I know it sounds a lazy question.. but I really have no much clue about how this scenario happened, neither can I find much information about it on Google.
Backg
The offending line can be found here: http://grepcode.com/file/repository.grepcode.com/java/ext/com.google.android/android/5.0.0_r1/com/android/server/am/ActivityManagerService.java/#4858 (Different line for different versions of Android L)
I assume you are using some form of Android L, since that particular error message wasn't added until then.
If you are running a ContentProvider in your process, these two comments in the ActivityManagerService might be of help:
9303 // NOTE: there is still a race here where a signal could be
9304 // pending on the process even though we managed to update its
9305 // adj level. Not sure what to do about this, but at least
9306 // the race is now smaller.
9307 if (!success) {
9308 // Uh oh... it looks like the provider's process
9309 // has been killed on us. We need to wait for a new
9310 // process to be started, and make sure its death
9311 // doesn't kill our process.
and then ...
9317 appDiedLocked(cpr.proc);
appDiedLocked can be called from some other source files too: ActiveServices.java and ActivityStackSupervisor.java, one depends on DeadObjectException being thrown, the other on RemoteException.
appDiedLocked looks like this
4853 final void appDiedLocked(ProcessRecord app, int pid, IApplicationThread thread) {
4854 // First check if this ProcessRecord is actually active for the pid.
4855 synchronized (mPidsSelfLocked) {
4856 ProcessRecord curProc = mPidsSelfLocked.get(pid);
4857 if (curProc != app) {
4858 Slog.w(TAG, "Spurious death for " + app + ", curProc for " + pid + ": " + curProc);
4859 return;
4860 }
4861 }
For some reason, the curProc isn't the same as the app ProcessRecord, and the appDiedLocked method is cut short. In your case curProc is null, again for some reason.
Long story short: Your process died, or was killed, and some state or condition is preventing appDiedLocked from continuing down to run the killProcess commands. You have some more investigation / logging to do to find out why that happened.
If you have a service that you want to keep alive, unless you are doing it already, I would recommend that you attach a status bar notification to it, that way the likeliness of it being killed is lowered.