Crash in [@ midir_impl_close_port]
Categories
(Core :: DOM: Device Interfaces, defect)
Tracking
()
Tracking | Status | |
---|---|---|
firefox-esr91 | --- | unaffected |
firefox95 | --- | unaffected |
firefox96 | --- | unaffected |
firefox97 | --- | affected |
People
(Reporter: calixte, Assigned: gsvelto)
References
(Blocks 1 open bug, Regression)
Details
(Keywords: crash, regression)
Crash Data
Maybe Fission related. (DOMFissionEnabled=1)
Crash report: https://crash-stats.mozilla.org/report/index/0aeec950-55c4-4371-99ec-4ab780211224
MOZ_CRASH Reason: called `Option::unwrap()` on a `None` value
Top 10 frames of crashing thread:
0 XUL RustMozCrash mozglue/static/rust/wrappers.cpp:18
1 XUL mozglue_static::panic_hook mozglue/static/rust/lib.rs:91
2 XUL core::ops::function::Fn::call library/core/src/ops/function.rs:70
3 XUL std::panicking::rust_panic_with_hook library/std/src/panicking.rs:628
4 XUL std::panicking::begin_panic_handler::{{closure}} library/std/src/panicking.rs:519
5 XUL std::sys_common::backtrace::__rust_end_short_backtrace library/std/src/sys_common/backtrace.rs:139
6 XUL rust_begin_unwind library/std/src/panicking.rs:517
7 XUL core::panicking::panic_fmt library/core/src/panicking.rs:100
8 XUL core::panicking::panic library/core/src/panicking.rs:50
9 XUL midir_impl_close_port dom/midi/midir_impl/src/lib.rs:262
There is 1 crash in nightly 97 with buildid 20211224095019. In analyzing the backtrace, the regression may have been introduced by patch [1] to fix bug 1201598.
[1] https://hg.mozilla.org/mozilla-central/rev?node=c5e98c30bbf2
Assignee | ||
Comment 1•2 years ago
|
||
I once stumbled upon this issue during local testing but couldn't reproduce it. I'll try it again, it's possible that we're shutting down the service twice (or at least attempting to do so).
Assignee | ||
Comment 2•2 years ago
|
||
There is something wrong with the object lifecycle. I managed to cause this crash on shutdown which doesn't match the one here but is likely related. In this crash we're closing the same port twice (something that shouldn't be possible), and in my crash we're closing the service twice which again shouldn't be possible. Additionally while testing I also managed to open the same port twice without closing it first, the second open operation failed. Last but not least on www.midimonitor.com my MIDI keyboard appears twice when I first loaded the page which again shouldn't be possible.
Updated•2 years ago
|
Updated•2 years ago
|
Assignee | ||
Comment 3•2 years ago
|
||
Actively working on this.
Assignee | ||
Comment 5•2 years ago
|
||
Yes, this turned out to be a duplicate of bug 1748647.
Updated•2 years ago
|
Description
•