ioctl与kernel binder通信时，binder kernel就能查询到调用进程的信息。BINDER_WRITE_READ是调用ioctl进程与Binder kernel通信一个非常重要的command。大家可以看到在IPCThreadState中的transact函数这个函数中call talkWithDriver发送的command就是BINDER_WRITE_READ。
static long binder_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
{
int ret;
struct binder_proc *proc = filp->private_data;
struct binder_thread *thread;
unsigned int size = _IOC_SIZE(cmd);
void __user *ubuf = (void __user *)arg;

/*printk(KERN_INFO "binder_ioctl: %d:%d %x %lx\n", proc->pid, current->pid, cmd, arg);*/
//将调用ioctl的进程挂起 caller将挂起直到 service 返回
ret = wait_event_interruptible(binder_user_error_wait, binder_stop_on_user_error < 2);
if (ret)
return ret;

mutex_lock(&binder_lock);
thread = binder_get_thread(proc);//根据当caller进程消息获取该进程线程池数据结构
if (thread == NULL) {
ret = -ENOMEM;
goto err;
}

switch (cmd) {
case BINDER_WRITE_READ: { //IPcThreadState中talkWithDriver设置ioctl的CMD
struct binder_write_read bwr;
if (size != sizeof(struct binder_write_read)) {
ret = -EINVAL;
goto err;
}
if (copy_from_user(&bwr, ubuf, sizeof(bwr))) {
ret = -EFAULT;
goto err;
}
if (binder_debug_mask & BINDER_DEBUG_READ_WRITE)
printk(KERN_INFO "binder: %d:%d write %ld at %08lx, read %ld at %08lx\n",
proc->pid, thread->pid, bwr.write_size, bwr.write_buffer, bwr.read_size, bwr.read_buffer);
if (bwr.write_size > 0) {
ret = binder_thread_write(proc, thread, (void __user *)bwr.write_buffer, bwr.write_size, &bwr.write_consumed);
if (ret < 0) {
bwr.read_consumed = 0;
if (copy_to_user(ubuf, &bwr, sizeof(bwr)))
ret = -EFAULT;
goto err;
}
}
if (bwr.read_size > 0) {//数据写入到caller process。
ret = binder_thread_read(proc, thread, (void __user *)bwr.read_buffer, bwr.read_size, &bwr.read_consumed, filp->f_flags & O_NONBLOCK);
if (!list_empty(&proc->todo))
wake_up_interruptible(&proc->wait); //恢复挂起的caller进程
if (ret < 0) {
if (copy_to_user(ubuf, &bwr, sizeof(bwr)))
ret = -EFAULT;
goto err;
}
}
.........................................
}

Int binder_thread_write(struct binder_proc *proc, struct binder_thread *thread,void __user *buffer, int size, signed long *consumed)
{
uint32_t cmd;
void __user *ptr = buffer + *consumed;
void __user *end = buffer + size;

while (ptr < end && thread->return_error == BR_OK) {
if (get_user(cmd, (uint32_t __user *)ptr))//从user空间获取cmd数据到内核空间
return -EFAULT;
ptr += sizeof(uint32_t);
if (_IOC_NR(cmd) < ARRAY_SIZE(binder_stats.bc)) {
binder_stats.bc[_IOC_NR(cmd)]++;
proc->stats.bc[_IOC_NR(cmd)]++;
thread->stats.bc[_IOC_NR(cmd)]++;
}
switch (cmd) {
case BC_INCREFS:
.........................................
case BC_TRANSACTION: //IPCThreadState通过writeTransactionData设置该cmd
case BC_REPLY: {
struct binder_transaction_data tr;

if (copy_from_user(&tr, ptr, sizeof(tr)))
return -EFAULT;
ptr += sizeof(tr);
binder_transaction(proc, thread, &tr, cmd == BC_REPLY);
break;
}
........................................
}

