Crash in [@ tokio_timer::wheel::Wheel<T>::poll]
Categories
(Core :: Audio/Video: cubeb, defect, P3)
Tracking
()
People
(Reporter: gsvelto, Assigned: kinetik)
References
Details
(Keywords: crash)
Crash Data
Crash report: https://crash-stats.mozilla.org/report/index/97cb4360-2bc6-47bb-9848-94a8d0200912
Top 10 frames of crashing thread:
0 libxul.so RustMozCrash /build/firefox/src/firefox-80.0.1/mozglue/static/rust/wrappers.cpp:17
1 libxul.so mozglue_static::panic_hook /build/firefox/src/firefox-80.0.1/mozglue/static/rust/lib.rs:89
2 libxul.so core::ops::function::Fn::call /build/rust/src/rustc-1.46.0-src/src/libcore/ops/function.rs:72
3 libxul.so std::panicking::rust_panic_with_hook /build/rust/src/rustc-1.46.0-src/src/libstd/panicking.rs:530
4 libxul.so rust_begin_unwind /build/rust/src/rustc-1.46.0-src/src/libstd/panicking.rs:437
5 libxul.so std::panicking::begin_panic_fmt /build/rust/src/rustc-1.46.0-src/src/libstd/panicking.rs:391
6 libxul.so tokio_timer::wheel::Wheel<T>::poll /build/firefox/src/firefox-80.0.1/third_party/rust/tokio-timer/src/wheel/mod.rs
7 libxul.so audioipc::core::spawn_thread::{{closure}} /build/firefox/src/firefox-80.0.1/third_party/rust/tokio-timer/src/timer/mod.rs:272
8 libxul.so std::sys_common::backtrace::__rust_begin_short_backtrace /build/rust/src/rustc-1.46.0-src/src/libstd/sys_common/backtrace.rs:130
9 libxul.so core::ops::function::FnOnce::call_once{{vtable.shim}} /build/rust/src/rustc-1.46.0-src/src/libcore/ops/function.rs:233
This is a Linux-specific low-volume crash happening within Tokio, in an audio thread. The crash reason is always in the following something like this:
elapsed=522005; when=522004
Usually the elapsed
value is just when+1
but sometimes the difference is bigger.
Assignee | ||
Updated•4 years ago
|
Comment 2•4 years ago
|
||
I can reproduce this issue reliably on my computer immediately after enabling microphone access to Google Meet. Is there useful information I can provide or a good way I could debug this issue? Maybe I could grab debug symbols and set breakpoints at appropriate things?
Reporter | ||
Comment 3•4 years ago
|
||
(In reply to Saagar Jha from comment #2)
I can reproduce this issue reliably on my computer immediately after enabling microphone access to Google Meet. Is there useful information I can provide or a good way I could debug this issue? Maybe I could grab debug symbols and set breakpoints at appropriate things?
An STR would be very useful. If you can also extract backtraces using gdb then that would be awesome. I can help you find the right debuginfo to do that.
Comment 4•4 years ago
|
||
This is my crash: https://crash-stats.mozilla.org/report/index/7896aa3d-9214-4d49-aa37-fc2af0201115; I can reliably get that by opening Google Meet to a meeting, then approving access to my camera and microphone–it crashes almost immediately after I do this.
Assignee | ||
Comment 5•4 years ago
|
||
(In reply to Saagar Jha from comment #2)
I can reproduce this issue reliably on my computer immediately after enabling microphone access to Google Meet. Is there useful information I can provide or a good way I could debug this issue? Maybe I could grab debug symbols and set breakpoints at appropriate things?
Thanks, that'd be much appreciated. Based on the described steps to reproduce, it's possible your crash was addressed by bug 1672275. As a first step, would you mind testing again with a Nightly build?
Comment 6•4 years ago
|
||
I installed firefox-trunk and symbols and can still reproduce this issue; I have it stopped in GDB right now (although working with it is a bit painful because this computer can barely handle loading the debug symbols). The backtrace looks to be identical. Is there something I can grab here that would be useful?
I found this: https://github.com/tokio-rs/tokio/issues/2473
This is anything but low-volume for me - attempting to play nearly any video now triggers a very similar crash for me.
See one of my crash reports here: https://crash-stats.mozilla.org/report/index/d3ae74f8-4ac1-4399-88f5-291d90210905
Please let me know if I can provide more information or test something - I'd very much like to be able to watch videos again.
Assignee | ||
Comment 9•1 year ago
|
||
Fixed via bug 1726279.
Description
•