X86-64 Assembly for Recursive Functions was just asked, and it is essentially a debugging help question. Our help center currently says that

Advice or support with specific coding problems or debugging help

is off-topic. However, the linked question is really about coding problems that arise in compiler implementation, not incorrect code generation. Therefore, I’m not sure that it covers questions like this one.

Is this question off-topic? And would the answer be the same for all questions about debugging incorrect codegen?


1 Answer 1



In my opinion this should definitely be on-topic. While I also support allowing specific coding problems, I believe this sort of question should be on-topic even if those were to be disallowed.

For one, this sort of question is less about debugging code (which would be on-topic in numerous other places), and more about debugging an algorithm or approach to a problem, which is as much a design/implementation choice as what comment syntax you use or what pattern you use for parsing.

Making this sort of question off-topic would be pointless; it's incredibly unlikely they'd contribute a meaningful amount of clutter (I don't anticipate too many of these being posted), and unlike specific code questions, we're the only place on the network (to my knowledge) where this would be on-topic.

Disallowing these would just hurt people who need help doing the one thing this site is meant for, and it would have virtually no upsides.

  • 2
    $\begingroup$ I agree with your answer but I feel the original question is off topic because it does not talk about how the code was generated only about the code that is actually generated. $\endgroup$ Commented Jul 19, 2023 at 15:17
  • 1
    $\begingroup$ @BruceAdams That's a good point. That almost makes it a weird sort of meta-"fix my code" problem, so the solution is likely to just put in place higher standards for this type of problem and ask OP for clarification $\endgroup$ Commented Jul 19, 2023 at 15:19
  • $\begingroup$ I think that makes it more on topic. I don't want to see questions about debugging the code for the compiler. Getting the compiler output right is an implementation topic, and that entails knowing what the correct output is. The point of a reference implementation isn't the compiler source code itself, but its behaviour, insofar as that fills in gaps in the specification. $\endgroup$ Commented Jul 25, 2023 at 8:00

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .