Re: emulating no-senders notifications in L4?

[Thomas Bushnell, BSG]

Espen Skoglund <address@hidden> writes:

> [Thomas Bushnell]
> > Espen Skoglund <address@hidden> writes:
> >> If the handler doesn't do so the task will merely loop on
> >> generating exception, but this is not much different from a task
> >> going into an endless programmed loop.
> 
> > Except that traditionally in Unix such tasks get blown away with
> > SIGILL, and we should do the same thing too.
> 
> The default handler for a SIGILL would of course be to blow away the
> task (with a core dump), just like traditional Unix systems.  The
> application may override this, though.  For unmaskable signals
> (SIGKILL, SIGSTOP) the signal syscall implementation would have to
> send a message to the task server to kill/stop the task.

You're not quite right.

In BSD Unix, if you destroy the ability to even deliver signals, then
the SIGILL is sent untrappably and generates a core dump.

Thomas