Commit Graph

92 Commits

Author SHA1 Message Date
Dmitry Petrov bbbca9fbbf Minor: mute inline classes tests in JVM_IR 2018-09-10 10:42:18 +03:00
Roman Elizarov e2713501ce Rename SuccessOrFailure to Result and hide Failure from ABI
* The members of Result are isSuccess, isFailure, exceptionOrNull, getOrNull
* The rest of API is implemented via inline-only extensions
* There are two internal functions to hide detailed mechanics of an internal
  Result.Failure class: createFailure and throwOnFailure
* Result.toString is explicit: either Success(v) or Failure(x)

See KT-26538
2018-09-09 11:34:31 +03:00
Dmitry Petrov 006c0aa740 Hide constructors accepting inline class parameters 2018-09-07 15:57:59 +03:00
Dmitry Petrov cafaa3e13c Mangle inline class members
<IMPL_SUFFIX> for method is a method signature hash,
if method value parameter types contain inline class types,
otherwise 'impl'.

Constructor methods are named as 'constructor-<IMPL_SUFFIX>'.

Synthesized 'box' and 'unbox' methods are named as
'<METHOD_NAME>-<IMPL_SUFFIX>'.

Erased implementations of overriding and non-overriding methods
are named as '<METHOD_NAME>-<IMPL_SUFFIX>'.

Fully specialized implementation of 'equals' will have a special suffix.
2018-09-07 10:25:53 +03:00
Dmitry Petrov aec173ac5c Minor: fix tests after rebase 2018-09-07 09:32:11 +03:00
Dmitry Petrov 427295525a Minor: mute test in JVM_IR 2018-09-07 09:31:33 +03:00
Dmitry Petrov ae0d980b54 Minor: mute some inline classes tests in JS_IR 2018-09-07 09:29:55 +03:00
Mikhail Zarechenskiy 690762d46c Inline classes ABI: rename equals--impl to equals-impl0 2018-09-06 17:01:01 +03:00
Mikhail Zarechenskiy 83975bd6ac Generate stub for specialized equals inside inline class 2018-09-06 10:39:33 +03:00
Mikhail Zarechenskiy b4674a172e Reserve secondary constructors with bodies inside inline classes
#KT-26575 Fixed
2018-09-06 10:22:01 +03:00
Mikhail Zarechenskiy 002a66fec1 Reserve box/unbox/equals/hashCode methods inside inline classes
#KT-26573 Fixed
2018-09-06 10:20:03 +03:00
Dmitry Petrov a2900282fd Call factory method for primary constructors of inner classes
We might want to add 'init' blocks later, so now, for the sake of
binary compatibility with 1.3-RC binaries, we have to generate these
'constructor' calls.

Note that in some tests inline class boxing is no longer redundant,
because resulting value is passed to 'constructor' as an argument.
2018-09-05 12:20:57 +03:00
Dmitry Petrov 05daa21657 Fix inline class secondary constructor call generation
Don't generate NEW+DUP for inline class constructor calls.
2018-09-05 12:20:57 +03:00
Dmitry Petrov 3080b65f7d Inline (wrapper) class IC extends erased inline class IC$Erased 2018-09-05 12:20:57 +03:00
Roman Artemev 8a871b3f0c Update tests 2018-08-31 15:34:18 +03:00
Ilmir Usmanov 2d04acba42 Update test data of compiler tests, except IR tests 2018-08-30 16:24:41 +03:00
Dmitry Petrov a205019156 Additional tests and fixes for function name mangling 2018-08-30 14:58:54 +03:00
Dmitry Petrov 99498eb7b8 Use 'name-hash' mangling scheme
'-' is allowed as a name character both in JVM and in Dalvik, but can't
be a part of a Java identifier.
2018-08-30 14:58:52 +03:00
Dmitry Petrov a56d1d3ce8 Mangle function names with inline class parameters
Avoid name clashes in cases such as

  inline class Login(val login: String)
  inline class Password(val password: String)

  fun validate(login: Login) { ... }
  fun validate(password: Password) { ... }
2018-08-30 14:58:50 +03:00
Anton Bannykh a83baee67b review fixes + DCE data update + test (un)muting 2018-08-30 14:57:27 +03:00
Anton Bannykh 944c5b6044 JS: modify a test to work around Kotlin/JS Double.toString problems 2018-08-30 14:57:18 +03:00
Anton Bannykh a11c4871ea JS: equals, hashCode, toString support in inline classes 2018-08-30 14:57:16 +03:00
Dmitry Petrov 2a524920a5 Don't remap inline function args requiring inline class boxing/unboxing
Same as for primitives: inline lambda expects to see a boxed value,
so, even if an argument is a local variable, it can't be remapped,
because it contains unboxed representation.
2018-08-24 14:52:29 +03:00
Dmitry Petrov 6e2d05cd94 Fix argument original type order in InlineCodegen
Arguments are put on stack in the direct order, and then stored into
local variables for inlining in the reversed order:

    <arg0>
    <arg1>
    <arg2>
    store <param2>
    store <param1>
    store <param0>

