1570145 - Crash in [@ cubeb_backend::capi::capi_stream_get_current_device]

Status: RESOLVED DUPLICATE of bug 1574284

(Core :: Audio/Video: cubeb, defect, P2)

Description

[Marcia Knous [:marcia]]

This bug is for crash report bp-b0665168-03b3-462c-b35f-dc3030190729.

Seen while looking at nightly crashes - crashes started in 20190729095501: https://bit.ly/2YFKKyU. Bug 1568182 landed in the build ID regression range.

Two of the URLs are from jsfiddle - https://jsfiddle.net/v6z1nk04/7/

Top 10 frames of crashing thread:

0 XUL gkrust_shared::panic_hook mfbt/Assertions.h:313
1 XUL core::ops::function::Fn::call src/libcore/ops/function.rs:69
2 XUL std::panicking::rust_panic_with_hook src/libstd/panicking.rs:478
3 XUL std::panicking::continue_panic_fmt src/libstd/panicking.rs:381
4 XUL std::panicking::begin_panic_fmt src/libstd/panicking.rs:336
5 XUL cubeb_backend::capi::capi_stream_get_current_device third_party/rust/cubeb-backend/src/capi.rs:246
6 XUL mozilla::AudioCallbackDriver::PanOutputIfNeeded dom/media/GraphDriver.cpp:1046
7 XUL mozilla::AudioCallbackDriver::DeviceChangedCallback_s dom/media/GraphDriver.cpp:813
8 XUL <futures_cpupool::MySender<F, core::result::Result<<F as futures::future::Future>::Item, <F as futures::future::Future>::Error>> as futures::future::Future>::poll third_party/rust/futures-cpupool/src/lib.rs:325
9 XUL std::sys_common::backtrace::__rust_begin_short_backtrace src/libstd/sys_common/backtrace.rs:136

Comment 1

[Paul Adenot (:padenot)]

Chunmin, this one is for you I think.

Comment 2

[C.M.Chang[:chunmin]]

It looks like it panic on an assertion, but the only assertion in code stack of the current_device in cubeb-coreaudio-rs is unlikely to be hit. Don't have idea about why it happens. Keep the needinfo for now.

Comment 3

[C.M.Chang[:chunmin]]

Just found the cause in bug 1574284. Clear the needinfo now.

Comment 4

[Matthew Gregan [:kinetik]]

Yep, given futures_cpupool is on the stack this must be with AudioIPC enabled, which isn't the default on macOS yet (bug 1425788).

Comment 5

[BugBot [:suhaib / :marco]]

:marcia, since this bug is a regression, could you fill (if possible) the regressed_by field?
For more information, please visit auto_nag documentation.

Comment 6

[Marcia Knous [:marcia]]

(In reply to Release mgmt bot [:sylvestre / :calixte / :marco for bugbug] from comment #5)

:marcia, since this bug is a regression, could you fill (if possible) the regressed_by field?
For more information, please visit auto_nag documentation.

I think :chunmin could probably answer this question better. I am not sure if the bug I referenced in the initial comment potentially caused it or not.

Comment 7

[C.M.Chang[:chunmin]]

(In reply to Marcia Knous [:marcia - needinfo? me] from comment #6)

I think :chunmin could probably answer this question better. I am not sure if the bug I referenced in the initial comment potentially caused it or not.

This crash was introduced with AudioIPC that would be turned on by a pref on MacOS (bug 1425788). Users won't be affected unless they turn on the pref. This crash will be fixed in bug 1574284. There are more details there.