linux-kernel - Re: [PATCH] kernel: panic: make panic() local lock raw

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]

Message-Id: <20160311120818.069a1c742540e1693bd9e60b@linux-foundation.org>
Date:	Fri, 11 Mar 2016 12:08:18 -0800
From:	Andrew Morton <akpm@...ux-foundation.org>
To:	Sebastian Andrzej Siewior <bigeasy@...utronix.de>
Cc:	Thomas Gleixner <tglx@...utronix.de>, linux-kernel@...r.kernel.org,
	Steven Rostedt <rostedt@...dmis.org>
Subject: Re: [PATCH] kernel: panic: make panic() local lock raw

On Fri, 11 Mar 2016 18:48:25 +0100 Sebastian Andrzej Siewior <bigeasy@...utronix.de> wrote:

> The lock is taken with disabled interrupts and therefore it must be
> a RAW lock. Otherwise -RT will convert it into a sleeping lock and
> complain about it.

panic_lock doesn't exist in current kernels.

However you may want to take a look at pause_on_oops_lock - strange
things are happening with IRQs there.

> --- a/kernel/panic.c
> +++ b/kernel/panic.c
> @@ -71,7 +71,7 @@ void __weak panic_smp_self_stop(void)
>   */
>  void panic(const char *fmt, ...)
>  {
> -	static DEFINE_SPINLOCK(panic_lock);
> +	static DEFINE_RAW_SPINLOCK(panic_lock);

This should have had a comment explaining why a raw spinlock was chosen.

>  	static char buf[1024];
>  	va_list args;
>  	long i, i_next = 0;
> @@ -95,7 +95,7 @@ void panic(const char *fmt, ...)
>  	 * stop themself or will wait until they are stopped by the 1st CPU
>  	 * with smp_send_stop().
>  	 */
> -	if (!spin_trylock(&panic_lock))
> +	if (!raw_spin_trylock(&panic_lock))
>  		panic_smp_self_stop();
>  
>  	console_verbose();