1344557 - Crash in mozalloc_abort | abort | rtc::FatalMessage::~FatalMessage

Status: RESOLVED INCOMPLETE

(Core :: WebRTC, defect, P2)

Description

[Calixte Denizet (:calixte)]

This bug was filed from the Socorro interface and is 
report bp-fbd8d3c4-bdea-4248-8003-204da2170305.
=============================================================

There 2 crashes on nightly 54 with buildid 20170304030205. In analyzing the backtrace, the regression may have been introduced by patch [1] to fix bug 1337777.


[1] https://hg.mozilla.org/mozilla-central/rev?node=cfa8acce82af1ce85b4cbd8f341a4d64e9f3fa81

Comment 1

[Marcia Knous [:marcia]]

[@ webrtc::ProcessThreadImpl::RegisterModule] may also be related - it shows up in the Mac crash stats with the same build ID, but with a slightly different stack.

https://crash-stats.mozilla.com/report/index/50150b52-e79a-4ecb-a540-052c02170305

Comment 2

[Nils Ohlmeier [:drno]]

So what this crash signature is about is webrtc.org code failing assertions.

The problem is this covers lots of different assertions. One example is/was bug 1363563.

Comment 3

[Bulk Bug Changes for mreavy's org]

Mass change P1->P2 to align with new Mozilla triage process

Comment 4

[Randell Jesup [:jesup] (needinfo me)]

I'm going to ask that they modify the crash-stats reporting so that it ignores rtc::FatalMessage (and looks up the stack); that will split this into N bugs

Comment 5

[Randell Jesup [:jesup] (needinfo me)]

We're now reporting the source of the call to FatalMessage, so this bug is no longer useful