Closed Bug 1869713 Opened 2 years ago Closed 2 years ago

[gpu-canvas] crash on any canvas webpage (MakeCurrent called on a thread other than the creating thread)

Categories

(Core :: Graphics: Canvas2D, defect)

Unspecified
Windows
defect

Tracking

()

RESOLVED DUPLICATE of bug 1869818
Tracking Status
firefox-esr115 --- unaffected
firefox120 --- unaffected
firefox121 --- unaffected
firefox122 --- affected

People

(Reporter: mayankleoboy1, Unassigned)

References

()

Details

(Keywords: regression)

Crash Data

Attachments

(1 file)

Attached file about:support

GP+[GFX1]: MakeCurrent called on a thread other than the creating thread!
(#39) GP+[GFX1]: Ignoring call to fDrawBuffers with failed mImplicitMakeCurrent.
(#40) GP+[GFX1]: MakeCurrent called on a thread other than the creating thread!
(#41) GP+[GFX1]: EGLContext#true is owned by/Current on thread#55032 but MakeCurrent requested on thread#56776!
(#42) GP+[GFX1]: Ignoring call to fReadBuffer with failed mImplicitMakeCurrent.
(#43) GP+[GFX1]: MakeCurrent called on a thread other than the creating thread!
(#44) GP+[GFX1]: EGLContext#true is owned by/Current on thread#55032 but MakeCurrent requested on thread#56776!
(#45) GP+[GFX1]: Ignoring call to fCheckFramebufferStatus with failed mImplicitMakeCurrent.
(#46) Error CompositorBridgeChild receives IPC close with reason=AbnormalShutdown
(#47) Error CompositorBridgeChild receives IPC close with reason=AbnormalShutdown
(#48) CP+[GFX1-]: CompositorBridgeChild receives IPC close with reason=AbnormalShutdown
(#49) CP+[GFX1-]: CompositorBridgeChild receives IPC close with reason=AbnormalShutdown
(#50) CP+[GFX1-]: CompositorBridgeChild receives IPC close with reason=AbnormalShutdown
(#51) CP+[GFX1-]: CompositorBridgeChild receives IPC close with reason=AbnormalShutdown
(#52) CP+[GFX1-]: CompositorBridgeChild receives IPC close with reason=AbnormalShutdown
(#53) CP+[GFX1-]: CompositorBridgeChild receives IPC close with reason=AbnormalShutdown

Flags: needinfo?(lsalzman)

Set release status flags based on info from the regressing bug 1829026

Crash Signature: [@ mozilla::ipc::MessageChannel::AssertWorkerThread | mozilla::ipc::MessageChannel::NextSeqno ] → [@ mozilla::ipc::MessageChannel::AssertWorkerThread | mozilla::ipc::MessageChannel::NextSeqno ] [@ mozilla::gfx::GPUParent::GetSingleton ]

apart from the crash, the browser and/or the tabs also stop responding to user inputs. Windows may also show the "spinning circle" thingy.

Dup. of bug 1869658?

This is not really supported or tested on Windows at all yet. Caveat emptor if you decide to enable it on Windows. But this is disabled by default.

Severity: -- → S4
Flags: needinfo?(lsalzman)
OS: Unspecified → Windows

(In reply to Lee Salzman [:lsalzman] from comment #6)

This is not really supported or tested on Windows at all yet. Caveat emptor if you decide to enable it on Windows. But this is disabled by default.

Right. Which then means that till this crash is fixed there is no way to test gpu-canvas on Windows.
Edit: I have disabled gpu-canvas on Windows nightly, so the crash is 'fixed'. Feel free to RESOLVED FIXED etc.

To avoid clutter, I will close this bug as INVALID and remove the regressed by bug. If/when windows support is actually developed, basic testing would catch this bug.

Status: NEW → RESOLVED
Closed: 2 years ago
No longer regressed by: 1829026
Resolution: --- → INVALID

Setting gfx.canvas.remote.worker-threads = 0 prevents the crashes
Briefly switching to d2d-canvas made me realize just how vastly superior gpu-canvas is. I would much rather prefer to continue to use gpu-canvas.

See Also: → 1869818
Duplicate of bug: 1869818
Resolution: INVALID → DUPLICATE
See Also: 1869818
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: