Closed Bug 1753189 Opened 3 years ago Closed 2 years ago

Crash in [@ free | webrender::scene_building::SceneBuilder::build]

Categories

(Core :: Graphics: WebRender, defect, P3)

Firefox 96
x86_64
Linux
defect

Tracking

()

RESOLVED WORKSFORME
Tracking Status
firefox96 --- affected

People

(Reporter: office, Unassigned)

References

(Blocks 1 open bug)

Details

(Keywords: crash)

Crash Data

Attachments

(1 file)

Was watching a video on https://www.youtube.com/watch?v=DZKuIdM0Zv4 when the whole browser and not only the single tab was sent to death...

Crash report: https://crash-stats.mozilla.org/report/index/122bbb80-3cd8-4284-845f-952a80220202

Reason: SIGSEGV / SEGV_MAPERR

Top 9 frames of crashing thread:

0 firefox-bin free memory/build/malloc_decls.h:54
1 libxul.so webrender::scene_building::SceneBuilder::build gfx/wr/webrender/src/scene_building.rs:523
2 libxul.so webrender::scene_builder_thread::SceneBuilderThread::process_transaction gfx/wr/webrender/src/scene_builder_thread.rs:595
3 libxul.so webrender::scene_builder_thread::SceneBuilderThread::run gfx/wr/webrender/src/scene_builder_thread.rs:311
4 libxul.so std::sys_common::backtrace::__rust_begin_short_backtrace library/std/src/sys_common/backtrace.rs:123
5 libxul.so core::ops::function::FnOnce::call_once{{vtable.shim}} library/core/src/ops/function.rs:227
6 libxul.so std::sys::unix::thread::Thread::new::thread_start library/std/src/sys/unix/thread.rs:106
7 libc.so.6 start_thread nptl/nptl/pthread_create.c:435
8 libc.so.6 __GI___clone 
Severity: -- → S3
Points: --- → 13
Priority: -- → P3
Blocks: wr-stability
Severity: S3 → --
Points: 13 → ---
Keywords: crash

Please open about:support, click on "Copy text to clipboard" and paste it here.
Which desktop environment are you using?

Attached file abbout.support (deleted) —

Please open about:config and reset these prefs back to their default:
gfx.webrender.compositor (experimental Wayland-only feature)
gfx.webrender.compositor.force-enabled (experimental Wayland-only feature)
gfx.webrender.fallback.software
gfx.webrender.fallback.software.requires-gpu-process
gfx.webrender.precache-shaders
gfx.webrender.svg-images
gfx.x11-egl.force-enabled
media.ffmpeg.vaapi.enabled
media.ffvpx.enabled
media.rdd-vpx.enabled
webgl.enable-ahardwarebuffer (Android-only feature)
webgl.enable-draft-extensions
webgl.enable-privileged-extensions
webgl.force-enabled

Please use https://nightly.mozilla.org for testing https://github.com/elFarto/nvidia-vaapi-driver.
Experimental features are more stable with Nightly than with Stable because not all patches get backported.
(How to create a custom .desktop file: bug 1751153 comment 19)

You might also want to try out https://firefox-source-docs.mozilla.org/tools/sanitizer/asan_nightly.html and set your email address (the same you use on bugzilla) in the asanreporter.clientid pref. Should it run into a security-relevant crash, you might get money.

no, that would mean loss of any kind of 3d acceleration and hardware ecoding

No, not any kind. You would still have hardware rendering. You would only disable experimental hardware video decoding with highly experimental nvidia-vaapi-driver.

Please use https://nightly.mozilla.org if you want to enable media.ffmpeg.vaapi.enabled on highly experimental nvidia-vaapi-driver.

no, before switching these flags and getting nvidia-vaapi-driver i had only use of cpu. its usage went down extremely since i use vaapi.
i know nvidia and linux is shit, but thats the only solution that works so far at all.

If you want to keep using Firefox Stable, please set these prefs back to their default.
(If "Compositing" on about:support is "WebRender", you have hardware rendering. Yes, you would still have software decoding for videos and that would cause higher CPU load.)

If you want to keep testing vaapi and egl, please use https://nightly.mozilla.org, but only set the prefs mentioned on https://github.com/elFarto/nvidia-vaapi-driver.
Experimental features are more stable in Nightly than in Stable because not all patches get backported.

Severity: -- → S4
Priority: -- → P3

The bug has a crash signature, thus the bug will be considered confirmed.

Status: UNCONFIRMED → NEW
Ever confirmed: true

Closing because no crashes reported for 12 weeks.

Status: NEW → RESOLVED
Closed: 2 years ago
Resolution: --- → WORKSFORME
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: