[PATCH 3.2 114/126] perf/core: Fix concurrent sys_perf_event_open() vs. 'move_group' race

[Ben Hutchings]

3.2.85-rc1 review patch.  If anyone has any objections, please let me know.

------------------

From: Peter Zijlstra <peterz@xxxxxxxxxxxxx>

commit 321027c1fe77f892f4ea07846aeae08cefbbb290 upstream.

Di Shen reported a race between two concurrent sys_perf_event_open()
calls where both try and move the same pre-existing software group
into a hardware context.

The problem is exactly that described in commit:

  f63a8daa5812 ("perf: Fix event->ctx locking")

... where, while we wait for a ctx->mutex acquisition, the event->ctx
relation can have changed under us.

That very same commit failed to recognise sys_perf_event_context() as an
external access vector to the events and thereby didn't apply the
established locking rules correctly.

So while one sys_perf_event_open() call is stuck waiting on
mutex_lock_double(), the other (which owns said locks) moves the group
about. So by the time the former sys_perf_event_open() acquires the
locks, the context we've acquired is stale (and possibly dead).

Apply the established locking rules as per perf_event_ctx_lock_nested()
to the mutex_lock_double() for the 'move_group' case. This obviously means
we need to validate state after we acquire the locks.

Reported-by: Di Shen (Keen Lab)
Tested-by: John Dias <joaodias@xxxxxxxxxx>
Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
Cc: Alexander Shishkin <alexander.shishkin@xxxxxxxxxxxxxxx>
Cc: Arnaldo Carvalho de Melo <acme@xxxxxxxxxx>
Cc: Arnaldo Carvalho de Melo <acme@xxxxxxxxxx>
Cc: Jiri Olsa <jolsa@xxxxxxxxxx>
Cc: Kees Cook <keescook@xxxxxxxxxxxx>
Cc: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
Cc: Min Chong <mchong@xxxxxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Cc: Stephane Eranian <eranian@xxxxxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Cc: Vince Weaver <vincent.weaver@xxxxxxxxx>
Fixes: f63a8daa5812 ("perf: Fix event->ctx locking")
Link: http://lkml.kernel.org/r/20170106131444.GZ3174@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Signed-off-by: Ingo Molnar <mingo@xxxxxxxxxx>
[bwh: Backported to 3.2:
 - Use ACCESS_ONCE() instead of READ_ONCE()
 - Test perf_event::group_flags instead of group_caps
 - Add the err_locked cleanup block, which we didn't need before
 - Adjust context]
Signed-off-by: Ben Hutchings <ben@xxxxxxxxxxxxxxx>
---
 kernel/events/core.c | 58 ++++++++++++++++++++++++++++++++++++++++++++++++----
 1 file changed, 54 insertions(+), 4 deletions(-)

--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -6474,6 +6474,37 @@ static void mutex_lock_double(struct mut
 	mutex_lock_nested(b, SINGLE_DEPTH_NESTING);
 }
 
+/*
+ * Variation on perf_event_ctx_lock_nested(), except we take two context
+ * mutexes.
+ */
+static struct perf_event_context *
+__perf_event_ctx_lock_double(struct perf_event *group_leader,
+			     struct perf_event_context *ctx)
+{
+	struct perf_event_context *gctx;
+
+again:
+	rcu_read_lock();
+	gctx = ACCESS_ONCE(group_leader->ctx);
+	if (!atomic_inc_not_zero(&gctx->refcount)) {
+		rcu_read_unlock();
+		goto again;
+	}
+	rcu_read_unlock();
+
+	mutex_lock_double(&gctx->mutex, &ctx->mutex);
+
+	if (group_leader->ctx != gctx) {
+		mutex_unlock(&ctx->mutex);
+		mutex_unlock(&gctx->mutex);
+		put_ctx(gctx);
+		goto again;
+	}
+
+	return gctx;
+}
+
 /**
  * sys_perf_event_open - open a performance event, associate it to a task/cpu
  *
@@ -6661,14 +6692,31 @@ SYSCALL_DEFINE5(perf_event_open,
 	}
 
 	if (move_group) {
-		gctx = group_leader->ctx;
+		gctx = __perf_event_ctx_lock_double(group_leader, ctx);
+
+		/*
+		 * Check if we raced against another sys_perf_event_open() call
+		 * moving the software group underneath us.
+		 */
+		if (!(group_leader->group_flags & PERF_GROUP_SOFTWARE)) {
+			/*
+			 * If someone moved the group out from under us, check
+			 * if this new event wound up on the same ctx, if so
+			 * its the regular !move_group case, otherwise fail.
+			 */
+			if (gctx != ctx) {
+				err = -EINVAL;
+				goto err_locked;
+			} else {
+				perf_event_ctx_unlock(group_leader, gctx);
+				move_group = 0;
+			}
+		}
 
 		/*
 		 * See perf_event_ctx_lock() for comments on the details
 		 * of swizzling perf_event::ctx.
 		 */
-		mutex_lock_double(&gctx->mutex, &ctx->mutex);
-
 		perf_remove_from_context(group_leader, false);
 
 		/*
@@ -6709,10 +6757,8 @@ SYSCALL_DEFINE5(perf_event_open,
 	++ctx->generation;
 	perf_unpin_context(ctx);
 
-	if (move_group) {
-		mutex_unlock(&gctx->mutex);
-		put_ctx(gctx);
-	}
+	if (move_group)
+		perf_event_ctx_unlock(group_leader, gctx);
 	mutex_unlock(&ctx->mutex);
 
 	event->owner = current;
@@ -6737,6 +6783,11 @@ SYSCALL_DEFINE5(perf_event_open,
 	fd_install(event_fd, event_file);
 	return event_fd;
 
+err_locked:
+	if (move_group)
+		perf_event_ctx_unlock(group_leader, gctx);
+	mutex_unlock(&ctx->mutex);
+	fput(event_file);
 err_context:
 	perf_unpin_context(ctx);
 	put_ctx(ctx);