Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Thomas Gleixner <tglx@linutronix.de>
To:     Mukesh Ojha <quic_mojha@quicinc.com>, linux-kernel@vger.kernel.org
Cc:     sboyd@kernel.org, rafael@kernel.org, johannes@sipsolutions.net,
        gregkh@linuxfoundation.org, Mukesh Ojha <quic_mojha@quicinc.com>
Subject: Re: [PATCH v2 ] devcoredump : Serialize devcd_del work
In-Reply-To: <1650892193-12888-1-git-send-email-quic_mojha@quicinc.com>
Date:   Mon, 25 Apr 2022 19:00:08 +0200
Message-ID: <87levt14kn.ffs@tglx>
MIME-Version: 1.0
Content-Type: text/plain
Precedence: bulk

On Mon, Apr 25 2022 at 18:39, Mukesh Ojha wrote:
> v1->v2:
>  - Added del_wk_queued to serialize the race between devcd_data_write()
>    and disabled_store().

How so?

Neither the flag nor the mutex can prevent the race between the work
being executed in parallel.

disabled_store()                                worker()    

  class_for_each_device(&devcd_class, NULL, NULL, devcd_free)
    ...
    while ((dev = class_dev_iter_next(&iter)) {
    						devcd_del()
                                                 device_del()
                                                 put_device() <- last reference
          error = fn(dev, data)                   devcd_dev_release()
            devcd_free(dev, data)                  kfree(devcd)
              mutex_lock(&devcd->mutex);
      

There is zero protection of the class iterator against the work being
executed and removing the device and freeing its data. IOW, at the
point where fn(), i.e. devcd_free(), dereferences 'dev' to acquire the
mutex, it might be gone. No?

If disabled_store() really needs to flush all instances immediately,
then it requires global serialization, not device specific serialization.

Johannes, can you please explain whether this immediate flush in
disabled_store() is really required and if so, why?

Thanks,

        tglx