Map over refs of ImportTypes in TypeMap
#20837
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
EugeneFlesselle
changed the title
ImportTypes in inliner tree type mapImportTypes in TypeMap
odersky
requested changes
Jul 2, 2024
| def apply(tp: Type): Type = tp match | ||
| case tp: TermRef if tp.symbol.isImport => mapOver(tp) | ||
| case tp => tp.substSym(substFrom, substTo) | ||
| typeMap.andThen(substMap).andThen(mapOwnerThis) |
There was a problem hiding this comment.
I believe andThen is not inlined. So in the interest of performance I suggest the less nice
def mapType(tp: Type): Type =
...
mapOwnerThis(substMap(typeMap(tp)))We allocate fewer closures that way, and empirically that matters (a bit)
There was a problem hiding this comment.
Indeed ! turns out it's even @annotation.unspecialized
odersky
approved these changes
Jul 2, 2024
The inliner replaces references to parameters by their corresponding proxys, including in singleton types. It did not, however, handle the mapping over import types, the symbols of which way have depended on parameters. Mapping imports correctly was necessary for i19493 since the `summonInline` resolves post inlining to a given imported within the inline definition. Fix scala#19493
Both i19493 and i19436 require mapping the type of the expr in an `ImportType` which is itself the info of a `TermRef`. In the first issue, for the substitution of an inline def parameter proxy. In the second issue, for the parameter of a lambda returned from an inline def. Both can be handled in `TypeMap` by mapping over references to `ImportType`s. The second case also requires modifying `TreeTypeMap#mapType` such that the logic mapping over imports is done within a `TypeMap` doing the symbol substitutions. Fixes scala#19436
EugeneFlesselle
force-pushed
the
fix-19493
branch
from
5535e13 to
ff003fd
Compare
WojciechMazur
added a commit
that referenced
this pull request
Jul 3, 2024
Backports #20837 to 3.5.1-RC1
WojciechMazur
added a commit
that referenced
this pull request
Jul 11, 2024
Backports #20837 to the LTS branch. PR submitted by the release tooling. [skip ci]
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
The inliner replaces references to parameters by their corresponding proxys, including in singleton types.
It did not handle the mapping over import types, the symbols of which way have depended on parameters.
Both i19493 and i19436 require mapping the type of the expr in an
ImportTypewhich is itself the info of aTermRef.In the first issue, for the substitution of an inline def parameter proxy.
In the second issue, for the parameter of a lambda returned from an inline def.
Both can be handled in
TypeMapby mapping over references toImportTypes.The second case also requires modifying
TreeTypeMap#mapTypesuch that the logic mapping over imports is done within aTypeMapdoing the symbol substitutions.Also note these mappings are only necessary for
summonInlines (which could have just been made non-inline) resolvingpost-inlining to givens imported within the inline definition.
Fix #19493
Fix #19436