1571352 - Crash in [@ nsComponentManagerImpl::GetServiceLocked(`anonymous namespace'::MutexLock&, `anonymous namespace'::EntryWrapper&, nsID const&, void**)]

Status: RESOLVED WORKSFORME

(Core :: XPCOM, defect)

Description

[Gabriele Svelto [:gsvelto]]

This bug is for crash report bp-4c95cdda-6aeb-4edf-a6e6-4ba4d0190803.

Top 10 frames of crashing thread:

0 xul.dll nsComponentManagerImpl::GetServiceLocked xpcom/components/nsComponentManager.cpp:1396
1 xul.dll nsComponentManagerImpl::GetServiceByContractID xpcom/components/nsComponentManager.cpp:1571
2 xul.dll static nsresult mozilla::dom::indexedDB::`anonymous namespace'::GetDatabaseFileURL dom/indexedDB/ActorsParent.cpp:3875
3 xul.dll static nsresult mozilla::dom::indexedDB::`anonymous namespace'::GetStorageConnection dom/indexedDB/ActorsParent.cpp:4593
4 xul.dll void mozilla::dom::indexedDB::`anonymous namespace'::TransactionDatabaseOperationBase::RunOnConnectionThread dom/indexedDB/ActorsParent.cpp:21793
5 xul.dll void mozilla::dom::indexedDB::`anonymous namespace'::Database::StartTransactionOp::RunOnConnectionThread dom/indexedDB/ActorsParent.cpp:13153
6 xul.dll nsresult mozilla::dom::indexedDB::`anonymous namespace'::TransactionDatabaseOperationBase::Run dom/indexedDB/ActorsParent.cpp:21986
7 xul.dll nsThread::ProcessNextEvent xpcom/threads/nsThread.cpp:1224
8 xul.dll NS_ProcessNextEvent xpcom/threads/nsThreadUtils.cpp:486
9 xul.dll nsresult mozilla::dom::indexedDB::`anonymous namespace'::ConnectionPool::ThreadRunnable::Run dom/indexedDB/ActorsParent.cpp:11829

I'm not sure what to make of these signatures: they're obviously some newish problem since they don't seem to happen on anything older than 69.0b. Apart from that I'm stumped: the stacks are often different and the cause is also different, sometimes a read access exception, sometimes a write and sometimes an execute one. The address also varies wildy and there's at least one that looks like an UAF. The vast majority of the crashes under this signature are 32-bit but there's a handful of 64-bit ones. The crashes seem to come from a limited number of installation so maybe it's flaky hardware but it's odd that it always happens in the same place.

Comment 1

[BugBot [:suhaib / :marco/ :calixte]]

Closing because no crashes reported for 12 weeks.