Bug 80447
| Summary: | Too many tests (200+ on SL, 80+ on Lion) tests are failing on Mac port | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Ryosuke Niwa <rniwa> |
| Component: | Tools / Tests | Assignee: | Nobody <webkit-unassigned> |
| Status: | NEW | ||
| Severity: | Normal | CC: | andersca, ap, darin, jberlin, jchaffraix, jhoneycutt, lforschler, mjs, robert, simon.fraser |
| Priority: | P2 | ||
| Version: | 528+ (Nightly build) | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Bug Depends on: | 67509, 80901, 81045, 80869, 80878, 80930 | ||
| Bug Blocks: | |||
Ryosuke Niwa
There are so many tests failing on Snow Leopard bots:
http://build.webkit.org/builders/SnowLeopard%20Intel%20Release%20%28Tests%29
http://build.webkit.org/builders/Lion%20Intel%20Release%20%28Tests%29
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Ryosuke Niwa
http://build.webkit.org/results/SnowLeopard%20Intel%20Release%20(Tests)/r110496%20(37999)/results.html
It seems like a large number of tedsts under java/lc3 are failing with the error like:
3FAIL TEST_CLASS.ambiguous(1) should be BYTE. Threw exception Error: java.lang.IllegalArgumentException: More than one method matching name ambiguous and arguments [java.lang.Double]
4 Method 1: public java.lang.String com.netscape.javascript.qa.lc3.number.Number_008.ambiguous(byte)
5 Method 2: public java.lang.String com.netscape.javascript.qa.lc3.number.Number_008.ambiguous(boolean)
Ryosuke Niwa
+andersca, +jhoneycutt for the java failures. I've seen this failure for a while now way before the bots started to see. If I remember correctly, it started happening after some system update.
Jessie Berlin
I don't know about the java failures yet, but a couple of us are trying to make headway on the Lion failures today.
Julien Chaffraix
Landed some rebaselines I forgot to disable after r110072 in http://trac.webkit.org/changeset/110599.
I have noticed that some table related failures are related to Robert's caption layout work (CC'ing him if he has some thoughts).
Robert Hogan
(In reply to comment #4)
> I have noticed that some table related failures are related to Robert's caption layout work (CC'ing him if he has some thoughts).
There are no actual failures related to captions (that I can see) in:
http://build.webkit.org/builders/Lion%20Intel%20Release%20%28Tests%29/builds/6339/steps/layout-test/logs/stdio
http://build.webkit.org/builders/SnowLeopard%20Intel%20Release%20%28Tests%29/builds/38042/steps/layout-test/logs/stdio
But there are a number of tests relating to the captions changes waiting to be rebaselined in Skipped/test_expectations.txt. They require image rebaselines, and although the bots don't use the pixel results, I have treated this as a blocker to adding just the rebaselined text results myself.