RESOLVED FIXED Bug 65579
Several Tests of http://test262.ecmascript.org/ fail
https://bugs.webkit.org/show_bug.cgi?id=65579
Summary Several Tests of http://test262.ecmascript.org/ fail
lars.sonchocky-helldorf
Reported 2011-08-02 16:21:18 PDT
Several Tests of http://test262.ecmascript.org/ fail What is test262? test262 is a test suite intended to check agreement between JavaScript implementations and the ECMA-262 Specification (currently 5th Edition). The test suite contains thousands of individual tests, each of which tests some specific requirements of the ECMAScript specification. currently (WebKit nightly Version 5.1 (6534.50, r92131)) 866 out of 10927 tests fail
Attachments
Results as XML as downloaded from the test site (without detailed description of the failures) (739.52 KB, application/xml)
2011-11-16 13:39 PST, lars.sonchocky-helldorf
no flags
Results as RTF as copied from the test site (failures only with detailed description) (566.71 KB, application/octet-stream)
2011-11-16 13:40 PST, lars.sonchocky-helldorf
no flags
Gavin Barraclough
Comment 1 2011-08-02 17:43:01 PDT
Gavin Barraclough
Comment 2 2011-08-08 12:20:59 PDT
*** Bug 41876 has been marked as a duplicate of this bug. ***
Gavin Barraclough
Comment 3 2011-08-12 01:14:39 PDT
*** Bug 36301 has been marked as a duplicate of this bug. ***
Gavin Barraclough
Comment 4 2011-09-22 15:34:19 PDT
Update: * 51 new tests have been added which put us at 866 out of 10978 fail. * Function.bind is now implemented, fixing 90 tests, putting us at 776 out of 10978 fail (775 when I fix a trivial issue that I set the number of args incorrectly!). Looks like the biggest causes of failure are still the inability to delete static properties, and incorrect handling of numeric accessor properties on arrays.
Gavin Barraclough
Comment 5 2011-10-14 18:34:51 PDT
This no longer depends on r61014. That bug is not fully fixed, but has been fixed in all aspects that would be visible within ES5.
Gavin Barraclough
Comment 6 2011-10-18 01:03:59 PDT
As of r97719 we're down to 568/11016 failing (or more optimistically, up to 10448 passing!)
lars.sonchocky-helldorf
Comment 7 2011-11-16 13:37:22 PST
One note on the performance: Running those tests in WebKit nightly is way slower than running them in the current Mozilla Firefox 8.0. Firefox seems to be about a magnitude faster than WebKit on that testsuite (same machine of course) Test Results btw.: Tests To Run: 11029 | Total Tests Ran: 11029 | Pass: 10479 | Fail: 550 | Failed To Load: 0 detailed results attached Tested with WebKit Nightly Version 5.1.1 (6534.51.22, r100296)
lars.sonchocky-helldorf
Comment 8 2011-11-16 13:39:02 PST
Created attachment 115439 [details] Results as XML as downloaded from the test site (without detailed description of the failures)
lars.sonchocky-helldorf
Comment 9 2011-11-16 13:40:28 PST
Created attachment 115440 [details] Results as RTF as copied from the test site (failures only with detailed description)
lars.sonchocky-helldorf
Comment 10 2012-01-11 19:15:01 PST
To be more specific: the testsuite at http://test262.ecmascript.org/ runs quite fast in the beginning it just slows down significantly during runtime. new results btw.: Tests To Run: 11108 | Total Tests Ran: 11108 | Pass: 10923 | Fail: 185 | Failed To Load: 0
Gavin Barraclough
Comment 11 2012-03-01 16:40:53 PST
Tests To Run: 11563 | Total Tests Ran: 11563 | Pass: 11563 | Fail: 0 | Failed To Load: 0 Fixed in r109461
Note You need to log in before you can comment on or make changes to this bug.