static void
binder_transaction(struct binder_proc *proc, struct binder_thread *thread,
struct binder_transaction_data *tr, int reply)
{
..............................................
if (reply) // cmd != BC_REPLY 不走这个case
{
......................................
}
else
{
if (tr->target.handle) { //对于service_manager来说这个条件不满足(handle == 0)
.......................................
}
} else {//这一段我们获取到了service_mananger process 注册在binder kernle的进程信息
target_node = binder_context_mgr_node; //BINDER_SET_CONTEXT_MGR 注册了service
if (target_node == NULL) { //manager
return_error = BR_DEAD_REPLY;
goto err_no_context_mgr_node;
}
}
e->to_node = target_node->debug_id;
target_proc = target_node->proc; //得到目标进程service_mananger 的结构
if (target_proc == NULL) {
return_error = BR_DEAD_REPLY;
goto err_dead_binder;
}
....................
}
if (target_thread) {
e->to_thread = target_thread->pid;
target_list = &target_thread->todo;
target_wait = &target_thread->wait; //得到service manager挂起的线程
} else {
target_list = &target_proc->todo;
target_wait = &target_proc->wait;
}
............................................
case BINDER_TYPE_BINDER:
case BINDER_TYPE_WEAK_BINDER: {
..........................
ref = binder_get_ref_for_node(target_proc, node); //在Binder kernel中创建
.......................... //查找到的service参考
} break;

............................................
if (target_wait)
wake_up_interruptible(target_wait); //唤醒挂起的线程 处理caller process请求
............................................//处理命令可以看svcmgr_handler
}
到这里我们已经通过getService连接到service manager进程了，service manager进程得到请求后，如果他的状态是挂起的话，将被唤醒。现在我们来看一下service manager中的binder_loop函数。
Service_manager.c
void binder_loop(struct binder_state *bs, binder_handler func)
{
.................................
binder_write(bs, readbuf, sizeof(unsigned));

for (;;) {
bwr.read_size = sizeof(readbuf);
bwr.read_consumed = 0;
bwr.read_buffer = (unsigned) readbuf;
res = ioctl(bs->fd, BINDER_WRITE_READ, &bwr); //如果没有要处理的请求进程将挂起
if (res < 0) {
LOGE("binder_loop: ioctl failed (%s)\n", strerror(errno));
break;
}
res = binder_parse(bs, 0, readbuf, bwr.read_consumed, func);//这里func就是
................................... //svcmgr_handler
}
}
接收到数据处理的请求，这里进行解析并调用前面注册的回调函数查找caller请求的service
int binder_parse(struct binder_state *bs, struct binder_io *bio,
uint32_t *ptr, uint32_t size, binder_handler func)
{
....................................
switch(cmd) {
......
case BR_TRANSACTION: {
struct binder_txn *txn = (void *) ptr;
if ((end - ptr) * sizeof(uint32_t) < sizeof(struct binder_txn)) {
LOGE("parse: txn too small!\n");
return -1;
}
binder_dump_txn(txn);
if (func) {
unsigned rdata[256/4];
struct binder_io msg;
struct binder_io reply;
int res;

bio_init(&reply, rdata, sizeof(rdata), 4);
bio_init_from_txn(&msg, txn);
res = func(bs, txn, &msg, &reply); //找到caller请求的service
binder_send_reply(bs, &reply, txn->data, res);//将找到的service返回给caller
}
ptr += sizeof(*txn) / sizeof(uint32_t);
break;
........
}

}
void binder_send_reply(struct binder_state *bs,
struct binder_io *reply,
void *buffer_to_free,
int status)
{
struct {
uint32_t cmd_free;
void *buffer;
uint32_t cmd_reply;
struct binder_txn txn;
} __attribute__((packed)) data;

data.cmd_free = BC_FREE_BUFFER;
data.buffer = buffer_to_free;
data.cmd_reply = BC_REPLY; //将我们前面binder_thread_write中cmd替换为BC_REPLY就可以知
data.txn.target = 0; //道service manager如何将找到的service返回给caller了
..........................
binder_write(bs, &data, sizeof(data)); //调用ioctl与binder kernel通信
}
从这里走出去后，caller该被唤醒了，client进程就得到了所请求的service的IBinder对象在Binder kernel中的参考，这是一个远程BBinder对象。

