>Indeed, good catch! Have you observed this, or is this patch due to code inspection?
I observed this. I was using perf_event_open instead of linux tools perf to monitor a process continuously forking new processes, generating a lot of mmap records. And a sample record happens while generating a mmap record. For missing barriers, I thought local_xxx provided implied barriers. But I found atomic_ops.txt said not. Thanks for reminding! I will add them in path v2. I also think it's a good idea to use READ_ONCE/WRITE_ONCE for rb->nest. Will use it in patch v2.
