Open Bug 1694671 Opened 5 months ago Updated 2 months ago

Crash in [@ rayon_core::job::{{impl}}::execute<T>]

Categories

(Core :: Layout, defect)

x86
Windows 7
defect

Tracking

()

REOPENED
88 Branch
Tracking Status
firefox-esr78 --- affected
firefox86 + wontfix
firefox87 + wontfix
firefox88 + wontfix
firefox89 --- wontfix

People

(Reporter: aryx, Assigned: emilio, NeedInfo)

References

Details

(Keywords: crash, topcrash)

Crash Data

Attachments

(1 file)

[Tracking Requested - why for this release]: Frequent crash

50+ crashes in a single day on release, all on Windows 7 with Firefox 32-bit executable. The signature was already active for previous releases but often <50 crashes for the whole cycle and those crashes also included amd64.

Emilio, can you investigated what's causing these crashes?

Crash report: https://crash-stats.mozilla.org/report/index/fde4fbe4-0f6a-4c77-8b24-546400210224

MOZ_CRASH Reason: called `Option::unwrap()` on a `None` value

Top 10 frames of crashing thread:

0 xul.dll RustMozCrash mozglue/static/rust/wrappers.cpp:16
1 xul.dll mozglue_static::panic_hook mozglue/static/rust/lib.rs:89
2 xul.dll core::ops::function::Fn::call<fn ../e1884a8e3c3e813aada8254edfa120e85bf5ffca/library/core/src/ops/function.rs:227
3 xul.dll std::panicking::rust_panic_with_hook ../e1884a8e3c3e813aada8254edfa120e85bf5ffca//library/std/src/panicking.rs:597
4 xul.dll std::panicking::begin_panic_handler::{{closure}} ../e1884a8e3c3e813aada8254edfa120e85bf5ffca//library/std/src/panicking.rs:497
5 xul.dll std::sys_common::backtrace::__rust_end_short_backtrace<closure-0, !> ../e1884a8e3c3e813aada8254edfa120e85bf5ffca//library/std/src/sys_common/backtrace.rs:141
6 xul.dll std::panicking::begin_panic_handler ../e1884a8e3c3e813aada8254edfa120e85bf5ffca//library/std/src/panicking.rs:495
7 xul.dll core::panicking::panic_fmt ../e1884a8e3c3e813aada8254edfa120e85bf5ffca//library/core/src/panicking.rs:92
8 xul.dll core::panicking::panic ../e1884a8e3c3e813aada8254edfa120e85bf5ffca//library/core/src/panicking.rs:50
9 xul.dll rayon_core::job::{{impl}}::execute<closure-0> third_party/rust/rayon-core/src/job.rs:168
Flags: needinfo?(emilio)
Severity: -- → S2

Rayon was updated in 86 (bug 1683294). Probably a rayon regression?

Flags: needinfo?(emilio)
See Also: → 1683294
Depends on: 1697842

This should help us isolate what rayon changes could've caused this
stability issue on 32-bit windows.

Automatically generated with:

$ cargo update -p rayon --precise 1.4.0 && cargo update -p rayon-core --precise 1.8.0 && ./mach vendor rust

Assignee: nobody → emilio
Status: NEW → ASSIGNED
Pushed by ealvarez@mozilla.com:
https://hg.mozilla.org/integration/autoland/rev/689dd6a2949c
Downgrade rayon-core to 1.8.0, rayon to 1.4.0. r=jrmuizel
Attachment #9208433 - Attachment description: Bug 1694671 - Downgrade rayon-core to 1.8.0, rayon to 1.4.0. r=nical,jrmuizel → Bug 1694671 - Downgrade rayon-core to 1.8.1, rayon to 1.4.1. r=nical,jrmuizel
Flags: needinfo?(emilio)
Pushed by ealvarez@mozilla.com:
https://hg.mozilla.org/integration/autoland/rev/a31f5ee91094
Downgrade rayon-core to 1.8.1, rayon to 1.4.1. r=jrmuizel
Status: ASSIGNED → RESOLVED
Closed: 4 months ago
Resolution: --- → FIXED
Target Milestone: --- → 88 Branch

The patch landed in nightly and beta is affected.
:emilio, is this bug important enough to require an uplift?
If not please set status_beta to wontfix.

For more information, please visit auto_nag documentation.

Flags: needinfo?(emilio)

We still don't know if/how this will affect the crash rate.

Status: RESOLVED → REOPENED
Flags: needinfo?(emilio)
Resolution: FIXED → ---

Is there anything else we can do here?

Flags: needinfo?(emilio)

We can keep bisecting / downgrading I guess. Other than that, we could really use the help of somebody with windows knowledge to see what might be going on, but that might be hard.

Trying to build a win32 rayon stress test and see whether we can reproduce the crash is another thing to do, but it will probably take a bit of time. I can try though.

You need to log in before you can comment on or make changes to this bug.