1102681 - Incomplete Webpage Rendering; Inconsistent

Status: RESOLVED DUPLICATE of bug 1067470

(Core :: Graphics, defect)

Description

[draco18s]

Attachment: mozilla-rendering.png

User Agent: Mozilla/5.0 (Windows NT 6.1; rv:33.0) Gecko/20100101 Firefox/33.0
Build ID: 20141106120505

Steps to reproduce:

This is happening on a number of sites, is inconsistent, and tends to "fix" itself by doing anything to cause a redraw of the window (scrolling the page, changing tabs, alt-tabbing to another program, etc.)  Occasionally changing tabs will cause the issue, i.e. I change from one tab to another and nothing updates at all (still displaying the old tab) until I scroll, hit refresh, or alt-tab.

It happens on multiple computers, though the Arcen Games Forum experiences the issue more often than other pages (http://www.arcengames.com/forums/index.php)

In the attached image I have provided 3 screenshots.  The first one is what the page looks like when it's rendering normally.  Then below are two more screenshots of the same page at the same scroll point where some aspect of the rendering failed: in one, its several content divs, in the other, its the background image (1,920px × 1,080px).  Once the background image stopped rendering I had to do a cache-avoid refresh (ctrl-F5) to re-display it as "it contains errors."

Holy bells, I'm even having trouble submitting this bug report because it wants a "component" and that selection doodad is NOT ON MY SCREEN.  Arg.


Actual results:

Whole sections of the webpage stop rendering.


Expected results:

Should have rendered the whole screen.

Comment 1

[:Gijs (he/him)]

Can you try restarting in safe mode? ( https://support.mozilla.org/kb/troubleshoot-firefox-issues-using-safe-mode - this will turn off a bunch of things that might be causing the issue, not just add-ons) Does that fix the problem? And can you provide the "Graphics" section of about:support?

Finally, assuming you have some familiarity with it, are the DOM nodes still there if you open them up in the builtin devtools inspector? (If you don't understand that question or some of the terms in it, don't worry about it and leave it for now)

Comment 2

[draco18s]

Here's the graphics details (original is all from home machine, this is my work machine, which shows the same behavior).

Both did update to 33.1.1 this morning though and I haven't had any issues, but as its inconsistent I can't be sure that it's fixed.  When I see it act up again I'll try safemode.

Graphics
Adapter Description	AMD Radeon HD 6570
Adapter Drivers	aticfx64 aticfx64 aticfx64 aticfx32 aticfx32 aticfx32 atiumd64 atidxx64 atidxx64 atiumdag atidxx32 atidxx32 atiumdva atiumd6a atitmm64
Adapter RAM	2048
Device ID	0x6759
DirectWrite Enabled	false (6.2.9200.16492)
Driver Date	12-6-2013
Driver Version	13.251.0.0
GPU #2 Active	false
GPU Accelerated Windows	0/2 Basic (OMTC)
Vendor ID	0x1002
WebGL Renderer	Google Inc. -- ANGLE (AMD Radeon HD 6570 Direct3D9Ex vs_3_0 ps_3_0)
windowLayerManagerRemote	true
AzureCanvasBackend	skia
AzureContentBackend	cairo
AzureFallbackCanvasBackend	cairo
AzureSkiaAccelerated	0

Comment 3

[draco18s]

Attachment: firefox-render.png

OK, 33.1.1 did not solve the problem.  Attached a screenshot of one of the types of rendering "corruption" that wasn't part of the original report.  Will be restarting in safe-mode after this update.

Comment 4

[draco18s]

And....Just got the same rendering type issue as the last attachment in safe mode.  Was unable to grab a screenshot, however.

Comment 5

[kenkon]

[bugday-20141124]

It is strange that your AzureCanvasBackend is skia. On my PC, it is direct2d. Check that gfx.canvas.azure.backends in about:config is set to default.

And your AMD graphics driver is old. You should download and install the latest stable version from http://support.amd.com/en-us/download.

Finally you can try setting layers.offmainthreadcomposition.enabled to false in about:config as a possible workaround.

