1601988 - Crash in [@ mozilla::dom::BrowserChild::DoesWindowSupportProtectedMedia]

Status: RESOLVED DUPLICATE of bug 1594794

(Core :: DOM: Content Processes, defect)

Description

[[:philipp]]

This bug is for crash report bp-a3f4b292-bdb4-4982-8219-a51a10191206.

Top 10 frames of crashing thread:

0 xul.dll mozilla::dom::BrowserChild::DoesWindowSupportProtectedMedia dom/ipc/BrowserChild.cpp:3930
1 xul.dll void mozilla::dom::MediaKeySystemAccessManager::Request dom/media/eme/MediaKeySystemAccessManager.cpp:87
2 xul.dll mozilla::dom::MediaKeySystemAccessManager::Request dom/media/eme/MediaKeySystemAccessManager.cpp:67
3 xul.dll mozilla::dom::Navigator::RequestMediaKeySystemAccess dom/base/Navigator.cpp:1914
4 xul.dll static bool mozilla::dom::Navigator_Binding::requestMediaKeySystemAccess_promiseWrapper dom/bindings/NavigatorBinding.cpp:2162
5 xul.dll mozilla::dom::binding_detail::GenericMethod<mozilla::dom::binding_detail::NormalThisPolicy, mozilla::dom::binding_detail::ConvertExceptionsToPromises> dom/bindings/BindingUtils.cpp:3198
6 xul.dll js::InternalCallOrConstruct js/src/vm/Interpreter.cpp:550
7 xul.dll Interpret js/src/vm/Interpreter.cpp:3112
8 xul.dll js::RunScript js/src/vm/Interpreter.cpp:425
9 xul.dll js::ExecuteKernel js/src/vm/Interpreter.cpp:812

this crash signature is starting to show up in fairly low volume since firefox 71, potentially related to the changes from bug 1586347.

Comment 1

[Liz Henry (:lizzard) (relman/hg->git project)]

Bryce, can you take a look at this crash? It's low volume but looks to be new in 71.

Comment 2

[Bryce Seager van Dyk [:bryce] (he/him) - Not reading bugmail]

I think this is bug 1594794, my working:

• Crash reports show bad access at 2 consistent low memory addresses. Looks like a bad member access on a nullptr, one address for 32bit, one for 64bit.
• Builds on the release channel will not have the nullptr assertion that caught this in bug 1594794, so we crash later and get a different signature than the one seen there.
• The fix for the 1594794 was landed in 72, but not uplifted. I think the original regression may be from bug 1581855, with my changes in 71 changing where the crash takes place by moving around some of the machinery involved.

I'm going to resolve this as such and set an alarm for myself to check the crash stats and ensure we don't see this in 72.