The regression range seems unrelated to your issue. Can you please rerun mozregression?
OK, done some tests with another computer and it seems not affected by my bug (nVidia GT 610). So, on the bugged computer (AMD R7 240), i can resolve the bug by disabling hardware acceleration. And now, i have an another regression range : https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=89ffef35daade74a97ad36134a78898a8cf3155b&tochange=01d20d2ebb57d1be52e87e4666e661577335c16e
@Reporter, Could you attach about:support results of the affected one?
Well, seems i can't reproduce the bug, but it's after disabling/re-enabling hardware acceleration that the bug resolved itself... Strange... sorry for the inconvenience.
For info, first Mozregression was on the AMD drivers 17.10.1 (2 GB memory leak max), second one was on the 17.10.3 (1.4 GB memleak max) Now i can't reproduce it with Mozregression since memory is stable about 500-600MB...
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:58.0) Gecko/20100101 Firefox/58.0 Firefox: 58.0a1, Build ID 20171031220132 I have not managed to reproduce this issue, using the steps provided in the description, on Windows 10 x64, Ubuntu 14.04 x64 and Mac 10.12.6 with the latest Firefox release (56.0.2) and latest Nightly (58.0a1) builds. @1for-matik, if you can no longer reproduce the issue, do you think we can close this as resolved - worksforme? And if you reproduce the issue in the future, could you please save a memory report using the about:memory page?
Yes we can close this bug. I still can't reproduce the bug and have no idea what was causing it even with Mozregression download builds (since Mozregression profile should be clean, so no webrender or others optional stuff activated).