1260040 - Black rectangles on Intel GPUs

Status: RESOLVED WORKSFORME

(Core :: Graphics, defect, P3)

Description

[akavel]

Attachment: Zrzut ekranu (1).png

User Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:45.0) Gecko/20100101 Firefox/45.0
Build ID: 20160315153207

Steps to reproduce:

Highly nondeterministic. Just browsing the Internet. That said, it seems to kinda happen more often after waking Windows from sleep, and after watching youtube. Also, it seems to not happen if I disable GPU acceleration in Fx (but only after I restart Fx after changing the option). Please note I've observed it already on 3 PCs. One is a laptop, where I've observed it since long ago (probably more than year), but thought it's maybe a hardware issue, so didn't report. Recently however I've observed it on 2 brand new MS Surface Pro 4 tablets. Note also I have NoScript installed on all of those Fx instances. 




Actual results:

 Black rectangles show on page. I'll try to attach a few images. They move with page when scrolling, and most often disappear when scrolled out of screen and back. But sometimes similar artifacts show up near the scrolling area too, and behave the same.

 I've also sometimes observed much bigger black rectangles, often covering whole DOM elements or whole page; not 100% sure if it's the same issue, but to me it seems to happen at similar times (and also observed this on all 3 PCs).


Expected results:

Normal page contents should show, without the artifacts.

Comment 1

[akavel]

Attachment: Zrzut ekranu (2).png

Comment 2

[akavel]

Attachment: Zrzut ekranu (3).png

Comment 3

[akavel]

Attachment: Zrzut ekranu (4).png

Comment 4

[akavel]

Attachment: Zrzut ekranu (5).png

Comment 5

[akavel]

Attachment: Zrzut ekranu (1).png

Comment 6

[akavel]

Attachment: Zrzut ekranu (2).png

Comment 7

[akavel]

Attachment: Zrzut ekranu (3).png

Comment 8

[Loic]

Type about:support in the location page and copy here the section "graphics".

Comment 9

[akavel]

Asynchroniczne przewijanie/powiększanie	brak
Data sterownika	1-12-2016
GPU #2 Active	false
ID dostawcy	0x8086
ID podsystemu	00151414
ID urządzenia	0x191e
Okna ze sprzętowym przyspieszeniem grafiki	1/1 Direct3D 11 (OMTC)
Opis adaptera	Intel(R) HD Graphics 515
RAM adaptera	Unknown
Renderer WebGL	Google Inc. -- ANGLE (Intel(R) HD Graphics 515 Direct3D11 vs_5_0 ps_5_0)
Sprzętowe dekodowanie H.264	Yes
Sterowniki adaptera	igdumdim64 igd10iumd64 igd10iumd64 igd12umd64 igdumdim32 igd10iumd32 igd10iumd32 igd12umd32
Wersja sterownika	20.19.15.4364
windowLayerManagerRemote	true
Włączone Direct2D	true
Włączone DirectWrite	true (10.0.10586.0)
AzureCanvasBackend	direct2d 1.1
AzureContentBackend	direct2d 1.1
AzureFallbackCanvasBackend	cairo
AzureSkiaAccelerated	0

Comment 10

[Loic]

I'm pretty sure disabling HWA will fix your issue.
https://support.mozilla.org/en-US/kb/forum-response-disable-hardware-acceleration

Do you have more recent drivers for your Intel GPU? (what's your model?)

Comment 11

[akavel]

As I wrote: "it seems to not happen if I disable GPU acceleration in Fx". So, yes, I already noticed. That said, with acceleration disabled, youtube gets awfully choppy when fullscreen, and WebGL doesn't work. So, unfortunately I can only see it as a kinda occasional  workaround, not a fix. I live with that knowledge for quite long already, so I can only say it's really not much of a news to me, I'm afraid.

The laptop is: http://www.notebookcheck.net/Sony-Vaio-SVS1511V9E.84976.0.html The other PCs, as I wrote: "2 [...] MS Surface Pro 4 tablets". Updated to most recent drivers. That's why I'm even bothering you with my reporting of some wonky non-deterministic behavior actually, because I noticed this issue on more than 1 PC already, and also they all seem to have Intel GPUs, which are somewhat popular too, I'd think.

Also, I'd be really happy to help debug this with some additional tools or something if possible, though I'm not a GPU guy so I don't know if there are any such ones at all for GPUs.

Comment 12

[akavel]

Attachment: Zrzut ekranu (7).png

Comment 13

[akavel]

Attachment: Zrzut ekranu (8).png

Comment 14

[akavel]

Attachment: Zrzut ekranu (9).png

Comment 15

[akavel]

Attachment: Zrzut ekranu (10).png

Comment 16

[u279076]

You first noted this bug with Firefox 45. Could you test Firefox 44 and see if this bug reproduces there? If it doesn't then this is likely a regression and I'll walk you through debugging this further.

Comment 17

[Andrew Overholt [:overholt]]

Let's just re-open this if it happens again. I was experiencing similar problems that went away with a driver update.