Fix screen taps for Sensorless Inline XRSessions

Currently, getPose for a screen tap on a sensorless inline session fails
because there is no "MojoFromViewer" pose available.  In a sensorless
session "Mojo" Space and "Viewer" space are the same, so this should
return identity.

Fixed: 1031596
Change-Id: Ic09590be6bbeff84cfaad894a2c1702ed7d3f896
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1978901
Auto-Submit: Alexander Cooper <alcooper@chromium.org>
Commit-Queue: Brandon Jones <bajones@chromium.org>
Reviewed-by: Brandon Jones <bajones@chromium.org>
Cr-Commit-Position: refs/heads/master@{#728051}

third_party/blink/renderer/modules/xr/xr_session.cc

@@ -1528,8 +1528,12 @@ base::Optional<TransformationMatrix> XRSession::MojoFromViewer() const {
   if (!CanReportPoses())
     return base::nullopt;
 
-  if (!mojo_from_viewer_)
+  if (!mojo_from_viewer_) {
+    if (sensorless_session_)
+      return TransformationMatrix();
+
     return base::nullopt;
+  }
 
   return *mojo_from_viewer_.get();
 }