ProcessSingleton now uses base::win::MessageWindow to create a message-only window.

ProcessSingleton now uses base::win::MessageWindow to create a message-only window.

Collateral changes:
  - base::win::MessageWindow registes a single window class used by all message-only windows it creates. The class is registered via base::LazyInstance.
  - Added base::win::MessageWindow::FindWindow() wrapper used to find other message-only windows, including the other created by a different process.
  - Removed chrome::kMessageWindowClass constant.

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=209047

[alexeypa (please no reviews), 2013-06-24 18:19:48]

PTAL at base/win/*

[alexeypa (please no reviews), 2013-06-26 16:19:54]

Carlos, ping. PTAL at base/win/* changes.

[cpu_(ooo_6.6-7.5), 2013-06-27 15:15:38]

lgtm

https://codereview.chromium.org/17615003/diff/3001/base/win/message_window.cc
File base/win/message_window.cc (right):

https://codereview.chromium.org/17615003/diff/3001/base/win/message_window.cc...
base/win/message_window.cc:24: ATOM atom() const { return atom_; }
member constness but not logical constness. I would remove the const here.

https://codereview.chromium.org/17615003/diff/3001/chrome/browser/process_sin...
File chrome/browser/process_singleton.h (right):

https://codereview.chromium.org/17615003/diff/3001/chrome/browser/process_sin...
chrome/browser/process_singleton.h:33: #endif  // defined(OS_WIN)
no need to have this // defined comment.

https://codereview.chromium.org/17615003/diff/3001/chrome/browser/process_sin...
File chrome/browser/process_singleton_win.cc (left):

https://codereview.chromium.org/17615003/diff/3001/chrome/browser/process_sin...
chrome/browser/process_singleton_win.cc:270: // another instance of chrome
running. Otherwise we may end up writing out
so we switched the order as the class gets unregistered, window destroyed after
the lock file gets closed. I don't think this is a problem but we need to notify
pastarmov@ of the change.

[cpu_(ooo_6.6-7.5), 2013-06-27 15:16:05]

.

[pastarmovj, 2013-06-27 15:25:54]

lgtm for the change of destruction order.

https://codereview.chromium.org/17615003/diff/3001/chrome/browser/process_sin...
File chrome/browser/process_singleton_win.cc (left):

https://codereview.chromium.org/17615003/diff/3001/chrome/browser/process_sin...
chrome/browser/process_singleton_win.cc:270: // another instance of chrome
running. Otherwise we may end up writing out
On 2013/06/27 15:15:38, cpu wrote:
> so we switched the order as the class gets unregistered, window destroyed
after
> the lock file gets closed. I don't think this is a problem but we need to
notify
> pastarmov@ of the change.

Thanks for informing me. I think this doesn't change the situation much for the
lock file. So I am fine with this change.

[alexeypa (please no reviews), 2013-06-27 17:10:58]

https://codereview.chromium.org/17615003/diff/3001/base/win/message_window.cc
File base/win/message_window.cc (right):

https://codereview.chromium.org/17615003/diff/3001/base/win/message_window.cc...
base/win/message_window.cc:24: ATOM atom() const { return atom_; }
On 2013/06/27 15:15:38, cpu wrote:
> member constness but not logical constness. I would remove the const here.

Done.

https://codereview.chromium.org/17615003/diff/3001/chrome/browser/process_sin...
File chrome/browser/process_singleton.h (right):

https://codereview.chromium.org/17615003/diff/3001/chrome/browser/process_sin...
chrome/browser/process_singleton.h:33: #endif  // defined(OS_WIN)
On 2013/06/27 15:15:38, cpu wrote:
> no need to have this // defined comment.

It matches the style used in the file. The majority of endif's in this file are
commented this way.

[alexeypa (please no reviews), 2013-06-27 17:13:20]

thestig@, please take a look at src/chrome/*
cpu@, it looks like you are an owner for src/win8/* too. Please rubber stamp the
changes in src/win8/* just in case. 

Thanks.

[cpu_(ooo_6.6-7.5), 2013-06-27 18:12:37]

yeah I looked at the win8 part. You don't need a lg for each directory

[Lei Zhang, 2013-06-27 21:16:37]

It looks like cpu@ already reviewed most of the chrome/ bits, so lgtm too.

[commit-bot: I haz the power, 2013-06-27 21:18:19]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/alexeypa@chromium.org/17615003/11001

[commit-bot: I haz the power, 2013-06-28 01:02:50]

Change committed as 209047