Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix google benchmark name and update version #425

Merged
merged 6 commits into from
Jun 8, 2023

Conversation

vyasr
Copy link
Contributor

@vyasr vyasr commented Jun 7, 2023

Description

The google benchmark package name was incorrect, making it impossible to find benchmark if it was previously installed. However, the currently pinned version of benchmark has a broken CMake config on conda-forge (it claims it is version 0.0.0) so it cannot be found even with the name fix. This is as good a time as any to go ahead and update to the latest version.

Resolves #424

Checklist

  • I am familiar with the Contributing Guidelines.
  • New or existing tests cover these changes.
  • The documentation is up to date with these changes.
  • The cmake-format.json is up to date with these changes.
  • I have added new files under rapids-cmake/
    • I have added include guards (include_guard(GLOBAL))
    • I have added the associated docs/ rst file and update the api.rst

@vyasr vyasr added improvement Improves an existing functionality non-breaking Introduces a non-breaking change labels Jun 7, 2023
@vyasr vyasr requested a review from a team as a code owner June 7, 2023 23:09
@vyasr vyasr self-assigned this Jun 7, 2023
@bdice
Copy link
Contributor

bdice commented Jun 7, 2023

Testing downstream: rapidsai/cudf#13528

@bdice
Copy link
Contributor

bdice commented Jun 7, 2023

Testing cuml: rapidsai/cuml#5463

@vyasr vyasr requested a review from bdice June 7, 2023 23:23
rapids-cmake/cpm/gbench.cmake Outdated Show resolved Hide resolved
@bdice
Copy link
Contributor

bdice commented Jun 8, 2023

cuDF passed using the conda package: rapidsai/cudf#13528
cuML passed using the CPM fetched package: rapidsai/cuml#5463

This should be good to merge if we can fix the failing test add-multi-jobs-single-gpu-makefile. Rerunning.

Details

696/816 Test #570: add-multi-jobs-single-gpu-makefile ..........................................................***Failed   17.09 sec
UpdateCTestConfiguration  from :/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/DartConfiguration.tcl
UpdateCTestConfiguration  from :/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/DartConfiguration.tcl
Test project /__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build
Constructing a list of tests
Done constructing a list of tests
Updating test list for fixtures
Added 0 tests to meet fixture requirements
Checking test dependency graph...
Checking test dependency graph end
test 1
    Start 1: concurrent_test_0

1: Test command: /opt/conda/envs/test/bin/cmake "-Dcommand_to_run=/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/verify_multi_concurrent" "-Dcommand_args=0" "-P=/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/rapids-cmake/./run_gpu_test.cmake"
1: Working Directory: /__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build
1: Test timeout computed to be: 10000000
test 2
    Start 2: concurrent_test_1

2: Test command: /opt/conda/envs/test/bin/cmake "-Dcommand_to_run=/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/verify_multi_concurrent" "-Dcommand_args=1" "-P=/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/rapids-cmake/./run_gpu_test.cmake"
2: Working Directory: /__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build
2: Test timeout computed to be: 10000000
test 3
    Start 3: concurrent_test_2

3: Test command: /opt/conda/envs/test/bin/cmake "-Dcommand_to_run=/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/verify_multi_concurrent" "-Dcommand_args=2" "-P=/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/rapids-cmake/./run_gpu_test.cmake"
3: Working Directory: /__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build
3: Test timeout computed to be: 10000000
test 4
    Start 4: concurrent_test_3

4: Test command: /opt/conda/envs/test/bin/cmake "-Dcommand_to_run=/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/verify_multi_concurrent" "-Dcommand_args=3" "-P=/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/rapids-cmake/./run_gpu_test.cmake"
4: Working Directory: /__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build
4: Test timeout computed to be: 10000000
test 5
    Start 5: concurrent_test_4

5: Test command: /opt/conda/envs/test/bin/cmake "-Dcommand_to_run=/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/verify_multi_concurrent" "-Dcommand_args=4" "-P=/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/rapids-cmake/./run_gpu_test.cmake"
5: Working Directory: /__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build
5: Test timeout computed to be: 10000000
test 6
    Start 6: concurrent_test_5

