Is there a single instruction to calculate the sum of all components of a float4, e.g., in OpenCL?

Example:
float4 v;
float desiredResult = v.x + v.y + v.z + v.w;

Is there an SSE-like set of instructions?