Skip to content

epic: Play store production crashes #986

Open
12 of 14 issues completed
Open

Description

@piotr-iohk

Elevated user-perceived crash rate on mainnet 2.2.0 (181). Tracks Play Vitals triage (Jun 2026).

Vitals: https://play.google.com/console/u/1/developers/5690237613698939049/app/4973032799696771282/vitals/crashes?days=28&isUserPerceived=true

Bundles: FGS (#980, #981, #443); native (#982 ± #984). DidNotStartInTime @ MainActivity (3/60d) → #981, no separate issue.

Done

P0 closed; P2 closed/deferred; AAB + native symbols each release; 28-day crash rate improved vs Jun 2026.

Activity

  1. self-assigned this
    on Jun 3, 2026
  2. modified the milestones: 2.3.0, 2.4.0 on Jun 3, 2026
  3. ovitrif commented on Jun 3, 2026

    @ovitrif
    Collaborator

    Milestones:

    • 2.3.0 for high priority issues
    • 2.4.0 for medium priority issues
  4. ovitrif commented on Jun 5, 2026

    @ovitrif
  5. ovitrif commented on Jun 22, 2026

    @ovitrif
  6. 10 remaining items

  7. changed the title [-][Epic] Play production crashes vitals triage[/-] [+]epic: Play production crashes[/+] on Aug 3, 2026
  8. changed the title [-]epic: Play production crashes[/-] [+]epic: Play store production crashes[/+] on Aug 3, 2026
  9. piotr-iohk commented on Sep 1, 2026

    @piotr-iohk
    CollaboratorAuthor

    2.4.1 / 188 (~24h, 188-only): leftover #982 Tokio SIGABRT (2 users / 3 events) and #983 UniFFI checksum (2 users / 2 events, API 28–30). Both reopened. No new clusters.

  10. modified the milestones: 2.5.0, 2.6.0, 2.7.0 on Sep 23, 2026
  11. piotr-iohk commented on Nov 16, 2026

    @piotr-iohk
    CollaboratorAuthor

    Newly observed graphics crash — 16 November 2026 review

    Monitoring evidence from Google Play Vitals: 3.1.1 (197), cluster 9ea717c3, 1 user-perceived crash event / 1 user. Window 19 Oct–16 Nov 2026, updated 15 Nov at 9:00 AM as displayed; timezone unstated.

    Sample: Samsung Galaxy A10, Android 11 / SDK 30, foreground, displayed as occurring 3 days ago. Abort message:

    Failed to set damage region on surface 0xb2b17200, error=EGL_BAD_ACCESS
    

    Relevant system frames:

    libhwui.so EglManager::damageFrame(...) +208
    libhwui.so SkiaOpenGLPipeline::draw(...) +30
    libhwui.so CanvasContext::draw() +974
    libhwui.so CanvasContext::prepareAndDraw(...) +140
    libhwui.so RenderThread::dispatchFrameCallbacks() +128
    libhwui.so WorkQueue::process() +162
    libhwui.so RenderThread::threadLoop() +72
    

    No app frames or reproducible trigger are present in the displayed sample. This is a different graphics signature from the older SkSurface::getCanvas sample, and distinct from the Tokio shutdown panic tracked in #1536. Keep as monitoring evidence here; a separate app bug is not warranted by this one system-only sample.

    Current build 197 totals 4 events across 2 clusters: 3 Tokio events / 2 users plus this graphics event. All versions: 24 events across 10 clusters; no ANR clusters in the selected view. These are rolling-window counts; crash rates were unavailable.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Labels

epicGroup of tickets

Type

No type

Projects

No projects

    Milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions