Commit Graph

4941 Commits

Author SHA1 Message Date
Dmitry Petrov 17b4d4a973 Differentiate accessors by FieldAccessorKind
Otherwise accessors for backing fields (as in '{ field }') clash with
accessors for properties (as in '{ prop }').

 #KT-21258 Fixed Target versions 1.2.30
2017-12-18 16:15:52 +03:00
Alexander Udalov 3a807cb39b Support Void.TYPE as underlying Class object for KClass
#KT-20875 Fixed
2017-12-18 11:57:05 +01:00
Denis Vnukov 52ccd67ec1 Remove duplicate parameter null checks in JvmStatic delegate methods.
Remove unnecessary non-null parameter checks inside static delegate methods
created for @JvmStatic companion object methods. Allows function generation
strategy decide if such checks need to be injected.

 #KT-7188 Fixed
2017-12-14 13:48:50 +03:00
Dmitry Petrov 1aab4e643c Add overflow-related tests for 'reversed' 2017-12-14 10:41:51 +03:00
Dmitry Petrov 65b5cdbb8d Maintain bounds evaluation order in intrinsics for 'reversed'
Makes sense for 'rangeTo', 'downTo', and 'until' with non-const bounds.
2017-12-14 10:41:51 +03:00
Dmitry Petrov 9fa9a8748b Check that qualified const expressions are recognized properly in 'for' 2017-12-14 10:41:51 +03:00
Dmitry Petrov bf97b332cf Support const-bounded for loop generation for reversed 'until' 2017-12-14 10:41:51 +03:00
Dmitry Petrov 54cceac99b Intrinsics for 'reversed': until 2017-12-14 10:41:51 +03:00
Dmitry Petrov 5f7460a8c7 Support const-bound counter loop generation for 'downTo' 2017-12-14 10:41:51 +03:00
Dmitry Petrov a4c29b3587 Support Long and Char in const-bounded counter loop generation
If the loop end value is a compile-time constant (best we can do now),
and it is safe to iterate over a given range using "naive" for loop
(using '<=' or '>=' in loop condition),
generate such loops for Longs and Chars as well Ints (Bytes, Shorts).
2017-12-14 10:41:51 +03:00
Dmitry Petrov 455a1c0f53 Intrinsics for 'reversed': downTo
#KT-21323 In Progress
2017-12-14 10:41:51 +03:00
Dmitry Petrov 64ba811b7f Intrinsics for 'reversed': CharSequence.indices
#KT-21323 In Progress
2017-12-14 10:41:51 +03:00
Dmitry Petrov 7ba73c1635 Intrinsics for 'reversed': collection.indices
#KT-21323 In Progress
2017-12-14 10:41:51 +03:00
Dmitry Petrov 5bcbe25469 Intrinsics for 'reversed': array.indices
#KT-21323 In Progress
2017-12-14 10:41:51 +03:00
Dmitry Petrov beff4a1b92 Intrinsics for 'reversed': support non-literal range expressions
#KT-21323 In Progress
2017-12-14 10:41:51 +03:00
Dmitry Petrov 821843e13f Intrinsics for 'reversed': generate in-const-bound ranges as countable
#KT-21323 In Progress
2017-12-14 10:41:51 +03:00
Dmitry Petrov 1775f294f4 Intrinsics for 'reversed': infrastructure & primitive range support
#KT-21323 In Progress
2017-12-14 10:41:51 +03:00
Mikhael Bogdanov edefb45585 Copy sam wrappers during inline
#KT-21671 Fixed
2017-12-07 12:57:43 +01:00
Dmitry Petrov 3d473f608e Add more tests for for-in-array
#KT-21354 Fixed Target versions 1.2.20
 #KT-21321 Fixed Target versions 1.2.20
2017-12-05 15:45:20 +03:00
Dmitry Petrov e4ecc13e13 Generate for-in-array loops with proper semantics in 1.3+
In Kotlin 1.3+, assignment to the for-in-array loop range variable in
the loop body doesn't affect loop execution (as if it was a loop on an
array iterator, or some other container).

