From:  Nicholas Nethercote <>
Date:  11 Nov 2016 06:39:23 Hong Kong Time

Re: PSA: Crash reports from crashes in SpiderMonkey now have moz crash reasons


[CC'ing the stability list]

Oh, nice! I hadn't even realized that this problem existed. 'moz crash
reason' is really helpful when diagnosing crash causes from crash reports,
so this is a great improvement. Thank you.

Here is a search that shows all crashes submitted in the past 7 days that
have a 'moz crash reason' field, faceted by that field:

Already some JS-related ones are showing up, e.g. 278 occurrences of
is someone touching JSAPI without an AutoJSAPI?)



On Fri, Nov 11, 2016 at 9:06 AM, Emanuel Hoogeveen <> wrote:

> With the landing of bug 1309573 (included in today's Nightly), crash
> reports from intentionally triggered crashes in SpiderMonkey now have a
> 'moz crash reason', which is the reason string or expression from a
> MOZ_CRASH() or failed assertion. This is something that most of core Gecko
> has had access to all along, but it was predicated on both
> MOZILLA_INTERNAL_API and MOZ_CRASHREPORTER - neither of which libjs had
> access to!
> The moz crash reason can be useful to understand stacks with heavy
> inlining at a glance, and might help aggregate crash signatures if the
> reason string or expression is unique enough. It can also be used to expose
> dynamic information such as an error code, though MOZ_CRASH() itself only
> works with string literals (several places in Gecko already work around
> this).
> As proof that this is working now, here are the first two crash reports
> from crashes in SM with a moz crash reason:
> a360-4f7f-8402-b07bb2161110
> 47fd-475f-ae51-4c4a82161110
> For context, we discovered that the moz crash reason was missing because
> the annotations from bug 1305360 weren't showing up. The protection added
> in that bug generates generic memory protection crashes that are very hard
> to pick out from the crowd without an annotation, so having the moz crash
> reason is essential.
> _______________________________________________
> dev-tech-js-engine-internals mailing list