Issue metadata
Sign in to add a comment
|
slither.io janky after C69 update
Reported by
kamil...@gmail.com,
Sep 4
|
||||||||||||||||||||||
Issue descriptionUserAgent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_13_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/69.0.3497.81 Safari/537.36 Example URL: http://slither.io/ Steps to reproduce the problem: 1. Install/Launch the latest Chrome 69.0.3497.81 2. Load http://slither.io/ in a new tab 3. Enter a nick name into the field and select "Play" 100% reproducible using the above STR. If you run the same website under 68.0.3440.106, you'll see the difference in performance. What is the expected behavior? Game/Content shouldn't be janking this badly. What went wrong? After updating from 68.0.3440.106 -> 69.0.3497.81, http://slither.io has become very janky. It used to be really smooth under 68.0.3440.106 so something has definitely regressed. Does it occur on multiple sites: N/A Is it a problem with a plugin? No Did this work before? Yes 68.0.3440.106 Does this work in other browsers? Yes Chrome version: 69.0.3497.81 Channel: stable OS Version: OS X 10.13.6 Flash Version: Also reproducible with other browsers that are using 69.0.3497.81, Example: * https://github.com/brave/browser-laptop/releases/tag/v0.23.203beta
,
Sep 5
,
Sep 5
Thanks for filing the issue! Able to reproduce the issue on reported chrome stable version 69.0.3497.81, Dev 70.0.3534.4 and canary 71.0.3543.0 using Windows 10 and Mac 10.13.1, when downloading the builds from https://www.chromium.org/getting-involved/dev-channel Note: Couldn't check the Linux behaviour as we are facing few technical difficulties while launching the game. But the same issue isn't seen when checking on builds(signed/unsigned) downloaded from build repository. In such case, it's out of scope for us to triage further to find the regression range. Hence, tentatively marking it as Untriaged.
,
Sep 6
Providing the list of the variations from Beta installation of 69.0.3497.81 where the issue is seen. Variations c134752e-95b424ac d01ab0d3-a1f91ada 3e006338-f23d1dea b7e2524c-3f4a17df a6674cf-4f295bf2 b1681d28-803f8fc4 cc20827f-dea79bd3 8fe39baa-3f4a17df 38eb801c-3f4a17df c27fec31-2d5b6ed9 7c1bc906-8122a015 47e5d3db-3d47f4f4 125b7f68-65bced95 d442dfb7-eeca42f7 9ca1387e-65bced95 41e765a5-f23d1dea 1149accc-65bced95 4dc30737-b8a5ea08 a582a1b8-ad75ce17 8ee5ed19-ca7d8d80 74658432-ca7d8d80 3042ad4b-ca54bb47 e56c5101-ad2fa222 249dd49a-28165b59 44827ee5-43146c13 edbcf7c5-961c461c 5485fc4d-3f4a17df de47491b-33c3eba5 93731dca-e89d496c 9e5c75f1-6943e6a2 2594bdf4-f1cfee3a 2b86fd96-f23d1dea 45df16ca-ca7d8d80 4ea303a6-9fba3b4c 6e6e0c7e-3f17a7d8 d92562a9-65bced95 7aa46da5-c946b150 4da5ae82-f23d1dea de52c077-65bced95 6973a1cf-3f4a17df cc54eb06-3f4a17df 58a025e3-36e97b2c ad6d27cc-1627c3cf df072bba-9a6c5085 f242806f-5810b593 4bc337ce-535cb40a 9a2f4e5b-3fe9c4dc 494d8760-52325d43 3ac60855-3ec2a267 f296190c-c4f4d776 4442aae2-6e597ede ed1d377-e1cc0f14 12e17bc5-e1cc0f14 75f0f0a0-6bdfffe7 e2b18481-e1cc0f14 e7e71889-e1cc0f14 5e60f31d-1410f10 3a4029d-ca7d8d80 94e68624-803f8fc4 cc73f8a1-a2d707c6 10a311eb-f23d1dea 8834fcca-cf4f6ead 81c6897f-672c76c9 Not 100% sure but suspecting VideoSurfaceLayer. mlamouri@: Is this enabled via Finch on Beta/Stable channel? kamiljoz@: Does disabling #enable-surfaces-for-videos under chrome://flags makes any difference?
,
Sep 6
ajha@, what makes you think it's VideoSurfaceLayer given that the game doesn't seem to be using any video.
,
Sep 6
ajha@: nope, disabled #enable-surfaces-for-videos via chrome://flags and attempted going through the same STR. Still pretty janky. Also restarted Chrome after disabling just incase, same results.
,
Sep 7
Thanks kamiljoz@ for the confirmation. As per C#3 the issue seems to be happening on canary/dev/beta/stable channel builds downloaded from https://www.chromium.org/getting-involved/dev-channel but doesn't repro on signed builds downloaded from build repository so probably this is related to some experiment or Finch trial on these channel. Somehow disabling #enable-surfaces-for-videos worked fine for me on first run and as VideoSurfaceLayer featured in Variations from C#4 hence suspected the same. But this looks unrelated due to no video on the page+issue persists as confirmed C#6. Not 100% sure about the other root cause of this. Will try a bisect of this based on CPU usage and Energy Impact under Activity Monitor as these seem to spike considerably when playing this game.
,
Sep 7
This is showing really inconsistent behavior when trying to bisect. Issue is not repro consistently on downloaded signed/unsigned build whereas this is easily reproducible on latest beta/stable(69.0.3497.81) build downloaded from https://www.chromium.org/getting-involved/dev-channel as tested on Windows-10. kamiljoz@: Could you please check if the issue is seen on Dev and Canary or not? You can download the same from here https://www.chromium.org/getting-involved/dev-channel, also please confirm whether there is any effect on jankiness when you turn off the hardware acceleration. Apprecite your help!
,
Sep 7
WIW, I can reproduce the regression in Beta but not Canary on Mac.
,
Sep 8
,
Sep 8
|
|||||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||||
Comment 1 by gov...@chromium.org
, Sep 4Labels: Needs-Triage-M69