We notice that the throughput is actually near to iperf benchmark. We found out that when we remove O_SYNC during open /dev/mem , the throughput can be achieved ...
確定! 回上一頁