I have been dealing with intermittent crashing in our Electron application for the last month and have not been able to arrive at a solution. Crashing causes the Electron window to turn blank because the renderer process is dead but the main process is still alive.
Crashing doesn't seem connected to any particular user action. The app lasts about half an hour on average, but could crash after only 5 minutes or run for a couple hours.
Our stack is pretty standard, with React, Redux, and Sagas.
Here is the exception from the crash dump:
In crash-43.dmp the assembly instruction at node!v8::internal::IncrementalMarking::RecordWriteSlow+1187 in C:\Users\Motorized Precision2\Desktop\mpstudio\node_modules\electron\dist\node.dll from Node.js has caused an access violation exception (0xC0000005) when trying to read from memory location 0x0000000b on thread 0
and the stack trace:
node!v8::internal::IncrementalMarking::RecordWriteSlow+1187
node!v8::internal::StoreBuffer::StoreBufferOverflow+1786
node!v8::internal::CancelableTaskManager::TryAbort+25f
electron!GetHandleVerifier+9e849
electron!IsSandboxedProcess+22d87e6
electron!IsSandboxedProcess+22d7a36
electron!IsSandboxedProcess+22d58b4
electron!GetHandleVerifier+9e849
electron!GetHandleVerifier+47367
electron!GetHandleVerifier+45dba
electron!GetHandleVerifier+a0e86
electron!GetHandleVerifier+4283e
electron!IsSandboxedProcess+3b4634
electron!GetHandleVerifier+3463da
electron!GetHandleVerifier+3462cf
electron!GetHandleVerifier+f6d68
electron+116365
electron!IsSandboxedProcess+2cd4efb
kernel32!BaseThreadInitThunk+14
ntdll!RtlUserThreadStart+21
I have examined the renderer code very carefully to make sure that all listeners/timeouts are handled, and now I'm honestly just not sure what to do next. Is this a memory leak, or something else? Someone save us!
User contributions licensed under CC BY-SA 3.0