Hello everyone: Im working with an detailed X86 Ruby MESI_CMP_Directory architecture and I did some changes to the *.sm files in order to implement two private and one third shared level of cache memory. So far I was able to debug all the "invalid transition" and "possible deadlock" errors and reviewed the source code several times to discard other bugs. However, when I try to run some PARSEC benchmarks the simulation get stuck after one hour and I see no progress or errors regardless of how much time I let it run (even days). The normal simulation time should be few hours.
The command I'm using is: ./build/X86/gem5.opt -d simout/ff_Blackscholes/ ./configs/example/ruby_fs.py -n 16 --kernel=x86_64-vmlinux-2.6.28.4-smp --script=runscript/blackscholes.rcS -F 437696510 --cpu-type=detailed --num-cpus=16 --num-dirs=16 --topology=Mesh --num-l3caches=16 --mesh-rows=4 --garnet-network=fixed These are the last lines that I got from the "m5out/system.pc.com_1.terminal" file hda: M5 IDE Disk, ATA DISK drive hdb: M5 IDE Disk, ATA DISK drive hda: UDMA/33 mode selected hdb: UDMA/33 mode selected ide0 at 0x1f0-0x1f7,0x3f6 on irq 14 isa bounce pool size: 16 pages ide1 at 0x170-0x177,0x376 on irq 15 ide_generic: please use "probe_mask=0x3f" module parameter for probing all legacy ISA IDE ports ide-gd driver 1.18 hda: max request size: 128KiB hda: 16514064 sectors (8455 MB), CHS=16383/16/63 hda: hda1 Since the simulation doesn't display any errors I have no idea how the changes I did affect the booting of the system? Any suggestions about how to get started with the debuging of this issue? Thank you so much -- Alberto Javier Naranjo-Carmona M.S. Student Computer Engineering _______________________________________________ gem5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
