From mboxrd@z Thu Jan  1 00:00:00 1970
From: Ingo Molnar <mingo2.kernel.org@gmail.com>
Subject: Re: [PATCH] signals: Generate warning when flush_signals() is called
 from non-kthread context
Date: Wed, 6 May 2015 12:19:27 +0200
Message-ID: <20150506101927.GA17291@gmail.com>
References: <1430502057.4472.255.camel@redhat.com>
 <CA+55aFziVpH1qyf8M+GdJxh3oJWd84pdJbo94kuMfSi1QscLew@mail.gmail.com>
 <20150501193813.GA2812@gmail.com>
 <20150503173401.GA22052@redhat.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: Linus Torvalds <torvalds@linux-foundation.org>,
	Neil Brown <neilb@suse.de>, Evgeniy Polyakov <zbr@ioremap.net>,
	Stephen Smalley <sds@tycho.nsa.gov>,
	Alex Williamson <alex.williamson@redhat.com>,
	linux-kernel <linux-kernel@vger.kernel.org>,
	kvm <kvm@vger.kernel.org>
To: Oleg Nesterov <oleg@redhat.com>
Return-path: <linux-kernel-owner@vger.kernel.org>
Content-Disposition: inline
In-Reply-To: <20150503173401.GA22052@redhat.com>
Sender: linux-kernel-owner@vger.kernel.org
List-Id: kvm.vger.kernel.org


* Oleg Nesterov <oleg@redhat.com> wrote:

> > --- a/kernel/signal.c
> > +++ b/kernel/signal.c
> > @@ -427,6 +427,10 @@ void flush_signals(struct task_struct *t)
> >  {
> >  	unsigned long flags;
> >
> > +	/* Only kthreads are allowed to destroy signals: */
> > +	if (WARN_ON_ONCE(!(current->flags & PF_KTHREAD)))
> > +		return;
> > +
> 
> But I am not sure this can't make some buggy driver even more buggy.
> Just suppose it does something
> 
> 	do {
> 		if (signal_pending())
> 			flush_signals();
> 	} while (wait_event_interruptible(...));
> 
> and this change will turn this into busy-wait loop.
> 
> So perhaps another change which just adds WARN_ON_RATELIMIT() 
> without "return" will be safer?

Yeah, absolutely.

Thanks,

	Ingo