Some fixes for CUDA runtime compile:

* When Baking wasn't used we got an error.
* On top of Volume Nodes (NODES_FEATURE_VOLUME), we now also check if we need volume sampling code,
so we can disable that as well and save some further compilation time.

intern/cycles/kernel/kernels/cuda/kernel.cu

@@ -193,6 +193,7 @@ kernel_cuda_shader(uint4 *input,
 	}
 }
 
+#ifdef __BAKING__
 extern "C" __global__ void
 CUDA_LAUNCH_BOUNDS(CUDA_THREADS_BLOCK_WIDTH, CUDA_KERNEL_MAX_REGISTERS)
 kernel_cuda_bake(uint4 *input, float4 *output, int type, int filter, int sx, int sw, int offset, int sample)
@@ -202,6 +203,7 @@ kernel_cuda_bake(uint4 *input, float4 *output, int type, int filter, int sx, int
 	if(x < sx + sw)
 		kernel_bake_evaluate(NULL, input, output, (ShaderEvalType)type, filter, x, offset, sample);
 }
+#endif
 
 #endif