1858995 - Firefox crashes often after outputting “wr_renderer_render: OutOfMemory”

Status: RESOLVED INACTIVE

(Core :: Graphics: WebRender, defect)

Description

[bugzilla.mozilla.org]

User Agent: Mozilla/5.0 (X11; Linux i686; rv:109.0) Gecko/20100101 Firefox/118.0

Steps to reproduce:

I used Firefox to browse the web.

Actual results:

Firefox crashes often (sometimes multiple times a day). This most often happens when navigating to a URL for an standalone photo-sized image or a PDF document.

The output to the terminal from Firefox looks like this:

[GFX1-]: wr_renderer_render: OutOfMemory
[GFX1-]: Handling webrender error 2
[GFX1-]: wr_renderer_render: OutOfMemory
[GFX1-]: Handling webrender error 2
[GFX1-]: wr_renderer_render: OutOfMemory
[GFX1-]: Handling webrender error 2
[GFX1-]: wr_renderer_render: OutOfMemory
[GFX1-]: Handling webrender error 2
[…]
[GFX1-]: CompositorBridgeChild receives IPC close with reason=AbnormalShutdown
Exiting due to channel error.
Exiting due to channel error.
Exiting due to channel error.
Exiting due to channel error.
Exiting due to channel error.
Exiting due to channel error.
Exiting due to channel error.
Exiting due to channel error.
[GFX1-]: CompositorBridgeChild receives IPC close with reason=AbnormalShutdown
Exiting due to channel error.
Exiting due to channel error.
Exiting due to channel error.
Exiting due to channel error.
Exiting due to channel error.
Exiting due to channel error.
Exiting due to channel error.

I have verified that I do have enough free RAM (my Thinkpad T60 has 2GB RAM).

Is this crash maybe related to GPU memory?

Expected results:

Firefox should not have crashed. It did not do that in the past.

Comment 1

[bugzilla.mozilla.org]

My graphics hardware, according to lspci:

00:02.0 VGA compatible controller: Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller (rev 03)
00:02.1 Display controller: Intel Corporation Mobile 945GM/GMS/GME, 943/940GML Express Integrated Graphics Controller (rev 03)

Do you need more information to triage this?

Comment 2

[Lee Salzman [:lsalzman]]

Can you please attach the contents of your "about:support" information?

Also, do you have any particular URL that reliably reproduces the issue for you? Without such, it would be almost impossible to determine what might be causing this, as out-of-memory situations can originate from a wide variety of causes, depending on user configuration too.

Comment 3

[bugzilla.mozilla.org]

No particular URL seems to cause this, it just happens after using Firefox for a while.

Here are the most recent crashes I submitted that I believe to be related to this bug:
https://crash-stats.mozilla.org/report/index/4a14e01a-2f41-4638-894a-348530231016
https://crash-stats.mozilla.org/report/index/7b97d1ef-cc98-4c44-8b0b-8ed440231015

Comment 4

[bugzilla.mozilla.org]

Attachment: about:support plain text information

Here is the debug information, with my home folder path replaced by `$HOME`:

Comment 5

[bugzilla.mozilla.org]

Here is the debug information, with my home folder path replaced by $HOME.

Comment 6

[bugzilla.mozilla.org]

Btw: Before crashing FIrefox sometimes stops rendering fonts on web pages, while everything else still seems to work. Not sure if this is related to the crash though.

Comment 7

[Ashley Hale [:ahale]]

Is this issue still occurring? I am concerned that 2GiB may not be enough depending on how you use Firefox, WebRender tends to allocate large textures and keep reusing them but never gets rid of them, it does reuse them but it's possible that you viewed enough large images or other web content that it ran out of memory and that would leave the browser unusable afterwards.

You can go to about:memory to save a memory report and post it here as an attachment, that may have more insights, but I fear this is the issue I just described where 2GiB is just not enough for the way you are using the browser. Making Firefox use less memory in this case may be a larger project.

Comment 8

[BugBot [:suhaib / :marco]]

Redirect a needinfo that is pending on an inactive user to the triage owner.
:gw, since the bug has recent activity, could you have a look please?

For more information, please visit BugBot documentation.