Provide lazy CommitPendingWrites in addition to eager SavePersistentPrefs.

chrome/browser/chromeos/cros/login_library.cc

Index: chrome/browser/chromeos/cros/login_library.cc
diff --git a/chrome/browser/chromeos/cros/login_library.cc b/chrome/browser/chromeos/cros/login_library.cc
index 026757f3b92226b9bdf99ae76e2eb3993e401f2f..82f32a137d421d0ad324b72e1c84771d78be5446 100644
--- a/chrome/browser/chromeos/cros/login_library.cc
+++ b/chrome/browser/chromeos/cros/login_library.cc
@@ -133,9 +133,9 @@ class LoginLibraryImpl : public LoginLibrary {
   bool RestartJob(int pid, const std::string& command_line) {
     if (g_browser_process && g_browser_process->local_state()) {
       // XXX: normally this call must not be needed, however it turned out that
-      // without this explicit call to SavePersistentPrefs it is possible for
-      // preferences to be lost. See http://crosbug.com/13102
-      g_browser_process->local_state()->SavePersistentPrefs();
+      // without explicit write it is possible for preferences to be lost.
+      // See http://crosbug.com/13102
+      g_browser_process->local_state()->CommitPendingWrites();
     }
     return chromeos::RestartJob(pid, command_line.c_str());

[Mattias Nissler (ping if slow), 2011/03/22 16:56:20]

On which thread does this run? The CommitPendingWrites runs on the file thread,
so we could race here, killing chrome before the write task finishes.

[Denis Lagno, 2011/03/22 16:58:12]

absoulutely.  I am adding notification to fix this.

[Denis Lagno, 2011/03/22 17:08:40]

OTOH it looks like RestartJob is invoked by sending dbus message, so there is
chance it is safe:)

[Mattias Nissler (ping if slow), 2011/03/22 17:27:02]

Does that mean you're willing to take chances? ;)

I don't feel strong about what we end up with as long as we actually fix the
problem in a reliable way.

[Denis Lagno, 2011/03/23 14:37:43]

Done.

   }