linux-kernel - Re: [PATCH 3/8] bdi: add a ->dev_name field to struct backing_dev

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: <20200419160651.GA18308@lst.de>
Date:   Sun, 19 Apr 2020 18:06:51 +0200
From:   Christoph Hellwig <hch@....de>
To:     Bart Van Assche <bvanassche@....org>
Cc:     Christoph Hellwig <hch@....de>, Jan Kara <jack@...e.cz>,
        axboe@...nel.dk, yuyufen@...wei.com, tj@...nel.org, tytso@....edu,
        gregkh@...uxfoundation.org, linux-block@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH 3/8] bdi: add a ->dev_name field to struct
 backing_dev_info

On Sun, Apr 19, 2020 at 08:29:21AM -0700, Bart Van Assche wrote:
> On 4/19/20 12:58 AM, Christoph Hellwig wrote:
>> On Sat, Apr 18, 2020 at 08:40:20AM -0700, Bart Van Assche wrote:
>>>> This can have a sideeffect not only bdi->dev_name will be truncated to 64
>>>> chars (which generally doesn't matter) but possibly also kobject name will
>>>> be truncated in the same way.  Which may have user visible effects. E.g.
>>>> for fs/vboxsf 64 chars need not be enough. So shouldn't we rather do it the
>>>> other way around - i.e., let device_create_vargs() create the device name
>>>> and then copy to bdi->dev_name whatever fits?
>>>
>>> How about using kvasprintf() instead of vsnprintf()?
>>
>> That is what v1 did, see the thread in response to that on why it isn't
>> a good idea.
>
> Are you perhaps referring to patch "[PATCH 3/8] bdi: add a ->dev_name field 
> to struct backing_dev_info" 
> (https://lore.kernel.org/linux-block/20200416071519.807660-4-hch@lst.de/) 
> and also to the replies to that patch? This is what I found in the replies: 
> "When driver try to to re-register bdi but without release_bdi(), the old 
> dev_name will be cover directly by the newer in bdi_register_va(). So, I am 
> not sure whether it can cause memory leak for bdi->dev_name."
>
> Has it been considered to avoid that leak by freeing bdi->dev_name from 
> unregister_bdi(), e.g. as follows?

We'd need some protection against concurrent accesses as unregister_bdi
can race with them.  But with RCU that could be handled, so let me try
that.