Re: [syzbot] possible deadlock in ktime_get_coarse_ts64

From: Dmitrii Banshchikov <me@ubique.spb.ru>
To: Thomas Gleixner <tglx@linutronix.de>
Cc: Alexei Starovoitov <alexei.starovoitov@gmail.com>,
	syzbot <syzbot+43fd005b5a1b4d10781e@syzkaller.appspotmail.com>,
	John Stultz <john.stultz@linaro.org>,
	LKML <linux-kernel@vger.kernel.org>,
	sboyd@kernel.org,
	syzkaller-bugs <syzkaller-bugs@googlegroups.com>,
	Peter Zijlstra <peterz@infradead.org>,
	Mark Rutland <mark.rutland@arm.com>,
	Steven Rostedt <rosted@goodmis.org>,
	Alexei Starovoitov <ast@kernel.org>,
	Daniel Borkmann <daniel@iogearbox.net>, bpf <bpf@vger.kernel.org>
Subject: Re: [syzbot] possible deadlock in ktime_get_coarse_ts64
Date: Sun, 7 Nov 2021 17:51:15 +0400	[thread overview]
Message-ID: <20211107135115.tqqx62sxsfeuzslb@amnesia> (raw)
In-Reply-To: <87zgqg2r4o.ffs@tglx>

On Sun, Nov 07, 2021 at 11:32:07AM +0100, Thomas Gleixner wrote:
> > 2) bpf_spin_lock/unlock have notrace attribute set.
> 
> How is that supposed to help?
> 
> You cannot take a spinlock from NMI context if that same lock can be
> taken by other contexts as well.
> 
> Also notrace on the public function is not guaranteeing that the inlines
> (as defined) are not traceable and it does not exclude it from being
> kprobed.
> 
> > 3) bpf_timer_* helpers fail early if they are in NMI.
> >
> > Why they have to be excluded?
> 
> Because timers take locks and you can just end up in the very same
> situation that you create invers lock dependencies or deadlocks when you
> use that from a tracepoint.
> 
> hrtimer_start()
>   lock_base();
>   trace_hrtimer...()
>     perf_event()
>       bpf_run()
>         bpf_timer_start()
>           hrtimer_start()
>             lock_base()         <- DEADLOCK
> 
> Tracepoints and perf events are very limited in what they can actually
> do. Just because it's BPF these rules are not magically going away.
> 

Thanks for the clarification.

-- 

Dmitrii Banshchikov