> If an m5 simulation terminates due to a simulation limit being reached, > is it then possible that buffered I/O will not be sent out to the > terminal? I noticed that output appears to buffered for quite a while > until I see it hit the output terminal. Is there a simple way to flush > the output to terminal more often or at least before teh simulation > terminates. I believe that it should be flushed before the simulator exits since it is exiting cleanly. Currently, python and C++ are not properly synchronized, so there can be a strange interleaving between them (fix in my tree). In python, you can do sys.stdout.flush. In C++, you can do cout.flush() or cout << flush. endl also causes a flush, and it is what \n translates to in a cprintf.
> The root of this question is that I am trying to add the ex_tpcb > database benchmark and in some cases I don't even see the unconditional > printf at the beginning of main(), while at other times the entire > benchmark works and I see all the printfs output to the terminal. What unconditional printf? A print in python main()? Is it something you've inserted? Nate _______________________________________________ m5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
