Create Binder 对象跨进程传递原理.md

blogs/Android/Framework/Interview/进程间通信相关/Binder 对象跨进程传递原理.md

@@ -0,0 +1,183 @@
+---
+Binder 对象跨进程传递的原理是怎样的？
+---
+
+1. binder 传递有哪些方式？
+2. binder 在传递过程中是怎么存储的？
+3. binder 对象序列化和反序列化过程？
+4. binder 对象传递过程中驱动层做了什么？
+
+先从 AIDL 入口：
+
+```java
+interface IRemoteCaller{
+	void publishBinder(ICallback callback);
+}
+```
+
+```java
+public void publishBinder(ICallback callback){
+    Parcel_data = Parcel.obtain();
+    Parcel _reply = Parcel.obtain();
+    try{
+        _data.writeInterfaceToken(DESCRIPTOR);
+        _data.writeStrongBinder(callback!=null?callback.asBinder():null);
+        mRemote.transact(Stub.TRANSACTION_publishBinder, _data, _reply, 0);
+        _reply.readException();
+    }finally{
+        _reply.recycle();
+        _data.recycle();
+    }
+}
+```
+
+```java
+public boolean onTransact(int code, Parcel data, Parcel reply, int flags){
+    String descriptor = DESCRIPTOR;
+    switch(code){
+        case TRANSACTION_publishBinder:{
+            data.enforceInterface(descriptor);
+            ICallback _arg0;
+            _arg0 = ICallback.Stub.asInterface(data.readStrongBinder());
+            this.publishBinder(_arg0);
+            reply.writeNoException();
+            return true;
+        }
+    }
+}
+```
+
+```java
+public final void writeStrongBinder(IBinder val){
+	nativeWriteStrongBinder(mNativePtr, val);
+}
+void Parcel writeStrongBinder(JNIEnv* env, jclass clazz, jlong nativePtr, jobject object){
+    Parcel* parcel = reinterpret_cast<Parcel*>(nativePtr);
+    if(parcel!=NULL){
+        const status_t err = parcel->writeStrongBinder(ibinderForJavaObject(env, object));
+        if(err!=NO_ERROR){
+            signalExceptionForError(env, clazz, err);
+        }
+    }
+}
+sp<IBinder> ibinderForJavaObject(JNIEnv* env, jobject obj){
+    if(env->IsInterfaceOf(obj, gBinderOffsets.mClass)){
+        env->GetLongField(obj, gBinderOffsets.mObject);
+        return jbh!=NULL?jbh->get(env, obj):NULL;
+    }
+    if(env->IsInstanceOf(obj, gBinderProxyOffsets.mClass)){
+        return (IBinder*)
+            env->GetLongField(obj, gBinderProxyOffsets.mObject);
+    }
+    return NULL;
+}
+```
+
+```c++
+status_t Parcel:writeStrongBinder(const sp<IBinder>&val){
+	return flatten_binder(ProcessState::self(), val, this);
+}
+status_t flatten_binder(sp<ProcessState>&proc, sp<IBinder>&binder, Parcel* out){
+    flat_binder_object obj;
+    IBinder *local = binder->localBinder();
+    obj.type = BINDER_TYPE_BINDER;
+    obj.cookie = reinterpret_cast<uintptr_t>(local);
+    return finish_flatten_binder(binder, obj, out);
+}
+status_t finish_flatten_binder(sp<IBinder>& binder, flat_binder_object& flat, Parcel* out){
+    return out->writeObject(flat, false);
+}
+status_t Parcel::writeObject(const flat_binder_object& val, bool nullMetaData){
+    *reinterpret_cast<flat_binder_object*>(mData+mDataPos)=val;
+    mObjectsSize++;
+    return finishWrite(sizeof(flat_binder_object));
+}
+status_t Parcel::finishWrite(size_t len){
+    mDataPos += len;
+    return NO_ERROR;
+}
+```
+
+Parce 到了驱动层是如何处理的？
+
+```c++
+static void binder_transaction(struct binder_proc *proc, ...){
+	for(;offp<off_end;offp++){
+        fp=(struct flat_binder_object *)(t->buffer->data+*offp);
+        switch(fp->type){
+            case BINDER_TYPE_BINDER:{
+                struct binder_node *node = binder_get_node(proc, fp->binder);
+                if(node==NULL){
+                    node=binder_new_node(proc,fp->binder,fp->cookie);
+                }
+                ref = binder_get_ref_for_node(target_proc,node);
+                if(fp->type==BINDER_TYPE_BINDER)
+                    fp->type=BINDER_TYPE_HANDLE;
+                fp->handle=ref->desc;
+            }break;
+        }
+    }
+}
+```
+
+读：
+
+```java
+public final IBinder readStrongBinder(){
+	return nativeReadStrongBinder(mNativePtr);
+}
+```
+
+```c++
+jobject Parcel_readStrongBinder(JNIEnv* env, jclass clazz, jlong nativePtr){
+	Parcel* parcel = reinterpret_cast<Parcel*>(nativePtr);
+	if(parcel!=NULL){
+		return javaObjectForIBinder(env, parcel->readStrongBinder());
+	}
+	return NULL;
+}
+sp<IBinder> Parcel::readStrongBinder() const{
+    sp<IBinder> val;
+    unflatten_binder(ProcessState::self(), *this, &val);
+    return val;
+}
+status_t unflatten_binder(sp<ProcessState>& proc, Parcel& in,sp<IBinder>* out){
+    const flat_binder_object* flat = in.readObject(false);
+    switch(flat->type){
+        case BINDER_TYPE_BINDER:
+        	*out = reinterpret_cast<IBinder*>(flat->cookie);
+            return ...;
+        case BINDER_TYPE_HANDLE:
+            *out = proc->getStrongProxyForHandle(flat->handle);
+            return ...;
+    }
+}
+sp<IBinder> ProcessState::getStrongProxyForHandle(int32_t handle){
+    sp<IBinder> result;
+    handle_entry* e = lookupHandleLocked(handle);
+    IBinder* b = e->binder;
+    if(b == NULL){
+        b = new BpBinder(handle);
+        e->binder = b;
+        result = b;
+    }
+    return result;
+}
+jobject javaObjectForIBinder(JNIEnv* env, const sp<IBinder>& val){
+    if(val->checkSubclass(&gBinderOffsets)){
+        jobject object = static_cast<JavaBBinder*>(val.get())->object();
+        return object;
+    }
+    object = env->NewObject(gBinderProxyOffsets.mClass, ...);
+    env->SetLongField(object, gBinderProxyOffsets.mObject, (jlong)val.get());
+    return object;
+}
+```
+
+#### binder 是怎么跨进程传输的？
+
+1. Parcel 的 writeStrongBinder 和 readStrongBinder
+2. binder 在 Parcel 中存储原理，flat_binder_object
+3. 说清楚 binder_node，binder_ref
+4. 目标进程根据 binder_ref 的 handle 创建 BpBinder
+5. 由 BpBinder 再往上到 BinderProxy 到业务层的 Proxy