My project is now counting clock cycles on the RP2350!  🎉

I'm noticing things are taking more clock cycles than I expect--eg. 237/115 cycles (debug/release) to read 2 32-bit registers (1 high word, 1 low word), combine and return them as a 64-bit result via an `#[inline(always)]` struct accessor method.

Does anyone know if ARM published details on clock-cycles per instruction for the M33 (`thumbv8m.main-none-eabihf`)?  I'm still searching, but so far, I haven't been able to find this information.

I believe this information is readily available for `RISC-V` so next I will target `riscv32imac-unknown-none-elf` on the RP2350 to better understand what's happening.