On Wed, 8 May 2024 14:53:05 GMT, Claes Redestad <redes...@openjdk.org> wrote:
> A rather large startup regression was introduced in 23-b13 from > [JDK-8309622](https://bugs.openjdk.org/browse/JDK-8309622). Some of that has > been dealt with as enhancements such as > [JDK-8330802](https://bugs.openjdk.org/browse/JDK-8330802), > [JDK-8330595](https://bugs.openjdk.org/browse/JDK-8330595) and > [JDK-8330681](https://bugs.openjdk.org/browse/JDK-8330681), which provide > both point fixes and reduce initialization overhead of certain constructs > more generally. The remaining issues stem from a set of lambdas added in code > for `java.util.Locale` and `jdk.internal.util.BaseLocale` causing early > bootstrapping of the lambda infrastructure and a bit of class generation. > > While the remaining overheads are relatively small and borderline acceptable > (< 2-3ms), I think it's still worth acting on them in this particular case > since the amount of added bootstrapping overhead is dependent on which locale > the system runs under, which complicates testing and comparisons due to > relatively large differences in paths taken on different systems. src/java.base/share/classes/java/util/Locale.java line 1001: > 999: = ReferencedKeyMap.create(true, > ReferencedKeyMap.concurrentHashMapSupplier()); > 1000: > 1001: private static final LocaleKey LOCALE_LOOKUP = new LocaleKey(); Reusing an existing class with a special instance for functional interface implementation is weird, I still wonder if we can do the local class + singleton (essentially same as non-capturing lambda) approach. src/java.base/share/classes/sun/util/locale/BaseLocale.java line 168: > 166: return CACHE.intern(new BaseLocale(language, script, region, > variant), > 167: // Avoid lambdas since this may be on the bootstrap path > in many locales > 168: new UnaryOperator<BaseLocale>() { Since this intern is called many times, can we convert this `new` to a local class in a static method and reuse its singleton? ------------- PR Review Comment: https://git.openjdk.org/jdk/pull/19140#discussion_r1594201610 PR Review Comment: https://git.openjdk.org/jdk/pull/19140#discussion_r1594199711
