Followup to bug 500491. The patch for bug 497060 broke setting JS options from the shell and browser. The patch "fixing" bug 500491 only fixed the shell part. Tests that set javascript.options from the browser are still broken.

The user.js sets content jit to false initially. This is for testing over a range of revisions where the range may include the revision which switched the default from jit off to jit on. It also enables XPConnect and friends so the test can turn javascript.options.* on/off as needed. By default, unless told otherwise the test framework itself will turn jit off before loading the test. If ;jit is appended to the test url, the framework will turn jit on prior to loading the actual test but after loading the test framework. to run a test with the jit off: http://test.bclary.com/tests/mozilla.org/js/js-test-driver-standards.html?test=ecma_3%2FObject%2F8.6.1-01.js;language=type;text/javascript to run a test with jit on: http://test.bclary.com/tests/mozilla.org/js/js-test-driver-standards.html?test=ecma_3%2FObject%2F8.6.1-01.js;language=type;text/javascript;jit javascript.options.jit.content: false, without ;jit : passes javascript.options.jit.content: true, without ;jit : fails javascript.options.jit.content: false, with ;jit : passes javascript.options.jit.content: true, with ;jit : passes You can open a new window with about:config and filter on javascript.options and run the test under Firebug 1.5 to watch that the preference setting correctly sets the values as reflected in about:config. However when the content jit is initially turned on, it appears that even though the strict pref is turned on, it is not activated.

Yeah, I know. the "correct" fix involves ... something a little bit involved. I'll link in the blocking bug. Sorry, this *is* on my list -- presently the top non-merge error -- I have just been deferring it while chasing the merge-blocking work :(

As near as I can tell, the patch for bug 503694 that just landed ( http://hg.mozilla.org/tracemonkey/rev/f5fd44e7cfdb ) fixes this. At least it seems to in my browser. Can you confirm, bc?

I'll get some fresh builds and let you know. thanks!

tests in bug 500491 comment 6 pass in tracemonkey/mac. I haven't got the results of a full run but it looks good so far. Thanks.

yes. everything is fixed on tracemonkey.

bug 503694 landed on mc http://hg.mozilla.org/mozilla-central/rev/f5fd44e7cfdb and 1.9.2 http://hg.mozilla.org/releases/mozilla-1.9.2/rev/2fbdb5c62928

no failures on 1.9.0 and later since 2009-09-06. v1.9.2