feat(ci,eof): include eofwrap in EOF prerelease

[pdobacz]

🗒️ Description

Follows up #896 with a new build step which adds eofwrapped ethereum/tests to the EOF tests prerelease. Depends on #961 , hence opening as draft.

🔗 Related Issues

na

✅ Checklist

• All: Set appropriate labels for the changes.
• All: Considered squashing commits to improve commit history.
• All: Added an entry to CHANGELOG.md.

[danceratopz]

Thanks for cracking on with this @pdobacz!

Looks good so far, just a minor comment below.

ICYMI: Did you know that you test this action locally before pushing using nectos/act? See:
https://ethereum.github.io/execution-spec-tests/main/dev/test_actions_locally/

[pdobacz]

ICYMI: Did you know that you test this action locally before pushing using nectos/act? See: https://ethereum.github.io/execution-spec-tests/main/dev/test_actions_locally/

I am already using nectos/act (it's very flaky though, breaks 80% of runs because of network hiccups 🤷 ). I have the artifacts locally on disk and have eyeballed them for correctness.

@danceratopz BTW, any clues what might have caused the cancel of the tag build? Can I rerun? I think it will draft the release under ipsilon fork so should not cause disruption?

[danceratopz]

I am already using nectos/act (it's very flaky though, breaks 80% of runs because of network hiccups 🤷 ). I have the artifacts locally on disk and have eyeballed them for correctness.

Shame about the flakiness, must admit I haven't used it for testing a release flow. Perhaps Mario has (but he's ooo at the mo).

@danceratopz BTW, any clues what might have caused the cancel of the tag build? Can I rerun? I think it will draft the release under ipsilon fork so should not cause disruption?

Feel free to make a release whereever you like for "shorter" lived releases. If it will be a longer lasting and communicated to other clients, it's probably doing it from ethereum/execution-spec-tests (main?) for transparency.

Below, I'm looking at this raw log from this run:
https://github.com/ipsilon/execution-spec-tests/actions/runs/11940860199/job/33284456025

2024-11-20T20:05:35.7024316Z [ 96%] Building CXX object test/unittests/CMakeFiles/evmone-unittests.dir/statetest_loader_tx_test.cpp.o
2024-11-20T20:05:35.7504170Z [ 96%] Building CXX object test/unittests/CMakeFiles/evmone-unittests.dir/statetest_logs_hash_test.cpp.o
2024-11-20T20:05:35.7986322Z [ 97%] Building CXX object test/unittests/CMakeFiles/evmone-unittests.dir/statetest_withdrawals_test.cpp.o
2024-11-20T20:05:35.9732743Z [ 97%] Building CXX object test/unittests/CMakeFiles/evmone-unittests.dir/tracing_test.cpp.o
2024-11-20T20:10:03.5046844Z c++: fatal error: Killed signal terminated program cc1plus
2024-11-20T20:10:03.5100459Z compilation terminated.
2024-11-20T20:10:03.5839370Z gmake[2]: *** [test/t8n/CMakeFiles/evmone-t8n.dir/build.make:79: test/t8n/CMakeFiles/evmone-t8n.dir/t8n.cpp.o] Error 1
2024-11-20T20:10:03.5914231Z gmake[1]: *** [CMakeFiles/Makefile2:1505: test/t8n/CMakeFiles/evmone-t8n.dir/all] Error 2
2024-11-20T20:10:03.5955009Z gmake[1]: *** Waiting for unfinished jobs....
2024-11-20T20:10:03.5977568Z gmake[2]: *** [test/statetest/CMakeFiles/evmone-statetest.dir/build.make:93: test/statetest/CMakeFiles/evmone-statetest.dir/statetest_runner.cpp.o] Terminated
2024-11-20T20:10:03.5993821Z gmake[2]: *** [test/unittests/CMakeFiles/evmone-unittests.dir/build.make:79: test/unittests/CMakeFiles/evmone-unittests.dir/analysis_test.cpp.o] Terminated
2024-11-20T20:10:03.6016019Z gmake[1]: *** [CMakeFiles/Makefile2:1373: test/statetest/CMakeFiles/evmone-statetest.dir/all] Terminated
2024-11-20T20:10:03.6042835Z gmake: *** [Makefile:166: all] Terminated
2024-11-20T20:10:03.6083982Z gmake[2]: *** [test/blockchaintest/CMakeFiles/evmone-blockchaintest.dir/build.make:79: test/blockchaintest/CMakeFiles/evmone-blockchaintest.dir/blockchaintest.cpp.o] Terminated
2024-11-20T20:10:03.6096468Z gmake[2]: *** [test/bench/CMakeFiles/evmone-bench.dir/build.make:93: test/bench/CMakeFiles/evmone-bench.dir/synthetic_benchmarks.cpp.o] Terminated
2024-11-20T20:10:03.6983245Z ##[error]The runner has received a shutdown signal. This can happen when the runner service is stopped, or a manually started runner is canceled.
2024-11-20T20:10:04.3089431Z ##[error]The operation was canceled.
2024-11-20T20:10:04.4374385Z Cleaning up orphan processes

