You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am trying to turn a retained HyperFrames headless-browser crash cluster into a useful reproducible bug report. I do not yet have a shareable minimal composition or the original render stderr, so I am asking here rather than claiming a reproducible upstream bug.
Recorded failure
On September 10, 2026 between 23:35 and 23:43 UTC+03, systemd recorded eleven cores from HyperFrames' bundled headless Chrome 152.0.7977.30. Renderer processes received SIGTRAP, including compositor and foreground thread-pool threads. Several records share executable offset 0x73fb6d8.
The executable is stripped. Local symbolization did not resolve the function, and its matching headless_shell.debug file was unavailable. The exact assertion remains unknown. The inspected journal records did not include a fatal assertion, thread-creation failure or OOM explanation.
None of these flags is established as the cause. The memory override does not mean each process allocated 16 GiB.
Environment and limits
Arch Linux with Omarchy 4.0.2; x86_64.
Running kernel during the cluster: 7.1.8-arch1-3.
Ryzen AI 9 HX 370 / Radeon 890M; approximately 60 GiB usable RAM.
Installed Mesa and Vulkan Radeon 26.2.1; AMD firmware package 20260810-2.
Hyprland 0.56.2 / Wayland desktop; the affected browser processes were headless.
Exact HyperFrames CLI version, composition, command and worker count for the failing invocations were not recovered. I have not established reproduction on the latest CLI.
Other browser automation was active that evening. Individual historical process-group memory peaks are not a simultaneous memory total.
Journald recorded memory pressure later, at 23:53. No accompanying OOM kill was found. A causal link to these SIGTRAPs is unproven.
Help requested
Is matching debug information available for the bundled Chrome version above, or is there a supported way to capture its fatal assertion and symbolized stack during a render? What is the recommended way to retain complete browser stderr and actual worker count for a single-worker attempt before comparing the original concurrency?
I understand the bug template requires a public minimal reproduction. The next step would be to capture these diagnostics and reduce the input before filing a bug. No raw core, private composition or browser profile is attached.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
I am trying to turn a retained HyperFrames headless-browser crash cluster into a useful reproducible bug report. I do not yet have a shareable minimal composition or the original render stderr, so I am asking here rather than claiming a reproducible upstream bug.
Recorded failure
On September 10, 2026 between 23:35 and 23:43 UTC+03, systemd recorded eleven cores from HyperFrames' bundled headless Chrome
152.0.7977.30. Renderer processes receivedSIGTRAP, including compositor and foreground thread-pool threads. Several records share executable offset0x73fb6d8.The executable is stripped. Local symbolization did not resolve the function, and its matching
headless_shell.debugfile was unavailable. The exact assertion remains unknown. The inspected journal records did not include a fatal assertion, thread-creation failure or OOM explanation.Recorded launch arguments include:
None of these flags is established as the cause. The memory override does not mean each process allocated 16 GiB.
Environment and limits
7.1.8-arch1-3.Help requested
Is matching debug information available for the bundled Chrome version above, or is there a supported way to capture its fatal assertion and symbolized stack during a render? What is the recommended way to retain complete browser stderr and actual worker count for a single-worker attempt before comparing the original concurrency?
I understand the bug template requires a public minimal reproduction. The next step would be to capture these diagnostics and reduce the input before filing a bug. No raw core, private composition or browser profile is attached.
All reactions