Message-ID: <9b0b9059-c0f6-df5b-863a-987970f3567d@suse.de>
Date:   Sun, 12 Feb 2023 12:36:52 +0100
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.6.1
Subject: Re: [PATCH -next RFC 2/3] block: factor out the setting of
 GD_NEED_PART_SCAN
Content-Language: en-US
To:     Yu Kuai <yukuai1@huaweicloud.com>, hch@infradead.org, jack@suse.cz,
        axboe@kernel.dk
Cc:     linux-block@vger.kernel.org, linux-kernel@vger.kernel.org,
        yukuai3@huawei.com, yi.zhang@huawei.com, yangerkun@huawei.com
References: <20230212092641.2394146-1-yukuai1@huaweicloud.com>
 <20230212092641.2394146-3-yukuai1@huaweicloud.com>
From:   Hannes Reinecke <hare@suse.de>
In-Reply-To: <20230212092641.2394146-3-yukuai1@huaweicloud.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk

On 2/12/23 10:26, Yu Kuai wrote:
> From: Yu Kuai <yukuai3@huawei.com>
> 
> In order to prevent scan partition for a device that is opened
> exclusively by someone else, new conditions will be added to
> disk_scan_partitions() in the next patch. Hence if device is opened
> exclusively between bdev_add() and disk_scan_partitions(), the first
> partition scan will fail unexpected. This patch factor out the setting
> of GD_NEED_PART_SCAN to prevent the problem.
> 
> Signed-off-by: Yu Kuai <yukuai3@huawei.com>
> ---
>   block/genhd.c | 2 +-
>   block/ioctl.c | 1 +
>   2 files changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/block/genhd.c b/block/genhd.c
> index 075d8da284f5..c0d1220bd798 100644
> --- a/block/genhd.c
> +++ b/block/genhd.c
> @@ -367,7 +367,6 @@ int disk_scan_partitions(struct gendisk *disk, fmode_t mode)
>   	if (disk->open_partitions)
>   		return -EBUSY;
>   
> -	set_bit(GD_NEED_PART_SCAN, &disk->state);
>   	bdev = blkdev_get_by_dev(disk_devt(disk), mode, NULL);
>   	if (IS_ERR(bdev))
>   		return PTR_ERR(bdev);
> @@ -493,6 +492,7 @@ int __must_check device_add_disk(struct device *parent, struct gendisk *disk,
>   		if (ret)
>   			goto out_unregister_bdi;
>   
> +		set_bit(GD_NEED_PART_SCAN, &disk->state);
>   		bdev_add(disk->part0, ddev->devt);
>   		if (get_capacity(disk))
>   			disk_scan_partitions(disk, FMODE_READ);
Usual caveat:
What happens if the flag is already set here?
Wouldn't that imply that another scan is underway?
And wouldn't it be better to use 'test_and_set()'?


> diff --git a/block/ioctl.c b/block/ioctl.c
> index 6dd49d877584..0eefcdb936a0 100644
> --- a/block/ioctl.c
> +++ b/block/ioctl.c
> @@ -528,6 +528,7 @@ static int blkdev_common_ioctl(struct block_device *bdev, fmode_t mode,
>   			return -EACCES;
>   		if (bdev_is_partition(bdev))
>   			return -EINVAL;
> +		set_bit(GD_NEED_PART_SCAN, &bdev->bd_disk->state);
>   		return disk_scan_partitions(bdev->bd_disk, mode & ~FMODE_EXCL);
>   	case BLKTRACESTART:
>   	case BLKTRACESTOP:
Similar here.

Cheers,

Hannes
-- 
Dr. Hannes Reinecke                Kernel Storage Architect
hare@suse.de                              +49 911 74053 688
SUSE Software Solutions GmbH, Maxfeldstr. 5, 90409 Nürnberg
HRB 36809 (AG Nürnberg), Geschäftsführer: Ivo Totev, Andrew
Myers, Andrew McDonald, Martje Boudien Moerman