> I think you're wrong on point 2... max_insts_all_threads runs each > thread for *at least* the specified number, and does not terminate > individual threads early. There's no option right now that does > terminate individual threads without terminating the entire > simulation, largely because in my opinion that's kind of meaningless > (for the reasons you explain in point 2). Ah, right. This should probably be listed as an unimplemented option #3.
> I can take a crack at updating the FAQ later if you don't get to it. I'll do it right now. > Overall I think our FAQ is way too big and should be reorganized into > normal topical documentation, but this particular question isn't any > worse than any other. I agree. We should probably have a meeting among all of those who are interested to discuss how to improve the wiki before we have the tutorial. Nate _______________________________________________ m5-users mailing list m5-users@m5sim.org http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
