linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] smpboot: add missing get_online_cpus() when register
@ 2014-07-31  3:30 Lai Jiangshan
  2014-08-01 21:54 ` David Rientjes
  2015-01-23 10:36 ` [tip:core/urgent] smpboot: Add missing get_online_cpus() in smpboot_register_percpu_thread() tip-bot for Lai Jiangshan
  0 siblings, 2 replies; 4+ messages in thread
From: Lai Jiangshan @ 2014-07-31  3:30 UTC (permalink / raw)
  To: linux-kernel
  Cc: Lai Jiangshan, Thomas Gleixner, Rusty Russell, Peter Zijlstra,
	Srivatsa S. Bhat, stable

If the smpboot_register_percpu_thread() is called after smpboot_create_threads()
but before __cpu_up(), the smpboot thread of the online-ing CPU is not created,
and it results a bug.  So we use get_online_cpus() to prevent it.

smpboot_unregister_percpu_thread() travels all possible CPU, it doesn't need
get_online_cpus() which is removed in the patch.

CC: Thomas Gleixner <tglx@linutronix.de>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
CC: stable@kernel.org
Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
---
 kernel/smpboot.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/kernel/smpboot.c b/kernel/smpboot.c
index eb89e18..8adab87 100644
--- a/kernel/smpboot.c
+++ b/kernel/smpboot.c
@@ -279,6 +279,7 @@ int smpboot_register_percpu_thread(struct smp_hotplug_thread *plug_thread)
 	unsigned int cpu;
 	int ret = 0;
 
+	get_online_cpus();
 	mutex_lock(&smpboot_threads_lock);
 	for_each_online_cpu(cpu) {
 		ret = __smpboot_create_thread(plug_thread, cpu);
@@ -291,6 +292,7 @@ int smpboot_register_percpu_thread(struct smp_hotplug_thread *plug_thread)
 	list_add(&plug_thread->list, &hotplug_threads);
 out:
 	mutex_unlock(&smpboot_threads_lock);
+	put_online_cpus();
 	return ret;
 }
 EXPORT_SYMBOL_GPL(smpboot_register_percpu_thread);
@@ -303,11 +305,9 @@ EXPORT_SYMBOL_GPL(smpboot_register_percpu_thread);
  */
 void smpboot_unregister_percpu_thread(struct smp_hotplug_thread *plug_thread)
 {
-	get_online_cpus();
 	mutex_lock(&smpboot_threads_lock);
 	list_del(&plug_thread->list);
 	smpboot_destroy_threads(plug_thread);
 	mutex_unlock(&smpboot_threads_lock);
-	put_online_cpus();
 }
 EXPORT_SYMBOL_GPL(smpboot_unregister_percpu_thread);
-- 
1.7.4.4


^ permalink raw reply related	[flat|nested] 4+ messages in thread

* Re: [PATCH] smpboot: add missing get_online_cpus() when register
  2014-07-31  3:30 [PATCH] smpboot: add missing get_online_cpus() when register Lai Jiangshan
@ 2014-08-01 21:54 ` David Rientjes
  2014-08-04  7:33   ` Lai Jiangshan
  2015-01-23 10:36 ` [tip:core/urgent] smpboot: Add missing get_online_cpus() in smpboot_register_percpu_thread() tip-bot for Lai Jiangshan
  1 sibling, 1 reply; 4+ messages in thread
From: David Rientjes @ 2014-08-01 21:54 UTC (permalink / raw)
  To: Lai Jiangshan
  Cc: linux-kernel, Thomas Gleixner, Rusty Russell, Peter Zijlstra,
	Srivatsa S. Bhat, stable

On Thu, 31 Jul 2014, Lai Jiangshan wrote:

> If the smpboot_register_percpu_thread() is called after smpboot_create_threads()
> but before __cpu_up(), the smpboot thread of the online-ing CPU is not created,
> and it results a bug.  So we use get_online_cpus() to prevent it.
> 

Do you have an example of the bug to include?  Maintainers are going to 
need to understand the implications of the problem before the 
stable@kernel.org annotation is warranted.

