Closed Bug 787892 Opened 13 years ago Closed 13 years ago

Huge slowdown in canvas demo from Ionmonkey landing

Categories

(Core :: JavaScript Engine, defect)

Product:
Core
Component:
JavaScript Engine
Version:
15 Branch
Platform:
x86
All
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

()

Status:
RESOLVED DUPLICATE of bug 786126
Tracking Flags:
Tracking Status
firefox15 - affected
firefox16 - affected
firefox18 - ---

People

(Reporter: matthew, Unassigned)

References

Details

(Keywords: qawanted, regression, verifyme, Whiteboard: [js:t])

User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_6_8) AppleWebKit/534.57.2 (KHTML, like Gecko) Version/5.1.7 Safari/534.57.2 Steps to reproduce: Go the following URL and scratch the image. It should another image beneath. This effect works in other browser I've tested and in previous version of FF (IE/MAC) but not after upgrading to v15. http://rabidflash.com/authoring/codecanyon/scratch/ Actual results: Scratching (holding mouse down and moving) has no visible effect. I then go to the firefox toolbar menu and click an item such as 'Find' and then I see the effect of my scratching. Expected results: The feedback of the scratch should be immediate. I'm not the developer of this plugin so sorry I can't be more specific technically but I'm led to beleive this is a bug as it was working in previous FF and works in other browsera.
I can only reproduce without HWA in Firefox15 and Firefox16Beta Windows7. I cannnot not reproduce in Ayrora17.a2 and Nightly18.a1 with/without HWA Windows7. Regression range(Beta channel) Good: http://hg.mozilla.org/releases/mozilla-beta/rev/791245980a6e Mozilla/5.0 (Windows NT 6.1; WOW64; rv:15.0) Gecko/20100101 Firefox/15.0 ID:20120808131812 Bad: http://hg.mozilla.org/releases/mozilla-beta/rev/5b309ad1a88a Mozilla/5.0 (Windows NT 6.1; WOW64; rv:15.0) Gecko/20100101 Firefox/15.0 ID:20120814224555 Pushlog: http://hg.mozilla.org/releases/mozilla-beta/pushloghtml?fromchange=791245980a6e&tochange=5b309ad1a88a Regression range(Aurora channel) Good: http://hg.mozilla.org/releases/mozilla-aurora/rev/4cbf6aa4f022 Mozilla/5.0 (Windows NT 6.1; WOW64; rv:16.0) Gecko/20120813 Firefox/16.0 ID:20120814111602 Bad: http://hg.mozilla.org/releases/mozilla-aurora/rev/c2d5393d005c Mozilla/5.0 (Windows NT 6.1; WOW64; rv:16.0) Gecko/20120814 Firefox/16.0 ID:20120814123301 Pushlog: http://hg.mozilla.org/releases/mozilla-aurora/pushloghtml?fromchange=4cbf6aa4f022&tochange=c2d5393d005c Suspected : Bug 773097
Status: UNCONFIRMED → NEW
tracking-firefox15: --- → ?
tracking-firefox16: --- → ?
Component: Untriaged → Canvas: 2D
Ever confirmed: true
Keywords: regression
OS: Mac OS X → All
Product: Firefox → Core
Dupe of bug 787623?
Blocks: 773097
Depends on: 787623
As things stand, I don't believe this regression is significant enough to warrant tracking for FF15. Tracking for FF16 and sending over to Joe for investigation, however.
Assignee: nobody → joe
tracking-firefox15: ? → -
tracking-firefox16: ? → +
Is this fixed in Nightly now that bug 787623 has landed?
Adding verifyme to find out.
Keywords: qawanted, verifyme
Assignee: joe → jbecerra
(In reply to David Baron [:dbaron] from comment #4) > Is this fixed in Nightly now that bug 787623 has landed? I tested in FF16 and with HWA off, the bubbling effect is very slow to appear. http://rabidflash.com/authoring/codecanyon/scratch/
Joe - can you take a look at this for FF16?
Assignee: jbecerra → joe
