Which ISA are you using? I'm assuming Alpha. Which m5.* did you use? If it was m5.fast, try running again with m5.opt. m5.fast compiles out asserts and that sort of thing, so you may get more information about what the problem is that way. You can run m5 under gdb and see where the segfault is occuring, and since it's not that far in I'd suggest m5.debug to get a better backtrace and fewer things optimized out.
Gabe Ying Zhang wrote: > Hi all, > > I am currently trying to simulate a small network consisting of 5 > ~6 computer systems. The "fs.py" gives an example to build a basic > server-client system with 2 computers. As a start point, I tried to > follow its way to build a network with a server and 2 clients, so I > did the following: > > adding three NICs to the test system: > ethernet0 = NSGigE(pci_bus=0, pci_dev=1, pci_func=0) > ethernet1 = NSGigE(pci_bus=0, pci_dev=2, pci_func=0) > > modifying the "makeDualRoot" function to make it containing two > clients, namely "drivesys0" and "drivesys1" > in the "Benchmark.py", I add an option for this system, e.g., > "Surge_multiclient = [Sysconfig(surver), Sysconfig(client), > Sysconfig(client)]". > > When running with "--benchmark=Surge_multiclient", the simulation > will exit with a segmentation fault as soon as it enters the "REAL > SIMULATION" mode. Since I am a newbie to M5 simulator, I have no idea > on how to fix it. I am not sure if I did this correctly, so could > anyone kindly give me some suggestions on this issue? I mean building > a small network with more than 2 computers. > > Another question is that how I can simulate a ring-like network, > where each computer can communicate with others without a coordinator > or switch. > > Thanks very much, > Ying Zhang > > ------------------------------------------------------------------------ > > _______________________________________________ > m5-users mailing list > [email protected] > http://m5sim.org/cgi-bin/mailman/listinfo/m5-users _______________________________________________ m5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