 #KT-21354 In Progress
 #KT-21321 In Progress
2017-12-05 15:45:20 +03:00
Dmitry Petrov 0db6e5c3f3 Fix SAM wrapper generation in secondary constructor calls
Should simply invoke 'checkSamCall' in corresponding visitor method to
check if call arguments should be wrapped.
2017-12-04 10:29:33 +03:00
Dmitry Petrov 3431a1445c Fix enum entry class delegating constructor call generation
When the enum entry requires a specific class, its constructor should
invoke proper supertype constructor (from the corresponding enum class).
Corresponding resolved call should be passed from the front-end in
CONSTRUCTOR_RESOLVED_DELEGATION_CALL slice.
In case of enum entries without explicit supertype initializer, this
information was missing.
2017-12-04 10:29:33 +03:00
Dmitry Petrov 5b8c0d4c3e Optimize for-in-string loops
For-in-string loop can be generated using specialized 'length' and
'charAt' method calls, and with cached string length.
Note that update of the string variable in loop body doesn't affect
loop execution semantics.

 #KT-21322 Fixed Target versions 1.2.20
2017-11-29 10:15:38 +03:00
Dmitry Petrov e2fa613b70 Cache array length in for-in-array loop if possible
If the range expression is not a local variable (which can be updated in
the loop body affecting loop behavior, see KT-21354), we can cache the
array length, thus turning a for-in-array loop into a simple optimizable
counter loop.

 #KT-21321 In Progress
2017-11-29 10:15:32 +03:00
Alexander Udalov 386a3fb5ce Fix tests after f4f5359725 2017-11-27 16:03:38 +01:00
Dmitry Petrov f586bd4a34 Generate proper visibility for companion object instance field in LV1.3+ 2017-11-27 17:15:16 +03:00
Dmitry Petrov 70d3e6592d Unwrap object member imported by name before determining receivers
Existing code for receiver generation accidentally worked in most cases
for object members imported by name. However, it generated strange
bytecode (such as
    GETFIELD AnObject.INSTANCE
    GETFIELD AnObject.INSTANCE
    POP
), and worked incorrectly for augmented assignments.

 #KT-21343 Fixed Target versions 1.2.20
2017-11-27 17:15:16 +03:00
Alexey Andreev 71b1591044 JS: replace suspend inline metadata after inlining
This fixes some issues on coroutine inlining, see tests
2017-11-27 17:01:18 +03:00
Alexey Andreev f8e7861ce6 JS: add partial tail-call optimization for suspend functions 2017-11-27 17:01:17 +03:00
Alexey Andreev 26843509c7 JS: disable suspend function tail-call optimization based on FE data
Partially fixes KT-21026
2017-11-27 17:01:16 +03:00
Alexey Andreev 6ab87ad66a Fix failing JS tests 2017-11-27 17:01:15 +03:00
Alexander Udalov 938fd1a57e Use ResolvedCall for callable reference in KCallableNameProperty intrinsic
Instead of manually inspecting the DOUBLE_COLON_LHS slice, which is a
bit more error-prone. Note that new tests were passing before this
change
2017-11-27 12:46:56 +01:00
Alexander Udalov f4f5359725 Fix exception on inlining callable reference with implicit this in LHS
Use ResolvedCall to determine the receiver type in the JVM codegen,
instead of manually inspecting the PSI

