linux-kernel - Re: [PATCH 09/10] n_tracerouter ldisc driver.

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:	Thu, 24 Feb 2011 20:20:36 +0100 (CET)
From:	Thomas Gleixner <tglx@...utronix.de>
To:	james_p_freyensee@...ux.intel.com
cc:	gregkh@...e.de, linux-kernel@...r.kernel.org,
	suhail.ahmed@...el.com, christophe.guerard@...el.com
Subject: Re: [PATCH 09/10] n_tracerouter ldisc driver.

On Thu, 24 Feb 2011, james_p_freyensee@...ux.intel.com wrote:
> +
> +/**
> + * tracerouter_alloc
> + *
> + * Allocates the structure needed for this ldisc.
> + */
> +static struct tracerouter_data *tracerouter_alloc(void)
> +{
> +	struct tracerouter_data *tptr = kzalloc(
> +					sizeof(struct tracerouter_data),
> +					GFP_KERNEL);
> +	if (tptr == NULL)
> +		return NULL;
> +	tptr->opencalled = 0;

Wheeee!

You first alloc it with kzalloc and then clear the already zero member
again ?

That whole crap condenses down to:

     return kzalloc(sizeof(*tptr), GFP_KERNEL);

Which hardly justfies a separate function

> +	return tptr;
> +}
> +
> +/**
> + * n_tracerouter_open() - Called when a tty is opened by a SW entity.
> + * @tty: terminal device to the ldisc.
> + *
> + * Return:
> + *      0 for success.
> + *
> + * Caveats: This should only be opened one time per SW entity.
> + */
> +static int n_tracerouter_open(struct tty_struct *tty)
> +{
> +	int retval = -EEXIST;
> +
> +	mutex_lock(&routelock);
> +	if (tr_data->opencalled == 0) {
> +
> +		tr_data->kref_tty = tty_kref_get(tty);
> +		if (tr_data->kref_tty == NULL)
> +			retval = -EFAULT;
> +		else {

Please use braces for the if as well. It's just irritating not to have
them before the else.

> +			tr_data->opencalled = 1;
> +			tty->disc_data      = tr_data;
> +			tty->receive_room   = RECEIVE_ROOM;
> +			tty_driver_flush_buffer(tty);
> +			retval = 0;
> +		}
> +	}
> +	mutex_unlock(&routelock);
> +	return retval;
> +}
> +
> +/**
> + * n_tracerouter_close() - close connection
> + * @tty: terminal device to the ldisc.
> + *
> + * Called when a software entity wants to close a connection.
> + */
> +static void n_tracerouter_close(struct tty_struct *tty)
> +{
> +	struct tracerouter_data *tptr = tty->disc_data;
> +
> +	WARN_ON(tptr->kref_tty != tr_data->kref_tty);
> +	tty_driver_flush_buffer(tty);

That code probably never run with lockdep as you would get a potential
deadlock warning. See n_tracerouter_open().

Though the deadlock cannot happen as you are protected by
tr_data->opencalled it does not make it more correct.

> +	mutex_lock(&routelock);
> +	tty_kref_put(tr_data->kref_tty);
> +	tr_data->kref_tty = NULL;
> +	tr_data->opencalled = 0;
> +	tty->disc_data = NULL;
> +	mutex_unlock(&routelock);
> +}
> +

Thanks,

	tglx
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/