- Materialize unit when its value is actually needed.
- Special-case Unit_getInstance return type at codegen. It should be a
proper Unit object instead of a "void"
Recursive results from checkForPrimitiveOrPattern were ignored. If it found a case/condition that could not be optimized the resulting "false" was not propagated. This would lead to a "optimized" when without all conditions.
- see KT-45738
- The return is now lifted out of the when to make it more obvious what is going on.
- New test for mixed multiple conditions in when
Before this commit we considered !isOverride as a sign that
function / field / accessor has no overridden symbols.
However, it's false for deserialized, because isOverride
is always false there.
This commit fixes 68 BB tests but breaks 25 BB tests (not yet muted)
Without this commit, JVM name mapping logic in BE does not work for FIR,
because FIR cannot use old BuiltInsPackageFragmentImpl descriptor.
In this commit we add our own implementation thus fixing
a pack of FIR black box tests.
See KT-36812. Aside from the problem stated there, D8 will throw out the
entire LVT if it sees a variable that has not been written to (and will
generate incorrect SSA if the slot is reused with a different type).
Note: this only fixes a FIR test because it's missing an `else -> throw`
branch, and default initialization satisfies the verifier and masks the
incorrect control flow.
Usually FIR enum entry is initialized by anonymous object,
which is the container for all enum entry' declarations.
However, for simple enum entries there is no need of initializer at all.
Library methods such as 'listOf' are resolved
to have the package fragments as their parents,
but JVM expects their containing file classes as parents.
This fix generates those file classes and
uses them as parent replacements for such library methods.
* fixed NoSuchMethod caused by mismatched signatures of the "invoke" method generated for lambda arguments
* added test cases in invoke.kt for KFunction and anonymous functions
* added a transformer to wrap the last expression in the bodies of lambdas with return
If "null" was the first entry in an optimizable "when" over enum,
mapRuntimeEnumEntry was called before mapConstEnumEntry, and the
$WhenMappings field was not created. Now both mapConstEnumEntry and
mapRuntimeEnumEntry create this field on the first access
* In blocks, discard the result of any statement that has a return
type other than void. This was previously done by wrapping each
statement into an "implicit Unit conversion" that was actually
compiled down to a stack pop instead. If an expression happened to
already have type Unit, however, such a conversion was not inserted,
resulting in a stray reference on the stack. These conversions are
now redundant and should probably be removed.
* In assignments and non-exhaustive conditionals, materialize a Unit
on the stack to avoid depth mismatches that trip up the bytecode
validator. Because such expressions are generally used at block level
(and, indeed, the frontend will reject a non-exhaustive conditional
used as an expression), combined with the above change this results
in no additional GETSTATIC opcodes, as they are immediately removed
by the peephole optimizer.
Effectively, the following when structure:
when (s) {
s1, s2 -> e1,
s3 -> e2,
s4 -> e3,
...
else -> e
}
is implemented as:
when (s.hashCode()) {
h1 -> {
if (s == s1)
e1
else if (s == s2)
e1
else if (s == s3)
e2
else
e
}
h2 -> if (s == s3) e2 else e,
...
else -> e
}
where s1.hashCode() == s2.hashCode() == s3.hashCode() == h1,
s4.hashCode() == h2.
A tableswitch or lookupswitch is used for the hash code lookup.
Change-Id: I087bf623dbb4a41d3cc64399a1b42342a50757a6
Code such as
```
val b = getBoolean()
if (b) 4
else if (b) 5
```
didn't generate a value on the stack always and therefore would
have control-flow paths leading to a pop instruction with nothing
on the stack.
Change-Id: I09d059f361e56a41880006e3f4e51e9acdbd167d
Sergej Jaskiewicz