linux-kernel - Re: [PATCH 13/13] xfs: convert xfs_sysfs attrs to use ->seq

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: <YUDCsXXNFfUyiMCk@kroah.com>
Date:   Tue, 14 Sep 2021 17:41:37 +0200
From:   Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To:     Christoph Hellwig <hch@....de>
Cc:     "Rafael J. Wysocki" <rafael@...nel.org>,
        Alexander Viro <viro@...iv.linux.org.uk>,
        Jens Axboe <axboe@...nel.dk>, Tejun Heo <tj@...nel.org>,
        linux-block@...r.kernel.org, linux-xfs@...r.kernel.org,
        linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 13/13] xfs: convert xfs_sysfs attrs to use ->seq_show

On Tue, Sep 14, 2021 at 05:30:11PM +0200, Christoph Hellwig wrote:
> On Tue, Sep 14, 2021 at 05:28:08PM +0200, Greg Kroah-Hartman wrote:
> > We can "force" it by not allowing buffers to be bigger than that, which
> > is what the code has always done.  I think we want to keep that for now
> > and not add the new seq_show api.
> 
> The buffer already is not larger than that.  The problem is that
> sysfs_emit does not actually work for the non-trivial attributes,
> which generally are the source of bugs.

They huge majority of sysfs attributes are "trivial".  So for maybe at
least 95% of the users, if not more, using sysfs_emit() is just fine as
all you "should" be doing is emitting a single value.

For those that are non-trivial, yes, that will be harder, but as the xfs
discussion shows, those are not normal at all, and I do not want to make
creating them easier as that is not the model that sysfs was designed
for if at all possible.

thanks,

greg k-h