I was looking at the IF.INC file and was wondering how expensive in regard to CPU instructions these macros are compared to regular jumping
Anyone know?

drobole,

Due to straightforward condition expression evaluator you can't expect optimized branching from those macros. For simple conditions they're as effective as hand-coded.

If they don't produce redundant instructions in circumstances like this

Im happy. That makes them very useful indeed

drobole,

That will be equivalent to What do you mean, redundant? While generated instructions are sub-optimal for this case (~ecx as the condition will generate shorter test ecx, ecx instead of cmp ecx, 0; entire conditional block can be rewritten using something like ?: C conditional operator, merging calls and stack adjustment), it's almost exactly what you've written.

I was thinking about how higher level language compilers tend to generate redundant assembly instructions. At least on the first pass. After all, optimizing a compiler is much about reducing, factoring and simplifying the assembly instructions generated.
I wasn't sure how assemblers and their macros go about stuff like that, but since assembly instructions for the most part correspond to a single machine instruction I guess there is not much room for redundant code being generated anyway.

Also, I'm not all that used to assembly at this point. I just dislike all the labels/jumps you need in order to facilitate a if,else if, else if etc. Especially when you get loops and stuff inside those.
Coming from C, this is kinda hard to swallow. But maybe I'll get used to it

drooble: Outside the extremely time critical pieces of code (eg. crypto algorithms, distributed computations, etc.) you can freely use them, difference would be hardly measurable .

But if you want to learn assembly, I'd suggest you to first learn how to go on without these macros (get accustomed to "linear" source code layout), and only start using macros after you understand well what they do.

Sounds like a good idea. Thanks guys