I definitely intend to do more along these lines, I haven't had a chance in a while and these have been sitting around. I want to get them moving now that the X86_FS O3 regression is in so the don't bit rot too badly.
One thing I didn't mention in the commit message for the decoder change was that this is also a transitional step to combine SE and FS. i'm thinking there will be some state in the decoder which will make it return SE or FS flavors of instructions, which more or less just boils down to the system call instructions and maybe one or two others. Right now these tend to be preprocessor based decisions "#if FULL_SYSTEM" but they could be compile time constant decisions "if (FULL_SYSTEM)" or even run time if FS and SE coexist "if (fullSystem)". fullSystem or something like it would live in the decoder and (I'm imagining) be in scope in the decode function. Gabe _______________________________________________ gem5-dev mailing list gem5-dev@gem5.org http://m5sim.org/mailman/listinfo/gem5-dev