diff options
| author | jam@chromium.org <jam@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98> | 2008-10-24 19:21:13 +0000 |
|---|---|---|
| committer | jam@chromium.org <jam@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98> | 2008-10-24 19:21:13 +0000 |
| commit | 3cdb7af81de506be730544edcfe4a5547f0fdaea (patch) | |
| tree | 223c1e000b0eccf637565345904fb4055e41c5ea /chrome/common/ipc_channel_proxy.cc | |
| parent | 8947da6c7cd943bcb4be80a258e263a84aa15006 (diff) | |
| download | chromium_src-3cdb7af81de506be730544edcfe4a5547f0fdaea.zip chromium_src-3cdb7af81de506be730544edcfe4a5547f0fdaea.tar.gz chromium_src-3cdb7af81de506be730544edcfe4a5547f0fdaea.tar.bz2 | |
Make IPC::SyncChannel not duplicate the underlying MessageLoop implementation by pumping messages on its own. This fixes the problem of windowless plugins not painting on right click, and generally makes this class almost ported, other than using a generic version of events/locks.Through this change I've also cleaned up the class and hopefully made it more understandable.
Review URL: http://codereview.chromium.org/8001
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@3934 0039d316-1c4b-4281-b951-d872f2087c98
Diffstat (limited to 'chrome/common/ipc_channel_proxy.cc')
| -rw-r--r-- | chrome/common/ipc_channel_proxy.cc | 24 |
1 files changed, 11 insertions, 13 deletions
diff --git a/chrome/common/ipc_channel_proxy.cc b/chrome/common/ipc_channel_proxy.cc index baac273..23d8ea9 100644 --- a/chrome/common/ipc_channel_proxy.cc +++ b/chrome/common/ipc_channel_proxy.cc @@ -52,14 +52,16 @@ bool ChannelProxy::Context::TryFilters(const Message& message) { // Called on the IPC::Channel thread void ChannelProxy::Context::OnMessageReceived(const Message& message) { // First give a chance to the filters to process this message. - if (TryFilters(message)) - return; + if (!TryFilters(message)) + OnMessageReceivedNoFilter(message); +} +// Called on the IPC::Channel thread +void ChannelProxy::Context::OnMessageReceivedNoFilter(const Message& message) { // NOTE: This code relies on the listener's message loop not going away while // this thread is active. That should be a reasonable assumption, but it // feels risky. We may want to invent some more indirect way of referring to // a MessageLoop if this becomes a problem. - listener_message_loop_->PostTask(FROM_HERE, NewRunnableMethod( this, &Context::OnDispatchMessage, message)); } @@ -82,7 +84,7 @@ void ChannelProxy::Context::OnChannelError() { } // Called on the IPC::Channel thread -void ChannelProxy::Context::OnOpenChannel() { +void ChannelProxy::Context::OnChannelOpened() { DCHECK(channel_ != NULL); // Assume a reference to ourselves on behalf of this thread. This reference @@ -99,7 +101,7 @@ void ChannelProxy::Context::OnOpenChannel() { } // Called on the IPC::Channel thread -void ChannelProxy::Context::OnCloseChannel() { +void ChannelProxy::Context::OnChannelClosed() { // It's okay for IPC::ChannelProxy::Close to be called more than once, which // would result in this branch being taken. if (!channel_) @@ -220,22 +222,18 @@ void ChannelProxy::Init(const std::wstring& channel_id, Channel::Mode mode, // complete initialization on the background thread context_->ipc_message_loop()->PostTask(FROM_HERE, NewRunnableMethod( - context_.get(), &Context::OnOpenChannel)); + context_.get(), &Context::OnChannelOpened)); } void ChannelProxy::Close() { // Clear the backpointer to the listener so that any pending calls to // Context::OnDispatchMessage or OnDispatchError will be ignored. It is // possible that the channel could be closed while it is receiving messages! - context_->clear(); + context_->Clear(); - if (MessageLoop::current() == context_->ipc_message_loop()) { - // We're being destructed on the IPC thread, so no need to use the message - // loop as it might go away. - context_->OnCloseChannel(); - } else { + if (context_->ipc_message_loop()) { context_->ipc_message_loop()->PostTask(FROM_HERE, NewRunnableMethod( - context_.get(), &Context::OnCloseChannel)); + context_.get(), &Context::OnChannelClosed)); } } |