Skip to content

GNOME Shell intermittently consumes a full CPU core; extension errors need attribution #997

Closed as not planned

Description

@hanthor

Observation

On a Utah machine running GNOME Shell 51.1, gnome-shell periodically consumed about 100% of one logical CPU across repeated 2-second samples. It had accumulated 6:39 CPU time after about 12 minutes uptime; the host reports 18 logical CPUs and remained mostly idle overall. The process later dropped to 0% briefly, then returned to about 100%, so this is sustained/intermittent single-core load rather than whole-host saturation.

Blur my Shell v76 was enabled. GNOME Shell logs also show JS errors in user-installed extensions (Copyous: No object to track signals; PaperWM: ImportError: Unable to load file) and a warning from Fuzzy Application Search. These may be unrelated; no stack profile established the CPU source. User-installed extensions are outside Utah's vendored extension set.

Utah is updating Blur my Shell to upstream v77, whose changes include reduced allocations while painting blur, but that alone does not prove it caused this CPU load.

Acceptance

Profile the busy interval and isolate the responsible extension or Shell code. With the fix, an idle GNOME 51.1 desktop should not repeatedly keep gnome-shell at a full core, and the relevant extension JS errors should be resolved or attributed.

Activity

  1. hanthor commented on Oct 28, 2026

    @hanthor
    MemberAuthor

    /kind bug

  2. github-actions commented on Oct 28, 2026

    @github-actions
    Contributor

    To accept this issue, comment /triage accepted. To categorize, comment /area <area> or /kind <kind>.

  3. hanthor commented on Oct 28, 2026

    @hanthor
    MemberAuthor

    PR #998 updates Blur my Shell to upstream v77, including reduced paint-time allocations. This is a plausible mitigation, not a proven cause; the CPU profile and other user-extension errors still need runtime verification after deployment.

  4. hanthor commented on Oct 28, 2026

    @hanthor
    MemberAuthor

    CPU attribution remains unresolved. The latest attempt to run a low-priority, temporary Blur package-build smoke on the GNOME 51.1 host failed because SSH timed out. Earlier host logs implicated errors from user-installed Copyous and PaperWM, but did not prove causality; no settings, processes, or installed extensions were changed. Keeping this open for a controlled runtime trace and extension isolation.

  5. hanthor commented on Oct 28, 2026

    @hanthor
    MemberAuthor

    Update: PR #998 updates the Blur source but does not resolve CPU attribution. Its remote contract check is pending, and the latest host connection attempt timed out. Earlier logs implicated errors from user-installed Copyous and PaperWM but did not prove causality. No settings or processes have been changed; issue remains open for controlled tracing.

  6. hivecommons-hive commented on Nov 10, 2026

    @hivecommons-hive
    Contributor

    🌱 Good first issue — bring your own coding agent

    Hi! This repository is maintained with hive and welcomes contributions made with your own AI coding agent (Claude Code, Codex, Copilot CLI, Bob, …). hosted-projectbluefin-knuckle-gjvq picked this issue as a good place to start because:

    • a concrete, reproducible bug report
    • has clear steps or acceptance criteria

    How to land it (no waiting on a human):

    1. Claim it through the relay: open https://hosted-projectbluefin-common-nmq5.hive.hivecommons.dev/contribute?issue=projectbluefin%2Futah%23997 and sign in with GitHub. Claiming stamps the issue so hive leaves it to you.
    2. Fork or branch, make the change with your agent, and sign each commit with DCO (git commit -s).
    3. Open a PR whose description says Fixes #997 — that closing keyword is how hive's contributor lane links the PR to your claim.
    4. Hive runs CI and merges green PRs from claimed contributors automatically.
    Paste this to your agent
    Fix #997 in github.com/projectbluefin/utah (https://github.com/projectbluefin/utah/issues/997). First open https://hosted-projectbluefin-common-nmq5.hive.hivecommons.dev/contribute?issue=projectbluefin%2Futah%23997 and claim the issue through the hive relay (sign in with GitHub). Work on a fork/branch; keep the change scoped to the issue; sign every commit with DCO (git commit -s); open a PR whose description contains the line 'Fixes #997' so the hive's contributor lane can auto-merge it once CI is green. Read and follow https://github.com/projectbluefin/utah/blob/HEAD/AGENTS.md before changing code.
    

    Prefer not to drive the loop by hand? Run your own hive (we call them clankers) — a relay on your machine that claims, opens the PR and handles review fixes for you, and earns you a contributor profile and a leaderboard place: see https://hosted-projectbluefin-common-nmq5.hive.hivecommons.dev/contribute.

    If nobody claims it, hive may pick it up itself — claiming first prevents that. Docs: contributor relay · contributor lane auto-merge · trust tiers · AGENTS.md.

  7. added
    good first issueDenotes an issue ready for a new contributor, according to the "help wanted" guidelines.
    help wantedDenotes an issue that needs help from a contributor. Must meet "help wanted" guidelines.
    stage/stalledIn its current stage longer than expected
    on Nov 10, 2026
  8. removed
    good first issueDenotes an issue ready for a new contributor, according to the "help wanted" guidelines.
    help wantedDenotes an issue that needs help from a contributor. Must meet "help wanted" guidelines.
    on Nov 11, 2026
  9. Danathar commented on Nov 11, 2026

    @Danathar
    Contributor

    Dropping good-first-issue/help-wanted: this can't be worked without the affected host. @hanthor #998 (Blur my Shell v77) has been in since 10-29 and the remaining errors were from user-installed extensions. Close this as mitigated and reopen with a profile if it recurs, or keep it for your trace?

  10. added
    good first issueDenotes an issue ready for a new contributor, according to the "help wanted" guidelines.
    help wantedDenotes an issue that needs help from a contributor. Must meet "help wanted" guidelines.
    on Nov 11, 2026
  11. Danathar commented on Nov 13, 2026

    @Danathar
    Contributor

    Closing as mitigated. #998 (Blur my Shell v77, with fewer allocations at paint time) has been in since 10-29. The only errors left in the logs came from user-installed extensions (Copyous, PaperWM), which Utah doesn't ship. Nobody has the affected host or a CPU profile, so no agent can take this further. If GNOME Shell pins a core again, reopen with a sysprof-cli or perf capture and the list of enabled extensions.

  12. removed
    stage/015-triage-neededActionable but held by the reporter-trust gate: awaiting a maintainer's triage label
    stage/stalledIn its current stage longer than expected
    on Nov 13, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    good first issueDenotes an issue ready for a new contributor, according to the "help wanted" guidelines.help wantedDenotes an issue that needs help from a contributor. Must meet "help wanted" guidelines.kind/bugSomething is broken.needs-triage

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions