In six weeks, when trunk merges to beta, the browser-chrome chunk that runs the pdf.js tests is going to have a permaleak like https://treeherder.mozilla.org/logviewer.html#?job_id=95033251&repo=try or https://treeherder.mozilla.org/logviewer.html#?job_id=95033462&repo=try on ASan, from bug 1348062. Should be possible to reproduce locally by just changing /config/milestone.txt from 55a1 to 55, but if it's some more subtle part of the central-to-beta changes, the stack of patches in https://treeherder.mozilla.org/#/jobs?repo=try&revision=bfaaa622d12692fedd9853e2e7ca051a19c7f285 should give you a full beta simulation on try.

[Tracking Requested - why for this release]: merge bustage, closed tree, delayed b1, tearing of hair and rending of garments.

Tracking 55+ for this permaorange.

I *think* this comes from the core of how the throttling (interdicted in bug 1312754 ) is currently implemented. I have plans to rewrite it a bit, specifically in relation to this bug - remove the parts that I suspect could cause this leak. Bug 1362071.

If this needs to be fixed quickly, then we may want to back the offending patch (bug 1348062) out. Otherwise see comment 3, bug 1362071 should be fixed in approx 2 weeks.

Still have 5 weeks until the merge (and yeah, I'm carrying a backout of bug 1348062 around in my merge simulation stack, to keep it rebased and handy just in case).

(In reply to Phil Ringnalda (:philor) from comment #5) > Still have 5 weeks until the merge (and yeah, I'm carrying a backout of bug > 1348062 around in my merge simulation stack, to keep it rebased and handy > just in case). Thanks. I will back bug 1348062 out should bug 1362071 be delayed.

Honza, with the soft code freeze coming up, do you think think 1362071 is in shape to land in time?

(In reply to Mike Taylor [:miketaylr] from comment #7) > Honza, with the soft code freeze coming up, do you think think 1362071 is in > shape to land in time? The interesting bug is bug 1365306. And yes, I'm about to r? and land it soon.

So, the probable fix has landed, can you please verify this bug is now fixed?

Indeed, fixed by bug 1365306.