iOS buildbot configurations.

webrtc/build/ios/client.webrtc/iOS32_Simulator_Debug.json

Index: webrtc/build/ios/client.webrtc/iOS32_Simulator_Debug.json
diff --git a/webrtc/build/ios/client.webrtc/iOS32_Simulator_Debug.json b/webrtc/build/ios/client.webrtc/iOS32_Simulator_Debug.json
new file mode 100644
index 0000000000000000000000000000000000000000..91e3d22f9bb1388c4199347d61c14f51faa840d2
--- /dev/null
+++ b/webrtc/build/ios/client.webrtc/iOS32_Simulator_Debug.json
@@ -0,0 +1,23 @@
+{
+  "owners": [
+    "kjellander"
+  ],
+  "comments": [
+    "Tests for 32-bit iOS 9.0 simulators."
+  ],
+  "xcode version": "7.0",
+  "GYP_DEFINES": {
+    "chromium_ios_signing": "0",
+    "target_arch": "ia32"

[smut, 2016/02/02 21:31:15]

I'm assuming you know your GYP_DEFINES are correct. Just FYI but on iOS we use
target_subarch instead of target_arch and we specify "arm32", "arm64", or
"both".

[kjellander_webrtc, 2016/02/02 22:28:29]

Yeah, this is different for WebRTC, we only use target_arch (arm,arm64,ia32 or
x64) and then let target_subarch be set in
https://chromium.googlesource.com/external/webrtc/+/master/webrtc/supplement....
I don't remember exactly the details behind it, but it's been working so far. 
I'm not familar with exactly how the fat binaries and such works for iOS, but I
guess this could be improved in the future (so we align with how Chrome builds
this).

+  },
+  "compiler": "ninja",
+  "configuration": "Debug",
+  "sdk": "iphonesimulator9.0",
+  "tests": [
+    {
+      "include": "common_tests.json",
+      "device type": "iPhone 5s",
+      "os": "9.0"
+    },
+  ]
+}