For me, I always do EQ into Comp (I remember it as Eric Clapton = EC = EQ into Comp) 🙂
But sometimes, if I feel that it still misses something, I might add EQ again after the Comp to compensate some annoying frequencies or even double Comp.
But my first attempt will always be EQ into Comp, just a preference.
I would think that it would be like any combination of EQ & Compression. If there is an annoying frequency that the compressor brings out, you would probably want to cut that before compression. My go to is EQ first, but I change it around if it isn't sounding good.
Warren talks about EQ (boost) -> compression on his desk for snare and guitar to give it spank. This way you're pushing up frequencies you want, then distorting them (in a good way!) by driving them harder into the compressor. It's actually pretty subtle but I've done something I assume is like this using the bx_console E.
But he's also usually doing something like EQ -> Comp -> deesser -> EQ -> Comp -> Comp -> deesser -> EQ -> etc... but only doing a little bit, so it is in that context.