See KT-48291 for example.
The rationale behind the change is that in "findExpectForActual" we want
to find a candidate to be a counterpart for the current (actual)
declaration. Fake overrides can not be a sensible counterpart in that
regard (you can't, nor you should, provide actual for expect which is
FAKE_OVERRIDE)
^KT-48291 Fixed
`ValueParameterData.hasDefaultValue` is now always equal to the original
value parameter's `declaresDefaultValue` flag, so `ValueParameterData`
usages can be simplified to just `KotlinType`.
The change in signatureEnhancement.kt in 432f581cb2 was incorrect.
Contrary to its name, the removed method `hasDefaultValueInAnnotation`
checked not only if the enhancement annotation has default value (which
is what the removed feature was about), but also if the parameter itself
declares default value. This was mistakenly substituted by just `false`
on line 234. The correct change is to use the `declaresDefaultValue`
flag of the original parameter.
It's kind of weird though that in case there's a nullability annotation
on the whole package (like in KT-48316) type enhancement is being done
on everything, including annotation constructors, whose parameter types
can't have any enhancement information. Maybe this should be improved
independently.
#KT-48316 Fixed
Shared Native Compilations seem to already implicitly add a
dependency on the stdlib. Adding it again will result in a
'library included more than once' warning.
^KT-46257 Verification Pending
For example, if a class has default constructor (`class Foo`),
both KtClassSymbol and KtConstructorSymbol will be pointing to the same
PSI - thus creating effectively the same `KtSymbolPointer`. Later
it will be impossible to deduce which symbol we had in mind
Currently, `Show Parameters Info` works incorrectly for
such generated declarations because of that - it throws CCE while
trying to cast class symbol to function (constructor) symbol
In 1.5.20 HostManager.host started to check host architecture and it
leads to regressing in case of os+arch combination that is not supported
by HostManager.
Before 1.5.20 HostManager.host was determined only by OS. It is
not correct, but Gradle plugin uses HostManager even if there are
no Native targets. We fallback to this behavior to workaround this
regression.
We can't apply "reuse loop variable as index variable" transformation
before local declarations lowering, otherwise it will affect captured
loop variable behavior, resulting in KT-48626.
Since it's JVM-specific, move it to JvmOptimizationLowering.
Elena Lepilkina