[Mac/GN] Print a warning if the SDK version is 10.11 but Xcode is <7.3.

build/config/mac/mac_sdk.gni

 # Copyright 2014 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.
 
 import("//build/config/chrome_build.gni")
 
 # See https://bugs.chromium.org/p/webrtc/issues/detail?id=5453.
 # We can drop the rtc_require_mac_10_7_deployment flag when Chromium
 # also requires a 10.7 deployment target.
 import("//build_overrides/build.gni")
@@ skipping 33 matching lines @@
 if (mac_sdk_path == "") {
   mac_sdk_path = find_sdk_lines[0]
 }
 
 script_name = "//build/config/mac/sdk_info.py"
 _mac_sdk_result = exec_script(script_name, [ mac_sdk_name ], "scope")
 mac_sdk_build = _mac_sdk_result.sdk_build
 xcode_version = _mac_sdk_result.xcode_version
 xcode_build = _mac_sdk_result.xcode_build
 machine_os_build = _mac_sdk_result.machine_os_build
+
+if (mac_sdk_version != mac_sdk_min_build_override &&
+    exec_script("//build/check_return_value.py",
+                [
+                  "test",
+                  xcode_version,
+                  "-ge",
+                  "0730",
+                ],
+                "value") != 1) {
+  print(
+      "********************************************************************************")
+  print(
+      " WARNING: The Mac OS X SDK is incompatible with the version of Xcode. To fix,")
+  print(
+      "          either upgrade Xcode to the latest (7.3.1) or install the Mac OS X")

[Mark Mentovai, 2016/06/25 20:31:38]

Unless you plan on keeping the number 7.3.1 in sync with whatever’s latest, get
rid of it.

[Robert Sesek, 2016/06/29 20:43:34]

Done.

+  print(
+      "          $mac_sdk_min_build_override SDK. For more information, see https://crbug.com/620127.")
+  print()
+  print(" Current SDK Version:   $mac_sdk_version")
+  print(" Current Xcode Version: $xcode_version ($xcode_build)")
+  print(

[Mark Mentovai, 2016/06/25 20:31:38]

Does this go to stderr? (If not, could it?)

[Robert Sesek, 2016/06/29 20:43:34]

It goes to stdout, as does the new assertion failure. I don't know if we can
change this easily.

+      "********************************************************************************")

[Mark Mentovai, 2016/06/25 20:31:38]

I like making this fatal, at least in the configuration where we know it won’t
work as you suggested.

[Robert Sesek, 2016/06/29 20:43:34]

Done.

+}