1984695 - Webrender blocked by presence of Intel DDX driver even if Intel GPU is not used

Status: RESOLVED FIXED

(Core :: Graphics, defect)

Description

[Mountain Dude]

User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:143.0) Gecko/20100101 Firefox/143.0

Steps to reproduce:

Play a YouTube video full screen.

Actual results:

CPU usage is high. Video playback is not smooth. And web UI is not very responsive. Once CPU thread probably maxing out.

gfx.color_management.enablev4 enabled is set to true, at least in Firefox Developer Edition.

Expected results:

Smooth playback. Low resource usage.

No issue using another browser (Brave) in X11.
No issue using same (Firefox) browser+profile in a Wayland environment.
Disabling gfx.color_management.enablev4 enabled fixes the resource usage issue (although still not as low as Wayland).

Comment 1

[BugBot [:suhaib / :marco/ :calixte]]

The Bugbug bot thinks this bug should belong to the 'Core::Graphics' component, and is moving the bug to that component. Please correct in case you think the bot is wrong.

Comment 2

[Timothy Nikkel (:tnikkel)]

Can you provide your about:support?

Comment 3

[Mountain Dude]

Attachment: about:support (bad)

<details>

</details>

Comment 4

[Mountain Dude]

Attachment: about:support (good)

Actually, I may have made a mistake, since disabling `gfx.color_management.enablev4` doesn't help anymore.

I also found out (I think) what's going on. Webrender hardware acceleration is disabled for some reason and "software" is used. Setting `gfx.webrender.all: true` fixes the issue fully.

Comment 5

[Jamie Nicol [:jnicol]]

Thanks for the bug report. I can see from your about:support that webrender is blocklisted with the code FEATURE_FAILURE_DDX_INTEL. Which means we think you have an Intel DDX driver, which we blocked in bug 1710400. But your about support indicates you have a radeon driver. Does your device have multiple GPUs?

Comment 6

[Mountain Dude]

(In reply to Jamie Nicol [:jnicol] from comment #5)

Does your device have multiple GPUs?

Yes. I enabled my iGPU recently for hardware encoding purposes. But it's not used for output/rendering/... at all. So presumably, the detection code should take into account what GPU will actually be doing the rendering!

Comment 7

[BugBot [:suhaib / :marco/ :calixte]]

:rmader, since you are the author of the regressor, bug 1710400, could you take a look?

For more information, please visit BugBot documentation.

Comment 8

[Mountain Dude]

Attachment: fix.patch

I gave the code a look. And unless inactive providers are taken into consideration on purpose, this patch to `glxtest` should fix the issue.

Comment 9

[Robert Mader [:rmader]]

Unfortunately don't have time to dive deeper into this, however on a quick look the fix looks good to me, thanks!

The intention back then was definitely to detect the active driver.

Comment 10

[Timothy Nikkel (:tnikkel)]

You can submit your patch using phabricator https://moz-conduit.readthedocs.io/en/latest/phabricator-user.html if you want to get it landed.

Comment 11

[Mountain Dude]

Attachment: Bug 1984695 - Fix glxtest to only add DDX_DRIVER entries for active providers

Check XRRProviderInfo outputs for connected crtcs before appending
DDX_DRIVER info.

Signed-off-by: Mountain Dude <gebirgsbrand@bmo>

Comment 12

[BugBot [:suhaib / :marco/ :calixte]]

Set release status flags based on info from the regressing bug 1710400

Comment 13

[Pulsebot]

Pushed by [e-mail]:
https://github.com/mozilla-firefox/firefox/commit/459bc3d47789
https://hg.mozilla.org/integration/autoland/rev/b594f17564c5
Fix glxtest to only add DDX_DRIVER entries for active providers r=rmader,emilio

Comment 14

[Narcis Beleuzu [:NarcisB]]

https://hg.mozilla.org/mozilla-central/rev/b594f17564c5

Comment 15

[BugBot [:suhaib / :marco/ :calixte]]

The patch landed in nightly and beta is affected.
:gebirgsbrand, is this bug important enough to require an uplift?

• If yes, please nominate the patch for beta approval.
  • See https://wiki.mozilla.org/Release_Management/Requesting_an_Uplift for documentation on how to request an uplift.
• If no, please set status-firefox143 to wontfix.

For more information, please visit BugBot documentation.

Comment 16

[Ryan VanderMeulen [:RyanVM]]

Is this something we should uplift to ESR140? Please nominate if yes.

Comment 17

[Emilio Cobos Álvarez [:emilio]]

301 Martin

Comment 18

[Martin Stránský [:stransky] (ni? me) (PTO till June 6)]

No need to uplift, I guess is pretty rare setup.