feat: support flamegraphs in Java walltime

[not-matthias]

No description provided.

[linear]

COD-2309 Walltime benchmarks WITH flamegraphs

What?

Builds on top of COD-2308, adds:

• Runner: Pass JVM flags
• Integration: Add root frame
• Backend: Process flamegraph, resolve symbols

Other stuff that might be needed:

• Node Origin
• Debug symbols (file, line)

How?

• Compare JFR (and other profilers) against perf

[avalanche-staging]

Merging this PR will improve performance by ×270

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

⚡ 3 improved benchmarks
✅ 6 untouched benchmarks

Performance Changes

	Benchmark	BASE	HEAD	Efficiency
⚡	bench.RleBenchmark.encode	274 µs	1 µs	×270
⚡	bench.RleBenchmark.decode	1,896.2 ns	737.4 ns	×2.6
⚡	bench.FibBenchmark.fib	9.5 ms	3.3 ms	×2.8

---

_{Comparing cod-2309-walltime-benchmarks-with-flamegraphs (a606260) with cod-2308-walltime-benchmarks-without-flamegraphs (580ee7c)}

[codspeed-hq]

Merging this PR will create unknown performance changes

🆕 125 new benchmarks
⏩ 9 skipped benchmarks¹

Performance Changes

	Benchmark	BASE	HEAD	Efficiency
🆕	binaryToGray[0]	N/A	12 ns	N/A
🆕	binaryToGray[42]	N/A	14 ns	N/A
🆕	encode[65536]	N/A	276.8 µs	N/A
🆕	findHighestSetBit[1024]	N/A	30 ns	N/A
🆕	binaryToGray[255]	N/A	14 ns	N/A
🆕	encode[1024]	N/A	4.2 µs	N/A
🆕	grayToBinary[255]	N/A	24 ns	N/A
🆕	grayToBinary[1024]	N/A	26 ns	N/A
🆕	binaryToGray[1024]	N/A	14 ns	N/A
🆕	decode[1024]	N/A	1.9 µs	N/A
🆕	sudokuSolver	N/A	174.4 µs	N/A
🆕	isPowerOfTwo[42]	N/A	14 ns	N/A
🆕	decode[65536]	N/A	189.8 µs	N/A
🆕	findHighestSetBit[65535]	N/A	35 ns	N/A
🆕	countSetBits[0]	N/A	14 ns	N/A
🆕	binaryToGray[65535]	N/A	14 ns	N/A
🆕	generateCombinations[5]	N/A	4.4 µs	N/A
🆕	grayToBinary[42]	N/A	20 ns	N/A
🆕	generateCombinations[9]	N/A	29.5 µs	N/A
🆕	isPowerOfTwo[0]	N/A	14 ns	N/A
...	...	...	...	...

ℹ️ Only the first 20 benchmarks are displayed. Go to the app to view all benchmarks.

---

_{Comparing cod-2309-walltime-benchmarks-with-flamegraphs (b43be0b) with cod-2308-walltime-benchmarks-without-flamegraphs (fd966b0)}

Footnotes

1. 9 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

[Copilot]

Pull request overview

Adds CodSpeed-flavored benchmark URIs and introduces a non-inlined “root frame” wrapper in the JMH codegen to make Java walltime flamegraphs more usable, while reorganizing examples and CI to exercise the new behavior.

Changes:

• Introduce io.codspeed.BenchmarkUri and propagate URI usage into result collection and instrumentation hooks.
• Modify JMH generator to emit a non-inlined root-frame method and route benchmark calls through it.
• Restructure examples under examples/ and update CI to run them (including walltime runs across multiple JDK distributions).

Reviewed changes

Copilot reviewed 29 out of 30 changed files in this pull request and generated 6 comments.

Show a summary per file

File	Description
settings.gradle.kts	Updates Gradle project include path to the new examples layout.
jmh-fork/jmh-core/src/main/java/io/codspeed/BenchmarkUri.java	Adds URI construction + source file resolution logic.
jmh-fork/jmh-core/src/test/java/io/codspeed/BenchmarkUriTest.java	Adds unit tests for URI pieces (bench name + source path fallback).
jmh-fork/jmh-core/src/main/java/org/openjdk/jmh/runner/Runner.java	Forces forks off when instrumented (for flamegraphs) and reports executed benchmark via URI.
jmh-fork/jmh-core/src/main/java/org/openjdk/jmh/runner/BaseRunner.java	Starts/stops CodSpeed hooks for embedded runs and reports executed benchmark via URI.
jmh-fork/jmh-core/src/main/java/org/openjdk/jmh/generators/core/BenchmarkGenerator.java	Emits a DONT_INLINE “root frame” wrapper and routes stub invocations through it.
jmh-fork/jmh-core/src/main/java/io/codspeed/result/CodSpeedResultCollector.java	Stores URI alongside benchmark name in walltime results.
jmh-fork/jmh-core/build.gradle.kts	Adds JUnit dependency for the new tests.
examples/example-gradle/**	Adds new Gradle example project and benchmarks (incl. TheAlgorithms-Java sources).
examples/example-maven/**	Adjusts Maven example layout and dependencies; adds a bench source link.
.github/workflows/ci.yml	Updates paths, adds caches, and expands walltime job to test multiple JDK distributions + installs a custom runner.
.gitmodules	Adds TheAlgorithms-Java submodule.
AGENTS.md / CLAUDE.md	Adds repo agent guidance / indirection file.

Comments suppressed due to low confidence (1)

examples/example-maven/src/main/java/bench:1

• This appears to be a symlink-style “link file” to a source directory. On Windows (and some CI setups), symlinks are often checked out as plain files, which would remove the bench/ directory from the compiler’s perspective and break the Maven example build. Prefer adding the shared benchmark sources via Maven configuration (e.g., add an additional source root pointing at ../example-gradle/src/jmh/java) rather than relying on symlink semantics.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

You can also share your feedback on Copilot code review. Take the survey.

[GuillaumeLagrange]

ogltm to me

[not-matthias]

Github is drunk, the PR has been merged onto main but it didn't detect it for some reason (probably because the base branch was set to cod-2308(?), not sure)