[jextract] Improve Cdecl lowering #238
Merged
Conversation
This file contains hidden or 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
rintaro
changed the title
rintaro
commented
May 30, 2025
| @@ -103,38 +76,30 @@ enum ConversionStep: Equatable { | |||
|
|
|||
| /// Convert the conversion step into an expression with the given | |||
| /// value as the placeholder value in the expression. | |||
| func asExprSyntax(isSelf: Bool, placeholder: String) -> ExprSyntax { | |||
| func asExprSyntax(placeholder: String, bodyItems: inout [CodeBlockItemSyntax]) -> ExprSyntax? { | |||
There was a problem hiding this comment.
Using bodyItems, ConversionStep can now introduce side effects, for example it can store the inner step into a variable, then return the DeclReferenceExprSyntax to it.
DougGregor
approved these changes
May 31, 2025
Comment on lines
+53
to
+54
| /// providing all of the mappings between the parameter and result types | ||
| /// of the original function and its `@_cdecl` counterpart. |
There was a problem hiding this comment.
Total nit, but there's no parameter or result type for a variable, just the type itself.
Comment on lines
+53
to
+55
| if let generics = node.genericParameterClause { | ||
| throw SwiftFunctionTranslationError.generic(generics) | ||
| } |
There was a problem hiding this comment.
Thank you for making it defensive like this
| hasGetter = true | ||
| case .keyword(.set), .keyword(._modify), .keyword(.unsafeMutableAddress): | ||
| hasSetter = true | ||
| default: // Ignore willSet/didSet and unknown accessors. |
There was a problem hiding this comment.
willSet/didSet imply that we have both getter and setter, don't they?
| switch (hasGetter, hasSetter) { | ||
| case (true, true): return [.get, .set] | ||
| case (true, false): return [.get] | ||
| case (false, true): return [.set] |
There was a problem hiding this comment.
Technically, this case is unreachable in well-formed code
There was a problem hiding this comment.
Ohh, I didn't know setter only variables are prohibited in Swift 😅
| expectedCDecl: """ | ||
| @_cdecl("c_takeString") | ||
| public func c_takeString(_ str: UnsafePointer<Int8>) { | ||
| takeString(str: String(cString: str)) |
* Newly introduce `lowerResult()` (splitted from `lowerParameter()` because result lowering can be differnt from parameter lowering. * Remove `ConversionStep.init(cdeclToSwift:)/.init(swiftToCdecl:)`. Instead embed the logic in `lowerParameter()`/`lowerResult()` so that we can implement type lowering and the actual conversion at the same place. * Support 'String', 'Void' and '() -> Void) types * Implement getter / setter lowering * Fix conversion for tuple returns (see: `lowerTupleReturns()` test) * Reference types are now returned indirectly, as the logic are being unifed. Thanks to this we don't need manual `retain()` before returning instances, which wasn't implemented correctly.
rintaro
force-pushed
the
jextract-re-lowering
branch
from
01b8081 to
bcd5e0e
Compare
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.
Part of #236 with some test cases.
Cdecl lowering improvements without integrating into the actual workflow.
lowerResult()(splitted fromlowerParameter()because result lowering can be differnt from parameter lowering.ConversionStep.init(cdeclToSwift:)/.init(swiftToCdecl:). Instead embed the logic inlowerParameter()/lowerResult()so that we can implement type lowering and the actual conversion at the same place.lowerTupleReturns()test)retain()before returning instances, which wasn't implemented correctly.