NFS and recalc_sigpending() semantics?

[Greg Banks <gnb@melbourne.sgi.com>]

G'day,

Can someone please explain to me why recalc_sigpending() fails
to clear the TIF_SIGPENDING flag if the pending signal is a
coredumping signal like SIGQUIT?  Is this deliberate?

I ask because there's code in the NFS client like this:

---> got here because signal_pending() is set <----
	spin_lock_irqsave(&current->sighand->siglock, flags);
	oldset = current->blocked;
	sigfillset(&current->blocked);
	recalc_sigpending();
	spin_unlock_irqrestore(&current->sighand->siglock, flags);
---> assumes signal_pending() is cleared <----

Have these semantics changed, or has NFS always been broken?

Greg.
-- 
Greg Banks, R&D Software Engineer, SGI Australian Software Group.
I don't speak for SGI.