Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:3 as permitted sender) client-ip=2620:137:e000::3:3;
Date:   Wed, 15 Nov 2023 22:14:05 +0100
From:   Thomas =?utf-8?Q?Wei=C3=9Fschuh?= <thomas@t-8ch.de>
To:     huangzaiyang <huangzaiyang@oppo.com>
Cc:     tj@kernel.org, jiangshanlai@gmail.com,
        linux-kernel@vger.kernel.org, ZaiYang Huang <hzy0719@163.com>,
        Mukesh Ojha <quic_mojha@quicinc.com>
Subject: Re: [PATCH] workqueue: Make a warning when a pending delay work
 being re-init
Message-ID: <54af48a0-2d3c-4888-953b-28760f129d33@t-8ch.de>
References: <20231115113427.1420-1-huangzaiyang@oppo.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <20231115113427.1420-1-huangzaiyang@oppo.com>
Precedence: bulk

On 2023-11-15 19:34:27+0800, huangzaiyang wrote:
> There is a timer_list race condition if a delay work is queued twice,
> this usually won't happen unless someone reinitializes the task before performing the enqueue operation,likeï¼?
> https://github.com/torvalds/linux/blob/master/drivers/devfreq/devfreq.c#L487
> A warning message will help developers identify this irregular usage.
> 
> Signed-off-by: ZaiYang Huang <hzy0719@163.com>
> ---
>  include/linux/workqueue.h | 33 ++++++++++++++++++---------------
>  1 file changed, 18 insertions(+), 15 deletions(-)
> 
> diff --git a/include/linux/workqueue.h b/include/linux/workqueue.h
> index 24b1e5070f4d..54102ed794e5 100644
> --- a/include/linux/workqueue.h
> +++ b/include/linux/workqueue.h
> @@ -266,6 +266,22 @@ static inline void destroy_delayed_work_on_stack(struct delayed_work *work) { }
>  static inline unsigned int work_static(struct work_struct *work) { return 0; }
>  #endif
> 
> +/**
> + * work_pending - Find out whether a work item is currently pending
> + * @work: The work item in question
> + */
> +#define work_pending(work) \
> +       test_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))
> +
> +/**
> + * delayed_work_pending - Find out whether a delayable work item is currently
> + * pending
> + * @w: The work item in question
> + */
> +#define delayed_work_pending(w) \
> +       work_pending(&(w)->work)
> +
> +
>  /*
>   * initialize all of a work item in one go
>   *
> @@ -310,6 +326,7 @@ static inline unsigned int work_static(struct work_struct *work) { return 0; }
> 
>  #define __INIT_DELAYED_WORK(_work, _func, _tflags)                     \
>         do {                                                            \
> +               WARN_ON(delayed_work_pending(_work));                   \

How does this work when the data _work points to is not yet initialized?
Reading uninitialized data is UB and may spuriously trigger the warning.

>                 INIT_WORK(&(_work)->work, (_func));                     \
>                 __init_timer(&(_work)->timer,                           \
>                              delayed_work_timer_fn,                     \
> @@ -318,6 +335,7 @@ static inline unsigned int work_static(struct work_struct *work) { return 0; }

[..]

> --
> 2.17.1
> 
> ________________________________
> OPPO

The gunk at the end of the mail will prevent your patch from being
considered at all.

> 
> ±¾µç×ÓÓÊ¼þ¼°Æä¸½¼þº¬ÓÐOPPO¹«Ë¾µÄ±£ÃÜÐÅÏ¢£¬½öÏÞÓÚÓÊ¼þÖ¸Ã÷µÄÊÕ¼þÈË£¨°üº¬¸öÈË¼°Èº×é£©Ê¹ÓÃ¡£½ûÖ¹ÈÎºÎÈËÔÚÎ´¾­ÊÚÈ¨µÄÇé¿öÏÂÒÔÈÎºÎÐÎÊ½Ê¹ÓÃ¡£Èç¹ûÄú´íÊÕÁË±¾ÓÊ¼þ£¬ÇÐÎð´«²¥¡¢·Ö·¢¡¢¸´ÖÆ¡¢Ó¡Ë¢»òÊ¹ÓÃ±¾ÓÊ¼þÖ®ÈÎºÎ²¿·Ö»òÆäËùÔØÖ®ÈÎºÎÄÚÈÝ£¬²¢ÇëÁ¢¼´ÒÔµç×ÓÓÊ¼þÍ¨Öª·¢¼þÈË²¢É¾³ý±¾ÓÊ¼þ¼°Æä¸½¼þ¡£

[..]