cef-binary: 142.0.10 -> 147.0.10

[thunze]

• Update CEF because we probably don't want to stay on such an outdated Chromium version forever
• Patch cef-rs of graphite to not check for an exact match of the CEF version (otherwise we could never bump CEF)

Closes #455980

Things done

• Built on platform:
  • x86_64-linux — cef-binary, graphite, stremio-linux-shell
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• Tested, as applicable:
  • NixOS tests in nixos/tests.
  • Package tests at passthru.tests.
  • Tests in lib/tests or pkgs/test for functions and "core" functionality.
• Ran nixpkgs-review on this PR. See nixpkgs-review usage.
• Tested basic functionality of all binary files, usually in ./result/bin/.
• Nixpkgs Release Notes
  • Package update: when the change is major or breaking.
• NixOS Release Notes
  • Module addition: when adding a new NixOS module.
  • Module update: when the change is significant.
• Fits CONTRIBUTING.md, pkgs/README.md, maintainers/README.md and other READMEs.

[thunze]

Hmm, I cannot get what graphite calls "accelerated UI rendering" to work:

ERR: Display.cpp:1097 (initialize): ANGLE Display::initialize error 0: Internal Vulkan error (-9): The requested version of Vulkan is not supported by the driver or is otherwise incompatible for implementation-specific reasons, in ../../third_party/angle/src/libANGLE/renderer/vulkan/vk_renderer.cpp, initialize:2601.
[309991:310005:0426/201919.936327:FATAL:content/browser/gpu/gpu_data_manager_impl_private.cc:417] GPU process isn't usable. Goodbye.
Trace/breakpoint trap      (core dumped) result/bin/graphite

Hardware acceleration works fine for stremio-linux-shell though, so that might not be a me problem.

@timon-schelling I noticed you're updating CEF upstream in GraphiteEditor/Graphite#4048. Do you have any insights on that, by any chance? Can you reproduce the UI acceleration issue?

[timon-schelling]

I'm relatively sure that things will break when running Graphite with a CEF version that is not compatible with the cef-rs crate. CLI flags we need to run Graphite with changed between 143 and 147.
I would prefer overriding the version for Graphite (like other CEF applications do).
Currently working on improving the Graphite update script. It could update CEF version in the graphite package and align it with the version tested upstream.

[thunze]

A few other packages based on CEF currently pin CEF, and therefore Chromium, to ancient versions because they haven't been updated in a long time. Imo we should work towards removing pins of security-critical components like Chromium where possible. In fact, I was explicitly discouraged from pinning CEF in a previous PR. I also expected issues when unpinning CEF in that PR, but that was unwarranted, minus a minor patch I had to vendor.

I'm not a CEF expert, but looking at upstream cef-rs, the compatibility guarantees between cef-rs and CEF don't seem to be documented, so maybe it's worth a try? Of course I'm not sure how much work this would be for graphite, specifically. Just wanted to get your opinion on this before I sink too much time into this, maybe you've seen this issue before.

[timon-schelling]

I completely get that, the problem is the setup we are running with Graphite, OSR with GPU acceleration etc. is relatively flaky while chromium is refactoring Linux rendering backends (I'm not a chromium expert but that's what my research while trying to update Graphites cef version suggests), 143-146 had issues that made them not useable for us.

And yes version incompatibility across cef versions is not that bad anymore, but still has some issues that are likely to break stuff in graphite unexpectedly.

[thunze]

I get that. From my experience in working on nixpkgs, the way we usually go about this is bump the dependency anyway and patch the dependents if needed and upstream doesn't catch up in time. Pinning is usually discouraged. In some cases we keep multiple versions of a dependency around, but not if they're EOL. But FWICT there are never two Chromium versions supported at the same time. I'd assume the same holds true for CEF.

Can you estimate when GraphiteEditor/Graphite#4048 will land? If it's going to land soon, I'd suggest we wait a bit until it does, then bump CEF and graphite here, then check if anything breaks on the next CEF update.

[timon-schelling]

I'll try to land it tomorrow, but it will for sure be merged this week.

[thunze]

Bumped graphite. UI acceleration seems to work now, so this should be good to go!

[timon-schelling]

LGTM

[timon-schelling]

Could you update to the latest commit? I merged a important Bug fix yesterday.

  version = "0-unstable-2026-05-02";
  rev = "ab7f59ca61004a1b11f9ae4b1c511cefc7a0f404";

  srcHash = "sha256-DV3/1dgtUiTmGkOm4z3GVJcWzvCjO/crzc/l8ovW0XA=";
  shaderHash = "sha256-uc6FU0df5Xqp6YXEwODULhgUjSQvjRFGvdk+uFB7II0=";
  cargoHash = "sha256-ZesLyXKjz2CSrAWUT5Hq6w97pR55I+C79qPwF0dqXXI=";
  npmHash = "sha256-AX5Jqk2E+WyQJyHbgvvq74MRsYmWUju4bOkabhYoeig=";

  brandingRev = "1509880500e32cb21235343ba865adcf0579133f";
  brandingHash = "sha256-ibTFiu4YawwWEKh3i9GZyb2vEZG0AG5rAH3ECkAvSqc=";

[thunze]

Could you update to the latest commit? I merged a important Bug fix yesterday.

Done! Also bumped brandingRev to latest.

[thunze]

nixpkgs-review result for #513755

Generated using nixpkgs-review-gha

Command: nixpkgs-review pr 513755
Commit: ed11b643dfd7e155068e44a69652475e757bf054 (subsequent changes)
Merge: e5b778239c0ca2e690038e7cc8ca6ce674999da8

Logs: https://github.com/thunze/nixpkgs-review-gha/actions/runs/25260344638

---

x86_64-linux

✅ 3 packages built:

• cef-binary
• graphite
• stremio-linux-shell

---

aarch64-linux

✅ 3 packages built:

• cef-binary
• graphite
• stremio-linux-shell

[nixos-discourse]

This pull request has been mentioned on NixOS Discourse. There might be relevant details there:

https://discourse.nixos.org/t/prs-already-reviewed/2617/3007