linux-ext4 - Re: [PATCH 06/13] xfs: close the RT and log block devices in xfs_free

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

Message-ID: <20230809161745.GB2346@lst.de>
Date:   Wed, 9 Aug 2023 18:17:45 +0200
From:   Christoph Hellwig <hch@....de>
To:     "Darrick J. Wong" <djwong@...nel.org>
Cc:     Christoph Hellwig <hch@....de>, Al Viro <viro@...iv.linux.org.uk>,
        Christian Brauner <brauner@...nel.org>,
        Namjae Jeon <linkinjeon@...nel.org>,
        Sungjong Seo <sj1557.seo@...sung.com>,
        Theodore Ts'o <tytso@....edu>,
        Andreas Dilger <adilger.kernel@...ger.ca>,
        Konstantin Komarov <almaz.alexandrovich@...agon-software.com>,
        linux-fsdevel@...r.kernel.org, linux-ext4@...r.kernel.org,
        ntfs3@...ts.linux.dev, linux-xfs@...r.kernel.org
Subject: Re: [PATCH 06/13] xfs: close the RT and log block devices in
 xfs_free_buftarg

On Wed, Aug 09, 2023 at 08:45:32AM -0700, Darrick J. Wong wrote:
> >  	blkdev_issue_flush(btp->bt_bdev);
> >  	fs_put_dax(btp->bt_daxdev, btp->bt_mount);
> > +	/* the main block device is closed by kill_block_super */
> > +	if (bdev != btp->bt_mount->m_super->s_bdev)
> > +		blkdev_put(bdev, btp->bt_mount->m_super);
> 
> Hmm... I feel like this would be cleaner if the data dev buftarg could
> get its own refcount separate from super_block.s_bdev, but I looked
> through the code and couldn't identify a simple way to do that. Soo...

blkdev_put doesn't really drop a refcount, it closes the device.
It just happens to be misnamed, but Jan is looking into a series that
will as a side effect end up with a better name for this functionality.