Render process freezes when opening multiple subpages on heavy sites
Reported by
blaz...@yandex-team.ru,
Mar 28 2017
|
|||||||||
Issue descriptionSteps to reproduce the problem: 1. Start Browser with --process-per-site flag (preferably debug version of the app, because it's slower) 2. Open a 'heavy' web-site (for example - rbc.ru, or some other heavy news site). 3. Start opening many subpages in new tabs. What is the expected behavior? Current pages doesn't freeze - all touches are handles immediately. What went wrong? Sometimes entire render-process may freeze (stop handling touches) - flag "--process-per-site" causes the render-process to load all subpages in the same process, and the touch events are not prioritized (even though it seems as they should). Note that the freeze is not easy reproducible. Did this work before? N/A Chrome version: 56.0.2924.0 Channel: canary OS Version: Flash Version: Created a fix, will upload it a bit later.
,
Mar 29 2017
Attached a screenrecord with freeze reproducing.
,
Apr 3 2017
Candidate patch: https://codereview.chromium.org/2782553002/
,
Apr 10 2017
,
Apr 11 2017
Could be related to https://bugs.chromium.org/p/chromium/issues/detail?id=498442.
,
Apr 11 2018
This issue has been Available for over a year. If it's no longer important or seems unlikely to be fixed, please consider closing it out. If it is important, please re-triage the issue. Sorry for the inconvenience if the bug really should have been left as Available. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Apr 11 2018
,
Aug 9
Chrome is increasingly moving toward isolating all pages into separate processes so this may become less of an issue over time. However for mid and low end devices I think it would still make sense to see if we can do better when scheduling multiple pages in the same process.
,
Aug 15
Could we check that lowering background page priority helps here?
,
Aug 15
,
Oct 3
I wasn't able to repro this using the last Canary, so I'll consider this fixed. I think that issue 806271 helped here. Feel free to reopen this bug (ideally with a trace) if you can repro this. |
|||||||||
►
Sign in to add a comment |
|||||||||
Comment 1 Deleted