Hi, I've been using gem5 (m5 + ruby) for the last couple of months. I've been using sys call emulation mode. Up until yesterday, I had been using TimingSimpleCPU. Yesterday I made an attempt to transition to DerivO3CPU and found that the program (m5.debug) seg faults pretty quickly.
Changing line 1092 in "m5/src/cpu/o3/lsq_unit_impl. hh" to a static cast (from a dynamic cast) resolves the segfault but causes the simulator to get stuck, printing the following message "Exiting @ tick 9223372036854775807 because simulate() limit reached". Note that it actually runs "hello world" fine when I make the described change, but when I run a binary slightly more complex I run into this issue. Is this a known issue? Does anyone have any idea how to resolve this? I copied m5's output below. Thanks for your help! Marc Orr ----------------------------- details: changed line 140 of m5/configs/examples/ruby_se.py to "class > CPUClass(DerivO3CPU): pass" > build_opts: > TARGET_ISA = 'x86' > CPU_MODELS = 'AtomicSimpleCPU,TimingSimpleCPU,O3CPU' > FULL_SYSTEM = 0 > PROTOCOL = 'MI_example' > RUBY = True > This is the output with the dynamic cast (lsq_unit_impl.hh unmodified): > [morr@mumble-12] (17)$ build/X86_SE_RUBY/m5.debug > configs/example/ruby_se.py -c tests/test-progs/hello/bin/x86/linux/hello > M5 Simulator System > > Copyright (c) 2001-2008 > The Regents of The University of Michigan > All Rights Reserved > > > M5 compiled Apr 28 2011 15:51:43 > M5 started Apr 28 2011 15:58:13 > M5 executing on mumble-12.cs.wisc.edu > command line: build/X86_SE_RUBY/m5.debug configs/example/ruby_se.py -c > tests/test-progs/hello/bin/x86/linux/hello > Global frequency set at 1000000000000 ticks per second > 0: system.remote_gdb.listener: listening for remote gdb #0 on port 7000 > **** REAL SIMULATION **** > info: Entering event queue @ 0. Starting simulation... > Segmentation fault
_______________________________________________ m5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
