| Summary: | build results should be reported to results database | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Aakash Jain <aakash_jain> |
| Component: | Tools / Tests | Assignee: | Jonathan Bedard <jbedard> |
| Status: | ASSIGNED --- | ||
| Severity: | Normal | CC: | aakash_jain, ap, ryanhaddad, webkit-bug-importer |
| Priority: | P2 | Keywords: | InRadar |
| Version: | Other | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
|
Description
Aakash Jain
2021-07-08 05:58:23 PDT
This should be pretty easy to do! My thinking on the best way to solve this problem was to have each project report to the results database as a "test". That way, the "suite" view would hold the total status of the build, and we would have a super clear (and machine parsable) history of not only when the build broke, but which projects broke it. This basically involves having the build script we use in automation (so build-webkit) upload a json dictionary to results.webkit.org, whose format is outlined in https://results.webkit.org/documentation#API-Uploads If we have this, we can even have EWS check results database and avoid doing re-builds (compile-webkit-without-change step) in various scenarios, resulting in further speedup of EWS builder queues. |