Unless you reject conservation and entropy once in motion the energy for the motion of the pendulum has to come from somewhere,
No, it doesn't. Conservation tells us that once in motion the pendulum will move forever without requiring any energy from anywhere. Rejecting conservation is what you must do in order to claim "once in motion the energy for the motion of the pendulum has to come from somewhere".
the energy to overcome losses in the pendulum.
Ah, now that's the exact opposite claim. You are talking about the energy for the motion of the pendulum being
lost somewhere.
But it's surely not being lost to kinetic energy of the Earth - that would imply that the pendulum would be causing the Earth to speed up.
In fact, the losses are minuscule for a sufficiently large pendulum with a low-friction hinge. What little energy is lost will mostly go to warming the air that the pendulum moves through, and a little to warming the hinge.
A well designed pendulum will keep going for several hours without any energy input to replace those tiny losses; Foucault pendulums in museums usually have an electromagnet to give the pendulum a tiny extra "kick" on each swing so to replace that lost energy. That energy mostly comes from burning coal.
As an aside, the pendulum in a clock typically gets its energy to replace frictional losses via an "escapement" mechanism that is powered from the winding of a spring or the lifting of weights, which is done by human muscle power, and therefore (as
@Bomb#20 points out above) ultimately from the Sun. It's an elegant design in which the clock's mechanism powers the pendulum, while in return the pendulum governs the rate at which the mechanism moves.