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: <20200916054157.GC825@sol.localdomain>
Date:   Tue, 15 Sep 2020 22:41:57 -0700
From:   Eric Biggers <ebiggers@...nel.org>
To:     Anant Thazhemadam <anant.thazhemadam@...il.com>
Cc:     linux-kernel-mentees@...ts.linuxfoundation.org,
        syzbot+4191a44ad556eacc1a7a@...kaller.appspotmail.com,
        Alexander Viro <viro@...iv.linux.org.uk>,
        linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] fs: fix KMSAN uninit-value bug by initializing nd in
 do_file_open_root

On Wed, Sep 16, 2020 at 10:56:56AM +0530, Anant Thazhemadam wrote:
> The KMSAN bug report for the bug indicates that there exists;
> Local variable ----nd@...file_open_root created at:
>  do_file_open_root+0xa4/0xb40 fs/namei.c:3385
>  do_file_open_root+0xa4/0xb40 fs/namei.c:3385
> 
> Initializing nd fixes this issue, and doesn't break anything else either
> 
> Fixes: https://syzkaller.appspot.com/bug?extid=4191a44ad556eacc1a7a
> Reported-by: syzbot+4191a44ad556eacc1a7a@...kaller.appspotmail.com
> Tested-by: syzbot+4191a44ad556eacc1a7a@...kaller.appspotmail.com
> Signed-off-by: Anant Thazhemadam <anant.thazhemadam@...il.com>
> ---
>  fs/namei.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/fs/namei.c b/fs/namei.c
> index e99e2a9da0f7..b27382586209 100644
> --- a/fs/namei.c
> +++ b/fs/namei.c
> @@ -3404,7 +3404,7 @@ struct file *do_filp_open(int dfd, struct filename *pathname,
>  struct file *do_file_open_root(struct dentry *dentry, struct vfsmount *mnt,
>  		const char *name, const struct open_flags *op)
>  {
> -	struct nameidata nd;
> +	struct nameidata nd = {};
>  	struct file *file;
>  	struct filename *filename;
>  	int flags = op->lookup_flags | LOOKUP_ROOT;

Looking at the actual KMSAN report, it looks like it's nameidata::dir_mode or
nameidata::dir_uid that is uninitialized.  You need to figure out the correct
solution, not just blindly initialize with zeroes -- that could hide a bug.
Is there a bug that is preventing these fields from being initialized to the
correct values, are these fields being used when they shouldn't be, etc...

- Eric

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