Lateinit local vars are guaranteed to be non-null after store.
So we mark such stores as storing non-null value
(could be useful for some other constructs, too),
and optimize null checks accordingly.
It's necessary for generic inline suspend as a codegen
for it uses binding slice SUSPEND_FUNCTION_TO_JVM_VIEW
to generate fake continuation parameter, so all the
descriptors that are used for body generation must be
obtained from the SUSPEND_FUNCTION_TO_JVM_VIEW
#KT-19528 Fixed
Otherwise, when completing all the unsuccessfull candiates,
resolution of each lambda-arguments starts repeatedly for each candidate
that leads to exponential time
NB: Changes in `completeArguments` are necessary because otherwise
nested lambdas will be analyzed twice:
once for the main resolved call, and then for all candidates
that again leads to exponential complexity
#KT-16672 Fixed
#KT-19457 Fixed
Note that the quick fix to implement header class works incorrectly when
that class has nested classes at the moment; this should be fixed
separately
#KT-15494 Fixed
#KT-18573 Fixed
Also support a quick fix to add 'impl' modifier (KT-18454), although it
doesn't work yet on classes because there's no error on them in the IDE
#KT-18087 Fixed
#KT-18452 Fixed
#KT-18454
Try to report most of the errors on the actual members of the impl
class. In many cases, there's a 1:1 mapping of header to impl class
members, so the error "some members are not implemented" on the class
declaration itself is redundant. Exceptions include functions/properties
from supertypes (there may be no other place to report a signature
mismatch error in this case), functions/properties not marked with
'impl' (the checker is only run for declarations explicitly marked with
'impl') and default constructors (the checker is not run for them)
#KT-18447 Fixed
Try to report most mismatch errors on the 'impl' declaration. Only
report a mismatch error on the 'header' declaration if no error would be
otherwise reported on any 'impl' declaration in the compilation unit.
Also render declaration kind in the message
#KT-18447 In Progress
In short, some of the bytecode analyzers assume that there could be
no stores instructions into parameter vars with value of different
types (even when the value type is a subtype)
See the issue for details
#KT-19713 Fixed
So (for the most often reproduction case) #KT-19433 Fixed
Before this commit, internal names for nested classes were written as test/Foo/Bar (comparing to test/Foo$Bar in the normal mode), as getting qualified names from such internal names was trivial. But, because of IC, we needed to write class files to the disk, so our decompiler could find such "broken" classes and read it in a wrong way.
Dmitry Petrov