Hi, folks.

1st up, is this the right forum to discuss assembly language, or should I be over in 'Bare Metal'?

2nd, I spent several hours at the weekend, away from my Pi-s (oh, no!), learning about doing multiplication by constants using addition/subtraction/shifting instead. Once I got back to my Pi-s, I tried a few constants on my Pi Zero and, for each, the technique was about the same speed as a MUL!

Does anyone know of any documentation of when (if ever) it's quicker to use ADD/SUB/RSB/LSL instead of MUL?

So far it looks like the weekend's reading was only of academic interest unless I stick to using them when I can do it in 1 instruction (since, unlike with MUL, I won't have to load the constant into a register).

Would Usenet be a better place to ask this? comp.sys.arm seems very quiet and a search didn't show anything about multiplication by constants.

The most recent post I could find about it in comp.compilers was from 2007 and used x86 for examples. Bleah! (It took me a while to figure out what 'LEA's were capable of.)

Any other ARM-related Usenet/Google groups?