[PATCH] bpf/perf: Fix suspicious RCU usage in get_callchain_entry()

[Qing Wang]

There is a patch intended to fix suspicious RCU usage in get_callchain_entry(),
but it is incorrect. Specifically, rcu_read_lock()/rcu_read_unlock() is not
called when may_fault == false.

Previous discussion:
https://lore.kernel.org/all/CAEf4BzaYL9zZN8TZyRHW3_O3vbHc7On+NSunrkDvDQx2=wwyRw@xxxxxxxxxxxxxx/#R

For perf's callchain, rcu_read_lock()/rcu_read_unlock() should be called when
trace_in == false.

Fixes: d4dd9775ec24 ("bpf: wire up sleepable bpf_get_stack() and bpf_get_task_stack() helpers")
Reported-by: syzbot+72a43cdb78469f7fbad1@xxxxxxxxxxxxxxxxxxxxxxxxx
Closes: https://syzkaller.appspot.com/bug?extid=72a43cdb78469f7fbad1
Tested-by: syzbot+72a43cdb78469f7fbad1@xxxxxxxxxxxxxxxxxxxxxxxxx
Signed-off-by: Qing Wang <wangqing7171@xxxxxxxxx>
---
 kernel/bpf/stackmap.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/kernel/bpf/stackmap.c b/kernel/bpf/stackmap.c
index da3d328f5c15..f97d4aa9d038 100644
--- a/kernel/bpf/stackmap.c
+++ b/kernel/bpf/stackmap.c
@@ -460,7 +460,7 @@ static long __bpf_get_stack(struct pt_regs *regs, struct task_struct *task,
 
 	max_depth = stack_map_calculate_max_depth(size, elem_size, flags);
 
-	if (may_fault)
+	if (!trace_in)
 		rcu_read_lock(); /* need RCU for perf's callchain below */
 
 	if (trace_in) {
@@ -474,7 +474,7 @@ static long __bpf_get_stack(struct pt_regs *regs, struct task_struct *task,
 	}
 
 	if (unlikely(!trace) || trace->nr < skip) {
-		if (may_fault)
+		if (!trace_in)
 			rcu_read_unlock();
 		goto err_fault;
 	}
@@ -494,7 +494,7 @@ static long __bpf_get_stack(struct pt_regs *regs, struct task_struct *task,
 	}
 
 	/* trace/ips should not be dereferenced after this point */
-	if (may_fault)
+	if (!trace_in)
 		rcu_read_unlock();
 
 	if (user_build_id)
-- 
2.34.1