Age | Commit message (Collapse) | Author |
|
This change removes all jobs that have set to be disabled
and marked to be deleted with a 'DELETE ME' displayname.
Signed-off-by: Kelley Spoon <kelley.spoon@linaro.org>
Change-Id: Ibaf8867461c114be0b594bce286ad90fec2cdcb2
|
|
The Android Runtime (ART) project has a expected last date of 30th
September 2022. This review will disable the ART jobs so we no longer run
them. However, VIXL is not part of the ART project and will stay
working.
Change-Id: Ic1e75e8db4e357ceb379469721882ba4b5456f9e
Signed-off-by: Benjamin Copeland <ben.copeland@linaro.org>
|
|
pollscm executes commands through the scm block
outside of a full build environment. As such
the wrappers settings aren't applied, which means
ssh-agent isn't setup, which means the identity
file isn't being loaded so the git request doesn't
present the specificed private key and it fails.
Switching over to using the internal jgit instance
instead will allow the git process to recognize
that it needs credentials, pull them in from the
jenkins credential store, and properly use them
to authenticate the connection.
jgit doesn't support the concept of a "shallow clone",
so let's also remove that setting when it's enabled.
Signed-off-by: Kelley Spoon <kelley.spoon@linaro.org>
Change-Id: I80eeb20a00300eaa787c771c60dc6ed5dac2a795
|
|
Removes explicit choice of armv8 lunch target for
Host and VIXL test job configs; art-build-script
scripts handles that on their own now.
Change-Id: Ie570bf0e42d67adcefa50e09ee28a1c75c9c0cd9
|
|
Change-Id: I2a30d71abe3d1838dace7fd1ad26d9f6b406904e
|
|
I3bb0035c6e2bce8d17b41f69dc0dac6fa4560789 CL refactored
the test_art_host.sh script so it needs an explicit
--default parameter. This patch adjusts the jenkins job
to account for that.
Change-Id: Ie322c4138d9e3215d13c3df4332c60ffa390896a
|
|
Change-Id: If5bd041bc0467384e6dc8148465e9add2b1c8e87
|
|
Change-Id: I53c8c5feb79d9fbeed93a68800e523a71c288248
|
|
Also use master-art for host tests.
Change-Id: I0455675c5a13457077b0a2a4026472379ee18e65
|
|
Try to debug issues that don't happen locally
Change-Id: I5e9ef46c6e54775aa4ab1ec0e88aeb72d639561f
|
|
Change-Id: Ib415df26a0a974f197eb053408b1e928d2d0683d
|
|
Change-Id: I5d7d44562a5bf44c808f7903b88926ef65f7307d
|
|
Change-Id: Ibf60b19ca2d0fd9eda51dcf2ac4ed176077ec064
|
|
We have observed false negatives when builds started from cache, hence
always clean the out/ directory before launching a new build.
Change-Id: Ifd2c41bf9cc8f8a56e1e226e045073402085d5e6
|
|
Change-Id: I1122925f68eb769f5a558d39347bb61b1c8fd647
|
|
Change-Id: Ida00e5e11d716aa45fc2bb6669c4b76f36338b2b
|
|
Signed-off-by: Fathi Boudra <fathi.boudra@linaro.org>
Change-Id: I8f22627865e1f10890d786523a094a0593ae807d
|
|
Both target test and host test require it.
Change-Id: I8283421a857401712e1b7655360e8c5e45476152
|
|
This reverts commit 0793b9c8b77d62a7bf0e5742854cf0e56ab655c6.
Change-Id: Ic0c51ad4afdea723204de96d99f792a174750e7e
|
|
Both target test and host test require it.
Change-Id: I85c296c6d03fcb018658f456f57568a633494526
|
|
Change-Id: I3cf4ab9d71372a63ddd815f6eeff605ab4fdd2e1
|
|
Change-Id: Id5720c08112fe201b9c5dfa6987d84dcb7451599
Signed-off-by: Vishal Bhoj <vishal.bhoj@linaro.org>
|
|
Change-Id: Id6907cfa5c36d1f07223042e3e742545c7b285d7
|
|
- Cleanup the out of date inline groovy code in some of the jjb's.
- Rename the postbuild.groovy script to tests.groovy as it relates
particularly to host and target tests.
- Add another groovy file: global.groovy that applies to all
linaro-art builds.
Change-Id: Icb624357f5b26a8fc999e00b71fe061297650097
|
|
- Add description for IncludeBenchmarks comment in benchmark job.
- Update and fix the git tree urls for the scripts the jobs are running
- Improve text readability by adding newlines
Change-Id: If5d90695e34e52db2c1bc4df7e4aadcf60b76083
|
|
Change-Id: Ic73fe3163cced5cfca5eb3911f306489fa39084e
Signed-off-by: Vishal Bhoj <vishal.bhoj@linaro.org>
|
|
Change-Id: I5b6cc1cae0a7dbfca1f4804a7e1e3b0d20b035d4
Signed-off-by: Fathi Boudra <fathi.boudra@linaro.org>
|
|
Building upon d11f7fb, it was noticed that a build step can
fail and Jenkins will not execute the rest of the build
steps, heading straight into the post-build steps.
