Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1633657

[tip:smp/hotplug] perf: Reorder cpu hotplug rwsem against cred_guard_mutex

Show all headers | View raw

Commit-ID:  1526eee294dd52b70804aa377579682cc4dcd9ad
Gitweb:     http://git.kernel.org/tip/1526eee294dd52b70804aa377579682cc4dcd9ad
Author:     Thomas Gleixner <tglx@linutronix.de>
AuthorDate: Mon, 1 May 2017 14:35:45 +0200
Committer:  Thomas Gleixner <tglx@linutronix.de>
CommitDate: Mon, 1 May 2017 14:54:40 +0200

perf: Reorder cpu hotplug rwsem against cred_guard_mutex

sys_perf_event_open() takes the hotplug rwsem before the
cred_guard_mutex. The exit() path has the reverse lock order.

The hotplug protection in sys_perf_event_open() is not required before
taking the cred_guard_mutex, so it can be reordered there.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Sebastian Siewior <bigeasy@linutronix.de>
Link: http://lkml.kernel.org/r/20170428142456.5xh44ef3fv7w2kkh@linutronix.de
---
 kernel/events/core.c | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

diff --git a/kernel/events/core.c b/kernel/events/core.c
index 997123c..71d8c74 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -9742,12 +9742,10 @@ SYSCALL_DEFINE5(perf_event_open,
 		goto err_task;
 	}
 
-	get_online_cpus();
-
 	if (task) {
 		err = mutex_lock_interruptible(&task->signal->cred_guard_mutex);
 		if (err)
-			goto err_cpus;
+			goto err_task;
 
 		/*
 		 * Reuse ptrace permission checks for now.
@@ -9765,11 +9763,13 @@ SYSCALL_DEFINE5(perf_event_open,
 	if (flags & PERF_FLAG_PID_CGROUP)
 		cgroup_fd = pid;
 
+	get_online_cpus();
+
 	event = perf_event_alloc(&attr, cpu, task, group_leader, NULL,
 				 NULL, NULL, cgroup_fd);
 	if (IS_ERR(event)) {
 		err = PTR_ERR(event);
-		goto err_cred;
+		goto err_cpus;
 	}
 
 	if (is_sampling_event(event)) {
@@ -10017,13 +10017,13 @@ SYSCALL_DEFINE5(perf_event_open,
 		perf_event_ctx_unlock(group_leader, gctx);
 	mutex_unlock(&ctx->mutex);
 
+	put_online_cpus();
+
 	if (task) {
 		mutex_unlock(&task->signal->cred_guard_mutex);
 		put_task_struct(task);
 	}
 
-	put_online_cpus();
-
 	mutex_lock(&current->perf_event_mutex);
 	list_add_tail(&event->owner_entry, &current->perf_event_list);
 	mutex_unlock(&current->perf_event_mutex);
@@ -10054,11 +10054,11 @@ err_alloc:
 	 */
 	if (!event_file)
 		free_event(event);
+err_cpus:
+	put_online_cpus();
 err_cred:
 	if (task)
 		mutex_unlock(&task->signal->cred_guard_mutex);
-err_cpus:
-	put_online_cpus();
 err_task:
 	if (task)
 		put_task_struct(task);

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[RFC PATCH] trace/perf: cure locking issue in perf_event_open()  error path Sebastian Siewior <bigeasy@linutronix.de> - 2017-04-28 16:30 +0200
  Re: [RFC PATCH] trace/perf: cure locking issue in perf_event_open()  error path Sebastian Siewior <bigeasy@linutronix.de> - 2017-04-28 16:40 +0200
  [tip:smp/hotplug] perf: Reorder cpu hotplug rwsem against  cred_guard_mutex tip-bot for Thomas Gleixner <tipbot@zytor.com> - 2017-05-01 15:10 +0200
  [tip:smp/hotplug] perf: Push hotplug protection down to callers tip-bot for Thomas Gleixner <tipbot@zytor.com> - 2017-05-01 15:10 +0200

csiph-web