> smpboot_unregister_percpu_thread() travels all possible CPU, it doesn't need
> get_online_cpus() which is removed in the patch.
> 
> CC: Thomas Gleixner <tglx@linutronix.de>
> Cc: Rusty Russell <rusty@rustcorp.com.au>
> Cc: Peter Zijlstra <peterz@infradead.org>
> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
> CC: stable@kernel.org
> Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
> ---
>  kernel/smpboot.c |    4 ++--
>  1 files changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/kernel/smpboot.c b/kernel/smpboot.c
> index eb89e18..8adab87 100644
> --- a/kernel/smpboot.c
> +++ b/kernel/smpboot.c
> @@ -279,6 +279,7 @@ int smpboot_register_percpu_thread(struct smp_hotplug_thread *plug_thread)
>  	unsigned int cpu;
>  	int ret = 0;
>  
> +	get_online_cpus();
>  	mutex_lock(&smpboot_threads_lock);
>  	for_each_online_cpu(cpu) {
>  		ret = __smpboot_create_thread(plug_thread, cpu);
> @@ -291,6 +292,7 @@ int smpboot_register_percpu_thread(struct smp_hotplug_thread *plug_thread)
>  	list_add(&plug_thread->list, &hotplug_threads);
>  out:
>  	mutex_unlock(&smpboot_threads_lock);
> +	put_online_cpus();
>  	return ret;
>  }

I think the {get,put}_online_cpus() pair should be nested inside the 
smpboot_threads_lock for better lock ordering since not all cases 
smpboot_threads_lock will require it.

That way, you can also do put_online_cpus() before 
smpboot_destroy_threads(), which you have already proven doesn't need it:

@@ -280,14 +280,17 @@ int smpboot_register_percpu_thread(struct smp_hotplug_thread *plug_thread)
 	int ret = 0;
 
 	mutex_lock(&smpboot_threads_lock);
+	get_online_cpus();
 	for_each_online_cpu(cpu) {
 		ret = __smpboot_create_thread(plug_thread, cpu);
 		if (ret) {
+			put_online_cpus();
 			smpboot_destroy_threads(plug_thread);
 			goto out;
 		}
 		smpboot_unpark_thread(plug_thread, cpu);
 	}
+	put_online_cpus();
 	list_add(&plug_thread->list, &hotplug_threads);
 out:
 	mutex_unlock(&smpboot_threads_lock);

>  EXPORT_SYMBOL_GPL(smpboot_register_percpu_thread);
> @@ -303,11 +305,9 @@ EXPORT_SYMBOL_GPL(smpboot_register_percpu_thread);
>   */
>  void smpboot_unregister_percpu_thread(struct smp_hotplug_thread *plug_thread)
>  {
> -	get_online_cpus();
>  	mutex_lock(&smpboot_threads_lock);
>  	list_del(&plug_thread->list);
>  	smpboot_destroy_threads(plug_thread);
>  	mutex_unlock(&smpboot_threads_lock);
> -	put_online_cpus();
>  }
>  EXPORT_SYMBOL_GPL(smpboot_unregister_percpu_thread);

This makes sense.

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH] smpboot: add missing get_online_cpus() when register
  2014-08-01 21:54 ` David Rientjes
@ 2014-08-04  7:33   ` Lai Jiangshan
  0 siblings, 0 replies; 4+ messages in thread
From: Lai Jiangshan @ 2014-08-04  7:33 UTC (permalink / raw)
  To: David Rientjes
  Cc: linux-kernel, Thomas Gleixner, Rusty Russell, Peter Zijlstra,
	Srivatsa S. Bhat, stable

On 08/02/2014 05:54 AM, David Rientjes wrote:
> On Thu, 31 Jul 2014, Lai Jiangshan wrote:
> 
>> If the smpboot_register_percpu_thread() is called after smpboot_create_threads()
>> but before __cpu_up(), the smpboot thread of the online-ing CPU is not created,
>> and it results a bug.  So we use get_online_cpus() to prevent it.
>>
> 
> Do you have an example of the bug to include?  

