----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: http://reviews.gem5.org/r/1123/#review2460 -----------------------------------------------------------
I appreciate the problem, but I don't think this is the right solution. Compiling all this stuff into gem5 just doesn't seem right. It seems like the real problem is the "from O3_ARM_v7a import *" lines in CacheConfig.py and Simulation.py. Basically the more specialized configuration should import the more general modules and not the other way around. Of course our current structure isn't really set up to make this easy. For now, can we just make the imports conditional on whether someone is actually using the corresponding options? configs/common/CacheConfig.py <http://reviews.gem5.org/r/1123/#comment2861> What about moving the import statement inside the if here? - Steve Reinhardt On April 4, 2012, 7:19 a.m., Korey Sewell wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > http://reviews.gem5.org/r/1123/ > ----------------------------------------------------------- > > (Updated April 4, 2012, 7:19 a.m.) > > > Review request for Default. > > > Description > ------- > > o3: move arm_v7a-like config to the o3 directory > when this is in the Simulation directory, it prevents the running of other > CPU models if O3 isn't > compiled in, which obviously isn't desirable. You can replicate this error by > just compiling > the InOrderCPU model (or SimpleTiming) and trying to run hello world. > There may be a cleaner way to do this without actually compiling in the > arm_v7a-like config into gem5 but > I dont know of one currently > > > Diffs > ----- > > configs/common/CacheConfig.py 570b44fe6e04 > configs/common/O3_ARM_v7a.py 570b44fe6e04 > configs/common/Simulation.py 570b44fe6e04 > src/cpu/o3/O3_ARM_v7a.py PRE-CREATION > src/cpu/o3/SConscript 570b44fe6e04 > > Diff: http://reviews.gem5.org/r/1123/diff/ > > > Testing > ------- > > > Thanks, > > Korey Sewell > > _______________________________________________ gem5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/gem5-dev
