Simplify building the @generated
function by return
ing the CodeInfo
#8
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Apparently using an actual
return code_info
line, instead of just having thecode_info
as the final expression, prevents Julia lowering pass from wrapping the return value in anExpr(:block, ...)
, meaning it can be used by Julia to build the generated function.Here is an example showing that difference:
With
return
:Without
return
:As far as I can tell, it's been like this since Julia 0.7.
Before this PR, I was manually lowering the function and unwrapping the return value to allow it to pass-through unmodified. Now, this syntax change somehow stops the lowering-pass from wrapping it in the first place.
That said, this seems like maybe a bug in Julia's lowering logic... I'm leaving a big note in the code, and we may need to re-evaluate this if that lowering logic changes in the future.