Commit Graph

11 Commits

Author SHA1 Message Date
Svyatoslav Kuzmich fdd7fa5aea [Wasm] Mute codegen box tests 2020-11-09 16:04:43 +03:00
Mikhail Glukhikh f3f7bf70f6 [FIR] Set delegate field receiver properly 2020-04-03 13:29:55 +03:00
Mark Punzalan 9df2f69f09 [FIR] Disable failing blackbox codegen tests for FIR. 2019-11-19 11:00:09 +03:00
pyos 8c55376f0c Unmute almost all JVM_IR tests that use property references 2019-03-19 12:00:29 +01:00
Mikhael Bogdanov e149cbe852 Mute failed jvm ir tests 2018-06-28 12:26:41 +02:00
Yan Zhulanow 9d1af5a17e Fix tests: "infix modifier required" and "operator modifier required" errors 2015-11-27 15:51:11 +03:00
Alexander Udalov 51bf68ce27 Update compiler tests to use KProperty instead of PropertyMetadata 2015-10-14 20:45:54 +03:00
Yan Zhulanow 1f2b4e20fe Replace get() and set() to getValue() and setValue() (property delegates) 2015-10-09 22:49:16 +03:00
Dmitry Jemerov 4bdf598bfe compiler testdata: s/trait/interface 2015-05-12 19:43:17 +02:00
Stanislav Erokhin 65c9ea2465 Fix variance problems. 2014-11-17 17:45:05 +03:00
Alexander Udalov 7041a10ecf Fix VerifyError on primitive override properties
Don't just blindly map property type in PropertyCodegen when generating getter:
we already have its full signature, so we can just as well take the type from
it. The former also isn't correct for properties which are overrides with a
primitive type. Simple mapType is safe for setter though, because the type of a
method parameter cannot change with override on JVM

The change in StackValue relates to call sites of properties which return Unit:
in that case StackValue's this.type is V, whereas the return type of its getter
is Ljet/Unit; -- so coerceTo was coercing from the wrong type in case of getter
calls

 #KT-4373 Fixed
 #KT-4383 Fixed
2014-01-14 17:52:27 +04:00