Seen on Socorro. There are two fixed bugs about this signature. Then it came back on 2017-12-01. (No crashes between 2017-07-30 and 2017-12-01.) bp-ab6db79e-e5d1-44ce-87a8-904bc0171211 20171210220040 Win10 > Crash Reason EXCEPTION_ACCESS_VIOLATION_READ > 0 xul.dll mozilla::wr::LockExternalImage(void*, mozilla::wr::WrExternalImageId, unsigned char) gfx/webrender_bindings/RendererOGL.cpp:27 > 1 xul.dll webrender_bindings::bindings::{{impl}}::lock gfx/webrender_bindings/src/bindings.rs:307 > 2 xul.dll webrender::renderer::Renderer::update_texture_cache gfx/webrender/src/renderer.rs:2926 > 3 xul.dll webrender_bindings::bindings::wr_renderer_render gfx/webrender_bindings/src/bindings.rs:506 > 4 xul.dll mozilla::wr::RendererOGL::Render() gfx/webrender_bindings/RendererOGL.cpp:149 > 5 xul.dll mozilla::wr::RenderThread::UpdateAndRender(mozilla::wr::WrWindowId) gfx/webrender_bindings/RenderThread.cpp:227 > 6 xul.dll mozilla::wr::RenderThread::NewFrameReady(mozilla::wr::WrWindowId) gfx/webrender_bindings/RenderThread.cpp:174 > 7 xul.dll mozilla::detail::RunnableMethodImpl<mozilla::wr::RenderThread*, void ( mozilla::wr::RenderThread::*)(mozilla::wr::WrWindowId), 1, 0, mozilla::wr::WrWindowId>::Run() xpcom/threads/nsThreadUtils.h:1192 > 8 xul.dll MessageLoop::DoWork() ipc/chromium/src/base/message_loop.cc:535 > 9 xul.dll base::MessagePumpDefault::Run(base::MessagePump::Delegate*) ipc/chromium/src/base/message_pump_default.cc:36 > 10 xul.dll MessageLoop::RunHandler() ipc/chromium/src/base/message_loop.cc:319 > 11 xul.dll MessageLoop::Run() ipc/chromium/src/base/message_loop.cc:299 > 12 xul.dll base::Thread::ThreadMain() ipc/chromium/src/base/thread.cc:181 > 13 xul.dll `anonymous namespace'::ThreadFunc ipc/chromium/src/base/platform_thread_win.cc:28 > 14 kernel32.dll BaseThreadInitThunk > 15 mozglue.dll patched_BaseThreadInitThunk mozglue/build/WindowsDllBlocklist.cpp:835 > 16 ntdll.dll __RtlUserThreadStart > 17 ntdll.dll _RtlUserThreadStart

The lookup failed and gave us a null texture pointer. I wonder if this is a result of image.mem.shared; there is the potential for a rare race condition that we release the shared surface before the render thread has finished with it. It isn't evident from the reports which type of texture is failing. Fixing the shared surface race is a WIP for me.