(In reply to David Baron [:dbaron] from comment #4) > Is this fixed in Nightly now that bug 787623 has landed? Build identifier: Mozilla/5.0 (Windows NT 6.1; rv:18.0) Gecko/18.0 Firefox/18.0 (18.0a1 (2012-09-17)) Built from http://hg.mozilla.org/mozilla-central/rev/3f0587ce1774 I can reproduce the issue on the latest Nightly build on Win 7 32bit, with and without HW acceleration enabled. Steps: 1. Go to http://rabidflash.com/authoring/codecanyon/scratch/ 2. Scratch the pictures using the mouse Result: Scratching effect is delayed, it is visible only after a few seconds, while the browser is in non-responsive state (can not open/close/switch tabs, select text, resize window, move focus to other browser areas, etc) I cannot reproduce with latest Aurora: Build identifier: Mozilla/5.0 (Windows NT 6.1; rv:17.0) Gecko/17.0 Firefox/17.0 (17.0a2 (2012-09-17)) Built from http://hg.mozilla.org/releases/mozilla-aurora/rev/88b15f553474 With latest beta build and HW acceleration disabled, the percentage updates instantly, but the image scratch is delayed 1-2 seconds, however, with no browser unresponsiveness. Enabling HW acceleration resolves the painting (scratching effect) delay - it appears "almost" instantly on first 2 pictures, and instantly on the 3rd. Mozilla/5.0 (Windows NT 6.1; rv:16.0) Gecko/20100101 Firefox/16.0 (beta 3) Built from http://hg.mozilla.org/releases/mozilla-beta/rev/de3f06a549de
I've tried twice to get a reasonable regression window for this bug, and twice failed; the most recent time I got a merge commit from mozilla-central to whatever branch Ionmonkey was being developed on. (97e8aecf1730) You need to do two things to get regression windows: set gfx.canvas.azure.enabled to false explicitly, and if compiling manually, enable optimization.
Keywords: regressionwindow-wanted
Regarding performance problem. #1 Scratching effect is delayed, but less than 1.0sec This is regression of Bug 550309 #2 Scratching effect is significantly delayed, more than 1.0-2.0sec This is regression of IonMonkey. #1 Regression window Good: http://hg.mozilla.org/mozilla-central/rev/e5f6caa40409 Mozilla/5.0 (Windows NT 6.1; WOW64; rv:14.0) Gecko/20120315 Firefox/14.0a1 ID:20120315214227 Bad: http://hg.mozilla.org/mozilla-central/rev/ee56787a20fb Mozilla/5.0 (Windows NT 6.1; WOW64; rv:14.0) Gecko/20120316 Firefox/14.0a1 ID:20120316025529 Pushlog: http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=e5f6caa40409&tochange=ee56787a20fb Last Good: 66bed1bee130 First Bad: 64daf7f88b87 Triggered by: 64daf7f88b87 Ms2ger — Bug 550309 - Part e: Remove custom quickstub for getImageData and return an actual ImageData; r=bz #2 Regression window Bad with #1 problem: http://hg.mozilla.org/mozilla-central/rev/d042ad078f43 Mozilla/5.0 (Windows NT 6.1; WOW64; rv:18.0) Gecko/18.0 Firefox/18.0 ID:20120911065552 Bad: http://hg.mozilla.org/mozilla-central/rev/fdfaef738a00 Mozilla/5.0 (Windows NT 6.1; WOW64; rv:18.0) Gecko/18.0 Firefox/18.0 ID:20120911103453 Pushlog: http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=d042ad078f43&tochange=fdfaef738a00 javascript.options.ion.content = false remove the #2 problem. gfx.canvas.azure.enabled = false does not help for both #1 and #2 regression..
Alice, you are incredible. Thank you. I'll file a follow-up bug for the #1 regression and morph this bug into the #2 regression. Also, I don't think this needs to be tracked for Firefox 16, considering it's totally usable (just a little slowdown) for #1 regression.
Assignee: joe → general
tracking-firefox16: + → ?
Component: Canvas: 2D → JavaScript Engine
Keywords: regressionwindow-wanted
Summary: HTML5/Canvas Effect not working → Huge slowdown in canvas demo from Ionmonkey landing
The follow-up bug is bug 792561.
The ion thing is likely bug 786126.
Depends on: 786126
Joe, Is still still an issue considering some landings in bug 786126 ?
It's significantly better (though not perfect) on Nightly now, yes.
How much of it is bug 792561?
When I visit this site on Firefox 17, the performance seems the same as on Firefox 19. Should we RESOLVE DUPE of bug 786126?
(In reply to David Anderson [:dvander] from comment #17) > When I visit this site on Firefox 17, the performance seems the same as on > Firefox 19. Should we RESOLVE DUPE of bug 786126? Sounds good - we'll continue to track that bug instead given the results of your testing.
tracking-firefox18: + → -
Whiteboard: [js:t]
Status: NEW → RESOLVED
Closed: 13 years ago
Resolution: --- → DUPLICATE
You need to log in before you can comment on or make changes to this bug.