Reminds me fondly of the time I was writing assembler for the ARM2/3... it had such a nice instruction set that made hand-writing assembler pleasant.

It had a "barrel shifter" that gave you free shifts of powers of two, so you could calculate screen byte offsets quickly:

  // offset = x + y * 320
  ADD R0, R1, R2, LSL #8
  ADD R0, R0, R2, LSL #5
  // = 2 cycles

It also had bulk loads and stores that made reading/writing RAM cheaper. The trick there was to spill as many registers as you possibly could, so that you could transfer as many words as possible per bulk load/store.

  LDMIA R10!, {R0-R9}
  STMIA R11!, {R0-R9}
  // Transfers 40 bytes from memory pointed to by R10 to memory pointed to by R11,
  // And updates both pointers to the new addresses,
  // And only takes (3+10)*2 = 26 cycles to do the lot.

Happy days...