1118695 - Google Maps shows distortion (shapes) with layer

Status: RESOLVED FIXED

(Core :: Graphics, defect)

Description

[lalande.xavier]

Attachment: First time at left and second time at right

User Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:34.0) Gecko/20100101 Firefox/34.0
Build ID: 20141126041045

Steps to reproduce:

1. Go to https://www.google.fr/maps/place/Paris/@48.8558942,2.3542268,14z/data=!4m2!3m1!1s0x47e66e1f06e2b70f:0x40b82c3688c9460?hl=fr

2. In first time, the map display is correct

3. But when the picture 'Earth' appears at left / down of screen the layers of the map appears strange shapes


Actual results:

All the layers of the map are transforms in strange shapes


Expected results:

Nothing deformation

Comment 1

[lalande.xavier]

The version of Mozilla is the last 34.0.5.

The problem appears at the penultimate version it seems.

Comment 2

[:Gijs (he/him)]

Can you go to about:support and copy-paste your graphics information here, please?

Comment 3

[lalande.xavier]

Accélération graphique
Date du pilote	11-16-2010
Description de la carte	AMD RADEON HD 6450
Direct2D activé	true
DirectWrite activé	true (6.2.9200.16492)
Fenêtres avec accélération graphique	1/1 Direct3D 11 (OMTC)
GPU 2 actif	false
ID du périphérique	0x6779
ID du sous-système	21201028
ID du vendeur	0x1002
Pilotes de la carte	aticfx64 aticfx64 aticfx64 aticfx32 aticfx32 aticfx32 atiumd64 atidxx64 atidxx64 atiumdag atidxx32 atidxx32 atiumdva atiumd6a atitmm64
RAM de la carte	1024
Rendu WebGL	Google Inc. -- ANGLE (AMD RADEON HD 6450 Direct3D9Ex vs_3_0 ps_3_0)
Version du pilote	8.783.2.2000
windowLayerManagerRemote	true
AzureCanvasBackend	direct2d
AzureContentBackend	direct2d
AzureFallbackCanvasBackend	cairo
AzureSkiaAccelerated	0

Comment 4

[:Gijs (he/him)]

Not sure if this is graphics or webgl, but either way... Milan, can you have a look? :-)

(Xavier, your GPU driver seems pretty old (2010) - it seems like AMD can provide you with newer ones from http://support.amd.com/en-us/download , which might fix the problem for you sooner than we can patch it)

Comment 5

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

Let's see if the updated driver fixes this problem - we may need to blacklist this combination if it does.

Comment 6

[lalande.xavier]

It's perfect the problem is resolved. 

Thank you very much and sorry for the no-problem of Mozilla.

Comment 7

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

Great news, thanks for the update.  Can you give us the new version of the driver you now have?

David, let's get the original one blacklisted

Comment 8

[David Anderson [:dvander] - inactive, e-mail if emergency]

Attachment: patch

Block D3D11 for ATI versions <= 8.783.2.2000 on Windows 7. This driver version is from November 2010.

The wiki page says we block < 8.741 (June 2010), but bug 792190 changed this so that it's now < 8.62. It looks like the intent was for D3D9 though, so this entry is D3D11 specific.

Comment 9

[Jeff Muizelaar [:jrmuizel]]

(In reply to David Anderson [:dvander] from comment #8)
> Created attachment 8555591 [details] [diff] [review]
> patch
> 
> Block D3D11 for ATI versions <= 8.783.2.2000 on Windows 7. This driver
> version is from November 2010.
> 
> The wiki page says we block < 8.741 (June 2010), but bug 792190 changed this
> so that it's now < 8.62. It looks like the intent was for D3D9 though, so
> this entry is D3D11 specific.

It's not obvious to me that this is a layers problem and not a WebGL problem. It would be nice to have some more information about the problem before blocking.

Comment 10

[David Anderson [:dvander] - inactive, e-mail if emergency]

(In reply to Jeff Muizelaar [:jrmuizel] from comment #9)
> It's not obvious to me that this is a layers problem and not a WebGL
> problem. It would be nice to have some more information about the problem
> before blocking.

Can we just block everything for this exact driver rev then? The drivers are very old, and the user upgraded theirs already. Unless we have a way of re-reproducing this.

Comment 11

[Jeff Muizelaar [:jrmuizel]]

(In reply to David Anderson [:dvander] from comment #10)
> (In reply to Jeff Muizelaar [:jrmuizel] from comment #9)
> > It's not obvious to me that this is a layers problem and not a WebGL
> > problem. It would be nice to have some more information about the problem
> > before blocking.
> 
> Can we just block everything for this exact driver rev then? The drivers are
> very old, and the user upgraded theirs already. Unless we have a way of
> re-reproducing this.

We could for sure. It would be good to have an idea how many users this will affect but maybe we should just do it. Do we need to block on more OS's than just Windows 7?

Comment 12

[lalande.xavier]

This is my new configuration after the update :

Date du pilote	11-20-2014
Description de la carte	AMD Radeon HD 6450
Direct2D activé	true
DirectWrite activé	true (6.2.9200.16492)
Fenêtres avec accélération graphique	2/2 Direct3D 11 (OMTC)
GPU 2 actif	false
ID du périphérique	0x6779
ID du sous-système	21201028
ID du vendeur	0x1002
Pilotes de la carte	aticfx64 aticfx64 aticfx64 aticfx32 aticfx32 aticfx32 atiumd64 atidxx64 atidxx64 atiumdag atidxx32 atidxx32 atiumdva atiumd6a atitmm64
RAM de la carte	1024
Rendu WebGL	Google Inc. -- ANGLE (AMD Radeon HD 6450 Direct3D9Ex vs_3_0 ps_3_0)
Version du pilote	14.501.1003.0
windowLayerManagerRemote	true
AzureCanvasBackend	direct2d
AzureContentBackend	direct2d
AzureFallbackCanvasBackend	cairo
AzureSkiaAccelerated	0

Comment 13

[Emanuel Hoogeveen [:ehoogeveen]]

FWIW, the latest 'legacy' drivers (for Radeon HD devices up to the 4xxx series) on AMD's website all have version 8.970.100.xxxx. The only OS they don't offer those for is Windows XP, which only supports up to D3D9.

Comment 14

[Jeff Muizelaar [:jrmuizel]]

Comment on attachment 8555591 [details] [diff] [review]
patch

Review of attachment 8555591 [details] [diff] [review]:
-----------------------------------------------------------------

Without understanding what the cause is, we should probably block all features.

Comment 15

[David Anderson [:dvander] - inactive, e-mail if emergency]

Attachment: patch

Okay, this just blocks the specific driver version for all features.

Comment 16

[David Anderson [:dvander] - inactive, e-mail if emergency]

https://hg.mozilla.org/integration/mozilla-inbound/rev/1fe938546be8

Comment 17

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/1fe938546be8