Sorry, no, I don't have.

> Maintainers are going to 
> need to understand the implications of the problem before the 
> stable@kernel.org annotation is warranted.

It is possible that smpboot_register_percpu_thread() can be called
any time in current kernel.  Repeating the module ehca and check while
repeating online/offline the CPUs, the bug is possible to hit.  I have not such
devices to test.

Let Thomas make the choice.

> 
>> smpboot_unregister_percpu_thread() travels all possible CPU, it doesn't need
>> get_online_cpus() which is removed in the patch.
>>
>> CC: Thomas Gleixner <tglx@linutronix.de>
>> Cc: Rusty Russell <rusty@rustcorp.com.au>
>> Cc: Peter Zijlstra <peterz@infradead.org>
>> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
>> CC: stable@kernel.org
>> Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
>> ---
>>  kernel/smpboot.c |    4 ++--
>>  1 files changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git a/kernel/smpboot.c b/kernel/smpboot.c
>> index eb89e18..8adab87 100644
>> --- a/kernel/smpboot.c
>> +++ b/kernel/smpboot.c
>> @@ -279,6 +279,7 @@ int smpboot_register_percpu_thread(struct smp_hotplug_thread *plug_thread)
>>  	unsigned int cpu;
>>  	int ret = 0;
>>  
>> +	get_online_cpus();
>>  	mutex_lock(&smpboot_threads_lock);
>>  	for_each_online_cpu(cpu) {
>>  		ret = __smpboot_create_thread(plug_thread, cpu);
>> @@ -291,6 +292,7 @@ int smpboot_register_percpu_thread(struct smp_hotplug_thread *plug_thread)
>>  	list_add(&plug_thread->list, &hotplug_threads);
>>  out:
>>  	mutex_unlock(&smpboot_threads_lock);
>> +	put_online_cpus();
>>  	return ret;
>>  }
> 
> I think the {get,put}_online_cpus() pair should be nested inside the 
> smpboot_threads_lock for better lock ordering since not all cases 
> smpboot_threads_lock will require it.
> 
> That way, you can also do put_online_cpus() before 
> smpboot_destroy_threads(), which you have already proven doesn't need it:
> 
> @@ -280,14 +280,17 @@ int smpboot_register_percpu_thread(struct smp_hotplug_thread *plug_thread)
>  	int ret = 0;
>  
>  	mutex_lock(&smpboot_threads_lock);
> +	get_online_cpus();

get_online_cpus() can't be nested in smpboot_threads_lock.


>  	for_each_online_cpu(cpu) {
>  		ret = __smpboot_create_thread(plug_thread, cpu);
>  		if (ret) {
> +			put_online_cpus();
>  			smpboot_destroy_threads(plug_thread);
>  			goto out;
>  		}
>  		smpboot_unpark_thread(plug_thread, cpu);
>  	}
> +	put_online_cpus();
>  	list_add(&plug_thread->list, &hotplug_threads);
>  out:
>  	mutex_unlock(&smpboot_threads_lock);
> 
>>  EXPORT_SYMBOL_GPL(smpboot_register_percpu_thread);
>> @@ -303,11 +305,9 @@ EXPORT_SYMBOL_GPL(smpboot_register_percpu_thread);
>>   */
>>  void smpboot_unregister_percpu_thread(struct smp_hotplug_thread *plug_thread)
>>  {
>> -	get_online_cpus();
>>  	mutex_lock(&smpboot_threads_lock);
>>  	list_del(&plug_thread->list);
>>  	smpboot_destroy_threads(plug_thread);
>>  	mutex_unlock(&smpboot_threads_lock);
>> -	put_online_cpus();
>>  }
>>  EXPORT_SYMBOL_GPL(smpboot_unregister_percpu_thread);
> 
> This makes sense.
> .
> 


