[XLA] Expose a way to control dot/conv precision

This adds a field to the proto so that we may serialize it.

On TPUs, we can simulate higher precision by splitting a float32 number into several bfloat16 numbers such that their sum closely approximates the original number.

A tensor contraction operation like convolution or a dot product can be computed by forming several partial products which approximate the correct answer to a closer margin.

PiperOrigin-RevId: 209720948