Original value parameter types were taken in direct order, though.
2018-08-24 14:52:29 +03:00
Dmitry Petrov 7d4dfc87b1 Use proper KotlinType in get/set methods for property reference 2018-08-21 08:43:12 +03:00
Dmitry Petrov b6e3218ca2 Use mapping mode for inline class underlying type without wrapping 2018-08-16 15:03:50 +03:00
Dmitry Petrov 8e95ecb821 Use underlying type when computing type mapping mode for inline classes 2018-08-15 15:37:46 +03:00
Svyatoslav Kuzmich 392ad521fd [JS IR BE] Reflection support 2018-08-15 13:35:14 +03:00
Dmitry Petrov ebf8ec455d Box/unbox nullable inline class values with null check
When we have a nullable inline class value with non-null underlying
type, corresponding value in unboxed representation is nullable. E.g.:

  inline class Str(val value: String)

  fun test(s: Str?) = listOf(s)

Here 'test(s: Str?)' accepts nullable 'java.lang.String' as a parameter.

When boxing/unboxing nullable values of such inline classes, take care
of nulls.

 #KT-26052 Fixed Target versions 1.3-M2
2018-08-14 10:22:07 +03:00
Dmitry Petrov 4f6aa50417 Fix == for inline classes with boxes
TODO generalize code generating object vs primitive equality

 #KT-25914 Fixed
 #KT-25981 Fixed
 #KT-25983 Fixed
2018-08-10 10:34:57 +03:00
Mikhail Zarechenskiy 55e7def3c1 Fix coercion for stack value property in case of inline classes
#KT-25771 Fixed
2018-08-06 17:45:20 +03:00
Mikhail Zarechenskiy 6d4d244c28 Generate function from Any for inline classes same as for data classes
#KT-24873 Fixed
 #KT-25293 Fixed
 #KT-25299 Fixed
2018-08-06 10:56:01 +03:00
Mikhail Zarechenskiy 043ce1cb27 Support secondary constructors for inline classes
#KT-25614 Fixed
 #KT-25246 Fixed

 KT-25599 Will be fixed after recompilation of unsigned classes
2018-08-06 10:55:57 +03:00
Mikhael Bogdanov 2884d728fd Mute/unmute jvm_ir tests 2018-08-01 12:29:24 +02:00
Mikhail Zarechenskiy 045058c29a Support inline class values inside string templates through boxing
#KT-25626 Fixed
 #KT-25613 Open
2018-07-20 13:58:31 +03:00
Mikhail Zarechenskiy ec9d8e580e Leave boxing for compareTo/areEqual methods for inline classes
Inline classes can override methods and thus introduce side effects
2018-07-20 11:51:34 +03:00
Mikhail Zarechenskiy bb575866cb Pass actual invoke parameters for inline functions with inline classes
#KT-25511 Fixed
2018-07-17 23:39:38 +03:00
Mikhail Zarechenskiy 77959247d2 Fix bridge generation for inline classes over Any type 2018-07-17 23:39:34 +03:00
Mikhail Zarechenskiy 820d168607 First, check for inline class type before boxing
The problem was that if `type` is of primitive type, but `KotlinType` is
 actually an inline class type, then anyway we boxed this type as primitive
2018-07-13 15:48:21 +03:00
Mikhail Zarechenskiy 4e3674b330 Fix for-in iterator over list of boxed inline class values
#KT-25325 Fixed
2018-07-12 18:53:30 +03:00
Mikhail Zarechenskiy 0308e10c11 Fix for-in iterator over array of boxed inline class values
#KT-25324 Fixed
2018-07-12 18:53:24 +03:00
Anton Bannykh 9c6452778f JS IR: unmute tests 2018-07-11 15:25:41 +03:00
Anton Bannykh ac1a97ad83 JS IR: unmute tests 2018-07-11 14:49:36 +03:00
Mikhail Zarechenskiy 6b73e528ce Ignore test for JS IR because of TODO() function 2018-07-11 11:41:24 +03:00
Mikhail Zarechenskiy 8d24ca65a3 Propagate KotlinType into when expression codegen
This commit removes unneeded boxing when result expression of `when` is
 value of inline class type
2018-07-10 23:10:57 +03:00
Anton Bannykh 07b3b66fd9 JS IR: unmute tests 2018-07-10 13:34:19 +03:00
Mikhail Zarechenskiy 2939d9c8c6 Use correct lhs KotlinType & AsmType for is check generation
There were two problems:
  - For asm type `OBJECT_TYPE` was used, which can be wrong in case of
  inline classes, because it can be an underlying value of some inline class
  - For KotlinType, type from rhs was used
2018-07-05 12:55:43 +03:00
Mikhail Zarechenskiy 4f490ac264 Fix generation of checkcast instructions for inline classes
Normalize LHS and RHS types to use only boxed ones and then let bytecode
 optmizer reduce redundant boxing
2018-07-05 12:55:39 +03:00
Mikhael Bogdanov e149cbe852 Mute failed jvm ir tests 2018-06-28 12:26:41 +02:00
Mikhail Zarechenskiy fcacdc1fc5 Fix bridge methods generation when inline class types are used 2018-06-19 20:09:35 +03:00