Skip to content

cef_host: point the resize capture refresh at CEF issue #3826 - #63

Merged
wenkaifan0720 merged 1 commit into
mainfrom
fix/capture-refresh-upstream-ref
Sep 29, 2026
Merged

wenkaifan0720 merged 1 commit into
mainfrom
fix/capture-refresh-upstream-ref

Conversation

@wenkaifan0720

Copy link
Copy Markdown
Collaborator

Comment and changelog only. The prebuilt isn't published yet, so it goes in first.

The stall behind #61 is upstream CEF issue #3826. Chromium's FrameSinkVideoCapturerImpl::MaybeDeliverFrame doesn't call VideoCaptureOracle::CompleteCapture for a dropped frame. num_frames_pending_ therefore stays up, and every capturer-side kRefreshRequest retry is refused. That is the "nothing pending" rejection our gpu.capture trace showed.

Invalidate(PET_VIEW) (kRefreshDemand) bypasses the check. It's the workaround used in that thread and by other embedders.

CEF fixed it in 0d05970 with viz_capture_3826.patch, on branch 8037 and master. We're on 7559. The comment now says the refresh can go once we're on such a CEF.

🤖 Generated with Claude Code

The capturer's retries never capture after a dropped frame because
Chromium's FrameSinkVideoCapturerImpl::MaybeDeliverFrame doesn't call
VideoCaptureOracle::CompleteCapture for it, so num_frames_pending_
stays up and every kRefreshRequest is refused (CEF issue #3826). That
is the "nothing pending" check our trace showed rejecting them. CEF
fixed it with viz_capture_3826.patch on branch 8037 and master; we're on
7559. Say so, and when the refresh can go. Comment and changelog only.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
@wenkaifan0720
wenkaifan0720 merged commit 7c7bcb0 into main Sep 29, 2026
2 of 3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant