Run tests on CI #6
Conversation
igchor
force-pushed
the
enable_tests
branch
5 times, most recently
from
dacd560 to
ba09746
Compare
| @@ -34,3 +34,5 @@ jobs: | |||
| uses: actions/checkout@v2 | |||
| - name: "build CacheLib using build script" | |||
| run: ./contrib/build.sh -j -v -T | |||
| - name: "run tests" | |||
| run: cd opt/cachelib && ../../run_tests.sh | |||
There was a problem hiding this comment.
Please add timeout for this step.
run_tests.sh
Outdated
| common-test-TimeTests | ||
| shm-test-test_page_size" | ||
|
|
||
| find -type f \( -not -name "*bench*" \) -executable | grep -vF "$BLACKLIST" | xargs -n1 bash -c |
There was a problem hiding this comment.
Will it be possible to show some output of this command so that we can see which tests were running / passing / failing?
There was a problem hiding this comment.
Reviewed 3 of 3 files at r1, all commit messages.
Reviewable status: all files reviewed, 3 unresolved discussions (waiting on @igchor)
run_tests.sh, line 2 at r1 (raw file):
BLACKLIST="allocator-test-NvmCacheTests common-test-TimeTests
isn't ""
needed after end of each line ?
There was a problem hiding this comment.
Reviewable status: 0 of 3 files reviewed, 3 unresolved discussions (waiting on @guptask and @victoria-mcgrath)
run_tests.sh, line 2 at r1 (raw file):
Previously, guptask (Sounak Gupta) wrote…
isn't ""
needed after end of each line ?
No, the idea here is to separate each test by a newline - I added a comment.
run_tests.sh, line 5 at r1 (raw file):
Previously, victoria-mcgrath wrote…
Will it be possible to show some output of this command so that we can see which tests were running / passing / failing?
The output is already visible here: https://github.com/pmem/CacheLib/runs/4118303221?check_suite_focus=true
I don't know why this does not work on centos...
.github/workflows/build-cachelib-centos.yml, line 38 at r1 (raw file):
Previously, victoria-mcgrath wrote…
Please add timeout for this step.
Done.
There was a problem hiding this comment.
I've also disabled navy tests for now to speed up the build - do you think it's OK or should we keep it? We can maybe run them periodically on master instead?
Reviewable status: 0 of 3 files reviewed, 3 unresolved discussions (waiting on @guptask and @victoria-mcgrath)
There was a problem hiding this comment.
I've added this change on top of my Config API integration PR to see if it builds here: #8
Reviewable status: 0 of 3 files reviewed, 3 unresolved discussions (waiting on @guptask and @victoria-mcgrath)
run_tests.sh, line 5 at r1 (raw file):
Previously, igchor (Igor Chorążewicz) wrote…
The output is already visible here: https://github.com/pmem/CacheLib/runs/4118303221?check_suite_focus=true
I don't know why this does not work on centos...
Ok, it works now, I just messed up the path.
|
Yes, let's disable all long tests and schedule periodic full test runs on develop. |
This change is