linux-kernel - Re: [CFT][PATCH] kernfs: Correct kernfs directory seeks.

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] [day] [month] [year] [list]

Date:   Thu, 7 Jun 2018 19:36:20 +0100
From:   Al Viro <viro@...IV.linux.org.uk>
To:     "Eric W. Biederman" <ebiederm@...ssion.com>
Cc:     "'tj@...nel.org'" <tj@...nel.org>,
        "Hatayama, Daisuke" <d.hatayama@...fujitsu.com>,
        "'gregkh@...uxfoundation.org'" <gregkh@...uxfoundation.org>,
        "Okajima, Toshiyuki" <toshi.okajima@...fujitsu.com>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "'ebiederm@...stanetworks.com'" <ebiederm@...stanetworks.com>
Subject: Re: [CFT][PATCH] kernfs: Correct kernfs directory seeks.

On Tue, Jun 05, 2018 at 12:47:33PM -0500, Eric W. Biederman wrote:

> Sigh,  I have found another issue with kernfs_fop_readdir.
> 
> We are not currently protecting file->private_data with the kernfs_mutex
> or any other kind of serialization.  Which means if two processes are
> calling readdir on the same file descriptor we might get unpredictable
> behavior.
> 
> It doesn't look too bad and easy enough to fix, but definitely something
> to be watchful of.

As discussed off-list - this is not a problem; getdents() et.al. are
serialized on per-struct-file basis by fdget_pos() in relevant syscalls,
since all directories automatically get FMODE_ATOMIC_POS in ->f_mode.