Call scoped_refptr<T>::get() rather than relying on implicit "operator T*"

This upates calls to bound temporary objects to also use get().
While it has the same semantic equivalence to the existing code, this generally
represents a dangerous pattern - indeed, part of the whole motivation for this
change is to make this anti-pattern very visible to authors.

This change simply updates all of the call sites, to allow the "operator T*"
to be removed and preventing new instances. The existing instances will then be
reviewed for "suspicious" changes and updated to use/pass scoped_refptr<T>
rather than T*, as appropriate.

BUG=110610
TBR=darin

Review URL: https://codereview.chromium.org/15984016

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@205560 0039d316-1c4b-4281-b951-d872f2087c98

1 files changed, 2 insertions, 2 deletions

diff --git a/ipc/ipc_sync_message_filter.cc b/ipc/ipc_sync_message_filter.cc
index 541d685..a534c44 100644
--- a/ipc/ipc_sync_message_filter.cc
+++ b/ipc/ipc_sync_message_filter.cc
@@ -46,8 +46,8 @@ bool SyncMessageFilter::Send(Message* message) {
     base::AutoLock auto_lock(lock_);
     // Can't use this class on the main thread or else it can lead to deadlocks.
     // Also by definition, can't use this on IO thread since we're blocking it.
-    DCHECK(MessageLoopProxy::current() != listener_loop_.get());
-    DCHECK(MessageLoopProxy::current() != io_loop_.get());
+    DCHECK(MessageLoopProxy::current().get() != listener_loop_.get());
+    DCHECK(MessageLoopProxy::current().get() != io_loop_.get());
     pending_sync_messages_.insert(&pending_message);
   }