[PATCH] kasan: Respect /proc/sys/kernel/traceoff_on_warning

[PATCH] kasan: Respect /proc/sys/kernel/traceoff_on_warning

[Peter Zijlstra]

After much waiting I finally reproduced a KASAN issue, only to find my
trace-buffer empty of useful information because it got spooled out :/

Make kasan_report honour the /proc/sys/kernel/traceoff_on_warning
interface.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
---
 mm/kasan/report.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/mm/kasan/report.c b/mm/kasan/report.c
index b82b3e2..6ff7efc 100644
--- a/mm/kasan/report.c
+++ b/mm/kasan/report.c
@@ -300,6 +300,8 @@ void kasan_report(unsigned long addr, size_t size,
 	if (likely(!kasan_report_enabled()))
 		return;
 
+	disable_trace_on_warning();
+
 	info.access_addr = (void *)addr;
 	info.access_size = size;
 	info.is_write = is_write;

Re: [PATCH] kasan: Respect /proc/sys/kernel/traceoff_on_warning

[Peter Zijlstra]

On Wed, Jan 25, 2017 at 03:25:24PM +0100, Peter Zijlstra wrote:
> 
> After much waiting I finally reproduced a KASAN issue, only to find my
> trace-buffer empty of useful information because it got spooled out :/
> 
> Make kasan_report honour the /proc/sys/kernel/traceoff_on_warning
> interface.
> 
> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
> ---

Seems to compile better when you add the required header files...

 mm/kasan/report.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/mm/kasan/report.c b/mm/kasan/report.c
index b82b3e2..3653ed1 100644
--- a/mm/kasan/report.c
+++ b/mm/kasan/report.c
@@ -24,6 +24,7 @@
 #include <linux/types.h>
 #include <linux/kasan.h>
 #include <linux/module.h>
+#include <linux/ftrace.h>
 
 #include <asm/sections.h>
 
@@ -300,6 +301,8 @@ void kasan_report(unsigned long addr, size_t size,
 	if (likely(!kasan_report_enabled()))
 		return;
 
+	disable_trace_on_warning();
+
 	info.access_addr = (void *)addr;
 	info.access_size = size;
 	info.is_write = is_write;

Re: [PATCH] kasan: Respect /proc/sys/kernel/traceoff_on_warning

[Alexander Potapenko]

On Wed, Jan 25, 2017 at 3:28 PM, Peter Zijlstra <peterz@infradead.org> wrote:
> On Wed, Jan 25, 2017 at 03:25:24PM +0100, Peter Zijlstra wrote:
>>
>> After much waiting I finally reproduced a KASAN issue, only to find my
>> trace-buffer empty of useful information because it got spooled out :/
>>
>> Make kasan_report honour the /proc/sys/kernel/traceoff_on_warning
>> interface.
>>
>> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
>> ---
>
> Seems to compile better when you add the required header files...
>
>  mm/kasan/report.c | 3 +++
>  1 file changed, 3 insertions(+)
>
> diff --git a/mm/kasan/report.c b/mm/kasan/report.c
> index b82b3e2..3653ed1 100644
> --- a/mm/kasan/report.c
> +++ b/mm/kasan/report.c
> @@ -24,6 +24,7 @@
>  #include <linux/types.h>
>  #include <linux/kasan.h>
>  #include <linux/module.h>
> +#include <linux/ftrace.h>
>
>  #include <asm/sections.h>
>
> @@ -300,6 +301,8 @@ void kasan_report(unsigned long addr, size_t size,
>         if (likely(!kasan_report_enabled()))
>                 return;
>
> +       disable_trace_on_warning();
> +
Who is going to enable tracing back after we're done?
Note that KASAN errors are not fatal, and the first one may possibly
occur quite early.
>         info.access_addr = (void *)addr;
>         info.access_size = size;
>         info.is_write = is_write;



-- 
Alexander Potapenko
Software Engineer

Google Germany GmbH
Erika-Mann-Straße, 33
80636 München

Geschäftsführer: Matthew Scott Sucherman, Paul Terence Manicle
Registergericht und -nummer: Hamburg, HRB 86891
Sitz der Gesellschaft: Hamburg

Re: [PATCH] kasan: Respect /proc/sys/kernel/traceoff_on_warning

[Peter Zijlstra]

On Wed, Jan 25, 2017 at 03:32:11PM +0100, Alexander Potapenko wrote:

> > @@ -300,6 +301,8 @@ void kasan_report(unsigned long addr, size_t size,
> >         if (likely(!kasan_report_enabled()))
> >                 return;
> >
> > +       disable_trace_on_warning();
> > +
> Who is going to enable tracing back after we're done?

