Expensive hit test while dragging when using transform
Reported by
klemenos...@gmail.com,
May 26 2017
|
||||||||||||
Issue descriptionUserAgent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_5) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.36 Steps to reproduce the problem: 1. Go to fiddle: https://jsfiddle.net/5063ngna/ 2. Drag red element over black boxes bellow while recording performance - Performance: OK 3. Now click "toggle performance" button 4. Repeat step 2. - Performance: BAD What is the expected behavior? Like in other browsers, this should not be an expensive rendering operation. What went wrong? - Profile No transform: from 811ms to 5.23s takes 377.1ms - Profile: With transform: from 245ms to 6.85s takes 5592.6ms Did this work before? N/A Chrome version: 58.0.3029.110 Channel: stable OS Version: OS X 10.12.5 Flash Version: Cannot attach profiles - rejected by server due to size limitation
,
May 29 2017
Tested the issue on mac os 10.12.4 using chrome M58 #58.0.3029.110 and followed below steps : !. Launched chrome and navigated to given jsfiddle. 2. opened dev tools , performance section and recorded performance on reload of page and while recording i dragged red element on black boxes and observed the performance rendering. 3. Repeated step 2, but clciked on toggle performance and observed performance rate. Attached screencast for reference. @klemenoslaj0-- Could you please check attached screnncast and confirm us if we had missed any steps in reproducing the issue and also please provide us the screencast of the expected and actual results. Thanks!
,
May 29 2017
Hi @hdodda, sorry for not being precise enough with my description. I was measuring performance while dragging "red div" around - so ~5s of dragenter, dragover and dragleave events. Please find the attached screencast.
,
May 29 2017
Thank you for providing more feedback. Adding requester "hdodda@chromium.org" to the cc list and removing "Needs-Feedback" label. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
May 30 2017
Tested the issue on Mac os 10.12.4 , windows 7 and ubuntu 14.04 suing chrome M58 # 58.0.3029.110 and M60 #60.0.3115.0 and issue is reproduced. Issues is seen from M52 #53.0.2744.0 and is a non-regression issue , as issue behavior is different in earlier versions of chrome. Marking it as untraiged for further inputs on this. Thanks!
,
May 30 2017
Is there any workaround available? Any setting in chrome flags maybe? Thanks for your quick responses.
,
May 30 2017
,
May 31 2017
The transform is placing some content in a different layer, which is apparently hammering hit test performance.
,
Jun 1 2017
Yes, that is probably the reason. Actually to my surprise even position absolute has some effect on performance, but not as dramatic as transform. You can test this scenario here: https://jsfiddle.net/5063ngna/2/
,
Oct 26 2017
Sorry for asking this, but is there any update on this issue?
,
Oct 26 2017
Sorry, no update yet. Can't promise anything right now.
,
Oct 29
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
,
Oct 29
|
||||||||||||
►
Sign in to add a comment |
||||||||||||
Comment 1 by kavvaru@chromium.org
, May 26 2017