diff mbox

kprobes: bugfix: set kprobes_all_disarmed earlier to enable re-optimization.

Message ID 1420461075-27012-1-git-send-email-wangnan0@huawei.com (mailing list archive)
State New, archived
Headers show

Commit Message

Wang Nan Jan. 5, 2015, 12:31 p.m. UTC
In original code, the probed instruction doesn't get optimized after

echo 0 > /sys/kernel/debug/kprobes/enabled
echo 1 > /sys/kernel/debug/kprobes/enabled

This is because original code checks kprobes_all_disarmed in
optimize_kprobe(), but this flag is turned off after calling that
function. Therefore, optimize_kprobe() will see
kprobes_all_disarmed == true and doesn't do the optimization.

This patch simply turns off kprobes_all_disarmed earlier to enable
optimization.

Signed-off-by: Wang Nan <wangnan0@huawei.com>
---
 kernel/kprobes.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

Comments

Masami Hiramatsu Jan. 12, 2015, 11:24 a.m. UTC | #1
(2015/01/05 21:31), Wang Nan wrote:
> In original code, the probed instruction doesn't get optimized after
> 
> echo 0 > /sys/kernel/debug/kprobes/enabled
> echo 1 > /sys/kernel/debug/kprobes/enabled
> 
> This is because original code checks kprobes_all_disarmed in
> optimize_kprobe(), but this flag is turned off after calling that
> function. Therefore, optimize_kprobe() will see
> kprobes_all_disarmed == true and doesn't do the optimization.

OK, a workaround is to disable/enable optimizing via sysctl.

> 
> This patch simply turns off kprobes_all_disarmed earlier to enable
> optimization.
> 

Anyway, thank you for finding that! :)

Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>


> Signed-off-by: Wang Nan <wangnan0@huawei.com>
> ---
>  kernel/kprobes.c | 7 ++++++-
>  1 file changed, 6 insertions(+), 1 deletion(-)
> 
> diff --git a/kernel/kprobes.c b/kernel/kprobes.c
> index bad4e95..b185464 100644
> --- a/kernel/kprobes.c
> +++ b/kernel/kprobes.c
> @@ -2320,6 +2320,12 @@ static void arm_all_kprobes(void)
>  	if (!kprobes_all_disarmed)
>  		goto already_enabled;
>  
> +	/*
> +	 * optimize_kprobe() called by arm_kprobe() checks
> +	 * kprobes_all_disarmed, so set kprobes_all_disarmed before
> +	 * arm_kprobe.
> +	 */
> +	kprobes_all_disarmed = false;
>  	/* Arming kprobes doesn't optimize kprobe itself */
>  	for (i = 0; i < KPROBE_TABLE_SIZE; i++) {
>  		head = &kprobe_table[i];
> @@ -2328,7 +2334,6 @@ static void arm_all_kprobes(void)
>  				arm_kprobe(p);
>  	}
>  
> -	kprobes_all_disarmed = false;
>  	printk(KERN_INFO "Kprobes globally enabled\n");
>  
>  already_enabled:
>
Wang Nan Jan. 20, 2015, 3:02 a.m. UTC | #2
On 2015/1/12 19:24, Masami Hiramatsu wrote:
> (2015/01/05 21:31), Wang Nan wrote:
>> In original code, the probed instruction doesn't get optimized after
>>
>> echo 0 > /sys/kernel/debug/kprobes/enabled
>> echo 1 > /sys/kernel/debug/kprobes/enabled
>>
>> This is because original code checks kprobes_all_disarmed in
>> optimize_kprobe(), but this flag is turned off after calling that
>> function. Therefore, optimize_kprobe() will see
>> kprobes_all_disarmed == true and doesn't do the optimization.
> 
> OK, a workaround is to disable/enable optimizing via sysctl.
> 
>>
>> This patch simply turns off kprobes_all_disarmed earlier to enable
>> optimization.
>>
> 
> Anyway, thank you for finding that! :)
> 
> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
> 

Hi Masami Hiramatsu,

What about this patch? Are you still thinking on it like
'[PATCH] kprobes: bugfix: checks kprobes_all_disarmed in unoptimized_kprobe()' ?

Thank you.

> 
>> Signed-off-by: Wang Nan <wangnan0@huawei.com>
>> ---
>>  kernel/kprobes.c | 7 ++++++-
>>  1 file changed, 6 insertions(+), 1 deletion(-)
>>
>> diff --git a/kernel/kprobes.c b/kernel/kprobes.c
>> index bad4e95..b185464 100644
>> --- a/kernel/kprobes.c
>> +++ b/kernel/kprobes.c
>> @@ -2320,6 +2320,12 @@ static void arm_all_kprobes(void)
>>  	if (!kprobes_all_disarmed)
>>  		goto already_enabled;
>>  
>> +	/*
>> +	 * optimize_kprobe() called by arm_kprobe() checks
>> +	 * kprobes_all_disarmed, so set kprobes_all_disarmed before
>> +	 * arm_kprobe.
>> +	 */
>> +	kprobes_all_disarmed = false;
>>  	/* Arming kprobes doesn't optimize kprobe itself */
>>  	for (i = 0; i < KPROBE_TABLE_SIZE; i++) {
>>  		head = &kprobe_table[i];
>> @@ -2328,7 +2334,6 @@ static void arm_all_kprobes(void)
>>  				arm_kprobe(p);
>>  	}
>>  
>> -	kprobes_all_disarmed = false;
>>  	printk(KERN_INFO "Kprobes globally enabled\n");
>>  
>>  already_enabled:
>>
> 
>
Masami Hiramatsu Jan. 20, 2015, 7:12 a.m. UTC | #3
(2015/01/20 12:02), Wang Nan wrote:
> On 2015/1/12 19:24, Masami Hiramatsu wrote:
>> (2015/01/05 21:31), Wang Nan wrote:
>>> In original code, the probed instruction doesn't get optimized after
>>>
>>> echo 0 > /sys/kernel/debug/kprobes/enabled
>>> echo 1 > /sys/kernel/debug/kprobes/enabled
>>>
>>> This is because original code checks kprobes_all_disarmed in
>>> optimize_kprobe(), but this flag is turned off after calling that
>>> function. Therefore, optimize_kprobe() will see
>>> kprobes_all_disarmed == true and doesn't do the optimization.
>>
>> OK, a workaround is to disable/enable optimizing via sysctl.
>>
>>>
>>> This patch simply turns off kprobes_all_disarmed earlier to enable
>>> optimization.
>>>
>>
>> Anyway, thank you for finding that! :)
>>
>> Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
>>
> 
> Hi Masami Hiramatsu,
> 
> What about this patch? Are you still thinking on it like
> '[PATCH] kprobes: bugfix: checks kprobes_all_disarmed in unoptimized_kprobe()' ?

