1967451 - 4% glterrain (Linux) regression on Wed May 14 2025

Status: RESOLVED WONTFIX

(Core :: Graphics: WebRender, defect)

Description

[Treeherder Bug Filer]

Perfherder has detected a talos performance regression from push 6be9c2492a2c6fca50a3bc6601dcde5116be244e. As author of one of the patches included in that push, we need your help to address this regression.

Please acknowledge, and begin investigating this alert within 3 business days, or the patch(es) may be backed out in accordance with our regression policy. Our guide to handling regression bugs has information about how you can proceed with this investigation.

If you have any questions or need any help with the investigation, please reach out to [e-mail]. Alternatively, you can find help on Slack by joining #perf-help, and on Matrix you can find help by joining #perftest.

Regressions:

Ratio	Test	Platform	Options	Absolute values (old vs new)
4%	glterrain	linux1804-64-shippable-qr	e10s fission stylo webrender	2.20 -> 2.29

Improvements:

Ratio	Test	Platform	Options	Absolute values (old vs new)
5%	tscrollx	windows11-64-24h2-shippable	e10s fission stylo webrender	0.57 -> 0.54
4%	tresize	windows11-64-24h2-shippable	e10s fission stylo webrender	4.71 -> 4.51
4%	tsvgx	linux1804-64-shippable-qr	e10s fission stylo webrender	221.16 -> 212.85
3%	tart	windows11-64-24h2-shippable	e10s fission stylo webrender	1.82 -> 1.77

Details of the alert can be found in the alert summary, including links to graphs and comparisons for each of the affected tests.

If you need the profiling jobs you can trigger them yourself from treeherder job view or ask [e-mail] to do that for you.

You can run all of these tests on try with ./mach try perf --alert 45152

The following documentation link provides more information about this command.

Comment 1

[BugBot [:suhaib / :marco]]

Set release status flags based on info from the regressing bug 1961724

Comment 2

[Jamie Nicol [:jnicol]]

Is the title of this bug innacurate? The comment seems to indicate an tsvgx improvement (amongst others), but a glterrain regression.

Seems the benefits likely outweigh the regressions, so I would assume we'd rather not back this out. But Martin, do you have any idea why it could have caused the glterrain regression?

Comment 3

[Martin Stránský [:stransky] (ni? me)]

Will look at it.

Comment 4

[Martin Stránský [:stransky] (ni? me)]

I wonder if it's related to GL fence we use in WebGL / dmabuf backend - I think the glFlush may speed up GL processing for WebGL so we don't wait at GL fence so long and we can recycle/paint faster then:

https://searchfox.org/mozilla-central/rev/02d33f4bf984f65bd394bfd2d19d66569ae2cfe1/gfx/gl/SharedSurfaceDMABUF.cpp#51

Comment 5

[Martin Stránský [:stransky] (ni? me)]

Hm, I'm unable to debug that glterrain locally as it needs python 3.11 but Fedora 42 ships 3.13. And Mozilla seems to ignore alternatives.

Comment 6

[Jamie Nicol [:jnicol]]

I usually use pyenv to force a specific python version whenever fedora's is too recent

Comment 7

[Martin Stránský [:stransky] (ni? me)]

Managed to test it locally, first numbers are with glFlush():

      "name": "glterrain",

         "name": "0.WebGL-terrain-alpha-no-AA-no",
         "value": 0.6824000000000001
          "value": 0.6852000000000005

          "name": "1.WebGL-terrain-alpha-no-AA-yes",
          "value": 1.0454999999999997
          "value": 1.0447000000000002

          "name": "2.WebGL-terrain-alpha-yes-AA-no",
          "value": 0.6995000000000005
          "value": 0.6663999999999999

          "name": "3.WebGL-terrain-alpha-yes-AA-yes",
          "value": 0.8072000000000026
          "value": 0.7852000000000021

      "value": 0.803075999509723
      "value": 0.7893149709856457

so I see improvement on my system without glFlush() (Fedora 42 / Wayland / Intel) as one may expect.
We may look why linux1804-64-shippable-qr has different results but it diverges from what we expect and what real scenario testing shows.

Comment 8

[Martin Stránský [:stransky] (ni? me)]

I wonder if the first GL run (0.WebGL-terrain-alpha-no-AA-no) is hit somehow by missing glFlush() as it's the only one which shows worse results.

Comment 9

[BugBot [:suhaib / :marco]]

Set release status flags based on info from the regressing bug 1961724

Comment 10

[BugBot [:suhaib / :marco]]

It has been over 7 days with no activity on this performance regression.

:stransky, since you are the author of the regressor, bug 1961724, which triggered this performance alert, could you please provide a progress update?

If this regression is something that fixes a bug, changes the baseline of the regression metrics, or otherwise will not be fixed, please consider closing it as WONTFIX. See this documentation for more information on how to handle regressions.

For additional information/help, please needinfo the performance sheriff who filed this alert (they can be found in comment #0), or reach out in #perftest, or #perfsheriffs on Element.

For more information, please visit BugBot documentation.

Comment 11

[Martin Stránský [:stransky] (ni? me)]

Still under investigation.

Comment 12

[Martin Stránský [:stransky] (ni? me)]

I can't reproduce and it's part of improved/optimized GPU usage so let's keep it as is.