6: Test command: /opt/conda/envs/test/bin/cmake "-Dcommand_to_run=/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/verify_multi_concurrent" "-Dcommand_args=5" "-P=/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/rapids-cmake/./run_gpu_test.cmake"
6: Working Directory: /__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build
6: Test timeout computed to be: 10000000
1: '/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/verify_multi_concurrent' '0'
1: 0 lock_state: 0 valid 1
1: 1 lock_state: 0 valid 0
1: 2 lock_state: 0 valid 0
1: 3 lock_state: 0 valid 0
1: 4 lock_state: 0 valid 0
1: 5 lock_state: 0 valid 0
2: '/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/verify_multi_concurrent' '1'
2: 0 lock_state: 0 valid 0
2: 1 lock_state: 0 valid 1
2: 2 lock_state: -1 valid 1
2: 3 lock_state: -1 valid 1
2: 4 lock_state: -1 valid 1
2: 5 lock_state: -1 valid 1
4: '/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/verify_multi_concurrent' '3'
4: 0 lock_state: 0 valid 0
4: 1 lock_state: -1 valid 1
4: 2 lock_state: -1 valid 1
4: 3 lock_state: 0 valid 1
4: 4 lock_state: -1 valid 1
4: 5 lock_state: -1 valid 1
3: '/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/verify_multi_concurrent' '2'
3: 0 lock_state: 0 valid 0
3: 1 lock_state: -1 valid 1
3: 2 lock_state: 0 valid 1
3: 3 lock_state: -1 valid 1
3: 4 lock_state: -1 valid 1
3: 5 lock_state: -1 valid 1
5: '/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/verify_multi_concurrent' '4'
5: 0 lock_state: 0 valid 0
5: 1 lock_state: -1 valid 1
5: 2 lock_state: -1 valid 1
5: 3 lock_state: -1 valid 1
5: 4 lock_state: 0 valid 1
5: 5 lock_state: -1 valid 1
6: '/__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/verify_multi_concurrent' '5'
6: 0 lock_state: 0 valid 0
6: 1 lock_state: -1 valid 1
6: 2 lock_state: -1 valid 1
6: 3 lock_state: -1 valid 1
6: 4 lock_state: -1 valid 1
6: 5 lock_state: 0 valid 1
1: CMake Error at rapids-cmake/run_gpu_test.cmake:34 (execute_process):
1:   execute_process failed command indexes:
1: 
1:     1: "Child return code: 1"
1: 
1: 
1: 
1/6 Test #1: concurrent_test_0 ................***Failed   13.72 sec
2: CMake Error at rapids-cmake/run_gpu_test.cmake:34 (execute_process):
2:   execute_process failed command indexes:
2: 
2:     1: "Child return code: 1"
2: 
2: 
2: 
2/6 Test #2: concurrent_test_1 ................***Failed   15.23 sec
4: CMake Error at rapids-cmake/run_gpu_test.cmake:34 (execute_process):
4:   execute_process failed command indexes:
4: 
4:     1: "Child return code: 1"
4: 
4: 
4: 
3/6 Test #4: concurrent_test_3 ................***Failed   15.85 sec
3: CMake Error at rapids-cmake/run_gpu_test.cmake:34 (execute_process):
3:   execute_process failed command indexes:
3: 
3:     1: "Child return code: 1"
3: 
3: 
3: 
4/6 Test #3: concurrent_test_2 ................***Failed   16.15 sec
6: CMake Error at rapids-cmake/run_gpu_test.cmake:34 (execute_process):
6:   execute_process failed command indexes:
6: 
6:     1: "Child return code: 1"
6: 
6: 
6: 
5/6 Test #6: concurrent_test_5 ................***Failed   16.29 sec
5: CMake Error at rapids-cmake/run_gpu_test.cmake:34 (execute_process):
5:   execute_process failed command indexes:
5: 
5:     1: "Child return code: 1"
5: 
5: 
5: 
6/6 Test #5: concurrent_test_4 ................***Failed   17.07 sec

0% tests passed, 6 tests failed out of 6

Total Test time (real) =  17.08 sec

The following tests FAILED:
	  1 - concurrent_test_0 (Failed)
	  2 - concurrent_test_1 (Failed)
	  3 - concurrent_test_2 (Failed)
	  4 - concurrent_test_3 (Failed)
	  5 - concurrent_test_4 (Failed)
	  6 - concurrent_test_5 (Failed)
Errors while running CTest
Output from these tests are in: /__w/rapids-cmake/rapids-cmake/build/test/add-multi-jobs-single-gpu-makefile-build/Testing/Temporary/LastTest.log
Use "--rerun-failed --output-on-failure" to re-run the failed cases verbosely.

@vyasr
Copy link
Contributor Author

vyasr commented Jun 8, 2023

/merge

@rapids-bot rapids-bot bot merged commit bf471a7 into rapidsai:branch-23.08 Jun 8, 2023
@vyasr vyasr deleted the fix/benchmark_name branch June 8, 2023 17:44
@harrism
Copy link
Member

harrism commented Jun 12, 2023

This must have been merged without testing RMM benchmarks. Nightly build is failing for rmm replay.cpp. Jumping from 1.5.3 to 1.8.0 is a huge jump -- previously public member variables were hidden behind accessors. This should have been tested with all libraries...

rapids-bot bot pushed a commit to rapidsai/rmm that referenced this pull request Jun 13, 2023
…1293)

rapidsai/rapids-cmake#425 updated the google benchmark version to a version that is more than 2 years newer, and the API changed. This PR just changes to use an accessor rather than a public (now private) data member to fix the compilation errors.

Authors:
  - Mark Harris (https://github.com/harrism)

Approvers:
  - Rong Ou (https://github.com/rongou)
  - Bradley Dice (https://github.com/bdice)

URL: #1293
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
improvement Improves an existing functionality non-breaking Introduces a non-breaking change
Projects
None yet
Development

Successfully merging this pull request may close these issues.

[BUG] rapids_cpm_gbench doesn't find benchmark conda package
3 participants