Received-SPF: pass (google.com: domain of linux-kernel+bounces-29657-linux.lists.archive=gmail.com@vger.kernel.org designates 139.178.88.99 as permitted sender) client-ip=139.178.88.99;
Date: Wed, 17 Jan 2024 21:26:46 -0500
From: Steven Rostedt <rostedt@goodmis.org>
To: Chen Zhongjin <chenzhongjin@huawei.com>
Cc: <linux-kernel@vger.kernel.org>, <linux-trace-kernel@vger.kernel.org>,
 <yangjihong1@huawei.com>, <naveen.n.rao@linux.ibm.com>,
 <anil.s.keshavamurthy@intel.com>, <davem@davemloft.net>,
 <paulmck@kernel.org>, <mhiramat@kernel.org>, <akpm@linux-foundation.org>,
 <tglx@linutronix.de>, <peterz@infradead.org>, <pmladek@suse.com>,
 <dianders@chromium.org>, <npiggin@gmail.com>, <mpe@ellerman.id.au>,
 <jkl820.git@gmail.com>, <juerg.haefliger@canonical.com>,
 <rick.p.edgecombe@intel.com>, <eric.devolder@oracle.com>, <mic@digikod.net>
Subject: Re: [PATCH v2] kprobes: Use synchronize_rcu_tasks_rude in
 kprobe_optimizer
Message-ID: <20240117212646.5f0ddf0c@gandalf.local.home>
In-Reply-To: <20240118021842.290665-1-chenzhongjin@huawei.com>
References: <20240118021842.290665-1-chenzhongjin@huawei.com>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit

On Thu, 18 Jan 2024 02:18:42 +0000
Chen Zhongjin <chenzhongjin@huawei.com> wrote:

> There is a deadlock scenario in kprobe_optimizer():
> 
> pid A				pid B			pid C
> kprobe_optimizer()		do_exit()		perf_kprobe_init()
> mutex_lock(&kprobe_mutex)	exit_tasks_rcu_start()	mutex_lock(&kprobe_mutex)
> synchronize_rcu_tasks()		zap_pid_ns_processes()	// waiting kprobe_mutex
> // waiting tasks_rcu_exit_srcu	kernel_wait4()
> 				// waiting pid C exit
> 
> To avoid this deadlock loop, use synchronize_rcu_tasks_rude() in kprobe_optimizer()
> rather than synchronize_rcu_tasks(). synchronize_rcu_tasks_rude() can also promise
> that all preempted tasks have scheduled, but it will not wait tasks_rcu_exit_srcu.
> 

Did lockdep detect this? If not, we should fix that.

I'm also thinking if we should find another solution, as this seems more of
a work around than a fix.

> Fixes: a30b85df7d59 ("kprobes: Use synchronize_rcu_tasks() for optprobe with CONFIG_PREEMPT=y")
> Signed-off-by: Chen Zhongjin <chenzhongjin@huawei.com>
> ---
> v1 -> v2: Add Fixes tag
> ---
>  arch/Kconfig     | 2 +-
>  kernel/kprobes.c | 2 +-
>  2 files changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/Kconfig b/arch/Kconfig
> index f4b210ab0612..dc6a18854017 100644
> --- a/arch/Kconfig
> +++ b/arch/Kconfig
> @@ -104,7 +104,7 @@ config STATIC_CALL_SELFTEST
>  config OPTPROBES
>  	def_bool y
>  	depends on KPROBES && HAVE_OPTPROBES
> -	select TASKS_RCU if PREEMPTION
> +	select TASKS_RUDE_RCU

Is this still a bug if PREEMPTION is not enabled?

-- Steve

>  
>  config KPROBES_ON_FTRACE
>  	def_bool y
> diff --git a/kernel/kprobes.c b/kernel/kprobes.c
> index d5a0ee40bf66..09056ae50c58 100644
> --- a/kernel/kprobes.c
> +++ b/kernel/kprobes.c
> @@ -623,7 +623,7 @@ static void kprobe_optimizer(struct work_struct *work)
>  	 * Note that on non-preemptive kernel, this is transparently converted
>  	 * to synchronoze_sched() to wait for all interrupts to have completed.
>  	 */
> -	synchronize_rcu_tasks();
> +	synchronize_rcu_tasks_rude();
>  
>  	/* Step 3: Optimize kprobes after quiesence period */
>  	do_optimize_kprobes();