We didn't end up using these, and profiles show that computing them can be expensive. (Well, the content blocking log to be precise.) I think we should just revert bug 1493563 part 4 and 11.
(That is backing out parts 4, 5 and 11) On a local tree, with those changes only I see about 30-40% improvement in the test case in bug 1510275.
We should do this before 65 merges, IMO.
Priority: -- → P2
I have a try push, may need some more test fixes before it can be reviewed: https://treeherder.mozilla.org/#/jobs?repo=try&revision=82fb40651bf0395161ea05edd9b92e5f415457d0
The straightforward backouts can land I think, no need for reviews here...
Pushed by firstname.lastname@example.org: https://hg.mozilla.org/integration/mozilla-inbound/rev/c259bf672187 Part 1: Backout changeset cdb8932d1d8d (bug 1493563 - Part 11) for regressing performance https://hg.mozilla.org/integration/mozilla-inbound/rev/82236b4be4c9 Part 2: Backout changeset f8849239da42 (bug 1493563 - Part 5) for regressing performance https://hg.mozilla.org/integration/mozilla-inbound/rev/4f08283a2e3a Part 3: Backout changeset d0997972e4d4 (bug 1493563 - Part 4) for regressing performance
You need to log in before you can comment on or make changes to this bug.