Bug 28397
| Summary: | [Chromium] Upstream Chromium port test baselines | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Dimitri Glazkov (Google) <dglazkov> |
| Component: | WebCore Misc. | Assignee: | Dirk Pranke <dpranke> |
| Status: | RESOLVED FIXED | ||
| Severity: | Normal | CC: | dpranke, michelangelo, mrowe, ojan, pam |
| Priority: | P2 | ||
| Version: | 528+ (Nightly build) | ||
| Hardware: | All | ||
| OS: | All | ||
| Bug Depends on: | |||
| Bug Blocks: | 28395, 28398 | ||
Dimitri Glazkov (Google)
If we want to run layout tests out of svn.webkit.org, we should have our baselines (and expectations) live upstream. This is a master bug to track the progress on this effort.
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Dimitri Glazkov (Google)
This probably shouldn't be tackled head-on: there are currently 24177 files weighing about 260MB in http://src.chromium.org/viewvc/chrome/trunk/src/webkit/data/layout_tests/platform/
I don't think adding these to LayoutTests/platform/ is prudent.
Perhaps they could be an external dependency?
Ojan Vafai
Wouldn't having them as an external dependency mean that people would need to check out chromium code in order to change a test's baseline for chromium?
Dimitri Glazkov (Google)
(In reply to comment #2)
> Wouldn't having them as an external dependency mean that people would need to
> check out chromium code in order to change a test's baseline for chromium?
You're right. With other ports starting to steadily check in their test expectations as well, how will we handle this?
Pam Greene (IRC:pamg)
I currently check out all the platform results for Mac, QT, etc., even though I don't use them. Granted that the platform/chromium results are much larger, it still seems better to me to include them in the repository like any other port. The first update afterward is painful, but the ongoing cost of ~250 MB isn't that bad.
I'd probably support a more general solution to letting people pick which parts of the repository they wanted to check out... but Ojan has a good point about baselines.
Dimitri Glazkov (Google)
We're uploading these straight-up.
Dirk Pranke
*** Bug 36330 has been marked as a duplicate of this bug. ***
Dimitri Glazkov (Google)
Expectations landed as
http://trac.webkit.org/changeset/56278,
http://trac.webkit.org/changeset/56277,
http://trac.webkit.org/changeset/56276,
http://trac.webkit.org/changeset/56274,
http://trac.webkit.org/changeset/56273,
http://trac.webkit.org/changeset/56272,
http://trac.webkit.org/changeset/56271,
http://trac.webkit.org/changeset/56270,
http://trac.webkit.org/changeset/56267,
http://trac.webkit.org/changeset/56259,
http://trac.webkit.org/changeset/56255,
http://trac.webkit.org/changeset/56253,
http://trac.webkit.org/changeset/56252,
http://trac.webkit.org/changeset/56251.