linux-kernel - Re: [PATCH v2] stop_machine: Disable preemption when waking two stopper threads

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <0a5741e3d244b9914667608df5b5db23@codeaurora.org>
Date:   Mon, 02 Jul 2018 22:53:45 -0700
From:   isaacm@...eaurora.org
To:     Peter Zijlstra <peterz@...radead.org>
Cc:     matt@...eblueprint.co.uk, mingo@...nel.org, tglx@...utronix.de,
        bigeasy@...utronix.de, linux-kernel@...r.kernel.org,
        psodagud@...eaurora.org, pkondeti@...eaurora.org
Subject: Re: [PATCH v2] stop_machine: Disable preemption when waking two
 stopper threads

Hi Peter,

Thanks for the feedback. I'll make sure to incorporate it into my next
patch, and send that soon.

Thanks,
Isaac Manjarres
On 2018-07-02 05:15, Peter Zijlstra wrote:
> On Fri, Jun 29, 2018 at 01:55:12PM -0700, Isaac J. Manjarres wrote:
>> When cpu_stop_queue_two_works() begins to wake the stopper
>> threads, it does so without preemption disabled, which leads
>> to the following race condition:
>> 
>> The source CPU calls cpu_stop_queue_two_works(), with cpu1
>> as the source CPU, and cpu2 as the destination CPU. When
>> adding the stopper threads to the wake queue used in this
>> function, the source CPU stopper thread is added first,
>> and the destination CPU stopper thread is added last.
>> 
>> When wake_up_q() is invoked to wake the stopper threads, the
>> threads are woken up in the order that they are queued in,
>> so the source CPU's stopper thread is woken up first, and
>> it preempts the thread running on the source CPU.
>> 
>> The stopper thread will then execute on the source CPU,
>> disable preemption, and begin executing multi_cpu_stop(),
>> and wait for an ack from the destination CPU's stopper thread,
>> with preemption still disabled. Since the worker thread that
>> woke up the stopper thread on the source CPU is affine to the
>> source CPU, and preemption is disabled on the source CPU, that
>> thread will never run to dequeue the destination CPU's stopper
>> thread from the wake queue, and thus, the destination CPU's
>> stopper thread will never run, causing the source CPU's stopper
>> thread to wait forever, and stall.
>> 
>> Disable preemption when waking the stopper threads in
>> cpu_stop_queue_two_works() to ensure that the worker thread
>> that is waking up the stopper threads isn't preempted
>> by the source CPU's stopper thread, and permanently
>> scheduled out, leaving the remaining stopper thread asleep
>> in the wake queue.
>> 
>> Co-developed-by: Pavankumar Kondeti <pkondeti@...eaurora.org>
>> Signed-off-by: Prasad Sodagudi <psodagud@...eaurora.org>
>> Signed-off-by: Pavankumar Kondeti <pkondeti@...eaurora.org>
>> Signed-off-by: Isaac J. Manjarres <isaacm@...eaurora.org>
> 
> That SoB chain is broken, if Prasad wrote the ptch then there needs to
> be a From: line somewhere.
> 
> But yes, that looks about right.
> 
>> ---
>>  kernel/stop_machine.c | 6 +++++-
>>  1 file changed, 5 insertions(+), 1 deletion(-)
>> 
>> diff --git a/kernel/stop_machine.c b/kernel/stop_machine.c
>> index f89014a..1ff523d 100644
>> --- a/kernel/stop_machine.c
>> +++ b/kernel/stop_machine.c
>> @@ -270,7 +270,11 @@ static int cpu_stop_queue_two_works(int cpu1, 
>> struct cpu_stop_work *work1,
>>  		goto retry;
>>  	}
>> 
>> -	wake_up_q(&wakeq);
>> +	if (!err) {
>> +		preempt_disable();
>> +		wake_up_q(&wakeq);
>> +		preempt_enable();
>> +	}
>> 
>>  	return err;
>>  }
>> --
>> The Qualcomm Innovation Center, Inc. is a member of the Code Aurora 
>> Forum,
>> a Linux Foundation Collaborative Project
>>