This is a regression? Any blame window? /be
yes, it is a regression. The blame window is probably a month. I have to step out for a few hours but will try to narrow it down later this evening.
I'm not sure it is a regression after all. Originally, with the first patch (attachment 383270 [details] [diff] [review]) and slightly different modification to reftest.js in bug 469718 I didn't see this behavior. With the new version of reftest.js, I see this behavior back in March. Prior to March crashes prevent the test from running long enough. The use of nsIDebug2 prevents testing prior to Jan 8.
The culprit is e4x/Regress/regress-474319.js. If that is disabled for debug builds, the test run will run to completion reasonably fast (for debug).
Assignee: general → bclary
Status: NEW → RESOLVED
Last Resolved: 9 years ago
Resolution: --- → FIXED
Summary: debug browser jsreftests slow to a crawl → e4x/Regress/regress-474319.js must reset gczeal.
You need to log in before you can comment on or make changes to this bug.