It's no longer needed since we going to start building libraries using the new BE,
so we have to be sure that everything works well in releases branches too.
Previously JS IR versions of stdlib and kotlin-test were build
by default using compiler previously built on a buildserver.
It had some issues:
- This required us to advance bootstrap every time we made any
incompatible IR changes. This happens often since IR ABI is
not quite stable yet.
- We never tested the exact combination of compiler and stdlib we publish
We tested:
- new compiler with new stdlib build by new compiler (in box tests)
- old compiler with new stdlib build by old compiler (in stdlib tests)
We published:
- new compiler with new stdlib build by old compiler
After this change JS IR compiler tests, builds and publishes
single configuration:
new compiler with new stdlib build by new compiler
JS IR stdlib and kotlin-test are now built using JavaExec of CLI instead
of Gradle plugin to avoid troubles of loading a freshly built plugin.
This also allows to have a granular dependencies: we don't rebuild klib
if we changed a lowering in a compiler backend, but we do rebuild it if
we changed IR serialization algorithm.
if it overrides functions with another return type.
Otherwise, we cannot determine on call site that the function returns Unit
and cannot { POP, PUSH Unit } in order to avoid the situation when callee's
continuation resumes with non-unit result. The observed behavior is that
suspend function, which should return Unit, suddenly returns other value.
#KT-35262: Fixed
Consider `fun <E : I> foo(a: Any?) = a as? E`, where I is an interface.
This check used to fail, because the `a == null` was missing, and
the `isInterface` stdlib method crashes if the first argument
is null. This change adds the null check.
Also this change prettifies the instance check in case of type parameter
left operand.
Since property accessor descriptors (unlike corresponding IR elements)
do not have type parameters, we need to take them from the corresponding
property to ensure the correct IR for delegated property accessors.
Alternatively, we could improve the lookup utilities and their usages to
always find the exact override of a symbol from
Collection/Iterable/CharSequence/etc, but since we need to load the
original symbol anyway in cases when the loop subject's type is a type
parameter, we might as well simplify everything and always reference the
original symbol.
Also improve exception message and removed unused declarations in
IrBackendUtils.kt.
For example, synthetic `$annotations` methods for properties were
previously mangled to `$annotations-...`, which breaks annotation
loader, and fails an assert in ClassCodegen.generateMethod.
Normally, the fact that is was Unit was not visible as enum constructors
are lowered to normal class constructors anyway. The exception is when
the arguments are reordered, causing the incorrect return type to leak
into the block that holds temporary variables.
In case of inline it should be same descriptor (except of fake override), In general case getter could be synthetic accessor and in such case it's not inline
In old BE we generate `create` for this kind of suspend lambdas, which,
like in simple suspend lambdas is responsible for putting arguments
(including extension receiver) to fields.
But, if number of parameters of suspend lambda (including receiver) >= 1,
there is no need to generate `create`, since `create` is called only by
`createCoroutine` and friends from stdlib, and there is no version of
`createCoroutine` for lambdas with multiple parameters.
Thus, in old BE we generate a redundant method, which affects method
count.
In JVM_BE we decided to 'inline' create into `invoke` for suspend lambdas
with multiple parameters.
Fix getting kType metadata in cases when corresponding jsClass value
is passed through temporary variable.
This can happen when jsClass expression is not consided trivial by
local variable optimizer. This happens for object declarations from
different modules, for example kotlin.Unit
Updated test checks, that large unsigned numbers are converted to corresponding
negative signed numbers properly. Using unsinged constants instead of signed
in test allows to remove supressed OI error as well as use test with NI.
It allows to fold these blocks and navigate to them in the IDEA.
Also, this commit adds compiler configuration keys to control generating comments with paths for file blocks,
and now generating such comments is enabled only in our tests.
Mikhael Bogdanov