Without this patch.
-----
[root@localhost tracing]# echo p do_fork+5 > kprobe_events
[root@localhost tracing]# echo 1 > events/kprobes/p_do_fork_5/enable
[root@localhost tracing]# cat ../kprobes/list
ffffffff810bd055  k  do_fork+0x5    [OPTIMIZED]
[root@localhost tracing]# echo 0 > /sys/kernel/debug/kprobes/enabled
[root@localhost tracing]# cat ../kprobes/list
ffffffff810bd055  k  do_fork+0x5
[root@localhost tracing]# echo 1 > /sys/kernel/debug/kprobes/enabled
[root@localhost tracing]# cat ../kprobes/list
ffffffff810bd055  k  do_fork+0x5
------
With this patch.
------
[root@localhost tracing]# echo p do_fork+5 > kprobe_events
[root@localhost tracing]# echo 1 > events/kprobes/enable
[root@localhost tracing]# cat ../kprobes/list
ffffffff810bc1c5  k  do_fork+0x5    [OPTIMIZED]
[root@localhost tracing]# echo 0 > ../kprobes/enabled
[root@localhost tracing]# cat ../kprobes/list
ffffffff810bc1c5  k  do_fork+0x5
[root@localhost tracing]# echo 1 > ../kprobes/enabled
[root@localhost tracing]# cat ../kprobes/list
ffffffff810bc1c5  k  do_fork+0x5    [OPTIMIZED]
------

So, looks OK to me too ;)

Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Tested-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>

Ingo,
Could you also pull this patch to your -tip tree?

Thank you,

> 
> Thank you.
> 
>>
>>> Signed-off-by: Wang Nan <wangnan0@huawei.com>
>>> ---
>>>  kernel/kprobes.c | 7 ++++++-
>>>  1 file changed, 6 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/kernel/kprobes.c b/kernel/kprobes.c
>>> index bad4e95..b185464 100644
>>> --- a/kernel/kprobes.c
>>> +++ b/kernel/kprobes.c
>>> @@ -2320,6 +2320,12 @@ static void arm_all_kprobes(void)
>>>  	if (!kprobes_all_disarmed)
>>>  		goto already_enabled;
>>>  
>>> +	/*
>>> +	 * optimize_kprobe() called by arm_kprobe() checks
>>> +	 * kprobes_all_disarmed, so set kprobes_all_disarmed before
>>> +	 * arm_kprobe.
>>> +	 */
>>> +	kprobes_all_disarmed = false;
>>>  	/* Arming kprobes doesn't optimize kprobe itself */
>>>  	for (i = 0; i < KPROBE_TABLE_SIZE; i++) {
>>>  		head = &kprobe_table[i];
>>> @@ -2328,7 +2334,6 @@ static void arm_all_kprobes(void)
>>>  				arm_kprobe(p);
>>>  	}
>>>  
>>> -	kprobes_all_disarmed = false;
>>>  	printk(KERN_INFO "Kprobes globally enabled\n");
>>>  
>>>  already_enabled:
>>>
>>
>>
> 
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>
Ingo Molnar Feb. 19, 2015, 12:10 a.m. UTC | #4
* Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> wrote:

> Ingo,
> Could you also pull this patch to your -tip tree?

Yeah, as long as you use Signed-off-by instead of Acked-by, 
which is really how patches written by others should be 
routed.

Please re-send any still pending patches that are not yet 
upstream.

Thanks,

	Ingo
diff mbox

Patch

diff --git a/kernel/kprobes.c b/kernel/kprobes.c
index bad4e95..b185464 100644
--- a/kernel/kprobes.c
+++ b/kernel/kprobes.c
@@ -2320,6 +2320,12 @@  static void arm_all_kprobes(void)
 	if (!kprobes_all_disarmed)
 		goto already_enabled;
 
+	/*
+	 * optimize_kprobe() called by arm_kprobe() checks
+	 * kprobes_all_disarmed, so set kprobes_all_disarmed before
+	 * arm_kprobe.
+	 */
+	kprobes_all_disarmed = false;
 	/* Arming kprobes doesn't optimize kprobe itself */
 	for (i = 0; i < KPROBE_TABLE_SIZE; i++) {
 		head = &kprobe_table[i];
@@ -2328,7 +2334,6 @@  static void arm_all_kprobes(void)
 				arm_kprobe(p);
 	}
 
-	kprobes_all_disarmed = false;
 	printk(KERN_INFO "Kprobes globally enabled\n");
 
 already_enabled: