Hi all, I am curious about the way *m_stall_time* stat is calculated in MessageBuffer. The *m_stall_time* is declared as statistics::Average, and MessageBuffer set the *m_stall_time* with the duration from enqueue to dequeue. At first, I thought this works like below: if three messages passed the MessageBuffer, and each stall time is 10, 15, 20 respectively, the average stall time per message is (10+15+20)/3.
However, statistics::Average uses AvgStor as storage and this works as per-tick average stat. I think AvgStor is well fit to *m_buf_msgs*, which stands for average number of messages in buffer, and it is also declared as statistics::Average. But for average stall time, isn't the calcuation of (sum of scalars) / (stall message count) right? For example, Garnet calculates queueing latency as m_avg_packet_vqueue_latency = m_packet_queueing_latency / m_packets_received. Thanks, Daecheol.
_______________________________________________ gem5-users mailing list -- gem5-users@gem5.org To unsubscribe send an email to gem5-users-le...@gem5.org %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
