Bug 36384 – REGRESSION: platform/mac/fast/loader/file-url-mimetypes-2.html failed on Snow Leopard Debug Bot

• Status: RESOLVED
• Resolution: FIXED
• Component: Tools / Tests

[Eric Seidel (no email)]

REGRESSION: platform/mac/fast/loader/file-url-mimetypes-2.html failed on Snow Leopard Debug Bot

http://build.webkit.org/results/SnowLeopard%20Intel%20Leaks/r56247%20(5005)/platform/mac/fast/loader/file-url-mimetypes-2-diffs.txt
--- /Volumes/Data/WebKit-BuildSlave/snowleopard-intel-leaks/build/layout-test-results/platform/mac/fast/loader/file-url-mimetypes-2-expected.txt	2010-03-19 11:36:55.000000000 -0700
+++ /Volumes/Data/WebKit-BuildSlave/snowleopard-intel-leaks/build/layout-test-results/platform/mac/fast/loader/file-url-mimetypes-2-actual.txt	2010-03-19 11:36:55.000000000 -0700
@@ -1,4 +1,4 @@
-example-js.js has MIME type application/x-javascript
+example-js.js has MIME type text/javascript
 example-jscript.jscript has MIME type text/javascript
 example-kar.kar has MIME type audio/midi
 example-l.l has MIME type text/plain

I'm not sure who would know what might be wrong here.  It's been failing every build for longer than the buildbot can remember.

Added by Brady, reviewed by Sam 9 months ago:
http://trac.webkit.org/browser/trunk/LayoutTests/platform/mac/fast/loader/file-url-mimetypes-2.html

[Brady Eidson]

It DEFINITELY has not been failing for a majority of those 9 months.

We need to know when it broke.

It doesn't fail locally for me on SnowLeopard.

[Eric Seidel (no email)]

Oh, I'm sure it hasn't been failing for 9 months either. :)

I also doubt it's related to your test, rather I expect it's a system configuration issue on the bot or some other change to WebCore.

Either way, the Snow Leopard Bot is current hamstrung by this failure.

[Mark Rowe (bdash)]

This is probably the Dashcode MIME type issue again.

[Mark Rowe (bdash)]

> I'm not sure who would know what might be wrong here.  It's been failing every
> build for longer than the buildbot can remember.

The historical results for test runs are available all the way back through r47483 for that particular bot, and as far back as r42302 for bots that existed prior to that date.  It’s pretty trivial to determine from that information that it was <http://trac.webkit.org/changeset/56014> that was the first to result in this particular test failing.

[Mark Rowe (bdash)]

It’s green again after working around the Dashcode MIME type issue (<rdar://problem/6240033>).

<http://build.webkit.org/results/SnowLeopard%20Intel%20Leaks/r56295%20(5031)/results.html>

[Eric Seidel (no email)]

(In reply to comment #4)
> > I'm not sure who would know what might be wrong here.  It's been failing every
> > build for longer than the buildbot can remember.
> 
> The historical results for test runs are available all the way back through
> r47483 for that particular bot, and as far back as r42302 for bots that existed
> prior to that date.  It’s pretty trivial to determine from that information
> that it was <http://trac.webkit.org/changeset/56014> that was the first to
> result in this particular test failing.

Thank you Mark!

I'd love to know more about how you traced through the results.  I was not aware we kept results that long.

I assume you used:
http://build.webkit.org/results/SnowLeopard%20Intel%20Leaks/
and just compared revisions?

Or did you use individual build pages?
http://build.webkit.org/builders/SnowLeopard%20Intel%20Leaks/builds/5081

I'm interested in your methodology as I'd like to write a tool to do this sort of diagnosis automatically.