Very strange... if it's not making progress, it must be livelocked, because otherwise you would hit the max_tick event and exit. You should be able to jump in with gdb, turn on a few trace flags, and see what's happening.
If any of these are benchmarks that used to complete, you could use 'hg bisect' to track down where things went wrong. Steve On Thu, Jul 4, 2013 at 6:14 AM, Andreas Sandberg <[email protected]>wrote: > Hi Everyone, > > Has anyone else noticed odd behaviour where gem5 stops making progress in > some applications? > > I'm currently experiencing this on an x86 system simulating an x86 system > in FS mode. I've seen this happening in some of the SPEC CPU2006 benchmarks > (450.soplex/ref, 454.calculix/ref, 445.gobmk/ref). I still haven't figured > out what's causing this to happen, but I know is that the simulator stops > committing instructions since I have events scheduled to happen every 10 > million instructions that stop happening after a while. There are other > benchmarks which I haven't gotten to terminate, but they seem to get stuck > because of other bugs (probably FP related things preventing proper > numerical convergence). > > //Anreas > ______________________________**_________________ > gem5-users mailing list > [email protected] > http://m5sim.org/cgi-bin/**mailman/listinfo/gem5-users<http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users> >
_______________________________________________ gem5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
