RFR: 8319220: Pattern matching switch with a lot of cases is unduly slow

[Chen Liang]

On Fri, 3 Nov 2023 08:41:12 GMT, Jan Lahoda <jlahoda at openjdk.org> wrote:

> Consider code like:
> 
> void test(Object o) {
>     switch (o) {
>         case X1 -> {}
>         case X2 -> {}
> ...(about 100 cases)
> ``` 
> 
> javac will compile the switch into a switch whose selector is an indy invocation to `SwitchBootstraps.typeSwitch`, with static arguments being the types in the cases.
> 
> `SwitchBootstraps.typeSwitch` will then create a chain of `MethodHandle`s performing `instanceof` checks between the switch's selector and the given case type. The problem is that when the number of cases is high enough, (more than ~40-50), the chain gets too long, and the tests won't inline anymore. This then leads to a very bad performance, when compared to manually written if-instanceof-else-if-instanceof- chain.
> 
> The proposal herein is to use bytecode (written using the ClassFile API/library) instead of the `MethodHandle`s chain. The overall performance of this seems to be similar to the manually written if-instanceof-else-if-instanceof- chain.
> 
> Using the benchmark from the bug, and this patch, I am getting:
> 
> MyBenchmark.testIfElse100  thrpt    5  521826.326 ± 7510.042  ops/s
> MyBenchmark.testSwitch100  thrpt    5  505440.170 ± 3757.178  ops/s
> 
> 
> The most tricky part of this new way to generate the tests is handling of non-type case labels, and in particular cases with enum constant labels. The resolution of enum constants is deferred as much as possible, by using an indirection through the `ResolvedEnumLabels`.
> 
> Further improvements may be possible, esp. for some specific cases (like all cases having a type, and the type being a final class).

src/java.base/share/classes/java/lang/runtime/SwitchBootstraps.java line 338:

> 336: 
> 337:         @Override
> 338:         public Object apply(Integer labelIndex, Object value) {

Since we already know the EnumDesc (i.e. `result`), we can just convert this to a `BiPredicate<Integer, Object>` for simplicity.

src/java.base/share/classes/java/lang/runtime/SwitchBootstraps.java line 346:

> 344:                     Class<?> clazz = label.constantType().resolveConstantDesc(lookup);
> 345: 
> 346:                     if (value.getClass() != clazz) {

Not related to this patch, but this appears to be wrong: we should do something like

if (!(value instanceof Enum<?> ev) || ev.getDeclaringClass() != clazz)
    return SENTINEL; // or false

src/java.base/share/classes/java/lang/runtime/SwitchBootstraps.java line 383:

> 381:               .withMethod("typeSwitch", MethodTypeDesc.ofDescriptor("(Ljava/lang/Object;ILjava/util/function/BiFunction;)I"), Classfile.ACC_STATIC, mb -> {
> 382:                   mb.withFlags(AccessFlag.PUBLIC, AccessFlag.FINAL, AccessFlag.STATIC)
> 383:                     .withCode(cb -> {

Can just use `clb.withMethodBody` and pass the access flags with `Classfile.ACC_PUBLIC | Classfile.ACC_STATIC | Classfile.ACC_FINAL`

src/java.base/share/classes/java/lang/runtime/SwitchBootstraps.java line 432:

> 430:                                 cb.aload(2);
> 431:                                 cb.constantInstruction(enumIdx);
> 432:                                 cb.invokestatic(Integer.class.describeConstable().get(),

You can use `ConstantDescs.CD_Integer` etc.

src/java.base/share/classes/java/lang/runtime/SwitchBootstraps.java line 434:

> 432:                                 cb.invokestatic(Integer.class.describeConstable().get(),
> 433:                                                 "valueOf",
> 434:                                                 MethodType.methodType(Integer.class,

I recommend `MethodTypeDesc.of()` instead.

-------------

PR Review Comment: https://git.openjdk.org/jdk/pull/16489#discussion_r1381409965
PR Review Comment: https://git.openjdk.org/jdk/pull/16489#discussion_r1381411165
PR Review Comment: https://git.openjdk.org/jdk/pull/16489#discussion_r1381384302
PR Review Comment: https://git.openjdk.org/jdk/pull/16489#discussion_r1381388594
PR Review Comment: https://git.openjdk.org/jdk/pull/16489#discussion_r1381391495