linux-kernel - Re: [PATCH] ftrace: disable tracing for suspend to ram

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: <200808271526.46868.rjw@sisk.pl>
Date:	Wed, 27 Aug 2008 15:26:45 +0200
From:	"Rafael J. Wysocki" <rjw@...k.pl>
To:	Steven Rostedt <rostedt@...dmis.org>
Cc:	LKML <linux-kernel@...r.kernel.org>, Pavel Machek <pavel@...e.cz>,
	Marcin Slusarz <marcin.slusarz@...il.com>,
	Ingo Molnar <mingo@...e.hu>,
	Nigel Cunningham <nigel@...el.suspend2.net>,
	Andrew Morton <akpm@...ux-foundation.org>,
	Linus Torvalds <torvalds@...ux-foundation.org>
Subject: Re: [PATCH] ftrace: disable tracing for suspend to ram

On Wednesday, 27 of August 2008, Steven Rostedt wrote:
> 
> I've been painstakingly debugging the issue with suspend to ram and 
> ftraced. The 2.6.28 code does not have this issue, but since the mcount 
> recording is not going to be in 27, this must be solved for the ftrace 
> daemon version.
> 
> The resume from suspend to ram would reboot because it was triple 
> faulting. Debugging further, I found that calling the mcount function 
> itself was not an issue, but it would fault when it incremented 
> preempt_count. preempt_count is on the tasks info structure that is on the 
> low memory address of the task's stack.  For some reason, it could not 
> write to it. Resuming out of suspend to ram does quite a lot of funny 
> tricks to get to work, so it is not surprising at all that simply doing a 
> preempt_disable() would cause a fault.
> 
> Thanks to Rafael for suggesting to add a "while (1);" to find the place in 
> resuming that is causing the fault. I would place the loop somewhere in 
> the code, compile and reboot and see if it would either reboot (hit the 
> fault) or simply hang (hit the loop).  Doing this over and over again, I 
> narrowed it down that it was happening in enable_nonboot_cpus.
> 
> At this point, I found that it is easier to simply disable tracing around 
> the suspend code, instead of searching for the particular function that 
> can not handle doing a preempt_disable.
> 
> This patch disables the tracer as it suspends and reenables it on resume.
> 
> I tested this patch on my Laptop, and it can resume fine with the patch.

Great, thanks for fixing that!

> Signed-off-by: Steven Rostedt <srostedt@...hat.com>

Acked-by: Rafael J. Wysocki <rjw@...k.pl>

> ---
>  kernel/power/main.c |    5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> Index: linux-compile.git/kernel/power/main.c
> ===================================================================
> --- linux-compile.git.orig/kernel/power/main.c	2008-08-27 08:53:11.000000000 -0400
> +++ linux-compile.git/kernel/power/main.c	2008-08-27 08:53:58.000000000 -0400
> @@ -21,6 +21,7 @@
>  #include <linux/freezer.h>
>  #include <linux/vmstat.h>
>  #include <linux/syscalls.h>
> +#include <linux/ftrace.h>
>  
>  #include "power.h"
>  
> @@ -310,7 +311,7 @@ static int suspend_enter(suspend_state_t
>   */
>  int suspend_devices_and_enter(suspend_state_t state)
>  {
> -	int error;
> +	int error, ftrace_save;
>  
>  	if (!suspend_ops)
>  		return -ENOSYS;
> @@ -321,6 +322,7 @@ int suspend_devices_and_enter(suspend_st
>  			goto Close;
>  	}
>  	suspend_console();
> +	ftrace_save = __ftrace_enabled_save();
>  	suspend_test_start();
>  	error = device_suspend(PMSG_SUSPEND);
>  	if (error) {
> @@ -352,6 +354,7 @@ int suspend_devices_and_enter(suspend_st
>  	suspend_test_start();
>  	device_resume(PMSG_RESUME);
>  	suspend_test_finish("resume devices");
> +	__ftrace_enabled_restore(ftrace_save);
>  	resume_console();
>   Close:
>  	if (suspend_ops->end)

Well, if that's enable_nonboot_cpus() that's faulting, I guess a similar change
in the hibernation code is needed.  I'll try to prepare a patch for that.

Thanks,
Rafael
--
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/