This bug was filed from the Socorro interface and is report bp-0e633b82-f98f-4647-8d33-b098a0171116. ============================================================= Top 10 frames of crashing thread: 0 nss3.dll PR_Close nsprpub/pr/src/io/priometh.c:104 1 xul.dll mozilla::layout::PRFileDescStream::Close layout/printing/DrawEventRecorder.h:38 2 xul.dll nsDeviceContextSpecProxy::EndPage widget/nsDeviceContextSpecProxy.cpp:228 3 xul.dll nsDeviceContext::EndPage gfx/src/nsDeviceContext.cpp:605 4 xul.dll nsSimplePageSequenceFrame::DoPageEnd layout/generic/nsSimplePageSequenceFrame.cpp:809 5 xul.dll nsPrintEngine::PrintPage layout/printing/nsPrintEngine.cpp:2896 6 xul.dll nsPagePrintTimer::Run layout/printing/nsPagePrintTimer.cpp:82 7 xul.dll mozilla::SchedulerGroup::Runnable::Run xpcom/threads/SchedulerGroup.cpp:396 8 xul.dll nsThread::ProcessNextEvent xpcom/threads/nsThread.cpp:1037 9 xul.dll mozilla::ipc::MessagePump::Run ipc/glue/MessagePump.cpp:97 ============================================================= these content process crashes show up in greater numbers after we started rolling out 58.0b4 to the general beta audience.

As is becoming a theme with print issues, this baffles me :-) The two patches of mine that are in beta right now do not alter control flow at all, all we did was swap one file implementation for another. A relevant difference here is that the old one silently handles double-closes in release builds: https://searchfox.org/mozilla-central/source/gfx/2d/DrawEventRecorder.cpp#70-76 while the new one crashes. I suppose it's possible that there's some corner case that works ok on release builds under the old implementation, even though there was a |MOZ_ASSERT| there.

The right thing to do here is going to be to make this fail loudly only on debug. I'll upload a patch shortly which we should be able to uplift easily.

Comment on attachment 8929095 [details] Bug 1417939 - do not crash in release builds when a PRFileDescStream is closed twice; https://reviewboard.mozilla.org/r/200390/#review205524

Pushed by ryanvm@gmail.com: https://hg.mozilla.org/integration/autoland/rev/da98c602b3c6 do not crash in release builds when a PRFileDescStream is closed twice; r=bobowen

Comment on attachment 8929095 [details] Bug 1417939 - do not crash in release builds when a PRFileDescStream is closed twice; Approval Request Comment [Feature/Bug causing the regression]: Caused by refactors to printing for sandboxing: bug 1319423 (part 3 was independently reverted) [User impact if declined]: In unknown cases, a content process will crash when printing [Is this code covered by automated tests?]: Not well, no. [Has the fix been verified in Nightly?]: Landed in last night's nightly [Needs manual test from QE? If yes, steps to reproduce]: No. [List of other uplifts needed for the feature/fix]: None [Is the change risky?]: No. [Why is the change risky/not risky?]: It makes a failure condition not crash, emulating the previous behavior before the bug which introduced this regression. [String changes made/needed]: none

Comment on attachment 8929095 [details] Bug 1417939 - do not crash in release builds when a PRFileDescStream is closed twice; The volume of crashes in 58.0b4 is bit high. Take this in Beta58. Beta58+.