Make GlobalErrorService's ownership model slightly less insane.

chrome/browser/sync/sync_global_error.cc

Index: chrome/browser/sync/sync_global_error.cc
diff --git a/chrome/browser/sync/sync_global_error.cc b/chrome/browser/sync/sync_global_error.cc
index 0e103dec8c47015111ef347c376b4d3da6cf7ed6..1d94f7460ed5b49b5ef42131a7278f957047f1ab 100644
--- a/chrome/browser/sync/sync_global_error.cc
+++ b/chrome/browser/sync/sync_global_error.cc
@@ -30,7 +30,7 @@ SyncGlobalError::SyncGlobalError(
   DCHECK(sync_service_);
   error_controller_->AddObserver(this);
   if (!switches::IsMaterialDesignUserMenu())
-    global_error_service_->AddGlobalError(this);
+    global_error_service_->AddUnownedGlobalError(this);

[Nico, 2016/10/10 14:03:37]

likewise:
https://cs.chromium.org/chromium/src/chrome/browser/sync/sync_global_error_fa...

[Avi (use Gerrit), 2016/10/10 16:52:36]

This is a profile-keyed service and is rather hard to kill. I'm going to have to
get back to you on this.

[Avi (use Gerrit), 2016/10/17 00:19:28]

Oh, boy, no.

This is a PKS. Lots of moving pieces depend on it, and I very much do not feel
comfortable fixing its broken ownership issues in this patch.

 }
 
 SyncGlobalError::~SyncGlobalError() {
@@ -40,7 +40,7 @@ SyncGlobalError::~SyncGlobalError() {
 
 void SyncGlobalError::Shutdown() {
   if (!switches::IsMaterialDesignUserMenu())
-    global_error_service_->RemoveGlobalError(this);
+    global_error_service_->RemoveUnownedGlobalError(this);
   error_controller_->RemoveObserver(this);
   error_controller_ = NULL;
 }