Performance dropped significantly when switching from Nested Patterns to Flattened Patterns (was "Sharing my experience with Switch Expressions")

[Angelos Bimpoudis]

Hello David,

Many thanks for the experience/bug report.

It would help us tremendously if you could provide a single, self-contained and concise test file, demonstrating this behaviour.

A simple .java file that can be compiled via javac speeds up the process, but it is also an interesting exercise for the bug reporter.

(this is actually how most bug reports are stored in our test suite, minimized and self-contained so that jtreg can invoke them: https://github.com/openjdk/jdk/tree/master/test/langtools/tools/javac; for you, no need to use jtreg, something that compiles with just javac is fine)

Looking forward for the test minimization.

Angelos


________________________________
From: amber-dev <amber-dev-retn at openjdk.org> on behalf of David Alayachew <davidalayachew at gmail.com>
Sent: 12 June 2025 21:55
To: Chen Liang <chen.l.liang at oracle.com>
Cc: amber-dev <amber-dev at openjdk.org>
Subject: Performance dropped significantly when switching from Nested Patterns to Flattened Patterns (was "Sharing my experience with Switch Expressions")

Woah, something new came up.

On a whim, I tried to run the performance numbers, to see if there were any performance differences between the nested pattern style and the flattened pattern style, and there was!

The nested pattern style finished in under 30-60 seconds. But the flattened pattern style took OVER AN HOUR.

I just wanted to bring this up, in case this is a bug, or maybe it's expected behaviour. Please let me know!

On Fri, May 23, 2025 at 7:10 AM David Alayachew <davidalayachew at gmail.com<mailto:davidalayachew at gmail.com>> wrote:
Please let me know if there are any details missing, or if I could better explain the situation.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.org/pipermail/amber-dev/attachments/20250612/c7748719/attachment-0001.htm>