@winsvega had similiar issues building evmone relatively recently. I think he suspected that the runner was running out of memory while trying to build evmone. I've got a feeling it was stopping at exactly the same point in the build every time. I think he solved it in his Github Workflow by reducing the scope of evmone targets built?

I managed to track down one of failing builds that he was struggling with:
https://github.com/ethereum/execution-spec-tests/actions/runs/11443961401/job/31838359104#step:14:1088

This failed in exactly the same line as your tag build above:

[ 96%] Building CXX object test/unittests/CMakeFiles/evmone-unittests.dir/tracing_test.cpp.o

Def worth discussing with @winsvega in the chat.

---

Probably unrelated and I know these are only deprecation warnings, but it's probably a good idea to clean these up? There's a lot of them ))

2024-11-20T20:04:20.4000445Z CMake Deprecation Warning at /home/runner/.hunter/_Base/Download/Hunter/0.25.3/0dfbc2c/Unpacked/cmake/modules/hunter_apply_gate_settings.cmake:4 (cmake_minimum_required):
2024-11-20T20:04:20.4001659Z   Compatibility with CMake < 3.10 will be removed from a future version of
2024-11-20T20:04:20.4002233Z   CMake.
2024-11-20T20:04:20.4002396Z 
2024-11-20T20:04:20.4002778Z   Update the VERSION argument <min> value or use a ...<max> suffix to tell
2024-11-20T20:04:20.4003908Z   CMake that the project does not need compatibility with older versions.
2024-11-20T20:04:20.4004515Z Call Stack (most recent call first):
2024-11-20T20:04:20.4005417Z   /home/runner/.hunter/_Base/Download/Hunter/0.25.3/0dfbc2c/Unpacked/cmake/modules/hunter_finalize.cmake:5 (include)
2024-11-20T20:04:20.4006771Z   /home/runner/.hunter/_Base/Download/Hunter/0.25.3/0dfbc2c/Unpacked/cmake/modules/hunter_add_package.cmake:7 (include)
2024-11-20T20:04:20.4007873Z   /home/runner/.hunter/_Base/Download/Hunter/0.25.3/0dfbc2c/Unpacked/cmake/modules/hunter_cache_run.cmake:6 (include)
2024-11-20T20:04:20.4009113Z   /home/runner/.hunter/_Base/Download/Hunter/0.25.3/0dfbc2c/Unpacked/cmake/modules/hunter_load_from_cache.cmake:5 (include)
2024-11-20T20:04:20.4010247Z   /home/runner/.hunter/_Base/Download/Hunter/0.25.3/0dfbc2c/Unpacked/cmake/modules/hunter_download.cmake:22 (include)
2024-11-20T20:04:20.4011315Z   /home/runner/.hunter/_Base/Download/Hunter/0.25.3/0dfbc2c/Unpacked/cmake/projects/ethash/hunter.cmake:8 (include)
2024-11-20T20:04:20.4012534Z   /home/runner/.hunter/_Base/Download/Hunter/0.25.3/0dfbc2c/Unpacked/cmake/modules/hunter_add_package.cmake:62 (include)
2024-11-20T20:04:20.4013339Z   CMakeLists.txt:39 (hunter_add_package)
``

[pdobacz]

runner was running out of memory while trying to build evmone. I've got a feeling it was stopping at exactly the same point in the build every time. I think he solved it in his Github Workflow by reducing the scope of evmone targets built?

Ah yes, that might be it, thanks! When working locally I need to limit the parallelism in cmake build command. I'll do that in the CI maybe as well, I already have it for nektos/act's sake

[winsvega]

You don't need to build unittests in evmone. Select only the desired target (t8n, statetest runner) it saves the time and avoid this issue

[danceratopz]

LGTM!

Glad you could clear up the evmone build issue so quickly. Thanks @winsvega for the help!

Feel free to merge, if/when you're happy with the releases generated!