连接建立后的client连接Service的通信过程：
virtual sp<ICamera> connect(const sp<ICameraClient>& cameraClient)
{
Parcel data, reply;
data.writeInterfaceToken(ICameraService::getInterfaceDescriptor());
data.writeStrongBinder(cameraClient->asBinder());
remote()->transact(BnCameraService::CONNECT, data, &reply);
return interface_cast<ICamera>(reply.readStrongBinder());
}
向前面分析的这里remote是我们得到的CameraService的对象，caller进程会切入到CameraService。android的每一个进程都会创建一个线程池，这个线程池用处理其他进程的请求。当没有数据的时候线程是挂起的，这时binder kernel唤醒了这个线程：
IPCThreadState::joinThreadPool(bool isMain)
{
LOG_THREADPOOL("**** THREAD %p (PID %d) IS JOINING THE THREAD POOL\n", (void*)pthread_self(), getpid());

mOut.writeInt32(isMain ? BC_ENTER_LOOPER : BC_REGISTER_LOOPER);

status_t result;
do {
int32_t cmd;
result = talkWithDriver();
if (result >= NO_ERROR) {
size_t IN = mIn.dataAvail(); //binder kernel传递数据到service
if (IN < sizeof(int32_t)) continue;
cmd = mIn.readInt32();
IF_LOG_COMMANDS() {
alog << "Processing top-level Command: "
<< getReturnString(cmd) << endl;
}
result = executeCommand(cmd); //service 执行binder kernel请求的命令
}

// Let this thread exit the thread pool if it is no longer
// needed and it is not the main process thread.
if(result == TIMED_OUT && !isMain) {
break;
}
} while (result != -ECONNREFUSED && result != -EBADF);
.......................
}

status_t IPCThreadState::executeCommand(int32_t cmd)
{
BBinder* obj;
RefBase::weakref_type* refs;
status_t result = NO_ERROR;

switch (cmd) {
.........................
case BR_TRANSACTION:
{
binder_transaction_data tr;
result = mIn.read(&tr, sizeof(tr));
LOG_ASSERT(result == NO_ERROR,
"Not enough command data for brTRANSACTION");
if (result != NO_ERROR) break;

Parcel buffer;
buffer.ipcSetDataReference(
reinterpret_cast<const uint8_t*>(tr.data.ptr.buffer),
tr.data_size,
reinterpret_cast<const size_t*>(tr.data.ptr.offsets),
tr.offsets_size/sizeof(size_t), freeBuffer, this);

const pid_t origPid = mCallingPid;
const uid_t origUid = mCallingUid;

mCallingPid = tr.sender_pid;
mCallingUid = tr.sender_euid;

//LOGI(">>>> TRANSACT from pid %d uid %d\n", mCallingPid, mCallingUid);

Parcel reply;
.........................
if (tr.target.ptr) {
sp<BBinder> b((BBinder*)tr. cookie ); //service中Binder对象即CameraService
const status_t error = b->transact(tr.code, buffer, &reply, 0);//将调用
if (error < NO_ERROR) reply.setError(error);//CameraService的onTransact函数

} else {
const status_t error = the_context_object->transact(tr.code, buffer, &reply, 0);
if (error < NO_ERROR) reply.setError(error);
}

//LOGI("<<<< TRANSACT from pid %d restore pid %d uid %d\n",
// mCallingPid, origPid, origUid);

if ((tr.flags & TF_ONE_WAY) == 0) {
LOG_ONEWAY("Sending reply to %d!", mCallingPid);
sendReply(reply, 0);
} else {
LOG_ONEWAY("NOT sending reply to %d!", mCallingPid);
}

mCallingPid = origPid;
mCallingUid = origUid;

IF_LOG_TRANSACTIONS() {
TextOutput::Bundle _b(alog);
alog << "BC_REPLY thr " << (void*)pthread_self() << " / obj "
<< tr.target.ptr << ": " << indent << reply << dedent << endl;
}
..................................
}
break;
}
..................................
if ((tr.flags & TF_ONE_WAY) == 0) {
LOG_ONEWAY("Sending reply to %d!", mCallingPid);
sendReply(reply, 0); //通过binder kernel返回数据到caller进程这个过程大家
} else { //参照前面的叙述自己分析一下
LOG_ONEWAY("NOT sending reply to %d!", mCallingPid);
}
if (result != NO_ERROR) {
mLastError = result;
}
return result;
}
调用CameraService BBinder对象中的transact函数：
status_t BBinder::transact(
uint32_t code, const Parcel& data, Parcel* reply, uint32_t flags)
{
.....................
switch (code) {
case PING_TRANSACTION:
reply->writeInt32(pingBinder());
break;
default:
err = onTransact(code, data, reply, flags);
break;
}
...................
return err;
}

