| commit | 563592dddd84bff1bb4cce87b90c45e08ce4db7c | [log] [tgz] |
|---|---|---|
| author | Oliver Newman <olivernewman@fuchsia.infra.roller.google.com> | Thu May 27 19:43:14 2021 +0000 |
| committer | Copybara-Service <copybara-worker@google.com> | Thu May 27 12:44:30 2021 -0700 |
| tree | 7563df91d03f3c569f7537bac4d112c88d041a35 | |
| parent | 2dd26dab800d31c17ef55ddb659e0b89f4594ae8 [diff] |
[roll] Roll fuchsia Revert "[fint] Analyze affected tests" This reverts commit d25ffaa55b8e0e8285abd63d972247310aa2bbe8. Reason for revert: breaks when trying to `touch` files in directories that have been deleted Original change's description: > [fint] Analyze affected tests > > Logic translated from the affected_tests.py script in recipes: > https://fuchsia.googlesource.com/infra/recipes/+/1d3ba67d813b21c10a37e785ad66abf9a643e742/recipe_modules/build/resources/affected_tests.py > > The ninja dry run output is now discarded after use and won't be > uploaded to Logdog, which will admittedly make it slightly harder to > understand the analysis. However, the dry run output is ~2 million lines > long for most builds, so it takes a long time to upload to Logdog and > isn't particularly useful anyway, because Chrome is extremely sluggish > to even scroll through the log, let alone search through it. > > One thing that could be described as a regression is that fint will do > this analysis even when running in a build-only build (where we won't > run any tests) or an integration change (where we'll run all tests even > if the build graph analysis indicates no work). If this becomes an issue > we can add a context.proto field to let recipes control whether we do > the analysis. However, it only takes ~10 seconds currently, and since > all the slowest builders in CQ run tests (so they use the results of the > analysis), skipping it in other builders would have basically zero > impact on CQ durations. > > Before updating the recipes to use the affected_tests and no_work > results emitted by fint, I plan to add a check that makes sure that the > results are the same as those emitted by affected_tests.py and only > switch over once the check has been running for a few days with no > failures. > > Original-Bug: 75371 > Original-Reviewed-on: https://fuchsia-review.googlesource.com/c/fuchsia/+/520500 TBR=shayba@google.com,ihuh@google.com,olivernewman@google.com No-Presubmit: true No-Tree-Checks: true No-Try: true Original-Bug: 75371 Original-Reviewed-on: https://fuchsia-review.googlesource.com/c/fuchsia/+/535923 Original-Revision: a3ddcd4008c0aefe42c78b772d106b7986ca01ca GitOrigin-RevId: bb4b74612ebd488fda508141b8abf289b7fcd584 Change-Id: Iba99542cf6508ea305dae931628df13cb99d24e4
This repository contains Fuchsia's Global Integration manifest files.
All changes should be made to the internal version of this repository. Our infrastructure automatically updates this version when the internal one changes.
Currently all changes must be made by a Google employee. Non-Google employees wishing to make a change can ask for assistance via the IRC channel #fuchsia on Freenode.
First install Jiri.
Next run:
$ jiri init $ jiri import minimal https://fuchsia.googlesource.com/integration $ jiri update
Third party projects should have their own subdirectory in ./third_party.