GH-34813: [C++] Improve GoogleTest detection #34920

kou · 2023-04-06T03:28:15Z

Rationale for this change

#34765 approach has some corner cases. They causes CI failures.

What changes are included in this PR?

If incompatible GoogleTest is detected, we can fallback to bundled GoogleTest automatically.

Are these changes tested?

Yes.

Are there any user-facing changes?

Yes.

Closes: [C++] Add an auto fallback solution when system Gtest is compiled with a C++ standard lower than 17. #34813

github-actions · 2023-04-06T03:28:43Z

Closes: [C++] Add an auto fallback solution when system Gtest is compiled with a C++ standard lower than 17. #34813

github-actions · 2023-04-06T03:28:45Z

⚠️ GitHub issue #34813 has been automatically assigned in GitHub to PR creator.

kou · 2023-04-06T03:35:08Z

@github-actions crossbow submit test-alpine-linux-cpp

kou · 2023-04-06T05:39:33Z

@github-actions crossbow submit test-alpine-linux-cpp

kou · 2023-04-06T06:13:46Z

@github-actions crossbow submit test-alpine-linux-cpp

If incompatible GoogleTest is detected, we can fallback to bundled GoogleTest automatically.

kou · 2023-04-06T06:52:14Z

@github-actions crossbow submit test-alpine-linux-cpp

github-actions · 2023-04-06T06:54:39Z

Revision: 3815328

Submitted crossbow builds: ursacomputing/crossbow @ actions-e991323147

Task	Status
test-alpine-linux-cpp

jorisvandenbossche · 2023-04-06T08:34:46Z

Confirming that the latest version of the PR seems to work for me (with a gtest from conda):

...
-- Found GTest: /home/joris/miniconda3/envs/arrow-dev/lib/cmake/GTest/GTestConfig.cmake (found suitable version "1.13.0", minimum required is "1.10.0")  
...

On a previous version I got an error like "CMake Error: The source directory "SOURCES/CMakeFiles/CMakeTmp" does not exist" from FindGTestAlt.cmake:53 (try_compile)

assignUser · 2023-04-06T13:25:51Z

cpp/CMakeLists.txt


-set(ARROW_TEST_LINK_TOOLCHAIN arrow::flatbuffers GTest::gtest_main GTest::gtest
-                              GTest::gmock)
+set(ARROW_TEST_LINK_TOOLCHAIN arrow::flatbuffers ${ARROW_GTEST_GTEST_MAIN}


IMHO this is a regression as using variables instead of targets is an anti-pattern but I currently don't see a cleaner approach either.

I think we should rethink the entire dependency management once we have bumped the cmake version to 3.16 as we will have much better tools available. (this is of course out of scope for this PR)

We can use add_library(ALIAS) with CMake 3.11: https://cmake.org/cmake/help/latest/command/add_library.html#alias-libraries

kou · 2023-04-06T20:08:27Z

Confirming that the latest version of the PR seems to work for me (with a gtest from conda):

Thanks for confirming this!

On a previous version I got an error like "CMake Error: The source directory "SOURCES/CMakeFiles/CMakeTmp" does not exist" from FindGTestAlt.cmake:53 (try_compile)

Yes. I needed to write a code to work with newer CMake and older CMake. On a previous version this implementation worked only with newer CMake.

kou · 2023-04-06T23:51:49Z

+1

ursabot · 2023-04-08T00:55:42Z

Benchmark runs are scheduled for baseline = 6f3d96e and contender = f7644ae. f7644ae is a master commit associated with this PR. Results will be available as each benchmark for each run completes.
Conbench compare runs links:
[Finished ⬇️0.0% ⬆️0.0%] ec2-t3-xlarge-us-east-2
[Failed ⬇️3.09% ⬆️1.16%] test-mac-arm
[Finished ⬇️0.0% ⬆️0.0%] ursa-i9-9960x
[Failed ⬇️1.14% ⬆️0.98%] ursa-thinkcentre-m75q
Buildkite builds:
[Finished] f7644ae8 ec2-t3-xlarge-us-east-2
[Finished] f7644ae8 test-mac-arm
[Finished] f7644ae8 ursa-i9-9960x
[Finished] f7644ae8 ursa-thinkcentre-m75q
[Finished] 6f3d96ec ec2-t3-xlarge-us-east-2
[Failed] 6f3d96ec test-mac-arm
[Finished] 6f3d96ec ursa-i9-9960x
[Failed] 6f3d96ec ursa-thinkcentre-m75q
Supported benchmarks:
ec2-t3-xlarge-us-east-2: Supported benchmark langs: Python, R. Runs only benchmarks with cloud = True
test-mac-arm: Supported benchmark langs: C++, Python, R
ursa-i9-9960x: Supported benchmark langs: Python, R, JavaScript
ursa-thinkcentre-m75q: Supported benchmark langs: C++, Java

zfoobar · 2023-04-08T21:15:15Z

This is a regression - build fails for me (macos/arm64/conda).

