Error on LUCI v8_presubmit |
||||
Issue descriptionSee: https://chromium-review.googlesource.com/c/590449/1 Experimental luci builder fails while non-luci succeeds. Repros locally with: ./scripts/slave/recipes.py run --properties-file - run_presubmit <<EOF {"recipe": "run_presubmit", "patch_ref": "refs/changes/49/590449/1", "patch_project": "v8/v8", "$recipe_engine/path": {"cache_dir": "/b/swarming/w/ir/cache", "temp_dir": "/b/swarming/w/ir/tmp/rt"}, "runhooks": true, "patch_gerrit_url": "https://chromium-review.googlesource.com", "dry_run": "true", "patch_repository_url": "https://chromium.googlesource.com/v8/v8", "swarming_run_id": "37a12e376b59e811", "attempt_start_ts": 1501230982000000, "blamelist": ["jgruber@chromium.org"], "master": "master.tryserver.v8", "repo_name": "v8", "category": "cq_experimental", "patch_set": 1, "repository": "https://chromium.googlesource.com/v8/v8", "buildername": "v8_presubmit", "mastername": "tryserver.v8", "patch_storage": "gerrit", "buildbucket": {"build": {"created_ts": 1501231361670390, "bucket": "luci.v8.try", "id": "8972838539663754592", "created_by": "user:luci-migration@appspot.gserviceaccount.com", "tags": ["buildset:patch/gerrit/chromium-review.googlesource.com/590449/1", "luci_migration_attempt:2", "luci_migration_buildbot_build_id:8972838919772159232", "user_agent:luci-migration"]}}, "reason": "CQ", "patch_issue": 590449, "$recipe_engine/step": {"prefix_path": ["/b/swarming/w/ir/cipd_bin_packages", "/b/swarming/w/ir/cipd_bin_packages/bin"]}, "path_config": "generic", "revision": "f4b2b9bef77c35491adc6c56be157c09294f79ba", "bot_id": "swarm915-c4"} EOF
,
Jul 31 2017
This actually looks like a bug in "presubmit_support.py". It's expecting the "Code-Review" label to be present in the JSON response, but it's not. Looking at the CL in question, the label is not present, likely because the change is abandoned. +agable@, who is working on Gerrit integration into our tooling. I suspect this is just adding code acknowledging that "Code-Review" is not always present.
,
Jul 31 2017
Oh actual error log: https://luci-logdog.appspot.com/v/?s=v8%2Fbuildbucket%2Fcr-buildbucket.appspot.com%2F8972838539663754592%2F%2B%2Fsteps%2Fpresubmit%2F0%2Fstdout DEBUG:root:Running CheckChangeOnCommit in /b/swarming/w/ir/kitchen-workdir/v8/PRESUBMIT.py Traceback (most recent call last): File "/b/swarming/w/ir/kitchen-checkout/scripts/slave/.recipe_deps/depot_tools/presubmit_support.py", line 1621, in <module> sys.exit(main()) File "/b/swarming/w/ir/kitchen-checkout/scripts/slave/.recipe_deps/depot_tools/presubmit_support.py", line 1605, in main options.dry_run) File "/b/swarming/w/ir/kitchen-checkout/scripts/slave/.recipe_deps/depot_tools/presubmit_support.py", line 1348, in DoPresubmitChecks results += executer.ExecPresubmitScript(presubmit_script, filename) File "/b/swarming/w/ir/kitchen-checkout/scripts/slave/.recipe_deps/depot_tools/presubmit_support.py", line 1258, in ExecPresubmitScript result = eval(function_name + '(*__args)', context) File "<string>", line 1, in <module> File "<string>", line 348, in CheckChangeOnCommit File "<string>", line 269, in _CommonChecks File "/b/swarming/w/ir/kitchen-checkout/scripts/slave/.recipe_deps/depot_tools/presubmit_canned_checks.py", line 864, in CheckOwners approval_needed=input_api.is_committing) File "/b/swarming/w/ir/kitchen-checkout/scripts/slave/.recipe_deps/depot_tools/presubmit_canned_checks.py", line 911, in GetCodereviewOwnerAndReviewers return func(input_api, email_regexp, approval_needed) File "/b/swarming/w/ir/kitchen-checkout/scripts/slave/.recipe_deps/depot_tools/presubmit_canned_checks.py", line 972, in _GerritOwnerAndReviewers r for r in input_api.gerrit.GetChangeReviewers(issue, approval_needed) File "/b/swarming/w/ir/kitchen-checkout/scripts/slave/.recipe_deps/depot_tools/presubmit_support.py", line 250, in GetChangeReviewers cr = self.GetChangeInfo(issue)['labels']['Code-Review'] KeyError: 'Code-Review'
,
Jul 31 2017
Ah, so it was a race between the non-luci bot and the experimental luci version and the CL got abandoned right in between?
,
Jul 31 2017
I think so, yeah.
,
Jul 31 2017
Yep, exactly. I can work on a change to PRESUBMIT to not fail in quite this way: https://chromium-review.googlesource.com/594839 But also it shouldn't be missing just because the change is abandoned. That has to be a bug. If it's a feature, it's a bad one: b/64207819
,
Jul 31 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/tools/depot_tools/+/8b478f04067599177cc5b064c81f0a1e915095f0 commit 8b478f04067599177cc5b064c81f0a1e915095f0 Author: Aaron Gable <agable@chromium.org> Date: Mon Jul 31 23:25:14 2017 PRESUBMIT: Be resilient to changes with no Code-Review label Bug: 750694 Change-Id: I69778c3ea0789795ad7e50d11bf677df8f7bc262 Reviewed-on: https://chromium-review.googlesource.com/594839 Reviewed-by: Daniel Jacques <dnj@chromium.org> Commit-Queue: Aaron Gable <agable@chromium.org> [modify] https://crrev.com/8b478f04067599177cc5b064c81f0a1e915095f0/presubmit_support.py [modify] https://crrev.com/8b478f04067599177cc5b064c81f0a1e915095f0/tests/presubmit_unittest.py
,
Sep 11 2017
|
||||
►
Sign in to add a comment |
||||
Comment 1 by machenb...@chromium.org
, Jul 31 2017