linux-kernel - Re: [patch 2/2] kernel/timer.c: use debugobjects to catch deletion of uninitialized timers

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

Message-ID: <alpine.LFD.2.02.1109211028060.2723@ionos>
Date:	Wed, 21 Sep 2011 10:50:48 +0200 (CEST)
From:	Thomas Gleixner <tglx@...utronix.de>
To:	akpm@...gle.com
cc:	cschan@...eaurora.org, john.stultz@...aro.org,
	LKML <linux-kernel@...r.kernel.org>
Subject: Re: [patch 2/2] kernel/timer.c: use debugobjects to catch deletion
 of uninitialized timers

Christine,

On Tue, 20 Sep 2011, akpm@...gle.com wrote:

> From: Christine Chan <cschan@...eaurora.org>
> Subject: kernel/timer.c: use debugobjects to catch deletion of uninitialized timers
> 
> del_timer_sync() calls debug_object_assert_init() to assert that a timer
> has been initialized before calling lock_timer_base().  lock_timer_base()
> would spin forever on a NULL(uninit-ed) base.  The check is added to
> del_timer() to prevent silent failure, even though it would not get stuck
> in an infinite loop.

Ok, that makes sense, but the changelog of the debugobjects code wants
a better explanation.
 
> +/*
> + * fixup_assert_init is called when:
> + * - an untracked/uninit-ed object is found
> + */
> +static int timer_fixup_assert_init(void *addr, enum debug_obj_state state)
> +{
> +	struct timer_list *timer = addr;
> +
> +	switch (state) {
> +	case ODEBUG_STATE_NOTAVAILABLE:
> +		if (timer->entry.prev == TIMER_ENTRY_STATIC) {
> +			/*
> +			 * This is not really a fixup. The timer was
> +			 * statically initialized. We just make sure that it
> +			 * is tracked in the object tracker.
> +			 */
> +			debug_object_init(timer, &timer_debug_descr);
> +			return 0;
> +		} else {
> +			WARN_ON(1);
> +			init_timer(timer);
> +			return 1;

Can we please remove the WARN_ON() and use the return value in the
debugobjects code? That needs a variant of debug_print_object() to give us
proper state output.

Also the fixup function should check for timer->function == NULL and
set a default stub callback.

Thinking more about that, we should do the same for the
timer_fixup_activate() case.

Can you have a stab at that, please?

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/