Add a basic implementation of the reduce kernel API to the CPU
reference implementation.

Bug: 22631253

For now, this just runs a serial reduction on one thread.

Change-Id: I34c96d24bb6f44274de72bb53160abcf79d143b0
diff --git a/cpu_ref/rsd_cpu.h b/cpu_ref/rsd_cpu.h
index 8e205d8..f2c7f19 100644
--- a/cpu_ref/rsd_cpu.h
+++ b/cpu_ref/rsd_cpu.h
@@ -58,6 +58,11 @@
                                    uint32_t usrLen,
                                    const RsScriptCall *sc) = 0;
 
+        virtual void invokeReduce(uint32_t slot,
+                                  const Allocation *ain,
+                                  Allocation *aout,
+                                  const RsScriptCall *sc) = 0;
+
         virtual void invokeInit() = 0;
         virtual void invokeFreeChildren() = 0;