Upstream: DexOptimizer and DexLoader classes

chrome/android/BUILD.gn

Index: chrome/android/BUILD.gn
diff --git a/chrome/android/BUILD.gn b/chrome/android/BUILD.gn
index cb1fbbace1503dec480c456869499b9fe0d104bf..e554f95d06991e24f3156813178459d00f3108d4 100644
--- a/chrome/android/BUILD.gn
+++ b/chrome/android/BUILD.gn
@@ -550,8 +550,10 @@ instrumentation_test_apk("chrome_public_test_apk") {
   deps = [
     ":chrome_public_test_apk_manifest",
     ":chrome_test_java",
+    "//chrome/android/webapk/shell_apk:shell_apk_javatests",

[Yaron, 2016/05/16 20:39:58]

This is sufficient for the tests to get picked up and run, right?

[jbudorick, 2016/05/16 20:57:41]

yep

   ]
   additional_apks = [
+    "//chrome/android/webapk/shell_apk/javatests/dex_optimizer:dex_optimizer_apk",

[Yaron, 2016/05/16 20:39:58]

I think this is fine and I think we discussed with the approach with John but
+jbudorick in case he has concerns. We could split out shell_apk_javatests as a
separate test target on the waterfall if needed

[jbudorick, 2016/05/16 20:57:41]

Yeah, I think it's fine for us to use chrome_public_test_apk as the default
target for instrumentation tests in //chrome/android. (sync_shell deviates
because it needs a different apk_under_test for some reason.)

     "//chrome/test/android/chrome_public_test_support:chrome_public_test_support_apk",
     "//net/android:net_test_support_apk",
   ]