The user..

> Note that KASAN errors are not fatal, and the first one may possibly
> occur quite early.

WARN*() isn't fatal either, and it does the same thing. Note that
traceoff_on_warning is default disabled. You have to explicitly request
this.

Re: [PATCH] kasan: Respect /proc/sys/kernel/traceoff_on_warning

[Alexander Potapenko]

On Wed, Jan 25, 2017 at 3:35 PM, Peter Zijlstra <peterz@infradead.org> wrote:
> On Wed, Jan 25, 2017 at 03:32:11PM +0100, Alexander Potapenko wrote:
>
>> > @@ -300,6 +301,8 @@ void kasan_report(unsigned long addr, size_t size,
>> >         if (likely(!kasan_report_enabled()))
>> >                 return;
>> >
>> > +       disable_trace_on_warning();
>> > +
>> Who is going to enable tracing back after we're done?
>
> The user..
>
>> Note that KASAN errors are not fatal, and the first one may possibly
>> occur quite early.
>
> WARN*() isn't fatal either, and it does the same thing. Note that
> traceoff_on_warning is default disabled. You have to explicitly request
> this.
Ah, that makes sense. Sorry, I wasn't aware of that interface before.



-- 
Alexander Potapenko
Software Engineer

Google Germany GmbH
Erika-Mann-Straße, 33
80636 München

Geschäftsführer: Matthew Scott Sucherman, Paul Terence Manicle
Registergericht und -nummer: Hamburg, HRB 86891
Sitz der Gesellschaft: Hamburg

Re: [PATCH] kasan: Respect /proc/sys/kernel/traceoff_on_warning

[Steven Rostedt]

On Wed, 25 Jan 2017 15:35:32 +0100
Peter Zijlstra <peterz@infradead.org> wrote:

> On Wed, Jan 25, 2017 at 03:32:11PM +0100, Alexander Potapenko wrote:
> 
> > > @@ -300,6 +301,8 @@ void kasan_report(unsigned long addr, size_t size,
> > >         if (likely(!kasan_report_enabled()))
> > >                 return;
> > >
> > > +       disable_trace_on_warning();
> > > +  
> > Who is going to enable tracing back after we're done?  
> 
> The user..

Yep,

 # echo 1 > /sys/kernel/debug/tracing/tracing_on

Re-enables tracing.

-- Steve

> 
> > Note that KASAN errors are not fatal, and the first one may possibly
> > occur quite early.  
> 
> WARN*() isn't fatal either, and it does the same thing. Note that
> traceoff_on_warning is default disabled. You have to explicitly request
> this.

Re: [PATCH] kasan: Respect /proc/sys/kernel/traceoff_on_warning

[Alexander Potapenko]

On Wed, Jan 25, 2017 at 3:25 PM, Peter Zijlstra <peterz@infradead.org> wrote:
>
> After much waiting I finally reproduced a KASAN issue, only to find my
> trace-buffer empty of useful information because it got spooled out :/
>
> Make kasan_report honour the /proc/sys/kernel/traceoff_on_warning
> interface.
>
> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Alexander Potapenko <glider@google.com>
> ---
>  mm/kasan/report.c | 2 ++
>  1 file changed, 2 insertions(+)
>
> diff --git a/mm/kasan/report.c b/mm/kasan/report.c
> index b82b3e2..6ff7efc 100644
> --- a/mm/kasan/report.c
> +++ b/mm/kasan/report.c
> @@ -300,6 +300,8 @@ void kasan_report(unsigned long addr, size_t size,
>         if (likely(!kasan_report_enabled()))
>                 return;
>
> +       disable_trace_on_warning();
> +
>         info.access_addr = (void *)addr;
>         info.access_size = size;
>         info.is_write = is_write;



