> I'm running an AtomicSimpleCPU using ALPHA_SE. Usually, at the end of > simulation I get the standard message: > > Exiting @ cycle 1297500 because target called exit() > > But, for one of my programs (gcc), I don't get this at all. The > simulation just stops when the program has finished. The stats and > config files are written just fine and the simulator exits normally, but > it doesn't print that message. It seems that the simulate() function > never returns so the simulator must be exiting elsewhere. I can't figure > out where though. > > Could anyone shed some light on why this might be happening or where I > should look to try to find out what's wrong with this? Is there some way > that the simulation could finish without control passing to the user's > python script? I don't think so. Not without some code calling exit() somewhere (or causing some sort of fault).
I think your best bet is to fire up a debugger and try to set some breakpoints. One of the last things that happens is the dumping of stats, so since that is happening, set a breakpoint there and see where it is getting called from. Alternatively, you can look in the stats file and figure out how many cycles are run, and then turn on all trace flags one cycle before the exit and see what comes out. Unfortunately, this may not help a ton since the exit process doesn't really have many DPRINTF statements. Nate _______________________________________________ m5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
