That sounds fine with me. I'm planning to augment how operands are declared to take care of the R15 issue (see my earlier email), but I expect that would be easy to merge in later if most of what's changing is indentation.
Gabe nathan binkert wrote: > For two reasons, I want to have scons import isa_parser.py instead of > having scons exec "python isa_parser.py" > > 1) Gabe wants to access some code in m5.util (and it might be nice to > move some of the stuff in isa_parser.py into m5.util) > > 2) I want to get rid of cpu_models.py and move that code into > SConstruct files (it just makes more sense that way). > > > The problem is that there are a whole lot of global variables in > isa_parser.py which makes things fail when you try to compile two > different ISAs at the same time. Fixing this more or less means > creating a class to wrap all of the global variables and functions. > The manipulation won't be that difficult, but it will most likely > result in a fair bit of churn in the code since I'll have to indent a > bunch of global functions. (In the process, I'll probably also move > some code into m5.util > > Does anyone care if I do this? By the way, a side benefit is that one > could turn isa_parser into a package instead of a single object if > you're getting sick of a 2000+ line file. > > > Nate > _______________________________________________ > m5-dev mailing list > [email protected] > http://m5sim.org/mailman/listinfo/m5-dev > _______________________________________________ m5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/m5-dev
