The StackOverflowError in parser is in a testcase that regressively tested a stack overflow in SpiderMonkey parser for a JS page that had a single monstruous statement few hundreds of KBytes long that appeared in some, I think Sony, website back in nineties. It was of course not manually coded, but autogenerated from something. You can look it up by the numeric ID of the test in Bugzilla for gory details if you wish. If we never fixed that, I won't necessarily feel bad about that :-)
Attila. On 2008.01.16., at 22:06, Marc Guillemot wrote: > Norris Boyd wrote: >> ... >> I've added the tests Attila mentions above to the skip list >> (incidentally, lc2/JavaToJS/char-002.js works for me on Linux). That >> leaves js1_5/Regress/regress-89443.js and js1_5/extensions/ >> regress-226507.js as tests that work for both me and Attila but fail >> for Marc. I don't know why those tests are problematic. I reran the >> tests and grepped for Failure (I usually use Calwell's tool) and >> didn't get any failures. > > On Linux using Java 1.6.0_03 (with empty JAVA_OPTS) I have > > - Testcase js1_5/Regress/regress-89443.js failed > Failure reason: > JavaScript errors: > null:0: exception from uncaught JavaScript throw: > java.lang.StackOverflowError > > - Testcase js1_5/extensions/regress-226507.js failed > Failure reason: > JavaScript errors: > ../tests/js1_5/extensions/regress-226507.js:305: Too deep recursion > while parsing > > this second failure denotes a StackOverflowError too. > > > Do you run the tests with extra parameters? > > Cheers, > Marc. > -- _______________________________________________ dev-tech-js-engine-rhino mailing list [email protected] https://lists.mozilla.org/listinfo/dev-tech-js-engine-rhino
