Hi all, I am having trouble with gem5 SE mode running a simple memory copy microbenchmark (https://github.com/BTone/cagbench) that uses a configurable number of worker threads that all wait to start and finish at a barrier.
I have found that using a pthread_barrier is not working, as once all worker threads reach the barrier, only a single thread leaves it while the rest wait. I have tried implementing the barrier using only a mutex and the same problem exists. I also tried implementing the barrier using only a semaphore, but all threads get stuck waiting at the second barrier (once all worker threads perform their respective memory copy operations) and none proceed past it. I know my implementation work correctly, because the microbenchmark runs perfectly fine with all variations of the pthread_barrier I implemented. Any help or advice would be greatly appreciated. Thank you.
_______________________________________________ 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