The only reason I could get it building is because I've been tracking this issue and knew to supply the proper BUNDLED flag. At least before the cmake module prompted you to put the flag in place if it detected the installed binaries weren't c++17.

It's back to the old behavior, build fails with the symbols it can't resolve.

kou · 2023-04-09T02:12:34Z

Could you open a new issue with how to reproduce and full error log?

At least I can't reproduce it with macOS/arm64/Homebrew(not conda).

zfoobar · 2023-04-09T03:08:28Z

Sure. Any specific flags you want me to pass to the build process for debugging/error log purposes?

…

On Sat, Apr 8, 2023 at 7:12 PM Sutou Kouhei ***@***.***> wrote: Could you open a new issue with how to reproduce and full error log? At least I can't reproduce it with macOS/arm64/Homebrew(not conda). — Reply to this email directly, view it on GitHub <#34920 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AC74BGOJK7CBX6ZWKIJDOGDXAILJ5ANCNFSM6AAAAAAWU3LK3E> . You are receiving this because you commented.Message ID: ***@***.***>

kou · 2023-04-09T04:22:50Z

cmake --debug-trycompile ... will help us: https://cmake.org/cmake/help/latest/manual/cmake.1.html#cmdoption-cmake-debug-trycompile

### Rationale for this change apache#34765 approach has some corner cases. They causes CI failures. ### What changes are included in this PR? If incompatible GoogleTest is detected, we can fallback to bundled GoogleTest automatically. ### Are these changes tested? Yes. ### Are there any user-facing changes? Yes. * Closes: apache#34813 Authored-by: Sutou Kouhei <kou@clear-code.com> Signed-off-by: Sutou Kouhei <kou@clear-code.com>

kou requested review from lidavidm and wjones127 as code owners April 6, 2023 03:28

github-actions bot added Component: C++ Component: FlightRPC Component: Parquet awaiting committer review Awaiting committer review labels Apr 6, 2023

kou force-pushed the cpp-gtest-fallback branch 2 times, most recently from d8ccf7e to 199647d Compare April 6, 2023 03:30

kou requested review from assignUser and raulcd as code owners April 6, 2023 03:30

kou force-pushed the cpp-gtest-fallback branch from 199647d to a48266e Compare April 6, 2023 03:33

This comment was marked as outdated.

Sign in to view

kou force-pushed the cpp-gtest-fallback branch from a48266e to 536af6a Compare April 6, 2023 05:39

kou force-pushed the cpp-gtest-fallback branch from 536af6a to a585e35 Compare April 6, 2023 05:39

This comment was marked as outdated.

Sign in to view

kou force-pushed the cpp-gtest-fallback branch 2 times, most recently from 0b9cf98 to 3440365 Compare April 6, 2023 06:13

This comment was marked as outdated.

Sign in to view

kou force-pushed the cpp-gtest-fallback branch from 3440365 to 3de38f0 Compare April 6, 2023 06:16

apacheGH-34813: [C++] Improve GoogleTest detection

3815328

If incompatible GoogleTest is detected, we can fallback to bundled GoogleTest automatically.

kou force-pushed the cpp-gtest-fallback branch from 3de38f0 to 3815328 Compare April 6, 2023 06:52

assignUser reviewed Apr 6, 2023

View reviewed changes

github-actions bot added awaiting changes Awaiting changes and removed awaiting committer review Awaiting committer review labels Apr 6, 2023

kou merged commit f7644ae into apache:main Apr 6, 2023

kou deleted the cpp-gtest-fallback branch April 6, 2023 23:52

GH-34813: [C++] Improve GoogleTest detection #34920

GH-34813: [C++] Improve GoogleTest detection #34920

Uh oh!

Conversation

kou commented Apr 6, 2023 • edited by github-actions bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Rationale for this change

What changes are included in this PR?

Are these changes tested?

Are there any user-facing changes?

Uh oh!

github-actions bot commented Apr 6, 2023

Uh oh!

github-actions bot commented Apr 6, 2023

Uh oh!

kou commented Apr 6, 2023

Uh oh!

This comment was marked as outdated.

kou commented Apr 6, 2023

Uh oh!

This comment was marked as outdated.

kou commented Apr 6, 2023

Uh oh!

This comment was marked as outdated.

kou commented Apr 6, 2023

Uh oh!

github-actions bot commented Apr 6, 2023

Uh oh!

jorisvandenbossche commented Apr 6, 2023

Uh oh!

assignUser Apr 6, 2023

Choose a reason for hiding this comment

Uh oh!

kou Apr 6, 2023

Choose a reason for hiding this comment

Uh oh!

kou commented Apr 6, 2023

Uh oh!

kou commented Apr 6, 2023

Uh oh!

ursabot commented Apr 8, 2023

Uh oh!

zfoobar commented Apr 8, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

kou commented Apr 9, 2023

Uh oh!

zfoobar commented Apr 9, 2023 via email

Uh oh!

kou commented Apr 9, 2023

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

kou commented Apr 6, 2023 •

edited by github-actions bot

Loading

zfoobar commented Apr 8, 2023 •

edited

Loading