Comment 6

[:Gijs (he/him)]

I don't think June 2013 is old enough that we should just tell people to turn off OMTC. Moving to Graphics and needinfo'ing Milan to see if we can do more.

Comment 7

[draco18s]

gfx.canvas.azure.backends is set to default ("direct2d,skia,cairo") here at home.  I don't know what would alter it and given my near-identical setup at work, I don't imagine that it would be different there.

In having safe-mode running all day at work, the issue seems to be most common when closing the active tab (which I do with a middle click) and then changing tabs away from the tab that dumps me to.  Not all the time, but I had it happen three times in quick succession that that might be an avenue to pursue in an attempt to replicate.

Also, while I've got this open, here's my about:settings -> Graphics for my home machine.  Video driver is older still (2012) so I'll go ahead and update that.

Adapter Description	ATI Radeon HD 5570
Adapter Drivers	aticfx32 aticfx32 aticfx32 atiumdag atidxx32 atiumdva
Adapter RAM	1024
Device ID	0x68d9
Direct2D Enabled	true
DirectWrite Enabled	true (6.2.9200.16571)
Driver Date	3-8-2012
Driver Version	8.951.0.0
GPU #2 Active	false
GPU Accelerated Windows	1/1 Direct3D 11 (OMTC)
Vendor ID	0x1002
WebGL Renderer	Google Inc. -- ANGLE (ATI Radeon HD 5570 Direct3D9Ex vs_3_0 ps_3_0)
windowLayerManagerRemote	true
AzureCanvasBackend	direct2d
AzureContentBackend	direct2d
AzureFallbackCanvasBackend	cairo
AzureSkiaAccelerated	0

Comment 8

[Milan Sreckovic [:milan] (needinfo for best results)]

It'd be interesting to see if our recent ATI carded arrival reproduces this.

Comment 9

[Jeff Muizelaar [:jrmuizel]]

I haven't seen anything like this on the AMD machine I have. Can you try reproducing the problem with Firefox Nightly?

Comment 10

[draco18s]

Got it to happen in Nightly (installed it this morning).  Attaching a new screenshot, the banners and central image are from one website, the remainder from another.

Comment 11

[draco18s]

Attachment: nightly.png

Comment 12

[Johan C]

This could be a dupe of bug 1067470. Very similar issue at least, you seem to run into it more than me.

Comment 13

[Johan C]

Oops, sent that before finishing it.

(In reply to Johan C from comment #12)
> Very similar issue at least, you seem to run into it more than me.

Very similar issue at least, only you seem to run into it more than me.

Comment 14

[draco18s]

Possible.  I haven't seen any issues with the tab bar as depicted in the screenshot on that bug, but that could be a result of having a person and/or using Classic Theme Restorer.

Comment 15

[draco18s]

Based on some experiences yesterday I am thinking this issue might be as a result of my system running low on available RAM (although with Firefox allocating 600 MB worth and still having 800+ MB "unused" according to Windows) I am skeptical on why that would be.

But anyway, last night I was doing a lot of image editing with a couple fairly beefy images (one 2048x2048, a 1024x1024, plus at least one more image of varying size) opened in GIMP and the repaint issues were getting worse and worse. To the point where an animated gif changing frames would cause the whole window to flicker black and repaint (badly).

Had this happen on Nightly this morning at work, as well.  An image was tweening (for a work project, it's being referred to as "Ken Burnsing the image") and causing a similar effect, although I am not sure what other program was using up a large portion of RAM other than Firefox.

Neither time did I look at my RAM usage, but closing other applications or restarting Firefox (and having fewer tabs loaded), did alleviate the symptoms.  Next time it gets that bad I'll take a look at my RAM breakdown.

Comment 16

[Virtual_ManPL [:Virtual] 🇵🇱 - (please needinfo? me - so I will see your comment/reply/question/etc.)]

Looking on screenshots (especially second one, where a new page content will be visible after mouse movement on this place) and description it's definitely a duplicate of bug #1067470.