[Webkit-unassigned] [Bug 58122] New: chromium rebaseline script did the wrong thing with snowleopard expectations

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=58122

           Summary: chromium rebaseline script did the wrong thing with
                    snowleopard expectations
           Product: WebKit
           Version: 528+ (Nightly build)
          Platform: PC
        OS/Version: Mac OS X 10.5
            Status: NEW
          Severity: Normal
          Priority: P2
         Component: Tools / Tests
        AssignedTo: webkit-unassigned at lists.webkit.org
        ReportedBy: ojan at chromium.org
                CC: tony at chromium.org, abarth at webkit.org,
                    victorw at chromium.org, dpranke at chromium.org


The script did the wrong thing here http://trac.webkit.org/changeset/83024, which was fixed by http://trac.webkit.org/changeset/83030.

I'll draw a hypothetical, simplified version of the error case for foo/baz.html.

1. The test has a result in platform/chromium-mac that actually matches Leopard, but not SnowLeopard.
2. Added "REBASELINE MAC : foo/baz.html = TEXT" to test_expectations.txt.
3. Ran the tool, resulting in r83024.

Here's what I think happened: Since the Leopard bot was passing, no results were downloaded. But the results from the SnowLeopard bot were put in chromium-mac. The results in platform/chromium-mac now matched the results in platform/mac, so they were deleted in r83024.

Instead, the chromium-mac results should have been moved to chromium-mac-leopard (which is was r83030 achieved).

FWIW, r83030 was accomplished by adding the lines back to test_expectations.txt and running the rebaseline tool again. So it worked well for the case of tests that pass on SnowLeopard but fail on Leopard.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.