Even ones inside inline functions. This was a backwards compatibility
hack for Kotlin 1.4, where the inliner would crash if it attempted to
regenerate an anonymous object with no SMAP; that has been fixed in 1.5,
and ever since then trivial SMAPs could be inferred from line number
markers in methods.
There are three kinds of changes to tests in this commit:
* Some SMAPs are gone entirely - self-explanatory.
* Some SMAPs have narrower line ranges - that's because the old SMAP
had the range for the entire file, while the new one only maps up to
the last line number used in the class. There should be no
difference in behavior.
* Some "source file name" markers are removed in continuation objects
- continuations don't have any line numbers, so there's no debugging
information anyway. The actual source information is in the
containing class.
Rename it to `deepCopyWithoutPatchingParents`, move to
`org.jetbrains.kotlin.ir.util`, make it inline+reified, and extract the
common implementation with `deepCopyWithSymbols` into `deepCopyImpl`.
`AbstractTypeChecker.isCommonDenotableType` calls a few functions which
check if the type is flexible, which is not cheap in case of JVM IR, see
`asFlexibleType`.
#KT-66281
When those classes become auto-generated, the order of constructor
parameters will differ.
This commit ensures they will still be resolved correctly.
^KT-65773 In Progress
This commit removes the default parameter in constructor,
as well as custom logic in this class, which would
make it harder for its subsequent auto-generation.
^KT-65773 In Progress
The problem was that in K2 for some top-level script declarations we
need to add a dispatch receiver parameter (because frontend do not
assign any, but representing script as a class requires it to be the
script class) and at the same time, calls to these declarations rely on
properly set dispatch receiver parameter.
The simplest solution found is to have an additional traversal on the
relevan top-level declarations and assigning the dispatch receiver,
before running the main transformation.
#KT-64502 fixed
The previous one was incorrect for K1 since parent of top-level function
is `IrClass`, not `IrPackageFragment`.
The change is non-functional, K1 still worked correctly, but had to do
some extra work when inlining `emptyArray` calls and produces less
performant bytecode.
- remove ENABLE/COMPATIBILITY because they can no longer be used
- remove forAllMethodsWithBody because its behavior is now equivalent to
isEnabled
- inline isCompatibility
- inline DEFAULT
- rename ALL_INCOMPATIBLE -> ALL
Substitution of type arguments to non-reified type parameters may lead
to accidental reification, which should not be done (see ^KT-60174 for
examples). So, we should erase them, except the few cases.
^KT-60174: Fixed
^KT-60175: Fixed
Motivation of using dispatch receiver type when calculating method owner
was discussed here: https://github.com/JetBrains/kotlin/pull/3054
However, this is incompatible with type erasure of non-reified type
parameters on inlining (which will be done in future). Consider the
code:
```
inline fun <T> f(arr: Array<T>, p: (T) -> Int): Int = p(arr[0])
fun box() = f(arrayOf("abacaba"), String::length)
```
After inlining and erasure, the type of `arr[0]` is `Any`. Thus, when
calculating owner of `String::length` we would have `Any` instead of
`String` if we used dispatch receiver type.
Note, that this change affects bytecode instruction that invokes
method, but does not change which method is being invoked.
Affects IR Evaluator in IDEA.
The problem is happening because
`ReflectiveAccessLowering#fieldLocationAndReceiver` returns the class
the field was called on, not the class the field was declared in.
Then the class is used for obtaining a field by using reflection's
`getDeclaredField` to make the field accessible after.
But `getDeclaredField` doesn't work for field declared in a superclass,
hence the error.
#KT-65012 fixed
Merge-request: KT-MR-13919
Merged-by: Alexander Kuznetsov <Aleksander.Kuznetsov@jetbrains.com>
This is an addition to the f5bb477459
commit. Earlier, some of the validation checks were moved from
`stickyPostconditions` into separate lowering. But these
post-conditions checks were launched only when several flags were
set to true. In this commit, we just return to such behavior.
#KT-64116
Alexander Udalov