-- 
Alexander Potapenko
Software Engineer

Google Germany GmbH
Erika-Mann-Straße, 33
80636 München

Geschäftsführer: Matthew Scott Sucherman, Paul Terence Manicle
Registergericht und -nummer: Hamburg, HRB 86891
Sitz der Gesellschaft: Hamburg

Re: [PATCH] kasan: Respect /proc/sys/kernel/traceoff_on_warning

[kbuild test robot]

[-- Attachment #1: Type: text/plain, Size: 1381 bytes --]

Hi Peter,

[auto build test ERROR on mmotm/master]
[also build test ERROR on v4.10-rc5 next-20170125]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]

url:    https://github.com/0day-ci/linux/commits/Peter-Zijlstra/kasan-Respect-proc-sys-kernel-traceoff_on_warning/20170125-230535
base:   git://git.cmpxchg.org/linux-mmotm.git master
config: x86_64-randconfig-x008-201704 (attached as .config)
compiler: gcc-6 (Debian 6.2.0-3) 6.2.0 20160901
reproduce:
        # save the attached .config to linux build tree
        make ARCH=x86_64 

All errors (new ones prefixed by >>):

   mm/kasan/report.c: In function 'kasan_report':
>> mm/kasan/report.c:303:2: error: implicit declaration of function 'disable_trace_on_warning' [-Werror=implicit-function-declaration]
     disable_trace_on_warning();
     ^~~~~~~~~~~~~~~~~~~~~~~~
   cc1: some warnings being treated as errors

vim +/disable_trace_on_warning +303 mm/kasan/report.c

   297	{
   298		struct kasan_access_info info;
   299	
   300		if (likely(!kasan_report_enabled()))
   301			return;
   302	
 > 303		disable_trace_on_warning();
   304	
   305		info.access_addr = (void *)addr;
   306		info.access_size = size;

---
0-DAY kernel test infrastructure                Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all                   Intel Corporation

[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 27451 bytes --]

[PATCH v2] kasan: Respect /proc/sys/kernel/traceoff_on_warning

[Andrey Ryabinin]

From: Peter Zijlstra <peterz@infradead.org>

After much waiting I finally reproduced a KASAN issue, only to find my
trace-buffer empty of useful information because it got spooled out :/

Make kasan_report honour the /proc/sys/kernel/traceoff_on_warning
interface.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Alexander Potapenko <glider@google.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Andrey Ryabinin <aryabinin@virtuozzo.com>
---
 mm/kasan/report.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/mm/kasan/report.c b/mm/kasan/report.c
index b82b3e2..f479365 100644
--- a/mm/kasan/report.c
+++ b/mm/kasan/report.c
@@ -13,6 +13,7 @@
  *
  */
 
+#include <linux/ftrace.h>
 #include <linux/kernel.h>
 #include <linux/mm.h>
 #include <linux/printk.h>
@@ -300,6 +301,8 @@ void kasan_report(unsigned long addr, size_t size,
 	if (likely(!kasan_report_enabled()))
 		return;
 
+	disable_trace_on_warning();
+
 	info.access_addr = (void *)addr;
 	info.access_size = size;
 	info.is_write = is_write;
-- 
2.10.2

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

[PATCH v2] kasan: Respect /proc/sys/kernel/traceoff_on_warning

[Andrey Ryabinin]

From: Peter Zijlstra <peterz@infradead.org>

After much waiting I finally reproduced a KASAN issue, only to find my
trace-buffer empty of useful information because it got spooled out :/

Make kasan_report honour the /proc/sys/kernel/traceoff_on_warning
interface.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Alexander Potapenko <glider@google.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Andrey Ryabinin <aryabinin@virtuozzo.com>
---
 mm/kasan/report.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/mm/kasan/report.c b/mm/kasan/report.c
index b82b3e2..f479365 100644
--- a/mm/kasan/report.c
+++ b/mm/kasan/report.c
@@ -13,6 +13,7 @@
  *
  */
 
+#include <linux/ftrace.h>
 #include <linux/kernel.h>
 #include <linux/mm.h>
 #include <linux/printk.h>
@@ -300,6 +301,8 @@ void kasan_report(unsigned long addr, size_t size,
 	if (likely(!kasan_report_enabled()))
 		return;
 
+	disable_trace_on_warning();
+
 	info.access_addr = (void *)addr;
 	info.access_size = size;
 	info.is_write = is_write;
-- 
2.10.2