Since during the build a fatal error may occur, the Gerrit
variables may not get defined at all before reaching the
email-ext plug-in.
This commit moves the recipient list to the beginning of
the job definition, so that it gets executed regardless of
the build result.
Change-Id: I8fe03cd5cbc7eebc176e151b71b890387bc26818
Signed-off-by: Daniel Díaz <daniel.diaz@linaro.org>
|
|
When a job is triggered for reasons other than Gerrit, those
variables are undefined, leading to annoying errors in the
mail logs, as mail notifications are sent to
$GERRIT_CHANGE_OWNER_EMAIL and $GERRIT_EVENT_ACCOUNT_EMAIL,
literally. Of course, those accounts do not exist.
This change defines those variables, to either nothing or
their prior values.
Change-Id: I2f9902b8a5f1263ff6eea1535c7d7cc0abd6894c
Signed-off-by: Daniel Díaz <daniel.diaz@linaro.org>
|
|
ANDROID_ADB_SERVER_PORT:
Since the setup_adb.sh script already uses 5037
as default, defining it here is not needed.
CURRENT_FAILING_ART_TESTS:
feature is not supported currently
Change-Id: I1913f1a8a1aa04e479661f4db0bd9763dfaf3620
|
|
The linaro-art-testing-presubmit-test job does the same with a more
maintainable version.
Change-Id: I1e3de90258026d22a965d3cfdcdec8079a5e5059
|
|
Change-Id: I8a233d76a8dacb9519be5d5a71d47ab74c6629b4
|
|
Change-Id: I252a0cbc54b7ea09a7a60ea58b8b43706a09de3b
Signed-off-by: Fathi Boudra <fathi.boudra@linaro.org>
|
|
Change-Id: I63d323eb8f45f8a6d4cf75b62e45e1b37054d7c0
Signed-off-by: Fathi Boudra <fathi.boudra@linaro.org>
|
|
Change-Id: I4168fe59da869b10ff99837d5c56c2e10003a89c
Signed-off-by: Fathi Boudra <fathi.boudra@linaro.org>
|
|
Change-Id: I22bb1672dfbca26f279ad5f5813febc09e695aff
Signed-off-by: Vishal Bhoj <vishal.bhoj@linaro.org>
|
|
Change-Id: Idd43ffa3571c44c793684e0b60d9c5baadf5846a
|
|
Change-Id: Id4d3441869446af01f5d3be32e9af8abd49c06fe
|
|
This change will send an email to owner of the change,
the person who started the test with "IncludeAllTests"
keyword and the ci maintainers whenever a build fails
indicating in which section it failed.
Change-Id: I92d250dae3305f37f476b498dd2f04a4b0d20b16
Signed-off-by: Vishal Bhoj <vishal.bhoj@linaro.org>
|
|
We are seeing failures on slower machine. This change would
allow us to test if host test pass on faster machine.
This reverts commit 6a690cfd7ba9cfe7b90538c6076c71a8b6276e76.
Change-Id: I8f739b55c0e6eabe26342f91382a16139faeacf0
|
|
Change-Id: Id9a23d73b6a99eba31498ac6caed073a9d7d5191
|
|
Change-Id: Idfd55cb7067bafd5bd0d6d4b0abd256875ac3b46
Signed-off-by: Vishal Bhoj <vishal.bhoj@linaro.org>
|
|
Builds get triggered once IncludeAllTests has been commented,
but they also get triggered after a merge.
If the patch is merged upstream, gerrit rebases the patch
automatically and marks it as merged and at that point builds get triggered which
should not happen.
Issue seen in this gerrit commit: https://dev-private-review.linaro.org/#/c/791/
Change-Id: I47ccd8313fbcea103a699cd573ada21b40847161
Signed-off-by: Vishal Bhoj <vishal.bhoj@linaro.org>
|
|
Change-Id: I3d3600e4addc53d209c02508741b222fae841d3c
Signed-off-by: Daniel Díaz <daniel.diaz@linaro.org>
|
|
This depends on
http://android-review.linaro.org/16895
Change-Id: Ic6a849325e566b9e96c2e7dae528e716a812960d
Signed-off-by: Daniel Díaz <daniel.diaz@linaro.org>
|
|
Change-Id: I04a557886e39b81272b8a69ba652a7d7c7bc3f5a
Signed-off-by: Daniel Díaz <daniel.diaz@linaro.org>
|
|
Change-Id: I6595331ddafb1209c07fe19211a7add285bccb8d
Signed-off-by: Daniel Díaz <daniel.diaz@linaro.org>
|
|
Change-Id: Ie1963ab9e45953a0573499193bf09deeb0069b93
Signed-off-by: Daniel Díaz <daniel.diaz@linaro.org>
|
|
Change-Id: I58236ba6eaf56d555aa13c03d2c864bd1bc5d437
|
|
Change-Id: I9b3b725d7ff14bd63661480ffb643e631fa58ea2
Signed-off-by: Vishal Bhoj <vishal.bhoj@linaro.org>
|