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;
Date:   Mon, 19 Sep 2022 21:03:23 +0800
From:   Ming Lei <ming.lei@redhat.com>
To:     ZiyangZhang <ZiyangZhang@linux.alibaba.com>
Cc:     axboe@kernel.dk, xiaoguang.wang@linux.alibaba.com,
        linux-block@vger.kernel.org, linux-kernel@vger.kernel.org,
        joseph.qi@linux.alibaba.com
Subject: Re: [PATCH V3 6/7] ublk_drv: add START_USER_RECOVERY and
 END_USER_RECOVERY support
Message-ID: <Yyhom5s4AxK+wSu4@T590>
References: <20220913041707.197334-1-ZiyangZhang@linux.alibaba.com>
 <20220913041707.197334-7-ZiyangZhang@linux.alibaba.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20220913041707.197334-7-ZiyangZhang@linux.alibaba.com>
Precedence: bulk

On Tue, Sep 13, 2022 at 12:17:06PM +0800, ZiyangZhang wrote:
> START_USER_RECOVERY and END_USER_RECOVERY are two new control commands
> to support user recovery feature.
> 
> After a crash, user should send START_USER_RECOVERY, it will:
> (1) check if (a)current ublk_device is UBLK_S_DEV_QUIESCED which was
>     set by quiesce_work and (b)the file struct is released.
> (2) reinit all ubqs, including:
>     (a) put the task_struct and reset ->ubq_daemon to NULL.
>     (b) reset all ublk_io.
> (3) reset ub->mm to NULL.
> 
> Then, user should start a new process and send FETCH_REQ on each
> ubq_daemon.
> 
> Finally, user should send END_USER_RECOVERY, it will:
> (1) wait for all new ubq_daemons getting ready.
> (2) update ublksrv_pid
> (3) unquiesce the request queue and expect incoming ublk_queue_rq()
> (4) convert ub's state to UBLK_S_DEV_LIVE
> 
> Note: we can handle STOP_DEV between START_USER_RECOVERY and
> END_USER_RECOVERY. This is helpful to users who cannot start new process
> after sending START_USER_RECOVERY ctrl-cmd.
> 
> Signed-off-by: ZiyangZhang <ZiyangZhang@linux.alibaba.com>
> ---
>  drivers/block/ublk_drv.c | 116 +++++++++++++++++++++++++++++++++++++++
>  1 file changed, 116 insertions(+)
> 
> diff --git a/drivers/block/ublk_drv.c b/drivers/block/ublk_drv.c
> index 4409a130d0b6..3a3af80ee938 100644
> --- a/drivers/block/ublk_drv.c
> +++ b/drivers/block/ublk_drv.c
> @@ -1961,6 +1961,116 @@ static int ublk_ctrl_set_params(struct io_uring_cmd *cmd)
>  	return ret;
>  }
>  
> +static void ublk_queue_reinit(struct ublk_device *ub, struct ublk_queue *ubq)
> +{
> +	int i;
> +
> +	WARN_ON_ONCE(!(ubq->ubq_daemon && ubq_daemon_is_dying(ubq)));
> +	/* All old ioucmds have to be completed */
> +	WARN_ON_ONCE(ubq->nr_io_ready);
> +	pr_devel("%s: prepare for recovering qid %d\n", __func__, ubq->q_id);
> +	/* old daemon is PF_EXITING, put it now */
> +	put_task_struct(ubq->ubq_daemon);
> +	/* We have to reset it to NULL, otherwise ub won't accept new FETCH_REQ */
> +	ubq->ubq_daemon = NULL;
> +
> +	for (i = 0; i < ubq->q_depth; i++) {
> +		struct ublk_io *io = &ubq->ios[i];
> +
> +		/* forget everything now and be ready for new FETCH_REQ */
> +		io->flags = 0;
> +		io->cmd = NULL;
> +		io->addr = 0;
> +	}
> +}
> +
> +static int ublk_ctrl_start_recovery(struct io_uring_cmd *cmd)
> +{
> +	struct ublksrv_ctrl_cmd *header = (struct ublksrv_ctrl_cmd *)cmd->cmd;
> +	struct ublk_device *ub;
> +	int ret = -EINVAL;
> +	int i;
> +
> +	ub = ublk_get_device_from_id(header->dev_id);
> +	if (!ub)
> +		return ret;
> +
> +	mutex_lock(&ub->mutex);
> +	if (!ublk_can_use_recovery(ub))
> +		goto out_unlock;
> +	/*
> +	 * START_RECOVERY is only allowd after:
> +	 *
> +	 * (1) UB_STATE_OPEN is not set, which means the dying process is exited
> +	 *     and related io_uring ctx is freed so file struct of /dev/ublkcX is
> +	 *     released.
> +	 *
> +	 * (2) UBLK_S_DEV_QUIESCED is set, which means the quiesce_work:
> +	 *     (a)has quiesced request queue
> +	 *     (b)has requeued every inflight rqs whose io_flags is ACTIVE
> +	 *     (c)has requeued/aborted every inflight rqs whose io_flags is NOT ACTIVE
> +	 *     (d)has completed/camceled all ioucmds owned by ther dying process
> +	 */
> +	if (test_bit(UB_STATE_OPEN, &ub->state) ||
> +			ub->dev_info.state != UBLK_S_DEV_QUIESCED) {
> +		ret = -EBUSY;
> +		goto out_unlock;
> +	}
> +	pr_devel("%s: start recovery for dev id %d.\n", __func__, header->dev_id);
> +	for (i = 0; i < ub->dev_info.nr_hw_queues; i++)
> +		ublk_queue_reinit(ub, ublk_get_queue(ub, i));
> +	/* set to NULL, otherwise new ubq_daemon cannot mmap the io_cmd_buf */
> +	ub->mm = NULL;
> +	ub->nr_queues_ready = 0;

I am wondering why you don't move the above(queue reinit, clearing ub->mm)
into ublk_ch_release(), and looks it is more clean to clear this stuff
there. Meantime I guess one control command might be enough for user
recovery.


Thanks, 
Ming