^ permalink raw reply	[flat|nested] 4+ messages in thread

* [tip:core/urgent] smpboot: Add missing get_online_cpus() in smpboot_register_percpu_thread()
  2014-07-31  3:30 [PATCH] smpboot: add missing get_online_cpus() when register Lai Jiangshan
  2014-08-01 21:54 ` David Rientjes
@ 2015-01-23 10:36 ` tip-bot for Lai Jiangshan
  1 sibling, 0 replies; 4+ messages in thread
From: tip-bot for Lai Jiangshan @ 2015-01-23 10:36 UTC (permalink / raw)
  To: linux-tip-commits
  Cc: srivatsa.bhat, rusty, laijs, tglx, linux-kernel, hpa, mingo,
	peterz, rientjes

Commit-ID:  4bee96860a65c3a62d332edac331b3cf936ba3ad
Gitweb:     http://git.kernel.org/tip/4bee96860a65c3a62d332edac331b3cf936ba3ad
Author:     Lai Jiangshan <laijs@cn.fujitsu.com>
AuthorDate: Thu, 31 Jul 2014 11:30:17 +0800
Committer:  Thomas Gleixner <tglx@linutronix.de>
CommitDate: Fri, 23 Jan 2015 11:33:51 +0100

smpboot: Add missing get_online_cpus() in smpboot_register_percpu_thread()

The following race exists in the smpboot percpu threads management:

CPU0	      	   	     CPU1
cpu_up(2)
  get_online_cpus();
  smpboot_create_threads(2);
			     smpboot_register_percpu_thread();
			     for_each_online_cpu();
			       __smpboot_create_thread();
  __cpu_up(2);

This results in a missing per cpu thread for the newly onlined cpu2 and
in a NULL pointer dereference on a consecutive offline of that cpu.

Proctect smpboot_register_percpu_thread() with get_online_cpus() to
prevent that.

[ tglx: Massaged changelog and removed the change in
        smpboot_unregister_percpu_thread() because that's an
        optimization and therefor not stable material. ]

Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Cc: David Rientjes <rientjes@google.com>
Cc: stable@vger.kernel.org
Link: http://lkml.kernel.org/r/1406777421-12830-1-git-send-email-laijs@cn.fujitsu.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 kernel/smpboot.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/kernel/smpboot.c b/kernel/smpboot.c
index f032fb5..40190f2 100644
--- a/kernel/smpboot.c
+++ b/kernel/smpboot.c
@@ -280,6 +280,7 @@ int smpboot_register_percpu_thread(struct smp_hotplug_thread *plug_thread)
 	unsigned int cpu;
 	int ret = 0;
 
+	get_online_cpus();
 	mutex_lock(&smpboot_threads_lock);
 	for_each_online_cpu(cpu) {
 		ret = __smpboot_create_thread(plug_thread, cpu);
@@ -292,6 +293,7 @@ int smpboot_register_percpu_thread(struct smp_hotplug_thread *plug_thread)
 	list_add(&plug_thread->list, &hotplug_threads);
 out:
 	mutex_unlock(&smpboot_threads_lock);
+	put_online_cpus();
 	return ret;
 }
 EXPORT_SYMBOL_GPL(smpboot_register_percpu_thread);

^ permalink raw reply related	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2015-01-23 10:37 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-07-31  3:30 [PATCH] smpboot: add missing get_online_cpus() when register Lai Jiangshan
2014-08-01 21:54 ` David Rientjes
2014-08-04  7:33   ` Lai Jiangshan
2015-01-23 10:36 ` [tip:core/urgent] smpboot: Add missing get_online_cpus() in smpboot_register_percpu_thread() tip-bot for Lai Jiangshan

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).