1970429 - Crash does not seem to have symbols (https://crash-stats.mozilla.org/report/index/d5f8e96f-a2f3-465f-a425-7ffb30250604)

Status: RESOLVED FIXED

(Toolkit :: Crash Reporting, defect)

Description

[Mayank Bansal]

https://crash-stats.mozilla.org/report/index/d5f8e96f-a2f3-465f-a425-7ffb30250604
https://crash-stats.mozilla.org/report/index/5a192345-3193-44e0-98b6-7162d0250604
https://crash-stats.mozilla.org/report/index/210a55f3-ae5d-4d77-9ace-289bd0250604
https://crash-stats.mozilla.org/report/index/fbb90b3b-bd88-4f7d-9fd0-ef8c20250604
https://crash-stats.mozilla.org/report/index/1ca3892b-ee4c-4ecf-acf4-bbda10250604

With the same STR, other crash signatures get generated that do have a signature. Example:
https://crash-stats.mozilla.org/report/index/424f30bb-cb53-4be4-a553-30f1a0250604
https://crash-stats.mozilla.org/report/index/385fefa2-c7f3-43bb-8515-b46fe0250604

Comment 1

[Mayank Bansal]

Rough STR is in bug 1970438

Comment 2

[Gabriele Svelto [:gsvelto]]

Attachment: Better stack trace

This is a bug in the stackwalker. The crash is happening within JIT'd code but we don't unwind properly from it. We have the fix available upstream so it's just a matter of cutting a new release and updating the version used in Socorro. In the meantime this is the proper stack trace for the crash. Note that the crash address contains the 0x4b pattern which is likely coming from the JS_SWEPT_TENURED_PATTERN. Nicolas, given that this seems reproducible can you have a look?

Comment 3

[Mayank Bansal]

Bug 1970164 is probably about the same thing.

Comment 4

[Nicolas B. Pierron [:nbp] {away}]

(In reply to Mayank Bansal from comment #3)

Bug 1970164 is probably about the same thing.

Bug 1970164 only landed in Firefox 140, and many of our crashes are coming from Firefox 139:

• 139.0.1 186 73.81 %
• 139.0 30 11.90 %
• 139.0b10 1 0.40 %

Comment 5

[Nicolas B. Pierron [:nbp] {away}]

(In reply to Gabriele Svelto [:gsvelto] from comment #2)

Nicolas, given that this seems reproducible can you have a look?

Thanks for the pointer, I will follow-up in Bug 1970164.

Comment 6

[Mayank Bansal]

This now has symbols, I think.

https://crash-stats.mozilla.org/report/index/1e235022-d3be-419b-b8d5-9fca50250610 is what i get today. Looks like soccoro was updated?

Comment 7

[Mayank Bansal]

gsvelto, should we close this bug as FIXED, as the crash signature has improved?

Comment 8

[Gabriele Svelto [:gsvelto]]

That was probably just luck. We need bug 1973186 for this to be fixed for good.

Comment 9

[BugBot [:suhaib / :marco]]

The bug is linked to a topcrash signature, which matches the following criterion:

• Top 20 desktop browser crashes on release (startup)

For more information, please visit BugBot documentation.

Comment 10

[BugBot [:suhaib / :marco]]

The severity field is not set for this bug.
:gsvelto, could you have a look please?

For more information, please visit BugBot documentation.

Comment 11

[Jens Stutte [:jstutte]]

(In reply to Gabriele Svelto [:gsvelto] from comment #8)

That was probably just luck. We need bug 1973186 for this to be fixed for good.

Should we make this bug depend on that one then?

Comment 12

[Gabriele Svelto [:gsvelto]]

Yes, this will fix this bug in terms of crash reporting even though it won't fix the root cause.

Comment 13

[BugBot [:suhaib / :marco]]

Based on the topcrash criteria, the crash signature linked to this bug is not a topcrash signature anymore.

For more information, please visit BugBot documentation.

Comment 14

[Gabriele Svelto [:gsvelto]]

The new stackwalker is in production and this particular type of crash is not producing a bad signature anymore (though we have more crashes under this signature, but they're different crashes).