[Webkit-unassigned] [Bug 231241] New: [webkitpy] The actual results reported for a flaky tests shouldn't include the expectation

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=231241

            Bug ID: 231241
           Summary: [webkitpy] The actual results reported for a flaky
                    tests shouldn't include the expectation
           Product: WebKit
           Version: WebKit Nightly Build
          Hardware: Unspecified
                OS: Unspecified
            Status: NEW
          Severity: Normal
          Priority: P2
         Component: Tools / Tests
          Assignee: webkit-unassigned at lists.webkit.org
          Reporter: clopez at igalia.com
                CC: aakash_jain at apple.com, ap at webkit.org,
                    jbedard at apple.com

This is about the following corner-case -> When a test is marked as flaky and fails the run (doesn't match the expectation) but passes the run on the retry.

In this case, at the end of the run we currently report as actual results of the tests the union between the first run and the expectations.

So it happens that, for example, when we have a test marked as  "[ Crash Timeout Pass ]" and the fail gives the results "Fail" and "Pass" we currently report the actual results of the tests to be "[ Fail Crash Timeout Pass ]" when "Crash" and "Timeout" didn't happened.

So this makes difficult to know when a tests stops giving a specific result. Once it is marked as flaky with a set of results there is no way to know if some of the results in the set not longer happen.

Here is a test-case to test this:

1. Apply this test patch <http://sprunge.us/MqyGkc> to add a test that fails the 50% of times and passes the other 50% and such test is marked with an expectation of

fast/random/fails-half.html [ Crash Timeout Pass ]

2. Run the test like:

$ Tools/Scripts/run-webkit-tests --no-build --no-show-results --release --results-directory layout-test-results --debug-rwt-logging fast/random/fails-half.html

Repeat the run above until you get the test to fail on the first run and pass on the second.

When that happens you will see this <http://sprunge.us/vwM9B7>:

Unexpected flakiness: text-only failures (1)
  fast/random/fails-half.html [ Crash Timeout Pass Failure ]


And on the json file at layout-test-results/full_results.json you will see:

ADD_RESULTS({"tests":{"fast":{"random":{"fails-half.html":{"report":"FLAKY","expected":"PASS TIMEOUT CRASH","actual":"TEXT PASS TIMEOUT CRASH"}}}}


So I think this is misleading, because the _actual_ result of this test was "FAIL/TEXT" and "PASS" but it didn't give TIMEOUT or CRASH

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20211005/b3620939/attachment.htm>