将调用CameraService的onTransact函数，CameraService继承了BBinder。
status_t BnCameraService::onTransact(
uint32_t code, const Parcel& data, Parcel* reply, uint32_t flags)
{
switch(code) {
case CONNECT: {
CHECK_INTERFACE(ICameraService, data, reply);
sp<ICameraClient> cameraClient = interface_cast<ICameraClient>(data.readStrongBinder());
sp<ICamera> camera = connect(cameraClient); //真正的处理函数
reply->writeStrongBinder(camera->asBinder());
return NO_ERROR;
} break;
default:
return BBinder::onTransact(code, data, reply, flags);
}
}
至此完成了一次从client到service的通信。

设计一个多客户端的Service
Service可以连接不同的Client,这里说的多客户端是指在Service中为不同的client创建不同的IClient接口，如果看过AIDL编程的话，应该清楚，Service需要开放一个IService接口给客户端，我们通过defaultServiceManager->getService就可以得到相应的service一个BpBinder接口，通过这个接口调用transact函数就可以与service通信了，这样也就完成了一个简单的service与client程序了，但这里有个缺点就是，这个IService是对所有的client开放的，如果我们要对不同的client做区分的话，在建立连接的时候所有的client需要给Service一个特性，这样做也未尝不可，但会很麻烦。比如对Camera来说可能不止一个摄像头，摄像头的功能也不一样，这样做就比较麻烦了。其实我们完全可以参照 Qt 中多客户端的设计方式，在Service中为每一个Client都创建一个IClient接口，IService接口只用于Serivce与Client建立连接用。对于Camera，如果存在多摄像头我们就可以在Service中为不同的Client打开不同的设备。
import android.os.IBinder;
import android.os.RemoteException;
public class TestServerServer extends android.app.testServer.ITestServer.Stub
{
int mClientCount = 0;
testServerClient mClient[];
@Override
public android.app.testServer.ITestClient.Stub connect(ITestClient client) throws RemoteException
{
// TODO Auto-generated method stub
testServerClient tClient = new testServerClient(this, client); //为Client创建
mClient[mClientCount] = tClient; //不同的IClient
mClientCount ++;
System.out.printf("*** Server connect client is %d", client.asBinder());
return tClient;
}

@Override
public void receivedData(int count) throws RemoteException
{
// TODO Auto-generated method stub

}
Public static class testServerClient extends android.app.testServer.ITestClient.Stub
{
public android.app.testServer.ITestClient mClient;
public TestServerServer mServer;
public testServerClient(TestServerServer tServer, android.app.testServer.ITestClient tClient)
{
mServer = tServer;
mClient = tClient;
}
public IBinder asBinder()
{
// TODO Auto-generated method stub
return this;
}
}
}
这仅仅是个Service的demo而已，如果添加这个作为system Service还得改一下android代码avoid permission check!

总结：
假定一个Client A 进程与Service B 进程要建立IPC通信，通过前面的分析我们知道他的流程如下：
1：Service B 打开Binder driver， 将自己的进程信息注册到kernel并为Service创建一个binder_ref。
2：Service B 通过Add_Service 将Service信息添加到service_manager进程
3：Service B 的Thread pool 挂起 等待client 的请求
4：Client A 调用open_driver打开Binder driver 将自己的进程信息注册到kernel并为Service创建一个binder_ref
5: Client A 调用defaultManagerService.getService 得到Service B在kernel中的IBinder对象
6：通过transact 与Binder kernel 通信，Binder Kernel将Client A 挂起。
7：Binder Kernel恢复Service B thread pool线程，并在 joinThreadPool 中处理Client的请求
8: Binder Kernel 挂起Service B 并将Service B 返回的数据写到Client A
9：Binder Kernle 恢复Client A
Binder kernel driver在Client A 与Service B之间扮演着中间代理的角色。任何通过transact传递的IBinder对象都会在Binder kernel中创建一个与此相关联的独一无二的BInder对象，用于区分不同的Client。