 #KT-20821 Fixed
2017-11-27 12:46:55 +01:00
Alexey Andreev ec8adfe7c4 JS: fix referencing outer class from secondary constructor
See KT-21041
2017-11-21 12:22:10 +03:00
Alexey Andreev 32a0221474 JS: refactor code that copies default methods in interfaces to classes
Als fixes KT-21245
2017-11-21 12:18:16 +03:00
Alexey Andreev a9548b1224 JS: fix translation of safe calls to suspend unit functions
See KT-21317
2017-11-20 19:19:24 +03:00
Ilmir Usmanov 0fbbe10143 Support suspendCoroutineUninterceptedOrReturn intrinsic
KT-17336: Fixed
2017-11-20 18:12:41 +03:00
Mikhael Bogdanov a547019ed0 Switch DEFAULT_METHOD_CALL_FROM_JAVA6_TARGET according to LL 2017-11-17 13:48:44 +01:00
Alexey Andreev 89db4dfe79 JS: translate when against enum to JsSwitch when possible 2017-11-17 11:07:41 +03:00
Alexey Andreev 678b4c67c1 JS: translate when to JsSwitch when possible
See KT-21160
2017-11-17 11:07:38 +03:00
Alexey Andreev 21de76f88d JS: fix identifier generation in delegate constructor calls
See KT-21093
2017-11-17 11:07:34 +03:00
Dmitry Petrov 4193fae9fa Fix floating point comparison generation for range literals 2017-11-16 10:54:25 +03:00
Dmitry Petrov 354d54aef6 Don't generate DUPX instructions for in-range-literal expressions
Store argument into a local variable instead.
2017-11-16 10:54:25 +03:00
Ilmir Usmanov f4e180556c Fix assignment codegen for suspend operators plus and plusAssign
KT-16079: Fixed
2017-11-14 14:49:26 +03:00
Dmitry Petrov 15ac471626 Treat accessors to JvmStatic methods as having no dispatch receiver
#KT-21246 Fixed
2017-11-14 09:59:16 +03:00
Dmitry Petrov 4365084645 Lookup for local variables taking into account uninitialized this
Consider a context with uninitialized this, e.g.:

  fun foo() {
    val x = "..."
    class Local(y: String) : Base(L@{ x + y })
  }

Lambda 'L' is an argument of a super class constructor call.
Here 'this@Local' is not initialized yet. Thus local variables captured
in 'Local' can't be used. Instead, they should be captured by lambda 'L'
itself.

Note that lambda 'L' sees both 'x' and 'y' as local variables that
should be captured.

When in context with uninitialized this (generating arguments for super
type constructor or delegating constructor call), and a variable in
question is not found in the current context, use enclosing local lookup
to determine whether a local variable should be captured by a closure.
2017-11-14 09:33:28 +03:00
Dmitry Petrov bbb0389c51 Fix determining enclosing class for closure
Enclosing class for closure is a class whose instance is captured by
closure as an outer 'this', and stored in a field 'this$0'.
Usually enclosing class for closure is an immediate outer class,
including classes for nested closures. For example:

  class C {
    fun foo() {}
    val example1 = L1@ { foo() }
    // Enclosing class for lambda 'L1' is 'C'
    val example2 = L2a@ { L2b@ { foo() } }
    // Enclosing class for nested lambda 'L2b'
    // is a closure class for outer lambda 'L2a'
  }

However, if the closure is created in a super type constructor call for
the outer class, corresponding instance is considered "uninitialized",
and can't be used as a proper class instance, and can't be referenced:
corresponding code is rejected by front-end.

  class Outer {
    fun foo() {}
    inner class Inner : Base(L3@ { foo() })
    // Enclosing class for lambda 'L3' is 'Outer',
    // because 'Inner' is uninitialized in super type constructor call.
  }

In CodegenAnnotatingVisitor, we maintain a stack of currently
uninitialized classes, and chose enclosing class for closure
as an inner-most surrounding class with initialized instance.

When generating code for this or outer class instance, we skip
contexts corresponding to classes with uninitialized instances.

This fixes a number of bytecode verification errors caused by incorrect
enclosing class for closure.

 #KT-4174 Fixed Target versions 1.2.20
 #KT-13454 Fixed Target versions 1.2.20
 #KT-14148 Fixed Target versions 1.2.20
2017-11-14 09:33:28 +03:00
Mikhael Bogdanov dff6e943bb Always do stack spilling during inline cause of dex problem
Dex issue: https://issuetracker.google.com/issues/68796377

  #KT-20844 Fixed
2017-11-13 16:50:24 +01:00
Ilmir Usmanov 0f9a21d429 Add regression test for KT-17640 2017-11-10 19:16:06 +03:00
Mikhael Bogdanov 416392bb74 Don't propagate reified markers for special enum functions
#KT-18254 Fixed
2017-11-08 15:35:44 +01:00