Removing nop from rtic::export

[romancardenas]

As discussed in #733, re-exporting nop in rtic::export may be unnecessary, and something like continue would do the trick in most cases. Personally, I'd keep the mandatory elements of rtic::export to the minimum. If anyone needs by chance the cortex_m::nop instruction, I'd say that they use it directly from the cortex-m crate instead of RTIC's re-export.

[korken89]

This sounds good to me. It's a remnant from when loop optimization was broken I'm the compiler.