If you think a bug might affect users in the 57 release, please set the correct tracking and status flags for Release Management.

Firefox WebGL handling issues

RESOLVED INCOMPLETE

Status

()

Core
Canvas: WebGL
RESOLVED INCOMPLETE
5 years ago
a year ago

People

(Reporter: Michael A. Bucko, Unassigned)

Tracking

({crash, testcase})

12 Branch
x86_64
Windows 7
crash, testcase
Points:
---

Firefox Tracking Flags

(Not tracked)

Details

(Whiteboard: webgl-correctness [gfx-noted], URL)

(Reporter)

Description

5 years ago
User Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:12.0) Gecko/20100101 Firefox/12.0
Build ID: 20120420145725

Steps to reproduce:

This (http://www.chromeexperiments.com/detail/swiss-addresses-in-3d/?f=webgl) has been launched using Firefox 12.0. The problem has been tested on Dell Vostro 1540, with Windows 7 Professional.


Actual results:

Device driver (Intel Graphics Accelerator Drivers) has been restarted. Firefox crashed. Other applications could be harmed too. Shockwave Flash plugin crashed. All applications that take advantage of the graphics driver may suffer from the issue and even become incapable of recovery. 

Due to the problems, the application loses WebGL context, which can be handled according to: http://www.khronos.org/webgl/wiki/HandlingContextLost



Expected results:

There should be an information that due to limitations of the device driver the request cannot be handled.
Component: Untriaged → Canvas: WebGL
Product: Firefox → Core
QA Contact: untriaged → canvas.webgl
Marking this as 'new' though I have not explicitly confirmed, lacking the right hardware.
Status: UNCONFIRMED → NEW
Ever confirmed: true
Is this a driver we need to blacklist? Does Chrome crash too, and if not why not since this seems like it's down in the driver?
Keywords: crash, testcase
(Reporter)

Comment 3

5 years ago
Chrome crushes too (although a couple of times a monit appeared (saying that thegraphics driver was unable to handle the very request) and did not crash). 

It might be the case the the roots of the problem lie in the graphics driver.
Over to Joe.
Assignee: nobody → joe
This is not security-sensitive; all drivers sometimes crash, unfortunately.
Group: core-security
(Reporter)

Comment 6

5 years ago
I agree the assessment is very likely to be true. Why the statement is so strict? (no cross-application scripting potential (interface: browser - device driver)).

Additionally, forced closing of all applications that use the very device driver may cause losing information from various opened 3rd party applications.
Sounds like we should blacklist this driver. What's the about:support graphics section say?
In addition to about:support, since comment 0 says that firefox crashes, could you please go to about:crashes and give us a crash link?
(Reporter)

Comment 9

5 years ago
The crash link for mozilla::WebGLContext::MakeContextCurrent() (thank you, Mr. Benoit, for explanation of the procedures) is as follows: 

https://crash-stats.mozilla.com/report/index/bp-df752c7f-fb9c-45bf-a27f-13ebd2120608

My extensions are:

Firebug1.9.2truefirebug@software.joehewitt.com
Live HTTP headers0.17true{8f8fe09b-0bd3-4470-bc1b-8cad42b8203a}
Web Developer1.1.9true{c45c406e-ab73-11d8-be73-000a95be3b12}

My settings are:

accessibility.typeaheadfind.casesensitive1accessibility.typeaheadfind.flashBar0browser.cache.disk.capacity1048576browser.cache.disk.smart_size.first_runfalsebrowser.cache.disk.smart_size_cached_value665600browser.display.background_color#C0C0C0browser.display.use_system_colorstruebrowser.places.smartBookmarksVersion3browser.startup.homepagehttp://google.com/enbrowser.startup.homepage_override.buildID20120601045813browser.startup.homepage_override.mstone13.0browser.tabs.warnOnClosefalseextensions.lastAppVersion13.0keyword.URLhttp://search.conduit.com/ResultsExt.aspx?ctid=CT3072253&SearchSource=2&q=network.cookie.prefsMigratedtrueplaces.database.lastMaintenance1339069380places.history.expiration.transient_current_max_pages76932places.history.expiration.transient_optimal_database_size49904353privacy.cpd.cachefalseprivacy.cpd.downloadsfalseprivacy.cpd.historyfalseprivacy.cpd.sessionsfalseprivacy.sanitize.migrateFx3Prefstruesecurity.warn_viewing_mixedfalse


My graphics:

Opis adapteraIntel(R) HD GraphicsID dostawcy0x8086ID urządzenia0x0046RAM adapteraUnknownSterowniki adapteraigdumd64 igd10umd64 igdumdx32 igd10umd32Wersja sterownika8.15.10.2342Data sterownika3-25-2011Włączone Direct2DtrueWłączone DirectWritetrue (6.1.7601.17789)Parametry ClearTypeGamma: 2200 Pixel Structure: RGB ClearType Level: 100 Enhanced Contrast: 50 Renderer WebGLGoogle Inc. -- ANGLE (Intel(R) HD Graphics) -- OpenGL ES 2.0 (ANGLE 1.0.0.963)Okna ze sprzętowym przyspieszeniem grafiki3/3 Direct3D 10 AzureBackend direct2d
We shouldn't blacklist drivers for single crashes, only for widespread stability issues. We *should* report this to Intel, though.
Thanks for the crash link. This resembles a bug that's been fixed in Firefox 14 (Beta): Bug 760415. Could you try Firefox Beta? Or maybe even go one step further and check if it's fixed in Nightly? (nightly.mozilla.org).
(Reporter)

Comment 12

5 years ago
Tested on Nightly. Crashes as well.
Assignee: joe → nobody
Whiteboard: webgl-correctness

Comment 13

2 years ago
Does this still occur on Firefox 39?  What about on Firefox 40 beta?
I am closing this bug since there's been no further reports of this happening and since comment 13 has gone unanswered for over a year. Please reopen this bug report if the issue still reproduces in a current version.
Status: NEW → RESOLVED
Last Resolved: a year ago
Resolution: --- → INCOMPLETE
Whiteboard: webgl-correctness → webgl-correctness [gfx-noted]
You need to log in before you can comment on or make changes to this bug.