Yup, Ali's right. InOrder isnt supporting FS mode right now.
It's pretty high on the to-do list (actually, it's next), but the recent time I've spent with Inorder has been with optimizing the simulation speed for inorder (check) and then adding microcode/register window support for InOrder so it can support the other ISAs M5 has description languages for (uncheck). The FS mode "glue logic" should be portable from the O3 and SimpleCPU models (at first pass, all the places where you see #if FULL_SYSTEM in the other models need to be implemented in InOrder) as the ISA specific code *should* be modular enough that it doesn't depend on what CPU model you choose. It's always tough to predict how long it will take to implement a particular, but I'd guess 3-4 weeks of focuded FS implementation would get it done (given you have the other models for reference). On Wed, Mar 2, 2011 at 12:55 PM, Ali Saidi <sa...@umich.edu> wrote: > > On Wed, 02 Mar 2011 12:20:52 +0000, Timothy M Jones < > timothy.jo...@cl.cam.ac.uk> wrote: > >> Hi Korey (primarily), >> >> Could you give me a brief summary of which ISAs work with inorder CPU >> in full-system mode, to the best of your knowledge? I'm just starting >> a new project and the inorder CPU would be a good fit for it, but I >> need to know the constraints on it. >> >> Cheers >> Tim >> > > Korey hasn't responded, but I believe the answer is none. MIPS almost > worked, but it doesn't now. > > Ali > > > _______________________________________________ > m5-dev mailing list > m5-dev@m5sim.org > http://m5sim.org/mailman/listinfo/m5-dev > -- - Korey _______________________________________________ m5-dev mailing list m5-dev@m5sim.org http://m5sim.org/mailman/listinfo/m5-dev
