Dmitry Petrov
7a43c2de79
This avoids an extra call to 'analyze', which is rather costly. Update debugger testData: Constant condition elimination now performs DCE more consistently.
78 lines
1.3 KiB
Kotlin
Vendored
78 lines
1.3 KiB
Kotlin
Vendored
// FILE: test.kt
|
|
|
|
fun foo() {
|
|
try {
|
|
mightThrow()
|
|
} finally {
|
|
"FINALLY"
|
|
}
|
|
|
|
val t = try {
|
|
mightThrow2()
|
|
} finally {
|
|
"FINALLY2"
|
|
}
|
|
}
|
|
|
|
var throw1 = false
|
|
var throw2 = false
|
|
|
|
fun mightThrow() {
|
|
if (throw1) throw Exception()
|
|
}
|
|
|
|
fun mightThrow2() {
|
|
if (throw2) throw Exception()
|
|
}
|
|
|
|
fun box() {
|
|
foo()
|
|
throw2 = true
|
|
foo()
|
|
// Never gets here.
|
|
throw1 = true
|
|
foo()
|
|
}
|
|
|
|
// The JVM backend steps back to line 11 when leaving the
|
|
// `mightThrow2` call. The JVM_IR backend does not. The
|
|
// JVM_IR behavior is consistent with what happens for the
|
|
// try-finally where the value is discarded which seems good.
|
|
|
|
// LINENUMBERS
|
|
// test.kt:29 box
|
|
// test.kt:4 foo
|
|
// test.kt:5 foo
|
|
// test.kt:21 mightThrow
|
|
// test.kt:22 mightThrow
|
|
// test.kt:7 foo
|
|
// test.kt:8 foo
|
|
// test.kt:10 foo
|
|
// test.kt:11 foo
|
|
// test.kt:25 mightThrow2
|
|
// test.kt:26 mightThrow2
|
|
// LINENUMBERS JVM
|
|
// test.kt:11 foo
|
|
// LINENUMBERS
|
|
// test.kt:13 foo
|
|
// test.kt:14 foo
|
|
// LINENUMBERS JVM
|
|
// test.kt:13 foo
|
|
// LINENUMBERS JVM_IR
|
|
// test.kt:10 foo
|
|
// LINENUMBERS
|
|
// test.kt:15 foo
|
|
// test.kt:30 box
|
|
// test.kt:31 box
|
|
// test.kt:4 foo
|
|
// test.kt:5 foo
|
|
// test.kt:21 mightThrow
|
|
// test.kt:22 mightThrow
|
|
// test.kt:7 foo
|
|
// test.kt:8 foo
|
|
// test.kt:10 foo
|
|
// test.kt:11 foo
|
|
// test.kt:25 mightThrow2
|
|
// test.kt:14 foo
|
|
// test.kt:13 foo
|