[patch 1/2] sched: Protect task->cpus_allowed access in sched_getaffinity()

[patch 1/2] sched: Protect task->cpus_allowed access in sched_getaffinity()

[Thomas Gleixner]

[-- Attachment #1: sched-fix-sched-get-affinity-racy-access.patch --]
[-- Type: text/plain, Size: 925 bytes --]

sched_getaffinity() is not protected against a concurrent modification
of the tasks affinity.

Serialize the access with task_rq_lock(task).

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 kernel/sched.c |    4 ++++
 1 file changed, 4 insertions(+)

Index: linux-2.6-tip/kernel/sched.c
===================================================================
--- linux-2.6-tip.orig/kernel/sched.c
+++ linux-2.6-tip/kernel/sched.c
@@ -6631,6 +6631,8 @@ SYSCALL_DEFINE3(sched_setaffinity, pid_t
 long sched_getaffinity(pid_t pid, struct cpumask *mask)
 {
 	struct task_struct *p;
+	unsigned long flags;
+	struct rq *rq;
 	int retval;
 
 	get_online_cpus();
@@ -6645,7 +6647,9 @@ long sched_getaffinity(pid_t pid, struct
 	if (retval)
 		goto out_unlock;
 
+	rq = task_rq_lock(p, &flags);
 	cpumask_and(mask, &p->cpus_allowed, cpu_online_mask);
+	task_rq_unlock(rq, &flags);
 
 out_unlock:
 	read_unlock(&tasklist_lock);

Re: [patch 1/2] sched: Protect task->cpus_allowed access in sched_getaffinity()

[Peter Zijlstra]

On Tue, 2009-12-08 at 20:24 +0000, Thomas Gleixner wrote:
> plain text document attachment
> (sched-fix-sched-get-affinity-racy-access.patch)
> sched_getaffinity() is not protected against a concurrent modification
> of the tasks affinity.
> 
> Serialize the access with task_rq_lock(task).
> 
> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
> ---
>  kernel/sched.c |    4 ++++
>  1 file changed, 4 insertions(+)
> 
> Index: linux-2.6-tip/kernel/sched.c
> ===================================================================
> --- linux-2.6-tip.orig/kernel/sched.c
> +++ linux-2.6-tip/kernel/sched.c
> @@ -6631,6 +6631,8 @@ SYSCALL_DEFINE3(sched_setaffinity, pid_t
>  long sched_getaffinity(pid_t pid, struct cpumask *mask)
>  {
>  	struct task_struct *p;
> +	unsigned long flags;
> +	struct rq *rq;
>  	int retval;
>  
>  	get_online_cpus();
> @@ -6645,7 +6647,9 @@ long sched_getaffinity(pid_t pid, struct
>  	if (retval)
>  		goto out_unlock;
>  
> +	rq = task_rq_lock(p, &flags);
>  	cpumask_and(mask, &p->cpus_allowed, cpu_online_mask);
> +	task_rq_unlock(rq, &flags);
>  
>  out_unlock:
>  	read_unlock(&tasklist_lock);
> 

Hehe, I wrote the exact same patch last night :-)

Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>

[tip:sched/urgent] sched: Protect task->cpus_allowed access in sched_getaffinity()

[tip-bot for Thomas Gleixner]

Commit-ID:  3160568371da441b7f2fb57f2f1225404207e8f2
Gitweb:     http://git.kernel.org/tip/3160568371da441b7f2fb57f2f1225404207e8f2
Author:     Thomas Gleixner <tglx@linutronix.de>
AuthorDate: Tue, 8 Dec 2009 20:24:16 +0000
Committer:  Ingo Molnar <mingo@elte.hu>
CommitDate: Wed, 9 Dec 2009 10:01:06 +0100

sched: Protect task->cpus_allowed access in sched_getaffinity()

sched_getaffinity() is not protected against a concurrent
modification of the tasks affinity.

Serialize the access with task_rq_lock(task).

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra <peterz@infradead.org>
LKML-Reference: <20091208202026.769251187@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 kernel/sched.c |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)

diff --git a/kernel/sched.c b/kernel/sched.c
index 281da29..c4635f7 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -6631,6 +6631,8 @@ SYSCALL_DEFINE3(sched_setaffinity, pid_t, pid, unsigned int, len,
 long sched_getaffinity(pid_t pid, struct cpumask *mask)
 {
 	struct task_struct *p;
+	unsigned long flags;
+	struct rq *rq;
 	int retval;
 
 	get_online_cpus();
@@ -6645,7 +6647,9 @@ long sched_getaffinity(pid_t pid, struct cpumask *mask)
 	if (retval)
 		goto out_unlock;
 
+	rq = task_rq_lock(p, &flags);
 	cpumask_and(mask, &p->cpus_allowed, cpu_online_mask);
+	task_rq_unlock(rq, &flags);
 
 out_unlock:
 	read_unlock(&tasklist_lock);