Fix double-initialization of CComModule.

ui/base/win/atl_module.h

+// Copyright (c) 2011 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef UI_BASE_WIN_ATL_MODULE_H_
+#define UI_BASE_WIN_ATL_MODULE_H_
+#pragma once
+
+#include <atlbase.h>
+#include <atlcom.h>
+
+namespace ui {
+namespace win {
+
+// Ensure that we have exactly one ATL module registered. It's safe to
+// call this more than once. ATL functions will crash if there's no
+// ATL module registered, or if you try to register two of them, so
+// dynamically registering one if needed makes it much easier for us
+// to support different build configurations like multi-dll without
+// worrying about which side of a module boundary each ATL module object
+// belongs on.
+//
+// This function must be implemented in this header file rather than a
+// source file so that it's inlined into the module where it's included,
+// rather than in the "base" module.

[Nico, 2011/11/14 16:21:13]

"ui"

[dmazzoni, 2011/11/14 16:27:41]

Done.

+static void CreateATLModuleIfNeeded() {
+  if (_pAtlModule == NULL) {
+    // This creates the module and automatically updates _pAtlModule.
+    static CComModule module;

[Nico, 2011/11/14 16:21:13]

Does CComModule have a destructor? Can you think of a way to not let this
introduce an exit destructor call? (If the destructor is not important, maybe
you can use CR_DEFINE_STATIC_LOCAL)

[dmazzoni, 2011/11/14 16:27:41]

My understanding is that the destructor is useful - it cleanly shuts down COM
connections with other processes. But I honestly don't know if there'd be any
harm in not calling it.

If it makes you feel any better, this code will never execute in a
statically-linked release build of chrome.exe, because everything is in one
module and ATL is initialized once in "content".

Because of that, I think it's safer to just let the destructor get called in
multi-dll builds and tests.

[Nico, 2011/11/14 16:34:28]

Can we try if that works then?
Good to know. But since the compiler can probably not prove that, i'd guess
it'll still emits code for "if static was initialized then call destructor" 
that runs at exit.

(Also, if we ever get to a state were I can make exit time destructors
compile-time errors on windows, this will need changing too.)

+  }
+}
+
+}  // namespace win
+}  // namespace ui
+
+#endif  // UI_BASE_WIN_ATL_MODULE_H_