Tweak examples to emphasize importance of retaining kernels

examples/demo.py

@@ -23,7 +23,8 @@ __kernel void sum(
 """).build()
 
 res_g = cl.Buffer(ctx, mf.WRITE_ONLY, a_np.nbytes)
-prg.sum(queue, a_np.shape, None, a_g, b_g, res_g)
+knl = prg.sum  # Use this Kernel object for repeated calls
+knl(queue, a_np.shape, None, a_g, b_g, res_g)
 
 res_np = np.empty_like(a_np)
 cl.enqueue_copy(queue, res_np, res_g)

examples/demo_array.py

@@ -22,6 +22,7 @@ prg = cl.Program(ctx, """
     }
     """).build()
 
-prg.sum(queue, a.shape, None, a_dev.data, b_dev.data, dest_dev.data)
+knl = prg.sum  # Use this Kernel object for repeated calls
+knl(queue, a.shape, None, a_dev.data, b_dev.data, dest_dev.data)
 
 print(la.norm((dest_dev - (a_dev+b_dev)).get()))