linux-kernel - Re: [PATCH 2/2] tty: add retry to tty_init

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]

Date:   Wed, 20 Nov 2019 16:44:50 +0100
From:   Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To:     Sudip Mukherjee <sudipm.mukherjee@...il.com>
Cc:     Jiri Slaby <jslaby@...e.com>, Rob Herring <robh@...nel.org>,
        linux-kernel@...r.kernel.org, linux-serial@...r.kernel.org
Subject: Re: [PATCH 2/2] tty: add retry to tty_init_dev() to workaround a
 race condition

On Wed, Nov 20, 2019 at 03:17:09PM +0000, Sudip Mukherjee wrote:
> There seems to be a race condition in tty drivers and I could see on
> many boot cycles a NULL pointer dereference as tty_init_dev() tries to
> do 'tty->port->itty = tty' even though tty->port is NULL.
> 'tty->port' will be set by the driver and if the driver has not yet done
> it before we open the tty device we can get to this situation. By adding
> some extra debug prints, I noticed that tty_port_link_device() is
> initialising 'driver->ports[index]' just few microseconds after I
> get the warning.
> So, add one retry so that tty_init_dev() will return -EAGAIN on its first
> try if 'tty->port' is not set yet, and then tty_open() will try to open
> it again.
> 
> Signed-off-by: Sudip Mukherjee <sudipm.mukherjee@...il.com>
> ---
>  drivers/tty/pty.c                   |  2 +-
>  drivers/tty/serdev/serdev-ttyport.c |  2 +-
>  drivers/tty/tty_io.c                | 20 ++++++++++++++------
>  include/linux/tty.h                 |  3 ++-
>  4 files changed, 18 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/tty/pty.c b/drivers/tty/pty.c
> index 00099a8439d2..22e8c40d9f9c 100644
> --- a/drivers/tty/pty.c
> +++ b/drivers/tty/pty.c
> @@ -842,7 +842,7 @@ static int ptmx_open(struct inode *inode, struct file *filp)
>  
>  
>  	mutex_lock(&tty_mutex);
> -	tty = tty_init_dev(ptm_driver, index);
> +	tty = tty_init_dev(ptm_driver, index, 0);

Horrible naming scheme for this new "flag".

Look at that call here, can you instantly tell what this call is doing
with "0"?  I sure can not :(

If you really want to do this, you make a different function,
tty_init_dev_retry() and then have that pass in a retry flag in the tty
core, so that any users always know what they are doing here.

But, this really feels like a race in the code somewhere:

> --- a/drivers/tty/tty_io.c
> +++ b/drivers/tty/tty_io.c
> @@ -1295,6 +1295,7 @@ static int tty_reopen(struct tty_struct *tty)
>   *	tty_init_dev		-	initialise a tty device
>   *	@driver: tty driver we are opening a device on
>   *	@idx: device index
> + *	@retry: retry count if driver has not set tty->port yet

Why would tty->port not be set up already?  The caller has control over
this, what is not happening correctly to cause this?

thanks,

greg k-h